Citation
Addiction is a brain disease : computational and functional neuroanatomy of substance use disorders using advanced MRI and neurostimulation-induced brain lessons

Material Information

Title:
Addiction is a brain disease : computational and functional neuroanatomy of substance use disorders using advanced MRI and neurostimulation-induced brain lessons
Creator:
Regner, Michael Francis
Place of Publication:
Denver, CO
Publisher:
University of Colorado Denver
Publication Date:
Language:
English

Thesis/Dissertation Information

Degree:
Doctorate ( Doctor of philosophy)
Degree Grantor:
University of Colorado Denver
Degree Divisions:
Department of Bioengineering, CU Denver
Degree Disciplines:
Bioengineering
Committee Chair:
Hunter, Kendall
Committee Members:
Tanabe, Jody L.
Tregellas, Jason R.
Kluger, Benzi
Kheyfets, Vitaly
Shandas, Robin

Record Information

Source Institution:
University of Colorado Denver
Holding Location:
Auraria Library
Rights Management:
Copyright Michael Francis Regner. Permission granted to University of Colorado Denver to digitize and display this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.

Downloads

This item has the following downloads:


Full Text
ADDICTION IS A BRAIN DISEASE:
COMPUTATIONAL AND FUNCTIONAL NEUROANATOMY OF SUBSTANCE USE DISORDERS USING ADVANCED MRI AND NEUROSTIMULATION-INDUCED BRAIN LESIONS
by
MICHAEL FRANCIS REGNER B.A., University of Wisconsin-Madison, 2009 M.D., University of Wisconsin-Madison, 2013
A dissertation submitted to the Faculty of the Graduate School of the University of Colorado in partial fulfillment of the requirements for the degree of Doctor of Philosophy Bioengineering
2019


Copyright © 2019
MICHAEL FRANCIS REGNER, M.D.
ALL RIGHTS RESERVED.
Medicine is an ever-changing science. Care has been taken to confirm the accuracy of the information presented and to describe generally accepted practices. However, the authors, editors, and publisher are not responsible for errors or omissions or for any consequences from application of the information in this book and make no warranty, expressed or implied, with respect to the contents of the publication.
Some drugs and/or medical devices presented in this publication have Food and Drug Administration (FDA] clearance for limited use in restricted research settings. It is the responsibility of the health care provider to ascertain the FDA status of each drug or device planned for use in their clinical practice.
11


This dissertation for the Doctor of Philosophy degree by Michael Francis Regner has been approved for the Bioengineering Program
by
Kendall Hunter (Chair)
Jody L. Tanabe (Co-Advisor)
Jason R. Tregellas (Co-Advisor)
Benzi Kluger Vitaly Kheyfets Robin Shandas
Approved: May 18, 2019


Regner, Michael Francis (Ph.D., Bioengineering)
Addiction is a Brain Disease: Computational and Functional Neuroanatomy of Substance Use Disorders Using Advanced MRI and Neurostimulation-Induced Brain Lesions
Dissertation directed by Professor Jody Tanabe and Professor Jason Tregellas
ABSTRACT
Substance use disorders (SUD) are a group of psychiatric diseases associated with significant mortality, morbidity, economic losses, social- and family-problems, and personal distress. Advances in the ability to image brain structure and function in vivo provide an opportunity to understand addictions. Novel techniques, such as transcranial magnetic stimulation, allow the ability to alter neurocircuitry function for potential diagnostic and therapeutic benefit This dissertation sought to study SUD as diseases of neurocircuitry affecting reward, craving, and goal-oriented behavior. We conducted advanced neuroimaging experiments involving two different populations of SUD: long-term abstinent, severely-dependent cocaine and methamphetamine addicts (Chapters 2 and, 3); and, acutely-abstinent, moderately-dependent cigarette smokers (Chapters 4 and 5 and Appendix 1).
Chapters 2 and 3 investigated structural and functional neuroimaging differences, respectively, in long-term abstinent cocaine and methamphetamine addicts compared to healthy controls. In Chapter 2, we observed that grey matter volumes in stimulant dependence differed compared to healthy controls (n=127), and that these changes differed by sex. We discuss the important implications with regards to sex-differences in the natural history of psychostimulant dependence. In Chapter 3, we investigated brain function as rest by functional connectivity changes amongst large-scale brain networks in a subset of the Chapter 2 population (n=100). We found that after long-term abstinence, large-scale brain networks
IV


thought to be involved in the pathogenesis of addiction followed a pattern of increased "top-down” cognitive control, which may reflect the necessary increased executive control over habit and reward systems to maintain disease remission. Stimulant dependence also demonstrated greater global efficiency and lower local efficiency amongst large-scale brain networks compared to healthy controls, suggesting abnormal brain organization despite long-term abstinence.
Chapters 4, 5, and Appendix 1 introduce and report results on our phase 1, sham-controlled, single-blinded, randomized clinical trial fwww.ClinicalTrials.gov identifier: NCT 02590640) to investigate the effect of inhibitoiy insular transcranial magnetic stimulation (TMS) on cigarette cravings and brain function in acutely abstinent, moderately-dependent smokers. Compared to sham treatment, inhibitory TMS targeting the right insula reduced self-reported cigarette craving and decreased brain activity responses to visual cigarette cues in primary sensorimotor cortices, supplementary motor area, and right anterior insula. These findings provide proof-of-concept of a potential neuroanatomical target for smoking cessation therapy. While the results should be considered preliminary, they provide hope that TMS could be developed as a treatment strategy to help reduce the burden of cigarette addiction.
Overall, these neuroimaging studies in two different populations of substance use disorders provide evidence that addiction is a brain disease with endogenous (i.e., increased top-down control, Chapter 3) and exogenous (i.e., therapeutically imposed, Chapter 5) mechanisms of remission and treatment
The form and content of this abstract are approved. I recommend its publication.
Approved: JodyL. Tanabe
Jason R. Tregellas
v


For Emilie -
who is my critic when I feel I have nothing left to perfect, and who sings my praises when I feel I have nothing worthy of praise.


ACKNOWLEDGEMENTS
Firstly, I would like to thank my dissertation advisor and mentor, Dr. Jody Tanabe. She gave me her time and an opportunity when I was a brand-new medicine intern with nothing to offer. Then, she challenged me to pursue new directions and pushed me to develop as a scientist as my skills developed. Over the past six years Jody has been a great role model and friend. For this, I am forever grateful.
I am particularly grateful to my dissertation advisor, Dr. Jason Tregellas, without whom this work would not have been possible. Jason has been invariably generous with his time, expertise, and resources.
I am grateful to several others without whom this work would not have been possible: Dr. Kendall Hunter, Dr. Benzi Kluger, Dr. Vitaly Khetfets, Dr. Robin Shandas, Dr. Deb Glueck, and Dr. Gerald Dodd III. I thank the Departments of Bioengineering and Radiology for their unwavering support I am particularly grateful to my radiology co-residents, who were invariably supportive of the call trades and schedule changes facilitating this research. I thank my first research mentor, Dr. Jack Jiang, and my good friend and medical school classmate/roommate, Dr. Matthew Hoffman.
I thank my wife, Dr. Emilie Regner, for her constant support We have both pursued unusual pathways as physician-scientists. I am ever-fortunate to take this journey with you.
I thank my late mother, Claire, who died from stage IV lung cancer (and indirectly, a lifetime of cigarette addiction) in December 2017. It is with a heavy heart that I wish you could have seen the final product of this effort I thank my family, particularly my father Scott and mother-in-law Kaye, for their constant support
Lastly, but most importantly, I would like to thank our patients and research volunteers. Addictions are often misguidedly considered a "moral failing,” conferring stigma and guilt that impairs awareness, treatment, and scientific progress. We must nevertheless persist and continue to shine light on these psychiatric diseases. Our field is indebted to those individuals who volunteered to walk into the light of scientific examination, if only for a few hours.
Vll


DECLARATION OF ORIGINAL WORK
I affirm that all work in this Doctoral Dissertation is my original work. Further, I confirm that all writing is my own writing. Work from others has been cited appropriately.
Michael Francis Regner oate
PhD Candidate
viii


TABLE OF CONTENTS
CHAPTER
1. INTRODUCTION......................................................1
1.1. Neuroimaging in Substance Use Disorders.....................3
1.2. Substance Use Disorders as a Class of Heterogeneous Neurocircuitry Diseases ...4
1.3. Dissertation Narrative Outline........................................6
2. SEX DIFFERENCES IN GREY MATTER CHANGES AND BRAIN-BEHAVIOR
RELATIONSHIPS IN STIMULANT DEPENDENCE.......................................9
2.1. Abstract..............................................................9
2.2. Introduction.........................................................10
2.3. Materials and Methods................................................12
2.4. Results..............................................................17
2.5. Discussion...........................................................23
2.6. Conclusion...........................................................26
2.7. Acknowledgements.....................................................27
3. TOP-DOWN NETWORK EFFECTIVE CONNECTIVITY IN ABSTINENT SUBSTANCE
DEPENDENT INDIVIDUALS.......................................................28
3.1. Abstract.............................................................28
3.2. Introduction.........................................................29
3.3. Materials and Methods................................................33
3.4. Results..............................................................40
3.5. Discussion...........................................................45
3.6. Conclusion...........................................................54
3.7. Acknowledgements.....................................................54
IX


4. THE INSULA IN NICOTINE USE DISORDER: FUNCTIONAL NEUROIMAGING AND
IMPLICATIONS FOR NEUROMODULATION.........................................55
4.1. Abstract...........................................................55
4.2. Introduction.......................................................56
4.3. Insular Role in Nicotine Use Disorder..............................58
4.4. Implications of Insular Role in Nicotine Use Disorder on Neuromodulatory
Therapeutic Development............................................71
4.5. Conclusion.........................................................77
4.6. Acknowledgements...................................................78
5. INSULAR INHIBITORY NEUROMODULATION IN SMOKERS DECREASES CIGARETTE
CRAVINGS AND BRAIN RESPONSES TO CIGARETTE CUES: A RANDOMIZED CONTROLLED TRIAL.........................................................79
5.1. Abstract...........................................................79
5.2. Introduction.......................................................80
5.3. Materials and Methods..............................................82
5.4. Results............................................................95
5.5. Discussion........................................................103
5.6. Conclusion........................................................108
5.7. Acknowledgements..................................................108
5.8. Supplement: Justification of Analysis Approach for Primary and Secondary
Outcomes..........................................................109
6. CONCLUSION..............................................................114
6.1. Specific Knowledge Gaps Addressed.................................114
6.2. Limitations.......................................................116
6.3. Future Work.......................................................117
x


6.4. Concluding Remarks..........................................118
REFERENCES...............................................................119
APPENDIX.................................................................138
xi


TABLE OF FIGURES
FIGURE
1-1. Simplified scheme illustrating how common drugs of abuse modulate the reward circuit.
Modified with permission from Nestler’s Molecular Neuropharmacology (Nestler et al., 2015).... 2
1- 2. Simplified illustration of the natural history of substance use disorders. Active substance
dependence (red) is characterized by sequential stages of binge/intoxication, withdrawal/negative affect, and craving/preoccupation (Koob and Volkow, 2010) that repeat a variable number of cycles, represented by m. Active disease episodes may be interrupted by periods of short- or long-term abstinence, represented by n. Relapse is indicated in blue. Note that at every stage of potential or active disease (top, in white) there is a possible avenue towards disease remission (bottom, in grey). Used with permission from Regner et al. (2016).5
2- 1. Sample population inclusion and exclusion criteria..........................................12
2-2. T-value map illustrating statistically significant greater grey matter volumes in healthy control
women compared to stimulant dependent women, after controlling for age and brain size
(p< 0.001)............................................................................................18
2-3. T-value map illustrating statistically significant greater regional grey matter volumes in control women compared to control men, after controlling for age and brain size (p<0.001)......................19
2-4. Left, T-value map of statistically significant negative correlations found on whole-brain level between stimulant dependence symptom count and grey matter volume within the bilateral nucleus accumbens (the so-called "reward nucleus”) in SDI. Right, negative correlation between the substance dependence symptom count and nucleus accumbens total volume as defined by
ROI................................................................................................20
2-5. Sex by group by behavior interactions on GMV. Left, scatterplots illustrating the sex by group interaction on the correlation between approach and GMV in the bilateral middle frontal gyri (dorsolateral prefrontal cortex). Right, scatterplots illustrating the sex by group interaction on the correlation between impulsivity and GMV in the left superior temporal gyrus and left insula.
xn


Middle, clusters of whole-brain significance demonstrating sex by group by impulsivity (red) and sex by group by approach (green) interactions..............................................21
2- 6. Boxplots illustrating the total grey matter volume in each subpopulation for each region-of-
interest. The top and bottom edges of the boxes indicate the third and first interquartiles, respectively. Lines in the middle of the boxes indicate the medians. Whiskers above and below the boxes indicate the 90th and 10th percentiles, respectively. Points above and below the whiskers indicate the 95th and 5th percentiles, respectively. Significance was tested using two-way ANCOVA. C = healthy control individuals. SDI = stimulant dependent individuals.........22
3- 1. Simplified illustration of the natural history of substance use disorders. Active substance
dependence is characterized by sequential stages of binge/intoxication, withdrawal/negative affect, and craving/preoccupation (Koob and Volkow, 2010)that repeat a variable number of cycles, represented by m. Active disease episodes may be interrupted by periods of short- or
long-term abstinence, represented by n.....................................................31
3-2. Diagram of the fMRI BOLD signal pre-processing, network-of-interest definition, and effective
connectivity analysis pipelines............................................................36
3-3. Effective connectivity network density graphs of SDI and controls. Thickness of each line
corresponds to the Granger causal strength and color corresponds to the efferent network. SDI
network density was significantly greater than healthy controls (p<0.001)...................42
3-4. Directed GC matrices for SDI (left) and controls (right). Colorbar corresponds to logarithm of F
values......................................................................................43
3-5. Effective connectivity matrix illustrating the FDR-corrected group differences between SDI and healthy controls. Colorbar corresponds to p-values. White arrows indicate Granger causal
direction................................................................................43
3-6. Left, NOI demonstrating significantly greater Granger causal relationships in SDI compared to
controls. Illustrated are the NOI relationships that differed between groups; colors correspond to NOI labels on right. Printed numbers in upper left corner of each slice correspond to Z-axis coordinates in MNI space. Right, Granger causal relationships demonstrating greater directed information flow in SDI compared to controls (FDR corrected). RECN, right executive control
xm


network; dDMN, dorsal default mode network; BGN, basal ganglia network Values indicate Granger causal p-values. Middle, whole brain illustrations of NOI identified. Arrows reflect
Granger causal influence....................................................................44
3-7. Correlations between mean beta value within the RECN and dDMN with impulsivity, approach, and negative affect metrics. Solid black lines indicate the linear regression, solid colored lines indicate the 95% confidence interval, and colored shaded regions indicate the prediction interval. Each point reflects a single participant’s mean beta within a given network, and their score on the given behavioral metric................................................................45
3- 8. Global efficiency (left) and local efficiency (right) in SDI and controls as a function of network
cost........................................................................................46
4- 1. Connectivity-based signal flow diagram of anterior insular control of bottom-up versus top-down
mechanisms of salience. The right dorsal anterior insula is involved in processing salience of externally-oriented stimuli and it is correlated with the executive control network (an externally-directed system). The right ventral anterior insula is involved in processing salience of internally-oriented stimuli and it is correlated with the default mode network (an internally-
directed system)...........................................................................57
4-2. Diagrammatic illustration of the natural history of nicotine use disorder. We attempt to
synthesize the neuroimaging literature of nicotine use disorder into four pharmacologically- and behaviorally-informed stages of nicotine use disorder. Acute nicotine exposure reflects the acute pharmacologic action on neural circuitry. Chronic nicotine exposure reflects pharmacologic dependency and results from repeated acute nicotine exposure; it manifests as maladaptive changes in reward, salience, and executive control circuitry. Acute abstinence provides a model for understanding the neural basis of the nicotine withdrawal syndrome and craving. Long-term abstinence serves as a model of neuroplastic recovery from nicotine use disorder. Relapse (bottom arrows) is the mechanism by which active disease is maintained. Associated
neuroimaging findings are summarized in Table 1............................................60
4-3. Differential resting-state connectivity of the dorsal (left) and ventral (right) right anterior insula, using the Human Connectome Project Connectome Workbench (n = 1206), uncorrected. Dorsal
xiv


right anterior insula is strongly connected with frontoparietal regions involved in executive control; in contrast, ventral right anterior insula connectivity is strongly connected with default mode network regions involved in internal feelings and self-referential processing. Note that the externally-directed system (dorsal anterior insula and executive control network) and internally-directed system (ventral anterior insula and default mode network) are inversely
correlated, consistent with mutual inhibition (see Figure 4-1)...............................70
4-4. Finite element model results of the predicted electromagnetic field produced by a standard
superficial planar transcranial magnetic stimulation 70mm figure-of-eight coil targeting the right anterior insula. Predicted current flux density (left) and normalized absolute value of the electric field (right) illustrating the pattern of energy deposition. Maximum energy is deposited in the scalp, superficial soft tissues, and cerebrospinal fluid due to high tissue conductivities. This illustrates the difficulty in targeting deep structures, such as the insula or anterior cingulate cortex. This is consistent with model results of insular targeting reported by Pollatos and colleagues (Pollatos and Rammer, 2017). Created using SimNIBS version 2.1.1 with right anterior
insular target [36,10, -6] in MNI space and default parameters (Thielscher et al., 2015)......75
5-1. CONSORT enrollment diagram for this phase 1 human trial. Careful attention was paid to
maintaining all participants’ blinding to group assignment throughout each study visit........84
5-2. Timeline of an individual participant’s research appointment. Treatment included either
rigorous sham or deep LF-rTMS targeting the right anterior insula. The time delay between the end of treatment and behavioral survey was <5 minutes. The time delay between the end of
treatment and post-treatment MRI was 11.5 ± 1.9 minutes (mean ± SD)...................85
5-3. Self-reported cigarette craving (QSU-Brief) by group and time. LF-rTMS causally reduced self-reported craving compared to sham by per protocol two-way repeated-measures ANOVA of group x time (p = 0.033). No statistically significant difference was observed in craving after
sham treatment, although there is a trend towards placebo effect (p = 0.084)..........99
5-4. Pre-treatment main effects of cigarette cue exposure compared to neutral cues, collapsed across groups. Brain activity significantly increased in the bilateral dorsolateral prefrontal cortex, primary visual cortex, and higher-level visual cortex during cigarette cues compared to neutral
xv


cues (red color bar). Brain activity significantly decreased in the bilateral posterior cingulate gyrus and precuneus during cigarette cues compared to neutral cues (blue, p < 0.001). Results were corrected for multiple comparisons using familywise p < 0.05, voxelwise p < 0.005, cluster
extent k > 464 voxels......................................................................100
5-5. Whole-brain interaction effect using a per protocol 2x2 mixed factorial interaction effect of time (within-subjects: post-treatment - pre-treatment) x group (between subjects: LF-rTMS - sham). Significantly decreased cigarette cue brain response after LF-rTMS compared to sham was observed in the right primary sensorimotor cortex, bilateral supplementary motor cortex (premotor), and right dorsal anterior insula (blue, p < 0.001). No significant increased cigarette cue brain response was observed after LF-rTMS compared to sham. Results were corrected for multiple comparisons using familywise p < 0.05, voxelwise p < 0.005, cluster extent k > 464
voxels.....................................................................................101
5-6. FEM-informed ROI analysis of brain activity responses using a per protocol analysis of
covariance. ROI beta values and significance levels were extracted using the MarsBar toolbox after accounting for nuisance covariates, using the SPM12 design used for whole-brain analysis.
...........................................................................................102
5-7. Correlation between absolute change in QSU-Brief self-reported craving (post-treatment > pretreatment) and change in brain activity responses to cigarette cues [(post-treatment > pretreatment) x (cigarette cues > neutral cues)] within the LF-rTMS group defined per protocol. Results were corrected for multiple comparisons using familywise p < 0.05, voxelwise p < 0.005, cluster extent k > 464 voxels.......................................................................103
xvi


CHAPTER I
INTRODUCTION
Human use of psychoactive drugs is described in some of the earliest human historical records. Pathologic use of ingested substances was described in early Greek and Roman literature; for example, the Greek scientist Aristotle (384 BC - 322 BC) provided the first detailed description of alcohol withdrawal, while the Roman physician Celsus (25 BC - 50 AD) precociously described alcohol dependence as a systemic disease (Crocq, 2007). Historically, human use of psychoactive substances can be categorized into four major categories:
1. Spiritual or religious ceremonial use, such as imbibing the wine during Catholic mass;
2. Medical or therapeutic use, such as modern prescription opioids for pain control;
3. Socially acceptable use for pleasure and/or socialization, such as cigarette smoking amongst construction workers;
4. Abuse or dependence of a substance, in a manner discordant with pro-social norms and/or despite negative consequences ("addiction”).
This dissertation focuses on the fourth category of use, addiction, which by the Diagnostic Statistical Manuel-5th edition (DSM-5) is categorized under "substance use disorders.” Substance use disorders (SUD) in the Diagnostic Statistical Manuel-5th edition (DSM-5) are a heterogeneous set of psychiatric diseases characterized by physical dependence, diminished control over substance use despite negative consequences, and cravings to use the substance (American Psychiatric Association, 2013). SUD are characterized in severity by the number of symptoms, with 2 or 3 symptoms representing mild and greater than 6 symptoms representing severe SUD (American Psychiatric Association, 2013).
1


All drugs of abuse cause the release of dopamine into the nucleus accumbens (Di Chiara and Imperato, 1988; Koob, 1992), producing a rewarding "high” that positively reinforces consumption (Figure 1-1). The mesocorticolimbic reward circuit consists of dopaminergic
Nicotine
Alcohol
Opiates —
Opioid
peptides
VTA__________
interneuron
lAlcohoH-^Tv
f + DA
I Nicotinel
Glutamate + inputs
â– GABA
|Stimulants|------ |Opiates|^_^
-DA-
Glutamate inputs (from cortex)
|Alcohol|
/1 Cannabinoids] IPCPl
VTA
NAc
Figure 1-1. Simplified scheme illustrating how common drugs of abuse modulate the reward circuit Modified with permission from Nestler’s Molecular Neuropharmacology (Nestler et al., 2015).
neurons with cell bodies in the ventral tegmental area of the midbrain and terminal projections in the ventral striatum (nucleus accumbens), prefrontal cortex, and limbic/paralimbic regions. When this circuit is stimulated, dopamine is released into synaptic clefts in these terminal regions. Dopamine binds to postsynaptic receptors and is subsequently rapidly re-sequestered within the presynaptic axonal boutons by the dopamine transporter (DAT), terminating the reward signal. Cocaine (studied in Chapters 2 and 3), for example, artificially enhances dopamine signaling by inhibiting the dopamine transporter, thus prolonging dopaminergic activity. Nicotine (studied in Chapters 4 and 5 and Appendix 1) acts directly on dopaminergic
2


reward neurons in the VTA as well as modulating interneurons. Over time, drug-use and drug-related stimuli become increasingly associated with the hedonic effects of drug induced dopamine transmission, and these stimuli serve as incentivized cues that trigger craving, anticipatory euphoria, or even symptoms of withdrawal (O'Brien et al., 1992). Drug dependence causes neurocircuitry adaptations that develop in response to repeated administration, as the brain’s effort to maintain overall neurotransmission homeostasis. Abrupt cessation of drug use after the development of addiction, in turn, unmasks withdrawal syndromes because the forces that keep these neurocircuitry alterations balanced are no longer present
1.1. Neuroimaging in Substance Use Disorders
Neuroimaging has fostered a paradigm shift in the understanding of SUD over the past two decades. Although once considered a "moral failure,” in recent decades neuroimaging has demonstrated that SUD are brain diseases with predictable alterations in neuro circuitry underlying reward, craving, learning, and cognitive control (Tanabe etal., 2019).
Neuroimaging provides both structural and functional observations of altered neuroanatomy. For example, quantitative structural neuroimaging using voxel-based morphology (see Chapter 2) has demonstrated lower grey matter tissue volumes compared to healthy adults in the anterior cingulate, dorsolateral prefrontal, medial prefrontal, and parietal cortex in people dependent upon cocaine (Connolly et al., 2013; Regneretal., 2015), amphetamine (Daumann et al., 2011; Mackey and Paulus, 2013; Tanabe et al., 2009b), and nicotine (Brody etal., 2004; Pan etal., 2013).
In addition to brain structural changes, neuroimaging has furthered our understanding of the functional mechanisms in the brain that promote and maintain SUD. For instance, intravenously administered nicotine (Stein et al., 1998) and cocaine (Breiter et al., 1997) both acutely increase blood-oxygen-level-dependent (BOLD) fMRI signal in the striatum, amygdala, and prefrontal cortex. This pattern of reward-circuitry activation after acute drug
3


administration has been shown to be altered even in the absence of drug administration. A large meta-analysis investigated the brain changes associated with monetary reward anticipation and reward outcome in 643 individuals with addictive behaviors (including all common drugs of abuse, gambling disorder, and gaming disorder) compared to 609 healthy control individuals (Luijten et al., 2017). They observed that during reward anticipation, people with substance, gambling, and gaming addictions exhibited ventral striatum hyperactivation. During reward outcome, they exhibited decreased dorsal striatum hypoactivation. This suggests that brain processing of reward signals is perturbed in addiction, even for rewards unrelated to the addiction. Neuroimaging has also demonstrated significant long-term shifts in the cognitive function in SUD in the absence of drug administration. A parametric meta-analysis of fMRI cue-reactivity studies discovered that nicotine, alcohol, and cocaine individuals have abnormally increased brain responsivity to drug cues compared to neutral cues in the ventral striatum, anterior cingulate cortex, and amygdala - regions involved in reward and emotional regulation (Kuhn and Gallinat, 2011). Although there are compelling common mechanisms of SUD across different substances and across individuals, significant heterogeneity exists.
1.2. Substance Use Disorders as a Class of Heterogeneous Neurocircuitry Diseases
Substance use disorders are understood to be chronic relapsing-remitting diseases that vary by drug, stage, and individual factors. First, SUD demonstrate brain and behavioral differences by drug and mode of use. For example, Hanlon et al. studied individuals dependent upon cocaine (n=55), alcohol (n=53), and nicotine (n=48) (Hanlon etal., 2018), and reported that each substance use disorder demonstrated overlapping yet anatomically distinct "hot spot” cue-induced craving fMRI brain response maps. This suggests that neuromodulatory efforts to therapeutically alter craving in SUD may need to be drug-specific. Second, there is significant evidence that each stage of addiction (Figure 1-2) is associated with distinct patterns of brain functional changes observed through neuroimaging compared to healthy controls (Koob and
4


Disease Progression / Natural History
Figure 1-2. Simplified illustration of the natural histoiy of substance use disorders. Active substance dependence (red) is characterized by sequential stages of binge/intoxication, withdrawal/negative affect, and craving/preoccupation (Koob and Volkow, 2010) that repeat a variable number of cycles, represented by m. Active disease episodes may be interrupted by periods of short- or long-term abstinence, represented by n. Relapse is indicated in blue. Note that at every stage of potential or active disease (top, in white) there is a possible avenue towards disease remission (bottom, in grey). Used with permission from Regner et al. (2016).
Volkow, 2010). Third, and lastly, there are numerous individual factors that contribute to and modulate the pathology of SUD. These include genetic predispositions, socio-demographics, problems of inhibitory control, co-morbid diagnoses (e.g., conduct disorder), sibling use of substances, family history of addiction (i.e., upbringing independent of genetic inheritance), neighborhood poverty and disorganization, and even locoregional laws and norms (American Psychiatric Association, 2013; Beyers etal., 2004; Conger, 1997; Hawkins etal., 1992; Kendler etal., 2013; Samhsa, 2018; Scaramella and Keyes, 2001; Stone etal., 2012; von Sydow etal., 2002).
5


1.3. Dissertation Narrative Outline
The overarching technical theme heavily influencing each chapter is the signal processing technique for structural and function brain MRI, including statistical parametric mapping to make hypothesis-based inferences. For each analysis, Tl-weighted images are segmented into tissue probability maps, including maps of grey matter, white matter, cerebrospinal fluid, bone, soft tissue, and air. These are non-linearly normalized into a standard space. Non-linear "warps” (forward and backward deformation vector fields across the field-of-view) are computed from native to standard space. Time-varying functional images are slicetime and motion-corrected, then forward-warped into standard space. Ultimately, generalized linear models are used to compute individual- and group-level contrast maps (T-statistic or F-statistic maps) for task-based fMRI, or alternatively connectivity maps for resting-state fMRI (either through spatiotemporal independent component analysis or seed-based whole-brain correlations).
The goal of this dissertation was to use advanced MRI-based neuroimaging and signal processing methods to address gaps in the literature regarding two different populations of SUD at different points along the spectrum of disease severity: long-term abstinent, severely-dependent cocaine and methamphetamine addicts (Chapters 2 and 3); and, acutely-abstinent, moderately-dependent cigarette smokers (Chapter 4 and 5 and Appendix 1). While these two populations differ in severity, they also differ in primary drug of abuse (psychostimulants versus nicotine) and stage of disease (chronic remission versus acute withdrawal).
Chapters 2 and 3 investigate structural and functional neuroimaging differences, respectively, in long-term abstinent psychostimulant (cocaine and methamphetamine) addicts compared to healthy controls. In Chapter 2, we report one of the largest samples (n=127) demonstrating that the differences in grey matter volumes in stimulant dependence compared to healthy controls after long-term abstinence differ by sex. Although limited by the cross-
6


sectional nature of the study, we discuss the important implications with regards to sex-differences in the natural history of psychostimulant dependence. In Chapter 3, we investigated resting-state functional connectivity changes in stimulant dependence compared to healthy controls in a subset of the population reported in Chapter 2 (n=100). We show that large-scale brain networks defined by independent component analysis follow a pattern of increased top-down cognitive control in long-term abstinent psychostimulant addicts compared to healthy controls, which may reflect increased executive control over habit and reward systems promoting remission. Stimulant dependence also demonstrated greater global efficiency and lower local efficiency amongst large-scale brain networks, suggesting abnormal brain organization despite long-term abstinence.
Chapters 4 and 5 and Appendix 1 report functional neuroimaging changes in acutely abstinent smokers after a phase 1, sham-controlled, single-blinded, randomized clinical trial fwww.ClinicalTrials.gov identifier: NCT 02590640) designed to test the hypothesis that inhibiting the right anterior insula will decrease cigarette cravings and alter brain activity and connectivity. Chapter 4 serves is a review introducing and motivating our trial of inhibitory TMS in acutely abstinent, moderate smokers. We specifically review compelling neuroimaging and neuroscience evidence that the insula may serve as an important therapeutic target for promoting smoking cessation, and we discuss possible mechanisms to pursue such neuromodulation. We introduce a useful method of non-invasively modulating brain activity: transcranial magnetic stimulation (TMS). Chapter 5 reports the main results of the clinical trial: inhibiting the right anterior insula in smokers using low-frequency deep TMS results in decreased cigarette craving and decreased brain fMRI responses to cigarette cues compared to sham. Appendix 1 reports some but not all exploratory data collected during the trial, including changes in resting-state connectivity after treatment compared to sham.
7


Each individual dissertation chapter addresses a specific yet important gap in the literature on the structural and functional computational neuroanatomy of SUD. Together, however, these neuroimaging studies at different points along the spectrum of substance use disorders provide support for a conclusion still controversial in the scientific literature (Lewis, 2018): addiction is a brain disease. We further extend this conclusion by investigating possible endogenous (i.e., increased "top-down” executive control, Chapter 3) and exogenous (i.e., therapeutically imposed, Chapter 5) mechanisms of disease remission and treatment While these results span different subtypes of substance use disorders, they provide hope that both endogenous and exogenous treatment strategies are possible to help reduce the burden of addiction.
8


CHAPTER II
SEX DIFFERENCES IN GREY MATTER CHANGES AND BRAIN-BEHAVIOR RELATIONSHIPS IN
STIMULANT DEPENDENCE
2.1. Abstract
This first chapter investigated whether sex modulates the effects of stimulant dependence on grey matter volumes (GMV) in long-term abstinence. We further sought to characterize how sex modulates brain-behavior relationships between GMV and specific behavioral measures, such as drug symptom count, behavioral approach, and impulsivity. In this prospective case-control study, 127 age- and sex-matched participants (68 controls [28F/40M] and 59 SDI [28F/31M]) underwentTl-weighted SPGR-IRbrain MRIs on a 3T system. Images were segmented using voxel-based morphometry MATLAB-based software. After adjusting for age, education, and head size, sex by group interactions and main effects were analyzed over the whole brain using ANCOVA, thresholded at p<0.05, corrected for multiple comparisons with family-wise cluster correction. Drug symptom count and behavioral measurements were correlated with whole brain GMV and five a priori regions-of-interest based on extant literature. Sex by group interactions on GMV were significant in numerous regions (p<0.001). Compared to female controls, female SDI had significantly less GMV in widespread brain regions (p<0.001). There were no significant GMV differences in male controls versus male SDI (p=0.625). Drug symptom count negatively correlated with nucleus accumbens GMV in women (left: r=-0.364, p=0.047; right: r=-0.407, p=0.031) but not men (left: r=-0.063, p=0.737; right: r=-0.174, p=0.349). Behavioral approach (p=0.002) and impulsivity (p=0.013) correlated negatively with frontal and temporal GMV changes in female SDI but not in other groups, demonstrating a sex by group interaction. Vast GMV changes in SDI were
9


observed in women but not men after prolonged abstinence. Sexual dimorphism in drug-related neuroanatomical changes and brain-behavior relationships may be a mechanism underlying the different clinical profiles of addiction in women compared to men. Future structural neuroimaging and clinical studies on substance use disorders should account for the modulatory effects of sex.
This chapter has been peer-reviewed and published: Regner MF, Dalwani M, Sakai JT, Yamamoto D, Perry RI, Honce J, Tanabe J. Sex modulates grey matter changes and brain-behavior relationships in substance dependence. Radiology. 2015; 277(3):801-812. PMID: 26133201.
2.2. Introduction
Substance use disorders are common, with lifetime prevalence estimated to be 10.3% in the United States (Miller and Hendrie, 2009). Understanding the neurobiology of substance dependence is requisite to advancing treatments. Neuroanatomical changes in drug addiction have been studied extensively using voxel-based morphometry (Ersche et al., 2013). Structural changes have been observed in the orbitofrontal cortex (OFC), medial frontal gyrus (MedFG), anterior cingulate gyrus (ACG), insula, and nucleus accumbens in individuals who abuse stimulants (Ersche etal., 2013; Koob and Volkow, 2010). In the largest meta-analysis of stimulant dependence to date, Ersche et al. (Ersche et al., 2013) reported significant decreases in gray matter (GM) in the insula, ventromedial prefrontal cortex, inferior frontal gyrus, ACG, and anterior thalamus. GM changes have also been studied adult sibling pairs, of whom one sibling was dependent on stimulants and the other had no stimulant dependence history with an age- and sex-matched control group (Ersche et al., 2012). That study revealed changes in limbic and sensory areas in both members of the sibling pair compared to controls, suggesting that GM volume changes may predate addiction and could potentially be an endophenotype for substance use disorder.
10


Few previous studies have investigated the role of sex on changes in brain structure in stimulant dependence. This is surprising considering the well-characterized sex differences in clinical presentation and natural history of stimulant addiction (Becker et al., 2012; Perry et al., 2013). Women exhibit a telescoping clinical course compared to men in that they begin cocaine or amphetamine use at earlier ages (Becker etal., 2012; Griffin etal., 1989; Mendelson etal., 1991), show accelerated escalation of drug use (Brady and Randall, 1999; Greenfield et al., 2010; Lynch, 2006), report more difficulty quitting (Back etal., 2005; Lynch, 2006), and upon seeking treatment report using larger quantities of these drugs compared to men (Becker et al., 2012; Kosten et al., 1993). Neuroendocrine factors have been hypothesized to underlie an accelerated clinical course (Becker et al., 2012). Another hypothesis is that compared to men, women respond differently to stress which influences drug related behavior (Potenza et al., 2012). However, scant evidence exists for a neuroanatomical correlate of these clinical differences. Many studies recruit primarily men to exclude confounding sex effects (Barros-Loscertales etal., 2011; Fein etal., 2002; Franklin etal., 2002), while other studies do not include sex as a factor in their analyses of GM in stimulant dependence (Ersche et al., 2013; Sim et al., 2007; Tanabe et al., 2009a). In fact, only two studies have described structural differences between sexes in stimulant dependence (Rando et al., 2013; Tanabe et al., 2013). Rando et al. (Rando etal., 2013) reported lower GMV in the left inferior frontal gyrus, insula, superior temporal gyrus, and hippocampus in female SDI compared to female controls while male SDI exhibited less GMV in the precentral gyrus and mid cingulate gyrus compared to male controls. However, this study was significantly limited by the potential effects of recent alcohol use (mean 87 drinks in the month prior to scanning) and lack of long-term abstinence (mean 3 weeks of abstinence prior to scanning), allowing acute effect of substances to skew results. Tanabe et al. (Tanabe et al., 2013) reported differential effects of sex on insular volumes in SDI: female SDI had smaller insulae, whereas male SDI had larger insulae. This study was limited by
11


the small sample size (28 SDI) and rudimentary methodology (using FreeSurfer to estimate GMV). The paucity of large, prospective, well-controlled studies to investigate long-term sex differences associated with abstinent stimulant dependence is addressed by this study.
This study investigated whether sex modulates the effects of stimulant dependence on grey matter volumes (GMV) in long-term abstinence. We further sought to characterize how sex modulates brain-behavior relationships between GMV and specific behavioral measures, such as drug symptom count, behavioral approach, and impulsivity.
2.3. Materials and Methods
2.3.1. Subjects
One hundred twenty seven individuals including 68 controls (28F/40M) and 59 (28F/31M) SDI were prospectively recruited (Figure 2-1). Controls were similar to SDI on age
Figure 2-1. Sample population inclusion and exclusion criteria.
and sex. Demographic information is reported in Table 2-1. SDI were recruited from a residential treatment program at the University of Colorado School of Medicine Addiction Research and Treatment Services. Inclusion criteria for SDI: lifetime dependence on stimulants
12


Table 2-1. Demographic description of the sample population. Data are presented as mean ± SD where appropriate.
Healthy Controls SDI p-value
Men Women Men Women Group Sex GxS
Sample Size* 40 28 31 28 NS NS NS
Age (yrj* 33.2 ±9.9 31.9 ±7.6 36.7 ± 7.6 33.7 ±7.9 NS NS NS
Right-handedness* 92.5% 89.3% 96.7% 89.3% NS NS NS
Education (yrj* 14.0 ± 1.6 15.4 ± 1.1 12.6 ± 1.5 12.2 ± 1.7 <0.001 NS 0.001
Years of Use (yr) - - 16.1 ± 7.0 15.2 ± 8.0 - NS -
Abstinence (mo)+ - - 17.5 ± 15.6 11.2 ± 7.9 - NS -
Drug symptom countt - - 33.6 ± 14.3 25.6 ± 10.2 - 0.015 -
Negative affect* 16.2 ± 6.7 14.3 ± 3.5 20.8 ± 9.0 22.1 ±7.1 <0.001 NS NS
Behavioral approach 39.3 ± 4.4 35.6 ±8.3 42.1 ± 5.0 44.3 ± 4.7 <0.001 NS 0.005
Impulsivity 59.1 ±8.7 55.2 ±7.1 69.6 ± 10.7 73.7 ± 11.3 <0.001 NS 0.022
Percent Satisfying Dependence Criteria*
Stimulants - - 100% 100% - NS -
Cocaine - - 67% 41% - NS -
Amphetamine - - 83% 89% - NS -
Nicotine - - 73% 74% - NS -
Alcohol - - 67% 67% - NS -
Cannabis - - 50% 37% - NS -
Opiates - - 37% 19% - NS -
Club drugs - - 10% 7% - NS -
Sedatives - - 10% 0% - NS -
Hallucinogens - - 10% 4% - NS -
(methamphetamine, cocaine, or amphetamine-class substances) (DSM-IV). Control subjects were recruited from the community and excluded if dependent on alcohol or other drugs of abuse excluding tobacco. Exclusion criteria for all subjects: depression within the last two months, psychosis, neurological illness, prior head trauma resulting in greater than fifteen minutes loss of consciousness, prior neurosurgery, positive HIV status, diabetes, hepatitis C, bipolar disorder, other major medical illness, inability to tolerate MRI, IQ < 80, positive urine screen (AccuTestâ„¢), or positive saliva screen (AlcoScreenâ„¢). All participants provided written informed consent approved by the Colorado Multiple Institutional Review Board.
13


2.3.2. Structured Interviews for Inclusions and Exclusions
Composite International Diagnostic Interview-Substance Abuse Module (CIDI-SAM) is a computerized structured interview that assesses substance dependence diagnoses and symptoms for 11 different drugs of abuse (Cottier et al., 1989). All subjects were administered the CIDI-SAM to verify stimulant dependence in the SDI group and to exclude controls with abuse or dependence on substances other than tobacco.
Diagnostic Interview Schedule version IV (CDIS-IV) is a computerized structured interview used to screen for psychiatric disorders (Robins et al., 1995a). All subjects were administered the CDIS-IV to exclude those with lifetime psychoses, lifetime bipolar disorder, or major depressive disorder in the last two months.
The CIDI-SAM interview (Gelhorn et al., 2008) computes four abuse (i.e., legal problems due to drug) and seven dependence (i.e., uncontrolled substance use escalation) symptoms for each drug class. Drug use severity was calculated by adding abuse and dependence symptom counts. This approach is consistent with the single set of clinically relevant criteria in DSM-V, which was released after data collection for the current study.
The Barratt Impulsiveness Scale is a 30-item self-reported questionnaire used to quantify impulsiveness (Patton and Stanford, 1995). Participants rate whether phrases and words describing aspects of impulsivity are self-descriptive. The Behavioral Activation System (BAS) scale is a 13-item self-reported questionnaire used to measure responsiveness of motivational systems (Campbell-Sills etal., 2004; Carver and White, 1994); this quantifies positive affective and approach response tendencies to appetitive stimuli.
2.3.3. MRI Examination
Brain MRI was performed using a 3T MR scanner (General Electric, Milwaukee, Wisconsin) and standard quadrature head coil. High resolution Tl-weighted SPGR-IR sequences were acquired for each subject using the following parameters: TR=45ms, TE=20ms,
14


flip angle=45°, 256x256 matrix, 240x240mm2 field-of-view (0.9x0.9mm2 in plane), 1.7mm slice thickness, and coronal plane acquisition. All images were evaluated by a board-certified neuroradiologist for structural abnormalities. No examinations were excluded on this basis.
2.3.4. Image Processing
Tl-weighted brain MR images were processed using the VBM8 toolbox (http://dbm.neuro.uni-jena.de/vbm8/) and SPM8 (http://www.fil.ion.ucl.ac.uk/spm/) software. Images were segmented into GM, WM, and CSF probability maps anatomically coregistered using Diffeomorphic Anatomical Registration Through Exponentiated Lie Algebra (DARTEL) (Ashburner, 2007). Custom DARTEL templates were created for our sample population and used to register the images. Segmented images were non-linearly modulated after registration to preserve relative regional volume, after correcting for different brain sizes. Segmented GM probability maps of each subject were visually inspected for quality control by a radiologist; no images were excluded. Normalized, modulated images were smoothed with an 8mm3 full-width at half-maximum Gaussian kernel.
2.3.5. Whole brain analyses
Morphometric analysis
Whole-brain analyses on grey matter volumes (GMV) were performed using a two-way ANCOVA testing for sex by group interactions and main effects of group and sex. All analyses were adjusted for age, education, and head size measured by total intracranial volume. Age did not differ between groups but is an important covariate as it directly affects global GMV (Good etal., 2001). Education differed by group (Table 2-1). Significance levels were setatp<0.05 corrected for multiple comparisons with family-wise error (FWE) using AlphaSim Monte Carlo simulations (10,000 simulations) and voxel-wise threshold p<0.005. Cluster threshold corresponded to 1202 voxels (each voxel 3.375 mm3) or 4056.8 mm3. Whole-brain analysis
15


interpretation was restricted to the supratentorial space given reported methodological difficulties of infratentorial space segmentation (Diedrichsen et al., 2009).
Behavioral-morphometric analysis
Within SDI, whole brain regression analyses examined GMV association with drug use severity. In exploratory analyses, behavior by sex by group interactions were regressed against whole brain GMV (M.F.R., radiology resident with two years’ experience studying brain morphometric methods; J.T., neuroradiology professor with 20 years’ experience studying brain morphometric methods). Significance was determined using the aforementioned cluster-based FWE of p<0.05 as well as threshold-free cluster enhancement (TFCE) with correction for multiple comparisons using FWE of p<0.05 (Smith and Nichols, 2009). TFCE was used to evaluate small structures less than 4.05cm3 such as the nucleus accumbens, which would otherwise be mathematically precluded from reaching significance (Radua et al., 2014; Smith and Nichols, 2009).
2.3.6. Regions of Interest (ROI)
While whole brain analyses offers statistical robustness, cross-validation with predefined ROIs using prior knowledge improves classification performance (Chu et al., 2012; Kerr et al., 2014; Nieto-Castanon et al., 2003). Whole brain voxel-level analysis is data-driven without regard to specific anatomy, while pre-defined ROI analysis is hypothesis driven for specific neuroanatomical structures based on prior knowledge. To confirm results from whole brain analyses, five a priori neuroanatomical structures were hypothesized to differ in SDI compared to controls based on their involvement in reward, learning, executive control, and affective processing, which are altered in SDI (Ersche et al., 2013; Koob and Volkow, 2010; Li and Sinha, 2008): OFC, MedFG, ACG, insula, and nucleus accumbens. Masks for these structures were created using the Automated Anatomic Labeling (AAL) atlas toolbox (Tzourio-Mazoyer et al., 2002). Total GMV of each structure was calculated by summing voxels in the ROI-masked
16


structural GM map multiplied voxel-wise by the voxel modulation. GMV in each ROI was analyzed using a two-way ANCOVA on group, sex, and sex by group interactions, after adjusting for age, education, and head size. To correct for multiple comparisons, results were considered significant at FWE p<0.05 with Bonferroni correction for five ROIs (pairwise comparison p<0.01). In exploratory analyses, GMV in ROI were regressed against drug symptom count Behavior by sex by group interactions were regressed against ROIs and considered significant with a Bonferroni correction for FWE across all structures (p<0.05; pairwise comparison p<0.01).
2.4. Results
2.4.1. Demographics, drug severity, and behavioral comparisons
SDI and controls were similar in age and sex (Table 2-1). There was no sex by group interaction on age. There was a sex by group interaction on education (Fi,i27=10.936, p<0.001) and main effects of group (Fi,i27=74.914, p<0.001). Female SDI had fewestyears followed by male SDI, male controls, and female controls with the most years of education. SDI had mean 2.2 fewer years of education than control subjects.
Within SDI, there were sex differences in drug use severity (p=0.015) with men having greater drug symptom count. There were no sex differences in drug exposure, abstinence duration, or years of drug abuse.
There were significant sex by group interactions in behavioral approach and impulsivity: female SDI had highest approach and impulsivity, followed by male SDI, male controls, and female controls with the least approach and impulsivity.
17


2.4.2. Whole Brain Analysis
Sex by group interaction
A significant sex by group interaction on GMV was found in widespread areas of cerebral cortex, thalamus, and basal ganglia. To further characterize the interaction, the effect of group was investigated in each sex separately. No GMV differences were observed between control and SDI men. In contrast, large, widespread differences in GMV were observed between control women and SDI women (Figure 2-2). Compared to control women, SDI women had
Figure 2-2. T-value map illustrating statistically significant greater grey matter volumes in healthy control women compared to stimulant dependent women, after controlling for age and brain size (p<0.001).
18


significantly less GMV in frontal lobe (OFC, MedFG, superior frontal gyrus), limbic regions (insula, amygdala, cingulate gyrus), temporal lobe (temporal pole, uncus, parahippocampal gyrus, hippocampus, occipitotemporal gyri, superior temporal gyrus, middle temporal gyrus), and inferior parietal lobule. Greater GMV observed in control women compared to SDI women showed anatomical congruence to the sex by group interaction effect.
Main effect of sex
A significant main effect of sex was found (Figure 2-3), with women exhibiting comparatively greater regional GMV than men widely throughout cerebral cortex, thalamus,
Figure 2-3. T-value map illustrating statistically significant greater regional grey matter volumes in control women compared to control men, after controlling for age and brain size (p<0.001).
19


and basal ganglia (p<0.001). Subgroup analysis demonstrated anatomically similar significant differences between healthy control men and women.
Main effect of group
A significant main effect of group was found throughout frontal, temporal, insular, and parietal regions (p<0.001). Controls had significantly greater GMV than SDI.
2.4.3. Whole Brain Correlations with Drug Use Characteristics
Significant negative correlations were found between drug use severity and bilateral nucleus accumbens GMV (Figure 2-4; left, 290 voxels, MNI coordinates [-9, 4, -2], p(FWE-
Figure 2-4. Left, T-value map of statistically significant negative correlations found on whole-brain level between stimulant dependence symptom count and grey matter volume within the bilateral nucleus accumbens (the so-called "reward nucleus”) in SDI. Right, negative correlation between the substance dependence symptom count and nucleus accumbens total volume as defined by ROI.
corr)<0.0001; right, 271 voxels, MNI coordinates [8,10, -6], p(FWE-corr)<0.0001). Years of substance use did not correlate with any GMV. Abstinence positively correlated with a small area of left superior frontal gyrus (171 voxels, MNI coordinates [-18, 38, 33], p(FWE-
20


corr)=0.006). Neither years of substance use, abstinence, or drug use severity were found to have significant sex interactions on GMV.
2.4.4. Exploratory Whole Brain Group by Sex by Behavior Interactions
Significant three-way interactions between behavioral approach, group, and sex were seen in the bilateral middle frontal gyri (Figure 2-5, green; Table 2-2). In these areas, in female
Figure 2-5. Sex by group by behavior interactions on GMV. Left, scatterplots illustrating the sex by group interaction on the correlation between approach and GMV in the bilateral middle frontal gyri (dorsolateral prefrontal cortex). Right, scatterplots illustrating the sex by group interaction on the correlation between impulsivity and GMV in the left superior temporal gyrus and left insula. Middle, clusters of whole-brain significance demonstrating sex by group by impulsivity (red) and sex by group by approach (green) interactions.
SDI behavioral approach correlated negatively with GMV, whereas female controls and both groups of men had positive correlation coefficients. Significant three-way interactions between impulsivity, group, and sex were seen in bilateral superior and middle temporal gyri, right insula, right superior temporal sulcus, and right inferior temporal gyrus (Figure 2-5, red; Table 2-2). In these areas in female SDI, impulsivity correlated negatively with GMV, whereas female controls and both groups of men had positive correlation coefficients.
Table 2-2. Significant sex by group interactions on correlations between behavior and GMV.
21


y z k fj-vuiuc fFWE-cor] Structure
Behavioral activation x -39 26 45 371 <0.001 Left middle frontal gyrus
Group x Sex, negative -46 18 46
correlation 48 12 40 353 <0.001 Right middle frontal gyrus
42 20 34
Impulsivity x Group x -56 -39 6 4475 <0.001 Right superior temporal gyrus, right
Sex, negative -57 6 -20 insula
correlation -51 5 -12
68 -24 -9 312 0.001 Left superior temporal gyrus
-27 -84 -9 1112 0.005 Right superior temporal gyrus
-39 -57 -14
-32 -72 -12
46 -42 7 176 0.006 Left middle temporal gyrus
-15 -91 -8 79 0.016 Left inferior frontal gyrus
-9 -81 -2
-6 -96 -5
2.4.5. ROIAnalysis
Two-way ANCOVA revealed statistically significant sex by group interactions and main effects of group for all structures except nucleus accumbens (Figure 2-6). Post-hoc pairwise
Orbitofrontal
Cortex
22000
20000
18000
*
o
>
^ 16000 E
> 14000
12000
10000
Group x Sex Interaction p = 0.003 Group Main Effect p < 0.001 Sex Main Effect p = 0.773
â–¡ Women â–  Men
T • S *
r
T
'I
p = 0.319
C SDI C SDI
Medial Frontal Gyrus
18000
16000
14000
12000
10000
8000
C SDI C SDI
p = 0.006 p< 0.001
p = 0.072
â–¡ Women â–  Men
p 0.274 ,______,
p < 0.001
Insula
13000 12000 11000 10000 9000 8000 7000 6000 5000
C SDI C SDI
p < 0.001 p< 0.001
p = 0.984
â–¡ Women
â–¡ Men
p = 0.416
p< 0.001
Anterior Cingulate Gyrus
9000 8000 7000 6000 5000 4000 3000 2000
â–¡ Women â–  Men

• • * T
------- •
p - 0.352 *-----*
p < 0.001
t------1-----r
Nucleus
Accumbens
700
600
500
400
300
200
â–¡ Women 13 Men
• • •
C SDI C SDI p = 0.005
p< 0.001
p = 0.634
C SDI C SDI p = 0.669 p = 0.337 p = 0.055
Figure 2-6. Boxplots illustrating the total grey matter volume in each subpopulation for each region-of-interest. The top and bottom edges of the boxes indicate the third and first interquartiles, respectively. Lines in the middle of the boxes indicate the medians. Whiskers above and below the boxes indicate the 90th and 10th percentiles, respectively. Points above and below the whiskers indicate the 95th and 5th percentiles, respectively. Significance was tested using two-way ANCOVA. C = healthy control individuals. SDI = stimulant dependent individuals.
22


comparisons revealed greater GMV in control women compared to SDI women in total volumes of each significant structure (p<0.001) but not between control men and SDI men, consistent with whole brain results.
2.4.6. ROI Correlations with Beh avior
Drug use severity correlated negatively with nucleus accumbens GMV (Figure 2-4, right). This correlation was driven by SDI women; a steeper and significant negative correlation was seen in women (left: r=-0.364, p=0.047; right: r=-0.407, p=0.031) compared to men (left: r=-0.063, p=0.737; right: r=-0.174, p=0.349) in which the correlations were not significant No other correlations between drug characteristics or behavioral metrics were significant No behavior by sex by group or behavior by group interactions were significant in ROI structures.
2.5. Discussion
The current finding of significantly lower GMV in abstinent stimulant dependent women compared to healthy control women is striking for two reasons: (1) no group differences were observed in men, and (2) the involved regions are anatomically vast and overlap substantially with pathways implicated in reward, learning, executive control, and affective processing (Koob and Volkow, 2010). The widespread anatomical extent to which men and women differ in relation to abstinent substance dependence has not been reported. These differences in SDI women compared to men could reflect a greater neuroanatomical endophenotype that predisposes them to stimulant dependence or a vulnerability to morphologic changes that result from stimulant dependence more so in women than men. Decreased GMV in female SDI compared to female controls was most striking in limbic regions, particularly the insula, further suggesting a functional role of these structures in mediating the clinical phenotype.
23


We expanded upon these structural results with brain-behavioral correlations. Nucleus accumbens volume was negatively correlated with drug use severity, consistent with its role in reward and salience. Previous studies have shown that ventral striatal activity, including nucleus accumbens, correlates with the intensity of received rewards (Sescousse et al., 2013). Persons abusing and dependent upon stimulants undergo pathologic overstimulation of this nucleus and may exhibit compensatory down-regulation of neuronal synapses with reduced dendritic branching, number of axonal boutons, and degree of axon myelination leading to reduced GMV (Draganski and Kherif, 2013; Fields, 2013). The negative relationship observed in this study between drug use severity and nucleus accumbens volume was significant in women but not in men, despite men exhibiting greater drug use severity. This suggests women may demonstrate greater susceptibility to drug use severity changes, possibly through neuroendocrine mechanisms which will be discussed below. Behavioral approach and impulsivity both interacted with group and sex to significantly correlate with GMV (Figure 2-4, Table 2-2). Higher behavioral approach and impulsivity were associated with lower GMV in female SDI. Behavioral approach characterizes level of arousal and response to cues toward favorable outcomes and positive affective states; higher approach motivates behavior. Impulsivity describes decreased inhibitory control over potential actions leading to reward. Previous studies have reported significant sex differences between approach and impulsivity characteristics in SDI (Perry et al., 2013); however, this is the first study to report structural neuroanatomical correlates of these findings. Higher approach in female SDI was correlated with lower GMV in the bilateral DLPFC and may reflect a deficit in top-down control over approach behaviors toward drug cues. The current structural and brain-behavioral relationship differences by sex may result from neuroendocrine factors. For example, sex and ovarian hormones affect the number, density, and firing rate of dopaminergic neurons, with women
24


showing enhanced dopaminergic system engagement during initial drug exposure and exacerbated negative affective state during drug withdrawal (Becker et al., 2012).
Few studies have investigated sexual dimorphism GMV in stimulant dependence; in fact, only two studies report structural sex differences in SDI: Rando et al. (Rando et al., 2013) and Tanabe et al. (Tanabe et al., 2013). Consistent with our findings, Rando et al. (Rando et al.,
2013) observed greater GMV in healthy compared to cocaine-dependent women in the left inferior frontal gyrus, left insula, left superior temporal gyrus, right temporo-occipital cortex, and left hippocampus. Tanabe et al. (Tanabe et al., 2013) observed a differential effect of sex on small regions of the insula. Consistent with our results, they reported that SDI women exhibited smaller insulae compared to controls. However, we report that the differences span nearly the entire insulae bilaterally. Both of these studies had modest patient sample sizes, 36 in Rando and 28 in Tanabe. A meta-analysis performed by Ersche etal. (Ersche etal., 2013) studied 494 stimulant dependent subjects (79% men) and 428 healthy control subjects (69% men) and reported smaller GMV in SDI compared to controls in the insulae, inferior frontal gyrus, ACG, and anterior thalamus; however, this study did not comment on any sex-effects. Other studies of drug effects on brain morphometry exclude women altogether (Barros-Loscertales etal., 2011; Fein etal., 2002; Franklin etal., 2002), pointing to the need for prospective studies to investigate effects of sex. Here we report significant neuroanatomical sexual dimorphism in the largest prospective sex by group sample of long-term abstinent stimulant dependence to date.
The lack of group differences in men was unexpected. Unlike our study, Rando et al. found small differences in men, with lower GMV in a small portion of the precentral and mid-cingulate gyrus in cocaine-dependent compared to healthy men. There are several possible explanations for this difference, such as recent large alcohol intake (mean 87 drinks in prior month), short length of abstinence (mean 3 weeks), and significantly older SDI than controls in
25


the Rando population. Our sample had much longer abstinence, mean 13.5 months. It has been reported that GMV "recovery” is associated with sustained abstinence (Connolly et al., 2013). For example, in the Ersche et al. (Ersche et al., 2013) meta-analysis of 494 stimulant dependent subjects, only four of the 13 studies included subjects abstinent for more than one month, with the majority of studies investigating active users. Thus, most studies examined acute drug effects. Because our study included subjects abstinent for at least 60 days, there may have been a "ceiling” effect. This hypothesis is consistent with results from Connolley etal. (Connolly et al., 2013) who found that in men, GMV positively correlated with early abstinence but tapered at 35 weeks to become equivalent to those of drug-naive controls. Given the average 13.5 months abstinence in our study, GMV recovery may have already reached a steady-state in men by the time of recruitment.
One limitation of this study is the polysubstance use characteristics of the SDI population. While this precludes us from relating structural changes to a single drug, our sample has biological and ecological validity as it reflects an important, real-world, clinical population of SDI. Epidemiological data demonstrate that stimulant dependence does not often occur in isolation; instead most stimulant dependence individuals meet dependence criteria for other substances (Sara et al., 2012; Stinson et al., 2005). Importantly, the GMV differences observed here were not due to differences in drug exposure or symptom severity. Another limitation is that our sample was referred from the justice system and we cannot exclude the possibility that antisocial personality traits contributed to the findings. Third, SDI and controls differed in years of education. Although we statistically covaried for this confounding variable in all analyses it is possible that education could influence the observed differences.
2.6. Conclusion
Vast neuroanatomical changes observed in abstinent SDI were present in women but not in men. In particular, structures involved in reward, learning, executive control, and
26


affective processing pathways were affected: insula, OFC, ACC, MedFG, and nucleus accumbens. These changes correlate with drug use and behavioral measures and may help to explain differences in the clinical course of stimulant dependence in women compared to men.
2.7. Acknowledgements
Several co-authors contributed to this chapter: Manish Dalwani MS, Dorothy Yamamoto PhD, Robert Perry MD, Joseph Sakai MD, Justin Honce MD, and Jody Tanabe MD. This study was supported by the National Institute of Drug Abuse (NIDA) grants DA024104 (JT) and DA 027748 (JT).
27


CHAPTER III
TOP-DOWN NETWORK EFFECTIVE CONNECTIVITY IN ABSTINENT SUBSTANCE
DEPENDENT INDIVIDUALS
3.1. Abstract
This chapter reports resting-state large-scale brain network connectivity changes in a subsample of the population presented in Chapter 2. We hypothesized that compared to healthy controls, long-term abstinent substance dependent individuals (SDI) will differ in their effective connectivity between large-scale brain networks and demonstrate increased directional information from executive control to interoception-, reward-, and habit-related networks. In addition, using graph theory to compare network efficiencies we predicted decreased smallworldness in SDI compared to controls. 50 SDI and 50 controls of similar sex and age completed psychological surveys and resting state fMRI. fMRI results were analyzed using group independent component analysis; 14 networks-of-interest (NOI) were selected using template matching to a canonical set of resting state networks. The number, direction, and strength of connections between NOI were analyzed with Granger Causality. Within-group thresholds were p<0.005 using a bootstrap permutation. Between group thresholds were p<0.05, FDR-corrected for multiple comparisons. NOI were correlated with behavioral measures, and group-level graph theory measures were compared. Compared to controls, SDI showed significantly greater Granger causal connectivity from right executive control network (RECN) to dorsal default mode network (dDMN) and from dDMN to basal ganglia network (BGN). RECN was negatively correlated with impulsivity, behavioral approach, and negative affect; dDMN was positively correlated with impulsivity. Among the 14 NOI, SDI showed greater bidirectional connectivity; controls showed more unidirectional connectivity. SDI demonstrated greater global efficiency
28


and lower local efficiency. Increased effective connectivity in long-term abstinent drug users may reflect improved cognitive control over habit and reward processes. Higher global and lower local efficiency across all networks in SDI compared to controls may reflect connectivity changes associated with drug dependence or remission and requires future, longitudinal studies to confirm.
This chapter has been peer-reviewed and published: Regner MF, Saenz N, Maharajh K, Yamamoto D, Mohl B, Wylie KP, Tregellas J, Tanabe J. Top-Down Network Effective Connectivity in Abstinent Substance Dependent Individuals. PLOS ONE. 2016 Oct24; 11(10): e0164818. PMID: 27776135.
3.2. Introduction
Substance dependence is a significant public health problem with an estimated 10.3% lifetime prevalence in the United States (Miller and Hendrie, 2009). Across substances of abuse, a generalizable pattern develops beginning with an initial stage of rewarding effects from occasional use and developing into a pathologic stage of loss of control, escalated use, compulsive drug seeking, and significant negative consequences (Wise and Koob, 2014). Individuals with substance dependence have been shown to exhibit higher levels of impulsivity, behavioral approach, and negative affect (Perry et al., 2013), and these differences have been associated with structural (Regner et al., 2015) and functional (Bell et al., 2014; Hyatt et al., 2012; Krmpotich et al., 2013; Wisner et al., 2013) brain differences compared to healthy controls. While task-based studies using fMRI and PET have contributed significantly to our understanding of functional brain changes in specific neuroanatomical areas (Jasinska et al., 2014), resting-state fMRI (rsfMRI) provides opportunity to explore large-scale networks and network interactions independent of task-specific neuropsychological constructs (Fox and Greicius, 2010). Advantages of rsfMRI include less confounding by differences in task paradigms, correlation of resting state networks (RSN) to specific tasks and neuropsychiatric
29


constructs (Smith et al., 2009), and reproducibility due to simplified experimental design and data acquisition (Chen et al., 2008).
Stimulant dependence is characterized by complex behaviors and, like other neuropsychiatric diseases, is thought to reflect pathology at the circuit-level rather than a single brain structure (Koob and Volkow, 2010). Moreover, activity and connectivity differences in stimulant dependence have been demonstrated using rsfMRI across disease stages and may explain the progressive behavioral phenotype changes across the natural history of the disorder (Sutherland etal., 2012). For example, active drug addiction stages include (I) binge/intoxication, (II) withdrawal/negative affect, and (III) preoccupation/anticipation (Koob and Volkow, 2010); involved circuits at these stages include (I) ventral tegmental area and striatum; (II) amygdala, bed nucleus of the stria terminalis, and ventral striatum; and (III) prefrontal cortex, hippocampus, basolateral amygdala, cingulate, and insula. Sequential cycling through these active disease stages is hypothesized to result in the neuroadaptive changes that give rise to compulsive drug-seeking and drug-taking (Figure 3-1).
Brain activity and connectivity at different disease stages have been correlated with individual differences in executive function, interoception, reward, and habit formation. For example, Gu et al. (Gu et al., 2010) observed decreased rsfMRI connectivity between nodes within the mesocorticolimbic reward pathway in active cocaine users compared to healthy controls. These findings are consistent with animal models, in which rats dependent upon and self-administering cocaine demonstrated decreased connectivity compared to control rats (Lu et al., 2014); affected pathways in this sample of rats included connections between the dorsolateral prefrontal cortex (PFC) and ventral striatum, as well as between the prelimbic cortex (homologous to anterior cingulate gyrus in humans) and entopeduncular nucleus (homologous to globus pallidus interna in humans) (Lu et al., 2014). These active disease findings stand in contrast to findings in disease remission. In short-term abstinent cocaine
30


Disease Progression / Natural History
c
o
E
DC
on
eti
QJ
Q
Figure 3-1. Simplified illustration of the natural history of substance use disorders. Active substance dependence is characterized by sequential stages of binge/intoxication, withdrawal/negative affect, and craving/preoccupation (Koob and Volkow, 2010)that repeat a variable number of cycles, represented by m. Active disease episodes may be interrupted by periods of short- or long-term abstinence, represented by n.
dependence (>3 days), Wilcox et al. (Wilcox et al., 2011) observed increased rsfMRI connectivity between the ventral striatum and ventromedial PFC. Camchong and colleagues (Camchong et al., 2014) measured resting state functional connectivity amongst reward processing regions in a cohort of stimulant dependent individuals at two time points, 5 weeks abstinence and 13 weeks abstinence, with comparison to a matched healthy control group. Abstinent stimulant dependent patients demonstrated increased functional connectivity compared to controls, consistent with prior studies in patients of 1.4 years of abstinence (Krmpotich et al., 2013) and 5.7 years of abstinence (Camchong et al., 2013). Although Camchong et al.(Camchong et al., 2014) found abstinent stimulant dependent patients demonstrated increased functional connectivity compared to controls, patients who relapsed between time points demonstrated decreased connectivity compared to patients who
31


maintained abstinence. The authors speculated that this reduction in functional connectivity from 5 to 13 weeks in relapsers compared to abstinent patients may be associated with these patients’ inability to maintain abstinence. These studies suggest that group differences in connectivity may be related to different stages of dependence/remission, possibly representing a transition from hypoconnectivity in limbic and subcortical regions during active dependence to increased top-down executive control in sustained abstinence.
Understanding differences in large-scale brain connectivity depends upon characterizing the relative activity within networks as well as between them. Two modes of functional interactions between brain regions include functional connectivity and effective connectivity. Functional connectivity is the simultaneous and temporally coherent activation of separate brain regions. Effective connectivity characterizes the directional flow of information. One method of characterizing effective connectivity is Granger causality (Seth, 2010), which is methodologically straightforward but requires careful application and interpretation.
To date, no study has investigated the effective connectivity differences in stimulant dependence. This is important because understanding the direction of information flow in large-scale brain networks may further elucidate mechanisms of abstinence and explain previously reported changes. To improve substance dependence treatments, a better understanding of the connectivity characteristics associated with long term remission are needed and may help to predict successful abstinence, evaluate treatment efficacy, and develop novel treatments. This study investigated the effective connectivity and graph theory characteristics of large scale networks in the resting brain in long-term abstinent SDI compared to healthy controls to provide holistic, organ-level measures of brain connectivity and organization for comparisons between groups. We hypothesized that compared to healthy controls, long-term abstinent SDI will demonstrate altered effective connectivity between large-
32


scale brain networks and increased directional information from executive control to interoception-, reward-, and habit-related networks.
3.3. Materials and Methods
3.3.1. Sample Population
Fifty substance dependent individuals (SDI) and 50 healthy controls matched in age and sex were prospectively recruited between October 2010 and June 2013. Demographic information is reported in Table 3-1. SDI were recruited from a residential treatment program at the University of
Colorado Denver Addiction Research Treatment Services. Inclusion criteria for SDI were lifetime DSM-IV psychostimulant dependence (methamphetamine, cocaine, or amphetamine-class substances) and abstinence from all drugs of abuse for a minimum 60 days, verified through close supervision and random urine screens. Participants were permitted to have previously met dependence criteria for substances other than psychostimulants due to the high prevalence of polysubstance use in people dependent upon psychostimulants. Average abstinence duration was 12.8 ± 12.4 months. Healthy controls were recruited from the community and excluded if dependent on alcohol or other drugs of abuse except tobacco. Exclusion criteria for all participants included major depression within the last two months, psychosis, neurological illness, prior head trauma with loss of consciousness exceeding 15 minutes, prior neurosurgery, HIV, bipolar disorder, other major medical illness, inability to tolerate MRI, positive urine or saliva screen (AccuTest™, AlcoScreen™), and IQ < 80. All participants provided written informed consent approved by the Colorado Multiple Institutional Review Board.
33


Table 3-1. Demographic, drug use, and behavioral characteristics of the sample population. BAS, Behavioral activation scale; BIS, Behavioral inhibition scale; BIS-11, Barratt impulsiveness scale version 11; PANAS-X, Positive and Negative Affect Schedule-Expanded Form.
SDI Control t-value p-value
Demographics N 50 (22F/28M) 50 (25F/25M) 0.869
Age 34.18±7.63 31.6±8.57 1.589 0.115
Education 12.48±1.42 14.66±1.47 -7.560 <0.001
Abstinence (mo) 12.8 ± 12.4
Drug Dependence Stimulants 50
Nicotine 35 8
Alcohol 35
Opioids 16
Cannabis 20
Other 9
Behavioral Metrics
BIS 20.64±3.26 20.10±3.56 0.792 0.430
BAS 76.74±6.54 69.40±9.27 4.575 <0.001
Drive 12.80±2.17 10.34±2.26 5.551 <0.001
Fun Seeking 13.16±1.845 11.38±2.18 4.414 <0.001
Reward 17.40±1.94 17.10±1.75 0.812 0.419
BIS-11 72.16±11.54 57.30±6.71 7.871 <0.001
Motor 27.30±4.72 22.78±3.02 5.707 <0.001
Non-Planning 27.08±4.96 20.48±3.33 7.809 <0.001
Attentional 17.78±3.81 14.04±3.14 5.361 <0.001
PANAS-X
Positive Affect 35.32±6.09 35.92±6.09 -0.488 0.626
Negative Affect 21.54±7.98 14.72±3.48 5.540 <0.001
3.3.2. Structured Interviews and Questionnaires
All participants received structured interviews and behavioral measures. Drug dependence was assessed using the computerized Composite International Diagnostic Interview-Substance Abuse Module (CIDI-SAM) (Cottier et al., 1989). IQ was estimated with matrix and verbal reasoning Wechsler Abbreviated Scale of Intelligence subtests (WASI,
34


Psychological Corporation, 1999) and was recorded to exclude subjects with low scores (IQ < 80). The Diagnostic Interview Schedule version IV is a computerized structured interview used to screen for psychiatric disorders. Participants completed this interview to exclude those with a history of psychiatric disorders as described above. Substance dependence severity was operationalized as the number of total substance dependence and abuse symptoms, quantified by the Diagnostic Interview Schedule version IV (Gelhorn et al., 2008; Robins et al., 1995b).
The Behavioral Inhibition and Activation Scale is a 20-item self-reported questionnaire used to measure responsiveness of motivational systems (Campbell-Sills etal., 2004; Carver and White, 1994). Behavioral approach and inhibition were operationalized as the total Behavioral Activation and Inhibition Scales, respectively.
The Barratt Impulsiveness Scale (BIS-11) is a 30-item self-reported questionnaire used to measure impulsivity; participants rated whether phrases and words describing aspects of impulsivity were self-descriptive (Patton et al., 1995). Impulsiveness was operationalized as the total Barratt score.
Positive and Negative Affect Schedule-Expanded Form (PANAS-X) quantifies a participant’s positive and negative affect using a series of 60 words and phrases that are rated on a scale of self-description (Crawford and Henry, 2004). Positive and negative affect were operationalized as the total PANAS-X score for positive and negative attributes.
3.3.3. MRI Examination and Image Analysis
MRI Acquisition
Brain MRI was performed using a 3T MR scanner (General Electric, Milwaukee, Wisconsin) and standard quadrature head coil. Head motion was minimized using a VacFix head-conforming vacuum cushion (Par Scientific A/S, Odense, Denmark). Any subjects with >2 mm of head motion were excluded. High resolution Tl-weighted SPGR-IR sequences (TR=45ms, TE=20ms, flip angle=70°, 256 * 256 matrix, 240 * 240mm2 field-of-view (0.9 * 0.9mm2 in
35


plane), 1.7mm slice thickness, and coronal plane acquisition) and resting-state functional scans (TR=2000ms, TE=30ms, flip angle=30°, axial acquisition, 64 * 64 matrix, 3.4 mm * 3.4 mm inplane voxel size, 3mm slice thickness, 1mm gap, 150 volumes) were acquired. During fMRI acquisition, participants were instructed to close their eyes, not think of anything in particular, and not fall asleep.
Image Preprocessing
Resting fMRI images were processed using the SPM8 toolbox in MATLAB. The first four volumes of each examination were excluded to avoid saturation effects (Figure 3-2). Standard
Figure 3-2. Diagram of the fMRI BOLD signal pre-processing, network-of-interest definition, and effective connectivity analysis pipelines.
pre-processing steps included slice timing correction, rigid realignment and motion correction (motion >1 voxel/TR was censored), spatial normalization, and de-noising. Motion parameters (three rotation and three translation parameters) for censorship were calculated for each time-point using corresponding SPM realignment pre-processing values. Anatomical volumes were segmented into gray matter, white matter, and CSF tissue maps, and the resulting binary masks were eroded (1 isotropic voxel) to mitigate partial volume effects. CSF and white matter time
36


series were obtained using the mean signals from voxels based on eroded CSF and white matter SPM template masks. Mask erosion and time series extraction were performed using functions contained in the CONN toolbox (Whitfield-Gabrieli and Nieto-Castanon, 2012). After linear trends were removed, time series of the motion parameters, WM signal, and CSF signal were removed from the resting-state BOLD data using linear regression, and the resultant residual BOLD time series were band-pass filtered (0.008 Hz Networks-of-Interest (NOI) Definition and Behavioral Correlations
Group independent component analysis (ICA) was performed using the GIFT toolbox as previously reported in the literature (Krmpotich etal., 2013; Tanabe etal., 2011; Tregellas et al., 2011a; Tregellas etal., 2011b) in order to define the networks-of-interest (NOI). For the purposes of this study, the term resting state networks (RSN) refers to the canonical spatial maps used to define the NOI. The term NOI refers to the independent components identified in our sample population and labeled by their corresponding RSN. The dimensionality of the data from each subject was first reduced to 100 components using principal component analysis. Subsequent group-level ICA yielded 34 components, the number of which was determined using the minimum description length (MDL) algorithm (Li et al., 2007). Fourteen canonical RSN templates (Table 3-2) were provided by Stanford's Functional Imaging in Neuropsychiatric Disorders (FIND) Laboratory (Shirer et al., 2012). Atthe group level, the 34 identified components were spatially correlated with the canonical RSN templates. Components with the highest spatial correlation to the canonical template were labeled with the corresponding standard RSN label. These labelled components formed the set of NOI for subsequent graph analysis. All components were visually inspected by a neuroradiology fellow (N.S.) and radiology resident (M.R.) independently to confirm accuracy with the canonical RSN templates.
37


Table 3-2. Canonical RSN included in the analysis and their corresponding symbol abbreviations. Spatial maps were provided by Stanford's Functional Imaging in Neuropsychiatric Disorders (FIND) Laboratory.
Resting State Network Symbol
Auditory Network AN
Anterior Salience Network aSN
Basal Ganglia Network BGN
Dorsal Default Mode Network dDMN
High Visual Network HVN
Left Executive Control Network LECN
Language Network LN
Precuneus Network PCN
Posterior Salience Network pSN
Primary Visual Network PVN
Right Executive Control Network RECN
Sensorimotor Network SMN
Ventral Default Mode Network vDMN
Visuospatial Network VSN
Concordance between inspectors was 100%. Subject-specific spatial maps and time courses were estimated using the GICA back-reconstruction function in GIFT.
For each subject, the strength (or coherence) of each NOI was operationalized as the mean beta value across spatial dimensions for that component in the mixing matrix. These values were regressed against impulsivity, approach, inhibition, and negative affect Regressions between the NOI strength and subjects’ behavioral metrics were used to interpret the neuroimaging findings within the context of measurable behavioral characteristics.
Effective Connectivity Analysis
For each individual, the time courses corresponding to the NOI were obtained from the back-reconstruction function in group ICA. These NOI signals were linear trend removed, normalized to zero mean and unit variance, and band-pass filtered at 0.008-0.15 Hz. The
38


resultant NOI time courses were temporally concatenated across individuals into SDI and control groups (Deshpande etal., 2010a; Ding and Lee, 2013). Effective connectivity between all 14 NOI time courses in the SDI and control graphs were calculated using Granger causality (GC) analysis implemented in the Granger Causal Connectivity Analysis (GCCA) MATLAB toolbox (Seth, 2010). Significance was estimated by comparing observed group difference to a randomized null hypothesis distribution, and the test statistic was determined by the percentile position of the observed difference (i.e., the proportion of randomizations with values greater than or equal to the observed value). To determine null hypothesis distributions, subjects’ group labels were randomized and GC connectivity differences estimated for each randomization permutation until the aggregate randomization distribution achieved statistical stability. Significance level was a = 0.05 using false discovery rate (FDR) q < 0.05 to correct for multiple comparisons.
Global Network Measures
To provide global network measures, graph theory measures were used to describe the topology of the graphs of NOI. The purpose was to provide holistic, organ-level measures of brain connectivity for comparisons between groups. These measures included total weighted network density, local efficiency (derived from clustering coefficients), and global efficiency (derived from path lengths). These measures’ derivations and their justification have been previously described in detail (Rubinov and Sporns, 2010).
To compute graph theory metrics the GC connectivity matrices were converted to a binary directed adjacency matrix, where connectivity strengths above and below a certain threshold costlevel are setto 1 and 0, respectively (Ginestet etal., 2011). The cost threshold,
K(Gx), was calculated by first sorting all elements other than the auto-correlating diagonals (identity axis) of the connectivity matrix in descending rank and keeping only the top x% (sub-
39


graph, Gx). Cost was then computed as the fraction of highest strength edges above the given
threshold divided by the total number of edges:
K(GX)
A |£(GJI
“ |£(GSat)|
where GSat represents an edge-saturated network with the same J\f and the function |£(G) | represents the cardinality of £(G). Therefore, a low value reflects a sparse network. The Brain Connectivity Toolbox (Rubinov and Sporns, 2010) was used to calculate global and local efficiency at each cost level. Since there was no a priori reason to select a particular network cost threshold, connectivity metrics as a function of cost were computed and integrated across the cost domain [0,1] (Ginestet et al., 2011):
[ K(Gx)dx
Jo
This approach abides by prior methodological recommendations to separate network cost from network topology (Ginestet et al., 2011). To test for significant differences between groups, a non-parametric randomized permutation test was performed. Subjects’ group labels were randomized and graph theory measures were calculated for each randomization permutation until the randomization distribution demonstrated statistical stability. These distributions formed the null hypothesis distributions representing no group differences for total weighted network density, integrated global efficiency, and integrated local efficiency. Significance of groups differences for each metric was determined by the percentile position within the null hypothesis distribution.
3.4. Results
3.4.1. Sample Population Demographics
There were no significant differences in age (p=0.12) or sex (p=0.87) between groups (Table 3-1). Years of education (p<0.01) and IQ (p<0.01) differed between groups. Controls had
40


higher IQ and more years of education than SDI. All SDI met DSM-IV dependence criteria for stimulants. Drug use characteristics are also summarized in Table 3-1. Eight controls met dependence criteria for tobacco. No controls met dependence criteria for drugs or alcohol.
3.4.2. Beh avioral Metrics
Behavioral characteristics are summarized in Table 3-1. No group difference in BIS inhibition was observed (p=0.43). A significant group difference in behavioral approach was observed, with SDI exhibiting higher total BAS scores than controls (p<0.001). Further analysis of BAS subscales showed that SDI had higher scores on "drive” (p<0.001) and "fun-seeking” (p<0.001), but not "reward-responsiveness” (p=0.42). As expected, SDI reported higher impulsivity than controls (p<0.001) as well as significant differences in the motor, non-planning, and attentional subscales (p<0.001). No significant difference in positive affect was observed (p=0.626); however, a large difference in negative affect was observed (p<0.001), with SDI demonstrating greater negative affect scores than controls.
41


3.4.3. Network Analysis
Directed Connectivity Analysis
SDI network density was significantly greater compared to controls (p<0.001, Figure 3-3). This measure reflects increased overall mean GC causal connectivity strength between all
SDI (n=50) Controls (n=50)
HVN dDMN HVN dOMM
SMN
LECN
aSN SMN
aSN
vDMN
PVIM KECW
BGN
VSN
Figure 3-3. Effective connectivity network density graphs of SDI and controls. Thickness of each line corresponds to the Granger causal strength and color corresponds to the efferent network. SDI network density was significantly greater than healthy controls (p<0.001).
14 NOI compared to controls. Specifically, GC analysis results show that among 182 possible between-network pairs (Figure 3-4), only three pairs differed significantly across group in the FDR corrected data (Figure 3-5). Compared to controls, SDI showed stronger effective connectivity from the RECN to the dDMN and from the dDMN to the BGN (Figure 3-6). In addition, SDI showed stronger effective connectivity from the SMN to the VSN. SDI showed
42


greater bidirectional connectivity (reciprocal GC connections) whereas controls showed more
unidirectional connectivity among the 14 network components.
SDI
Controls
0.03 VSN
vDMN
RECN
â– SC PVN
0.02 O 3 PCN
0) z PSN
SMN
*- LECN
c LN

0.01 HVN
dDMN
< BGN
AN
0 â–  ASN
^ < (J s >
Efferent Network
Efferent Network
Figure 3-4. Directed GC matrices for SDI (left) and controls (right). Colorbar corresponds to logarithm of F values.
SDI > Controls
o
5
+â– >
a)
C
a)
a>
<
VSN
vDMN
RECN
PVN
PCN
PSN
SMN
LECN
LN
HVN
dDMN
BGN
AN
ASN
i/i < <
>
x
>
>
Efferent Network
I
0.005 T
0.010
0.015
0.020 I
0.0251 p-value
Figure 3-5. Effective connectivity matrix illustrating the FDR-corrected group differences between SDI and healthy controls. Colorbar corresponds to p-values. White arrows indicate Granger causal direction.
43
PCN
PVN
RECN
vDMN
VSN


Figure 3-6. Left, NOI demonstrating significantly greater Granger causal relationships in SDI compared to controls. Illustrated are the NOI relationships that differed between groups; colors correspond to NOI labels on right. Printed numbers in upper left corner of each slice correspond to Z-axis coordinates in MNI space. Right, Granger causal relationships demonstrating greater directed information flow in SDI compared to controls (FDR corrected). RECN, right executive control network; dDMN, dorsal default mode network; BGN, basal ganglia network. Values indicate Granger causal p-values. Middle, whole brain illustrations of NOI identified. Arrows reflect Granger causal influence.
Network-Behavioral Correlations
The strength of the RECN correlated negatively with impulsivity (p<0.001), behavioral approach (p<0.001), and negative affect across the population (p=0.006) (Figure 3-7). In contrast, the strength of the dDMN correlated positively with impulsivity (p<0.001), but not behavioral approach (p=0.034) or negative affect (p=0.030) after correcting for multiple comparisons (Figure 3-7). No NOI correlated with positive affect or educational attainment in years.
Global Graph Measures
Group comparison results for local and global efficiency across the domain of cost functions are illustrated in Figure 3-8. Global efficiency was significantly higher in SDI than in
44


BIS-11 BAS PANAS-X (negative)
40 50 60 70 80 90 100 60 70 80 90 10 20 30 40
6.8
6.6
6.4 73
m
o
6.2 z
£
6.0 CD SU
3
5.8 ta
5.6
7.0
6.8 Q.
6.6 o
£
z
6.4 £
(D
6.2 B> 3
6.0 'Ca
5.8
Figure 3-7. Correlations between mean beta value within the RECN and dDMN with impulsivity, approach, and negative affect metrics. Solid black lines indicate the linear regression, solid colored lines indicate the 95% confidence interval, and colored shaded regions indicate the prediction interval. Each point reflects a single participant’s mean beta within a given network, and their score on the given behavioral metric.
controls (p<0.01), suggesting greater global integration. Local efficiency was higher in controls (p<0.05), suggesting greater local specialization. These findings in conjunction suggest reduced small-worldness in SDI compared to healthy controls.
3.5. Discussion
This study revealed greater effective connectivity network patterns in abstinent substance dependent individuals compared to healthy controls. Specifically, in drug users who have been abstinent for on average over one year, effective connectivity analysis revealed increased information flow from the RECN to dDMN and dDMN to BGN compared to controls.
45


Figure 3-8. Global efficiency (left) and local efficiency (right) in SDI and controls as a function of network cost.
The areas of increased effective connectivity observed in our study correspond to regions involved in executive control (RECN), interoception (dDMN), reward (BGN), and habit (BGN). The mean strength of the RECN component correlated negatively with impulsivity, behavioral approach, and negative affect In contrast, the mean strength of the dDMN component correlated positively with impulsivity and trended towards positive correlations with behavioral approach and negative affect. Given the prolonged abstinence of our SDI sample population, these findings are consistent with the hypothesis that successful long-term abstinence is associated with increased top-down cognitive control.
3.5.1. Increased Effective Connectivity from RECN to dDMN
The pattern of increased effective connectivity from RECN to dDMN is consistent with increased top-down executive control in long-term abstinence. Previous work has demonstrated both task-related and resting state hyperactivity within executive control and default mode cortices in abstinent stimulant dependence associated with heightened behavioral
46


monitoring (Connolly etal., 2012; Mayer etal., 2013; Wilcox etal., 2011); however, this is the first study to suggest that these neural signals follow a directional flow of information from RECN to dDMN. Connolly et al. (Connolly et al., 2012) conducted a cognitive control task-based fMRI study of short- (2.4 ± 1.34 weeks) and long-term (69 ± 17.49 weeks) abstinent cocaine addicts. Abstinent cocaine users demonstrated increased activity in PFC, cingulate, and inferior frontal gyri compared to healthy controls. Moreover, short-term abstinent individuals showed right dorsolateral PFC (corresponding to RECN in our study) hyperactivity positively correlating with inhibitory control. Long-term abstinent individuals showed the same finding as well as anterior and mid cingulate (corresponding to part of the dDMN in our study) hyperactivity positively correlating with cognitive errors and heightened behavioral monitoring in abstinence. The present study showed that the RECN strength was negatively correlated with subjects’ impulsiveness, while dDMN strength was positively correlated with impulsiveness.
Our results advance our understanding of neural network changes during substance use disorder remission: as abstinence progresses, cortices within the RECN and dDMN may become hyperactive to exert top-down executive control in a directed fashion; this neuroadaptive change may be associated with decreases in impulsivity and increases in inhibitory control.
However, the top-down cognitive control hypothesis is not straightforward because in addition to executive function and cognitive control, affect plays an important role. Albein-Urios etal. (Albein-Urios et al., 2014) showed that short-term abstinent (2.5 ± 5.5 months) cocaine dependent individuals had increased right dorsolateral PFC and bilateral temporoparietal cortex activation during negative emotion experiences without a concomitant increase in the subjective negative experience itself, suggesting an exaggerated neural response in these regions is required to produce normal levels of emotional salience. The regions reported closely resemble by visual comparison the RECN identified by our analysis. Albein-Urios et al. posited that these areas demonstrate increased sensitization toward negative emotions in SDI. If
47


increased RECN top-down control is a durable feature of long-term abstinence, the literature thus far suggests that its manifestations in human behavior are complex and not reducible to a single neuropsychological construct Our finding that RECN strength is negatively correlated to negative affect while dDMN strength trended towards positive correlation with negative affect provides further evidence that the top-down control model may involve affective components as well, possibly through reciprocal connectivity with limbic areas.
Right-sided lateralization of our ECN findings is not unexpected given the asymmetric functional specialization of cerebral hemispheres in healthy humans. Cocaine dependent patients exhibit reduced resting state interhemispheric connectivity compared to healthy controls in prefrontal and parietal cortices (Meunier et al., 2012), suggesting increased lateralization of function. Connolly et al. (Connolly et al., 2012) reported that hyperactivity in the inferior frontal gyrus correlated with inhibitory control was greater in the left hemisphere in short-term abstinent individuals and greater in the right hemisphere in long-term abstinent individuals. They hypothesized that a shift from left to right inferior frontal gyrus for inhibitory control may reflect a transition from short-term to long-term abstinence. Our results of increased RECN effective connectivity in long-term abstinent stimulant dependence are consistent with this hypothesis, although future longitudinal studies are required for substantiation.
3.5.2. Increased Effective Connectivity from dDMN to BGN
Although the DMN is incompletely understood, growing evidence demonstrates its roles in internally directed tasks such as spontaneous cognition (Mantini and Vanduffel, 2013), self-referential (Vessel etal., 2013) and autobiographical thought (Buckner etal., 2008), and social understanding of others (Li et al., 2014). We demonstrated increased effective connectivity from the dDMN to BGN; however, this finding must be interpreted in the context of the structures within the NOI identified as BGN (Figure 3-4). This NOI included basal ganglia,
48


thalamus, amygdala, hippocampus, hypothalamus, midbrain, and pons. Thus, BGN included several key regions of the bottom-up mesocorticolimbic circuit including the ventral tegmental area, nucleus accumbens, amygdala, and striatum.
Prior studies have demonstrated hypoactivity in stimulant users in the dDMN and in BGN as well as decreased connectivity between these networks. In active cocaine users, Tomasi et al. (Tomasi et al., 2015) showed that cocaine cues disengaged fMRI activity in the ventral striatum, hypothalamus, and DMN in proportion to density of striatal dopamine receptors by PET. DMN activation has been shown to predict performance errors, is diminished in active cocaine dependence, and the extent of altered error-preceding activation has been reported to correlate with years of cocaine use (Bednarski etal., 2011). Gu et al. (Gu et al., 2010) used a seed-based fMRI paradigm in active cocaine users and found significantly decreased functional connectivity between multiple regions of the DMN and BGN. McHugh et al. (McHugh et al., 2014) showed that individuals successfully abstinent 30 days after detoxification had stronger functional connectivity between the amygdala, ventromedial PFC, and anterior cingulate cortex compared to those who had relapsed. By visual comparison, these regions correspond to structures within the NOI identified as dDMN and BGN in our study. Connolly et al. (Connolly et al., 2012) demonstrated increased anterior and mid cingulate activity in long-term abstinent compared to short-term abstinent individuals, activity which correlated with heightened behavioral monitoring. Together, these prior studies suggest that DMN activity may change during abstinence. Initial hypoactivation during early abstinence may transition to hyperactivation and increased connectivity with long-term abstinence. One interpretation is that findings of increased effective connectivity from dDMN to BGN in long-term abstinence may be a compensatory mechanism related to behavioral monitoring not seen in active users. However, longitudinal studies are needed to demonstrate this.
49


3.5.3. Increased Global and Decreased Local Integration
Our findings of increased bidirectional connectivity, increased global efficiency, and decreased local efficiency in long-term abstinent SDI compared to healthy controls suggests pathologically greater global integration and lower local integration in SDI; that is, a connectomic decrease in small-worldness. Similar findings in humans have only been reported using EEG data in 1-3 week abstinent methamphetamine dependent persons. Ahmadlou et al. (Ahmadlou etal., 2013) showed thatthese patients demonstrated a deviation from small-worldness and increased global hypersynchronization in the gamma frequency band, the EEG band most reactive to cognitive information processing. In contrast, active cocaine users demonstrated less global connectivity compared to healthy controls during a Stroop task; however, after adjusting for individual connectivity, cocaine dependent individuals showed greater intrinsic connectivity in the ventral striatum, putamen, inferior frontal gyrus, anterior insula, thalamus and substantia nigra (Mitchell et al., 2013).
Several animal studies provide important context for the interpretation of our findings. Schwarz et al. (Schwarz et al., 2012) used a pharmacological challenge design which revealed that rats under the acute effects of amphetamine compared to a saline vehicle exhibited less clustering (small-worldness) and increased connectedness within somatosensory, motor, cingulate, prefrontal, and insular cortices. In the rhesus monkey model, active cocaine selfadministration was associated with decreased global functional connectivity that selectively affected top-down prefrontal circuits and control behavior while sparing limbic and striatal areas (Murnane et al., 2015). Interestingly, impaired connectivity between prefrontal and striatal areas during abstinence predicted cocaine intake when these monkeys were again provided access to cocaine (i.e., prediction of relapse), consistent with the connectivity pattern associated with relapse in humans as reported by Camchong et al. (Camchong et al., 2014).
50


Together these findings suggest there is globally decreased connectivity in active users and short-term abstinent with a transition to globally increased connectivity in long-term abstinent users. These findings may improve clinical management if global connectivity patterns can be used to predict abstinence success or trajectory in humans. Future longitudinal studies comparing global connectivity in active, short-term, and long-term abstinent drug users must be performed to address this question. Another approach could involve correlating abstinence duration with global connectivity across individuals, an approach we could not implement due to the group-level nature of our statistical design.
3.5.4. Limitations
Controversy Surrounding Granger Causality
While our study provides several important novel findings, it has limitations. Influences between specialized neural systems exist on a spectrum of temporal lag. Functional connectivity using temporal correlation reflect influences with causal latencies that are below the temporal resolution of the repetition time. These influences are not truly contemporaneous in vivo, but appear so by fMRI as a result of low temporal sampling and temporal blurring induced by the hemodynamic response function. Time-lag based measures such as Granger causality reflect slower influences with greater causal latencies that occur on the order of hundreds of milliseconds, which may provide greater power in predicting cause-effect relationships atthe timescale of conscious thought (Tononi et al., 2016).
Neural signals between two nodes may have significantly different physiologic functions depending upon the directionality. As a result, segregating neural influences according to their directionality is necessary in order to properly examine brain function. Methods of examining effective connectivity using fMRI data include structural equation modelling (Buchel and Friston, 1997) and dynamic causal modelling (Friston et al., 2003). These methods require a priori hypotheses describing the theoretical connectivity structure and are limited to models
51


consisting of a small number of nodes. We used an alternative method, Granger causality, which is based on time-lag regressions and is more data-driven.
Granger causality is increasingly used in fMRI-based neuroscience (Chiong et al., 2013; Cohen Kadosh etal., 2016; Feng etal., 2016; Wen etal., 2013; Zhang etal., 2017) and has been previously applied specifically to independent component analysis as in our study (Demirci et al., 2009; Diez etal., 2015; Ding and Lee, 2013; Stevens etal., 2009; Zhongetal., 2012).
However, criticisms of the application of Granger causality to fMRI data have included (Deshpande etal., 2010b): (1) lack of evidence that Granger causality in fMRI-level time series reflects causality in neuronal-level time series, (2) insufficient temporal sampling relative to the timescale of neuronal events, and (3) the possibility that spurious findings may result from systematic differences in hemodynamic response functions. Several recent developments have provided evidence that fMRI Granger causality reliably reflects neuronal causality (Deshpande and Hu, 2012; Deshpande etal., 2010b; Schippers etal., 2011; Wen etal., 2013). Seth and colleagues (Seth et al., 2013) demonstrated that Granger causality is reliably invariant to interregional differences in the hemodynamic response function, including the time-to-peak. However, they reported significant effects of temporal resolution on their results. Wen and colleagues (Wen et al., 2013) demonstrated that fMRI-based Granger causality is a monotonic function of neural Granger causality. Importantly, they showed that this relationship can be reliably detected using conventional fMRI temporal resolution and noise levels as was used here. However, they cautioned that differences in the hemodynamic response could lead to spurious results.
The impact of hemodynamic response variability is currently debated. Schippers and colleagues (Schippers etal., 2011) demonstrated that hemodynamic response variability was minimized by multisubject group inference. Statistically, this is intuitive because population averaging will augment systematic differences (e.g., true neuronal differences) while
52


suppressing random or pseudorandom differences (e.g., hemodynamic response variability). Some authors speculate that HRF variability could be systematic (Smith et al., 2012), and indeed this is a confound that by design exists in the majority of between-group fMRI studies using independent samples (D'Esposito etal., 2003; Hillman, 2014; Murphy etal., 2013). Accordingly, we cannot exclude that systematic differences in the neurovascular response to neural activity between groups may have contributed to our findings.
Other Limitations
Network resolution was limited by the manner in which independent component analysis identifies temporally coherent signals across the brain. For example, the network component identified as BGN included several non-basal ganglia structures, such as the thalamus, amygdala, hippocampus, and midbrain. Additionally, concatenation across individuals precluded correlation of individual psychological measures to resting state network Granger causality; as such, correlations between the strength of each NOI and behavioral metrics were used to provide psychological context for the findings. Lastly, polysubstance use and low educational attainment among psychostimulant users may be viewed as potential confounds or representation of real world clinical features. There is significant literature describing the correlation between drug use and low educational attainment; it is debated whether low educational attainment is the cause or result of drug use disorders (Fergusson et al., 2003; Swaim et al., 1997; Yamada et al., 1996). More recently, however, authors have reported that this correlation is due in part to shared genetic factors (Bergen et al., 2008) while others report that it is due to shared environmental or non-genetic familial risk factors (Grant et al., 2012; Verweij et al., 2013). These studies suggest that low educational attainment is a behavioral component of the pathology of substance use disorders. With regard to polysubstance use among SDI, while this limitation prevents our findings from being attributed to a single drug, it strengthens our results by providing biological and ecological validity.
53


Epidemiologic studies have demonstrated that psychostimulant dependence does not naturally occur in isolation; rather, most patients meet dependence criteria for other drugs of abuse (Sara et al., 2012; Stinson etal., 2005). Our sample population thus reflects the real-world, clinical population of patients with stimulant dependence.
3.6. Conclusion
Increased effective connectivity in long-term abstinent drug users may reflect improved cognitive control and behavioral monitoring (ECN) over self-referential thought (DMN), habit (BGN), and reward (BGN) processes in long-term abstinent drug users. Higher global and lower local efficiency across all networks in SDI compared to healthy controls may reflect connectivity changes associated with drug dependence or remission. Future, longitudinal studies are necessary to definitively characterize connectomic changes across the natural history of substance use disorders.
3.7. Acknowledgements
Several co-authors contributed to this chapter: Naomi Saenz MD, Keeran Maharajh PhD, Dorothy Yamamoto PhD, Brianne Sutton (nee Mohl) PhD, Korey Wylie, Jason Tregellas PhD, and Jody Tanabe MD. This work was supported by the National Institute of Drug Abuse (NIDA) grants DA024104 (JT), DA 027748 (JT), and DA 041011 (MFR). The authors report no biomedical financial interests or potential conflicts of interest
54


CHAPTER IV
THE INSULA IN NICOTINE USE DISORDER: FUNCTIONAL NEUROIMAGING AND IMPLICATIONS FOR NEUROMODULATION
4.1. Abstract
In this chapter, we review the literature and provide motivation for our randomized clinical trial of inhibitory TMS targeting the insula in smoker, presented in Chapter 5. Animal and human literature suggest that the insula is necessary for nicotine use disorders. Yet, much remains unknown about how insular function drives nicotine use. Insular subdivisions show distinct patterns of connectivity with large-scale brain networks, and each subdivision is associated with different functions and behaviors. Saliency, including "bottom-up” sensations and "top-down” sensitivity control mechanisms, is the common theme across insular functions. During acute withdrawal, the insula arbitrates bottom-up versus top-down saliency mechanisms to guide behavior - either to pursue smoking or to avoid relapse - and this arbitration is associated with craving and the nicotine withdrawal syndrome. The purpose of this narrative review article is to synthesize neuroimaging evidence of the insula’s role in nicotine use disorder and present evidence for suitability as a neuromodulation target to promote cessation. Given the limited efficacy of standard-of-care treatments for nicotine use disorder, insular neuromodulation may contribute to the next generation of cessation treatments by offering what henceforth has not been available: a minimally-invasive, anatomically-driven approach to smoking cessation therapy.
This chapter is currently under review for publication in Neuroscience & Biobehavioral Reviews•, pending revisions.
55


4.2. Introduction
Given the serious health problems caused by smoking, it is not surprising that the majority of smokers, nearly 7 in 10 in 2015, wantto quit (Babb, 2017). Unfortunately, however, smoking cessation treatments are largely ineffective, with most abstinence attempts failing within the first 24 hours. Approximately 80% of patients relapse by six months, despite combined pharmacologic and behavioral therapies (Tobacco, 2008). Whereas only 1 patient out of 10 receiving behavioral therapy alone remains abstinent, only 2 patients out of 10 receiving pharmacological (e.g. nicotine replacement) and behavioral therapy remain abstinent after six months (Stead and Lancaster, 2012). These low success rates underscore the need for more effective interventions and improved prognostication of individuals most likely to benefit from a given intervention. To achieve these goals, a better understanding of the neural underpinnings of compulsive nicotine use is required.
A significant advance in our understanding of the neurobiology of smoking behavior was made in 2007, when Naqvi and colleagues observed that lesions to the insula disrupt cigarette smoking (Naqvi et al., 2007). These findings and subsequent studies that will be described in greater detail below convincingly demonstrated that the insula plays a critical role in smoking maintenance and cravings, and that the region may be a therapeutic target for smoking cessation. (Abdolahi et al., 2015a, b, 2017; Forget etal., 2010; Naqvi et al., 2007; Pushparaj etal., 2013; Suner-Soler etal., 2012).
The insula is involved in a wide array behaviors and functions, including salience, interoception, awareness, affect, anticipation, uncertainty, self-recognition, prediction error, perception, attention, and cognitive processing (Nieuwenhuys, 2012). Recent meta-analyses of functional MRI studies suggest that the insula contains two to seven functionally-distinct regions (Cauda etal., 2012; Chang etal., 2013; Deen etal., 2011; Kelly etal., 2012; Kurthetal., 2010). Current insula models suggest that the region plays a role in three broad categories of
56


function. (Deen et al., 2011; Uddin et al., 2014). The dorsal anterior insula is associated with cognitive control functions, such as attention, inhibitory control, and goal-directed cognitive tasks (Dosenbach et al., 2007). The ventral anterior insula is involved in with emotional-limbic functions, including peripheral physiological responses to emotional experiences, as measured by heart rate or galvanic skin response (Mutschler etal., 2009). Finally, the posterior insula mediates sensorimotor-interoceptive functions, and receives rich afferents from spinothalamocortical pathway carrying nociceptive, thermal, and other interoceptive information (Craig, 2002). Despite this functional diversity within insular subunits, several closely-related theories about the functional role of the insula in craving and smoking behaviors have emerged. While differing somewhat in their models and interpretations, the theories all point to involvement of the anterior insula in processing and balancing of "bottom- up” sensations versus "top-down” cognitive control processes (Figure 4-1). In this context, the
Bottom-Up
(Sensations)
Salience Arbitrator
Top-Down
(Cognitive Sensitivity Control)
Viscerosensory, Somatosensory, and Special Sensory (Auditory, Visual)
Salience Network
r ' Posterior , > Dorsal Anterior
Insula ' w Insula
Affective, 1 Motivational, tJ f
Hedonic, and r 1 Ventral Anterior
jpccial jtiiSui^ ^ (Olfaction, â–  1 Insula , j
Gustation, Visual, \ V
Executive Control Network
Default Mode Network
Figure 4-1. Connectivity-based signal flow diagram of anterior insular control of bottom-up versus top-down mechanisms of salience. The right dorsal anterior insula is involved in processing salience of externally-oriented stimuli and it is correlated with the executive control network (an externally-directed system). The right ventral anterior insula is involved in processing salience of internally-oriented stimuli and it is correlated with the default mode network (an internally-directed system).
57


insula may serve as a sensory signal bottleneck, such that insular lesions impair cognitive processing of craving and nicotine withdrawal sensations.
The following sections of this review describe and discuss: (1) evidence of insular involvement in nicotine use disorder pathophysiology and (2) possible therapeutic strategies to target the region for smoking cessation using neuromodulation, a non-invasive technique capable of altering the function of specific brain regions and networks. Given the limited efficacy of standard-of-care treatments for nicotine use disorder, insular neuromodulation may contribute to cessation treatments by offering a non-invasive, anatomically-driven approach to smoking cessation therapy.
4.3. Insular Role in Nicotine Use Disorder
4.3.1. Introduction: Insular Lesions Disrupt Smoking Behaviors
Converging evidence strongly implicates the insula in the maintenance of smoking behaviors and cigarette craving. Naqvi and colleagues (Naqvi et al., 2007) reported that smokers with cerebrovascular damage to the right insula were able to stop smoking easily without cravings or relapse, supporting a role of the insula in addiction. A subsequent, large, prospective study over a one-year period also found that insular lesions in smokers were strongly associated with becoming a non-smoker (Suner-Soler et al., 2012). Abdolahi and colleagues (Abdolahi et al., 2015b) conducted a prospective cohort study with three-month follow-up in 156 smokers hospitalized for acute ischemic stroke, of which 38 were insular strokes. They reported insular damage was associated with increased odds of three-month continuous abstinence as well as cessation from all nicotine products at three months. Insular damage in the same cohort was also associated with fewer nicotine withdrawal symptoms and cravings compared to those with non-insular strokes (Abdolahi etal., 2015a, 2017). These findings have been corroborated in animal models of nicotine dependence. For example, insular inactivation in rat models significantly reduced nicotine motivation, nicotine seeking-, and
58


nicotine taking-behaviors, with no effect on food behaviors (Forget et al., 2010; Pushparaj et al., 2013). These findings will be discussed in the context of animal-model neuromodulation in Section 4.4: Implications of Insular Role in Nicotine Use Disorder on Neuromodulatory Therapeutic Development (page 71). These human and animal studies together demonstrate that insular lesions disrupt smoking behaviors and underscore the need to understand insular function in smokers in vivo.
4.3.2. Neuroimaging the Insula in Nicotine Use Disorder
Nicotine exposure causes two distinct sets of effects based on timing: acute states and chronic effects. Acute states refer to the short-term behavioral changes and altered brain function independent of dependency. Chronic effects refer to pharmacologic dependence and the associated neuroadaptations, independent of acute exposure to nicotine. Nicotine’s acute pharmacodynamics are distinguished from the durable neural changes caused by chronic use; that is, nicotine use disorder reflects chronic effects resulting from repeated acute exposure to nicotine. We synthesize the neuroimaging literature into four distinct stages of nicotine use disorder and recovery (Figure 4-2). First, we review studies of the acute effects of nicotine on neurobiology (page 60). Second, we review studies comparing chronic smokers to controls to understand dependence (page 62). Third, we review studies of nicotine-dependent individuals during acute abstinence to understand the mechanisms of the nicotine withdrawal syndrome and craving (page 65). Fourth, we review long-term abstinence as a model of neuroplastic recovery from nicotine use disorder (page 67). Long-term abstinence provides neuroimaging biomarkers of recovery may serve as useful indicators of treatment efficacy. Finally, we attempt to synthesize the findings from these four stages of the disease into a single model (Section 4.3.3: Putting it All Together: Unified Models of the Role of the Insula in Nicotine Use Disorder Pathogenesis, page 68). Our findings are summarized in Table 4-1.
59


Nicotine Use Disorder Natural History
Active Use
/
Remission of Use
Antedating Predisposition, Endophenotype, and Risk Factors
Stage 1:
Acute Nicotine Exposure
(Immediate "High")
Stage 2:
Chronic Nicotine Exposure
(Pharmacologic
Dependency)
Stage 3:
Acute Abstinence
(Nicotine Withdrawal Syndrome)
Stage 4: Chronic Abstinence
(Neuroplasticity of Recovery)
Insular Lesions, Insular Neuromodulation
Figure 4-2. Diagrammatic illustration of the natural history of nicotine use disorder. We attempt to synthesize the neuroimaging literature of nicotine use disorder into four pharmacologically-and behaviorally-informed stages of nicotine use disorder. Acute nicotine exposure reflects the acute pharmacologic action on neural circuitry. Chronic nicotine exposure reflects pharmacologic dependency and results from repeated acute nicotine exposure; it manifests as maladaptive changes in reward, salience, and executive control circuitry. Acute abstinence provides a model for understanding the neural basis of the nicotine withdrawal syndrome and craving. Long-term abstinence serves as a model of neuroplastic recovery from nicotine use disorder. Relapse (bottom arrows) is the mechanism by which active disease is maintained. Associated neuroimaging findings are summarized in Table I.
Stage 1: Acute Nicotine Exposure (Neural Pharmacodynamics)
Like other drugs of abuse, nicotine acts on the brain’s reward circuit and induces dopamine release from ventral tegmental area neurons into the nucleus accumbens and prefrontal cortex (Volkow et al., 2012). Nicotine acts as a ligand at nicotinic acetylcholine receptors (nAChRs), a family of ligand-gated ion channels involved in three major circuits: (1) diffusely projecting cholinergic neurons from the brainstem’s ascending arousal system that synapse on dopaminergic neurons in the ventral tegmental area, (2) widely-projecting cholinergic neurons from the basal forebrain (nucleus basalis of Meynert) involved in attention, and (3) fast-acting excitatory post-synaptic potentials in autonomic ganglia associated with
60


autonomic and visceral sensations (Nestler et al., 2015). Through these molecular mechanisms, acute nicotine exposure affects circuits involved in arousal, reward, attention, and autonomic regulation. Neuroimaging allows scientists to image in vivo the downstream effects of nicotine-induced changes in these circuits.
Table 4-1. Summary of large-scale brain network neuroimaging findings and associated role of the insula across different stages of nicotine use disorder. The neuropharmacologic mechanisms associated with each disease stage are listed in italics.
Disease Stage
Dmg-Induced Mechanism
Acute Nicotine Exposure
Neural Phannacodynamics
Chronic Nicotine Exposure
Pharmacologic Dependency
Acute Abstinence
Nicotine Withdrawal Syndrome
Neuroimaging Findings
â–¼ Default mode network activity (Tanabe 2011, Sutherland 2015)
A ECN activity (Sutherland 2015)
A SN-DMN-ECN connectivity (Lerman 2014)
A Anterior insula activity (Sutherland
2015)________________________________
A Insulo-cingulate connectivity
• Positively associated with:
FTND, successful future abstinence
• Negatively associated with: incongruent errors on Stroop task, lifetime nicotine consumption (pack-years), future relapse (Janes et al., 2010; Lin et al., 2017)
A Associated striatal adaptations, progressively ventral to dorsal reflecting habit formation___________
â–¼ SN-DMN-ECN connectivity (Lerman 2014)
A Granger causality from insula to other brain regions (Ding and Lee, 2013)
A Right Anterior Insula - DMN connectivity associated with craving magnitude (Moran-Santa Maria et al., 2015)________________________________
Chronic Abstinence
Neuroplastic Recoveiy
â–² Right anterior insular activation in response to cue exposure, associated with lifetime nicotine consumption (Nestor et al., 2011; Zanchi et al., 2015)
^ Salience Network Coherence (Zanchi et al., 2015)
Role of the Insula
• Mediating a higher-order representation of positive somatosensory, viscerosensory, and interoceptive sensations associated with drug reward
• New homeostatic set point for visceral sensations associated with drug reward
• Cigarette-related memory retrieval (Janes et al., 2015b)
• Representing the negative somatosensory, viscerosensory, and interoceptive sensations associated with cravings (Abdolahi et al., 2015a, b, 2017)
• Representing the relative hyper-saliency of drug cues for further monitoring and decision-making
• Hyper-saliency of drugs cues is associated with insular activity and is durable up to 1 year
61


Acute nicotine exposure - the nicotine "high” - alters neural activation and connectivity patterns observed by fMRI in both healthy adults and individuals with nicotine use disorder.
For example, nicotine administration relative to placebo in non-smokers led to decreased default mode network activity (Hahn et al., 2007; Tanabe et al., 2011) and significantly increased local efficiency of connectivity by whole-brain graph theory analysis, particularly in right-sided limbic and paralimbic areas (Wylie et al., 2012). Sutherland and colleagues (Sutherland et al., 2015) conducted a large Activation Likelihood Estimation (ALE) metaanalysis of acute effects of nicotinic agonists on brain activity changes in smokers as measured by fMRI or PET. The sample population included 796 participants spanning 77 different contrasts and experiments, including the resting state. Compared to placebo, nicotinic agonist administration was associated with decreased activity in the bilateral anterior insulae but mixed effects in the left middle insula. Nicotinic agonists also resulted in significantly decreased activity within default mode network regions and increased activity in executive control network regions. When comparing vehicles of nicotine administration (pure nicotine pharmacologic administration versus cigarette smoking), decreased left middle insula activity was common by conjunction analysis to both manipulations, while decreased right anterior insula activity was specific to cigarette smoking compared to pharmacologic administration. Importantly, this study did not evaluate or control for effects of satiation versus abstinence and heterogeneity of tasks (including resting state), which confound interpretation. Relative to placebo, however, acute nicotine exposure overall consistently leads to decreased anterior insular activity, decreased default mode network activity, and increased executive control network activity.
Stage 2: Chronic Nicotine Exposure in the Cigarette-Sated State (Pharmacologic Dependency)
Several studies have examined individuals with a nicotine use disorder during a resting state fMRI scan. One study found that reduced circuit strength between the insula and dorsal
62


anterior cingulate cortex, the two principal nodes of the salience network, was associated with increased addiction severity (Moran et al., 2012). These associations were observed when participants were scanned both after smoking or after acute abstinence, suggesting that decreased salience network coherence reflects a chronic effect of nicotine use disorder rather than an acute pharmacologic effect Salience network coherence has been consistently associated with severity of nicotine use disorder across studies (Bi et al., 2017; Lin et al., 2017; Moran et al., 2012; Wilcox et al., 2017; Zhou et al., 2017). For example, Zhou and colleagues (Zhou et al., 2017) reported that reduced connectivity between the insula and anterior cingulate cortex was associated with increased nicotine use disorder severity. Li and colleagues (Lin et al., 2017) extended these findings, showing that reduced circuit strength between right insula and anterior cingulate cortex was associated with higher number of incongruent errors during a cognitive-control task, implicating this circuit in top-down cognitive control of saliency. More importantly, diminished circuit strength between these regions was associated with greater lifetime nicotine consumption. Overall, these studies provide converging evidence that reduced salience network coherence at rest is a marker of chronic nicotine use and reflects addiction severity.
Insular connectivity may also have prognostic importance related to vulnerability to relapse during future cessation attempts. For example, decreased connectivity between the insula and brain regions involved in cognitive control, including the dorsal anterior cingulate and dorsolateral prefrontal cortices, was associated with greater risk of future relapse after attempted cessation (Janes et al., 2010), possibly reflecting a mechanism of reduced top-down control. In another study, circuit strength between the insula and dorsal anterior cingulate cortex was significantly associated with enhanced smoking cue reactivity in areas involved in attention and motor planning, such as the right ventrolateral prefrontal cortex and dorsal striatum (Janes et al., 2015a). Interestingly, the authors reported that insular-anterior cingulate
63


connectivity in smokers was durable over a one-hour period and not associated with subjective craving or exhaled carbon monoxide, suggesting that increased salience network coherence may represent a chronic effect (i.e., a neural signature of hypersensitive cue reactivity in nicotine use disorder). Recent, larger studies have corroborated these findings that salience coherence is important in mediating chronic effects of nicotine use disorder. Wilcox and colleagues studied 144 individuals with nicotine use disorder during the resting state and reported that decreased circuit strength between the insula and dorsal anterior cingulate was significantly correlated with higher cigarette consumption (Wilcox et al., 2017). After controlling for addiction severity, increased circuit strength between these regions was associated with greater likelihood of successful abstinence. Similarly, a 10-week longitudinal study (Addicott et al., 2015) found that increased insular connectivity to executive control and striatal regions was seen in non-relapsers (i.e., successful abstainers) compared to relapsers. This suggests lower insular connectivity may be associated with relapse vulnerability. Together, these studies suggest that circuit strength between the insula and both (1) anterior cingulate, and (2) regions involved in cognitive control are not only markers of nicotine use disorder, but are also meaningful for prognosis, since it is associated with ability to quit smoking.
Insula activation during various tasks is also a potentially useful biomarker of nicotine use disorder. For example, Janes and colleagues. (Janes etal., 2017) studied 23 smokers during a cessation attempt, 10 of whom remained abstinent during a two-week follow up. Relative to successful abstainers, smokers who relapsed demonstrated increased right insular activation in response to cigarette cues, suggesting that this activation predicted likelihood of future use. In another study using a subsample of smokers from the Human Connectome Project dataset, individuals who smoked more cigarettes had greater right anterior insular activation in response to viewing faces expressing negative emotions such as anger (Dias et al., 2016). These studies suggest that greater insula activation in response to both smoking cues and emotional
64


cues may indicate a higher propensity for smoking and relapse. Neuroimaging smokers in nicotine withdrawal and experiencing cigarette cravings provides a possible mechanism for these observations.
Stage 3: Acute Abstinence (Nicotine Withdrawal Syndrome)
Acute abstinence in heavy smokers invariably causes the nicotine withdrawal syndrome, characterized by cigarette craving, hedonic dysregulation, cognitive difficulties, and increased negative affect (Jackson et al., 2015). Craving is a negatively reinforcing aspect of nicotine use disorder and is important for conferring relapse vulnerability (Ferguson and Shiffman, 2009). In a longitudinal study of smokers during abstinence, the strength of urges to smoke showed an exponential decline over 12 months of abstinence (Ussher et al., 2013).
Six months after cessation, 13% of ex-smokers still reported "strong urges,” but after 12 months, no ex-smokers reported "strong urges,” although 34% reported "some urges.” Since the nicotine withdrawal syndrome and craving in particular are remarkably durable over time, the effects of acute abstinence on brain activity and connectivity could provide insights into nicotine use disorder and its refractoriness to treatment.
Several studies have examined how brain circuits are altered during acute withdrawal, suggesting altered large-scale brain network dynamics between salience, executive control, and default mode networks. For example, a within-subject study of the effect of 24-hour abstinence compared to satiation on resting-state connectivity in smokers demonstrated that abstinence compared to satiety was associated with weaker mutual inhibition between the default mode and salience networks (Lerman et al., 2014). Weaker between-network coupling predicted abstinence-induced cravings to smoke and less suppression of default mode network activity during a working memory task (Lerman et al., 2014). The insula specifically may be involved in causing the altered network connectivity observed in withdrawal. To investigate this, Ding and colleagues (Ding and Lee, 2013) studied 21 heavy smokers in cigarette-sated and abstinent
65


conditions. After smoking-replenishment, directed connectivity from salience network to default mode network was significantly reduced and directed connectivity from both executive control and default mode networks to the salience network was enhanced. Moreover, the insula showed significantly increased directed connectivity with salience, default mode, and executive control regions in cigarette abstinence compared to satiation. This suggests that directed information flow from the insula to other brain regions is increased in abstinent compared to sated heavy smokers, possibly reflecting increased signaling of withdrawal symptoms and craving. Moran-Santa Maria and colleagues (Moran-Santa Maria et al., 2015) studied acutely abstinent smokers using an fMRI visual craving-cue task. Psychophysiologic interaction with a seed in the right anterior insula was used to infer directed connectivity. Results demonstrated significantly greater effective connectivity from the right anterior insula to the bilateral precuneus, a key node of the default mode network, during smoking compared to neutral cues. Insula-to-precuneus effective connectivity showed a significant positive correlation with craving magnitude, providing further evidence that this circuit between salience and default mode networks plays an important role in cue-induced craving.
Causal effects of cigarette cues on brain function during acute withdrawal were also investigated by Claus and colleagues (Claus et al., 2013). The investigators studied neural responses to cigarette cues in 116 smokers abstinent for >3 hr using a psychophysiological interaction centered on a left dorsal anterior insular seed. Results suggested that smoking cues compared to neutral cues caused stronger connectivity between the left insula and multiple nodes, including right insula (anterior and posterior), amygdala, somatosensory cortex, orbitofrontal cortex, and striatum. In addition, during smoking video exposure significant positive correlations were observed between insula activity and dependence severity; and again, salience network coherence was associated with dependence severity. The authors speculated that the anterior insula may contribute to the initial evaluation of cigarette cue
66


value, interoceptive processing of withdrawal symptoms, and engagement of motor circuits in preparation for drug-seeking behavior.
Drug expectancy, or prior beliefs about impending acute nicotine administration, has been shown to be a factor that modulates the effects of acute withdrawal. Gu and colleagues (Gu et al., 2016) studied 24 overnight abstinent smokers who performed a sequential reward learning task immediately after a cigarette-smoking intervention. Smokers received either a 0.6 mg nicotine cigarette or a de-nicotinized cigarette and were either told that the cigarette contained "nicotine” or "no nicotine”. All subjects completed all four intervention conditions. Only when smokers received a cigarette with nicotine and were told that it contained nicotine, significant activation in the ventral anterior insula was observed during a reward learning task, which was positively correlated with craving magnitude. This suggests that the anterior insula is not only involved in interoceptive processing, but that anterior insula processing of craving and reinforcement learning is modulated by drug expectancy, presumably through top-down cognitive influences.
Stage 4: Chronic Abstinence (Neuroplastic Recovery)
It is unclear whether neural function returns to healthy levels following long-term abstinence, or if the differences associated with nicotine use disorder are durable even after years of abstinence. Few studies have investigated the neuroimaging correlates of long-term abstinence in nicotine use disorder. This is unfortunate, because although chronic nicotine exposure results in upregulation of nicotinic acetylcholine receptors throughout the brain (Breese et al., 1997; Gentry and Lukas, 2002), former smokers exhibit nicotinic acetylcholine receptors concentrations similar to non-smokers (Breese etal., 1997), suggesting that pathologic upregulation is reversible. Similar evidence of neuroplastic recovery is suggested by behavioral changes during chronic abstinence. Measures of impulsivity have been shown to be abnormally elevated in active smokers, but former smokers show levels similar to never-
67


smokers (Bickel et al., 1999). Studies of ex-smokers thus may provide important insights into the successful maintenance of smoking cessation.
Despite limited literature, several small studies have examined insula connectivity dynamics in chronic abstinence. Zanchi and colleagues (Zanchi et al., 2015) studied non-smokers, active smokers, and ex-smokers during a craving-cue task fMRI scan and reported several findings supporting insular role in recovery. First, ex-smokers with greater right anterior insular activity in response to cigarette cues also had higher lifetime nicotine consumption. Second, ex-smokers demonstrated decreased circuit strength between the right anterior insula and anterior cingulate compared to non-smokers, but no significant difference was observed between ex-smokers and active smokers. This suggests that insular function may not completely recover in long-term abstinence, possibly reflecting a mechanism of persistent craving. Another fMRI study (Nestor et al., 2011) of smokers, ex-smokers, and healthy controls used an attentional bias paradigm with neutral cues, emotionally-evocative cues, and smoking cues. Across all cue conditions, ex-smokers exhibited significantly greater activation in the right anterior insula compared to active smokers and controls. In a separate experiment employing a go/no-go paradigm to investigate motor response inhibition and cognitive error monitoring, the ex-smokers had significantly greater error-related activation than both controls and smokers in the left insula. Taken together, these results suggest that heightened insular monitoring of cues and errors contribute to the successful maintenance of abstinence. Higher right anterior insular activity in ex-smokers compared to healthy controls may reflect a hypervigilance against smoking cues necessary for successful long-term abstinence.
4.3.3. Putting it All Together: Unified Models of the Role of the Insula in Nicotine Use Disorder Pathogenesis
Insular function in the context of nicotine use disorder illustrates this brain region’s important role in higher cognitive function in normal, non-addicted persons. One of the primary
68


functions of the anterior insula is salience detection (Seeley et al., 2007), such as identifying stimulus features that stand out or are of instinctual or learned importance. Saliency involves the selection of stimuli from a continuous stream of internal and external sensory inputs for additional processing. Another theory is that the anterior insula serves as the "apex of a predictive cortical hierarchy” that spans all sensory systems (Barrett and Simmons, 2015; Chanes and Barrett, 2016), selecting goal-relevant stimuli for attention and cognitive processing. The insula is unique amongst cortical areas in that it contains sequential yet overlapping maps from all exteroceptive and interoceptive senses (Craig, 2009, 2011). These higher-order maps are successively re-represented from posterior insula to middle insula to anterior insula, progressively acquiring additional sensory input maps, interoceptive signals, and reward signals along the way. The anterior insula then provides a single cortical representation of how an individual is feeling at a given time: the "global emotional moment” (Craig, 2009) or "cinemascopic awareness” (Craig, 2011). Despite these slightly differing models and interpretations, the evidence suggests broad involvement of the anterior insula in both polysensory processing and negotiating "bottom-up” sensations versus "top-down” sensitivity control mechanisms of salience. Normal insular physiology thus provides a framework for understanding insular pathophysiology in nicotine use disorders.
Several hypotheses have been proposed to explain the insula’s role in nicotine use disorder. One hypothesis centers on the salience network, comprised of the insula and anterior cingulate cortex. The insula is believed to serve as a toggle, directing brain function towards internal or external stimuli, in order to maintain homeostasis of cognitive resources and guide goal-directed behavior (Sutherland et al., 2012). Internal focus is reflected by greater default mode network (endogenous-oriented) activity, whereas external focus is reflected by greater
69


executive control network (exogenous-oriented) activity (Figure 4-3). The insula’s function of toggling between these two networks is hypothesized to be usurped in nicotine use disorder.
The concept that the insula directs attention towards the most homeostatically relevant
Right Dorsal Anterior Insula
Right Ventral Anterior Insula
Normalized Correlation Coefficient [1]
Precuneus / Posterior Cingulate Cortex (Major Default Mode Network Node)
Dorsolateral Prefrontal Cortex (Major Executive Control Network Node)
Figure 4-3. Differential resting-state connectivity of the dorsal (left) and ventral (right) right anterior insula, using the Human Connectome Project Connectome Workbench (n = 1206), uncorrected. Dorsal right anterior insula is strongly connected with frontoparietal regions involved in executive control; in contrast, ventral right anterior insula connectivity is strongly connected with default mode network regions involved in internal feelings and self-referential processing. Note that the externally-directed system (dorsal anterior insula and executive control network) and internally-directed system (ventral anterior insula and default mode network) are inversely correlated, consistent with mutual inhibition (see Figure 4-1).
stimuli - internal or external - provides a neurobiological model to explain both cognitive changes and functional connectivity findings of acute nicotine ingestion, nicotine satiety in dependence, and nicotine withdrawal syndrome. This review of the evidence suggests that the critical role of the insula in maintaining nicotine use disorder is related to its function in providing conscious awareness of craving and withdrawal symptoms. Lerman and colleagues (Lerman et al., 2014) provided supporting evidence for this model, reporting decreased
70


between-network coherence amongst salience, default mode, and executive control networks in abstinence compared to satiety. They reported that weaker between-network coupling predicted abstinence-induced cravings and less suppression of default mode activity during performance of a subsequent working memory task, possibly reflecting a mechanism of cognitive and attentive impairments commonly observed during the nicotine withdrawal syndrome.
In summary, the evidence suggests that insular function is disrupted compared to healthy controls across all stages of nicotine use disorder (Table 4-1). Salience network coherence between insular and anterior cingulate nodes is particularly important for craving-induced behaviors, reflects disease severity, and has prognostic value. However, large-scale longitudinal studies are needed to understand the altered connectivity profiles of the insula with salience, default mode, and executive control regions at different disease stages. Although the evidence is still largely comprised of single-site, small-population studies, it provides a compelling neurobiological argument for future work. In the following section, we explore in detail the possibility of targeting the insula with neuromodulation as a therapeutic strategy for promoting abstinence.
4.4. Implications of Insular Role in Nicotine Use Disorder on Neuromodulatory
Therapeutic Development
While pharmacotherapy can be used with some efficacy to diffusely modulate dysregulated brain circuits with the goal of promoting abstinence, new treatment strategies clearly are needed. One possibility is a targeted circuit-node approach, aligned with current understanding of the underlying pathology. Such a neurocircuit-based approach may improve successful cessation by intervening upon or modulating specific neuroanatomical structures that serve as key nodes within behaviorally relevant circuits, such as the insula in the case of nicotine use disorder. A promising candidate for this approach is neuromodulation.
71


4.4.1. Therapeutic Neuromodulation in Animal Models of Nicotine Use Disorder
Aside from systemic pharmacologic approaches, animal models have shown significant benefit of targeted neuromodulation. For example, one study (Forget et al., 2010) reported that insular inactivation via intracranial GABA agonist microinfusion in nicotine-dependent rats significantly reduced nicotine motivation, nicotine seeking-, and nicotine taking-behaviors, with no effect on food behaviors. These findings were further confirmed using an alternative lesioning method, bilateral insular deep brain stimulation, in a rat model of nicotine dependence (Pushparaj etal., 2013). Kutlu and colleagues (Kutlu etal., 2013) extended these findings by showing that locally infused Di but not D2 antagonists into the rostral anterior insular cortex decreased rats’ nicotine self-administration acutely by more than 50%, with repeated Di antagonist infusions resulting in continued decreases in consumption without evidence of tolerance. The cause-and-effect relationship between decreased Di activity in the insula and decreased nicotine self-administration suggests that mesocorticolimbic dopaminergic afferents onto the anterior insula are critically involved in promoting and maintaining nicotine dependence (Kutlu et al., 2013). Disrupting this insular mechanism leads to diminished nicotine consumption, possibly through diminished interoception of reward (or lack of reward) signals.
4.4.2. Therapeutic Neuromodulation in Humans with Nicotine Use Disorder
Transcranial magnetic stimulation (TMS) is a neuromodulation technique that shows promise as a means to target insula function. TMS is a noninvasive intervention in which extracorporeal current-carrying electrical coils are used to induce rapid, transient, focal magnetic fields targeting a specific brain region. These transient magnetic field fluxes cause electromagnetic induction in underlying neural tissues that alter neural transmembrane potentials and in turn affect neural activity. Applying a sequence of TMS pulses causes longterm effects that either facilitate or inhibit neuronal excitability, depending upon multiple
72


factors, including pulse parameters and stimulation site. Based on studies of the corticospinal motor tract, low-frequency (< 1 Hz) repetitive TMS is inhibitory and high-frequency (> 5 Hz) repetitive TMS is faciliatory, with aftereffects closely paralleling long-term depression and longterm potentiation mechanisms of neuroplasticity (Hoogendam et al., 2010).
Since 2003, several studies in the English language literature have investigated the role of high- or low-frequency TMS targeting the dorsolateral prefrontal cortex in cigarette craving mitigation (Sailing and Martinez, 2016). The focus in the literature on high-frequency TMS of the dorsolateral prefrontal cortex in smokers is likely related to its demonstrated efficacy in major depressive disorder (Brunoni et al., 2017; Milev et al., 2016) and relative accessibility of this region as a superficial target site compared to other, deeper structures. Overall, these studies demonstrate that both single-session and repeated-sessions of TMS to the dorsolateral prefrontal cortex reduces cigarette craving, although in some studies the cigarette consumption effects were mixed, highlighting the need for objective endpoints and response measures (as opposed to self-reported craving measures) in clinical trials.
Given the efficacy of dorsolateral prefrontal cortex neuromodulation in promoting smoking cessation, it stands to reason that other cortical areas involved in nicotine use disorder, such as the insula, may be useful therapeutic targets. Enhancing top-down control of craving-related behaviors has been shown effective. However, this is arguably an indirect method - by augmenting top-down control through excitation of dorsolateral prefrontal cortex, this approach is thought to enhance suppression of bottom-up craving urges. Alternatively, inhibiting the bottom-up craving urges from the cortical source itself may be a more direct and effective treatment Human stroke and animal neuromodulation studies reviewed here implicate a crucial cause-and-effect role of the insula in maintaining bottom-up craving sensations and nicotine-consuming behaviors. In chapter 5, we report the main results of a randomized sham-controlled clinical trial fwww.ClinicalTrials.gov identifier: NCT 02590640) in
73


active smokers to investigate the efficacy of insular inhibitory neuromodulation in directly reducing cigarette craving at the presumed neural source.
One of the major technical limitations in TMS is the limited spatial depth of electromagnetic induction and its inverse relationship with focality. Theoretical simulations have demonstrated that commercially available figure-of-eight coils can penetrate the cortex 1.0 to 3.5 cm within normal safety parameters (Deng et al., 2013). Across different coil geometries, stimulation of deeper brain targets necessitates greater electrical field spread (reduced focality). This tradeoff between electric field depth and focality poses important physical challenges to stimulating the anterior insula. Moreover, non-target brain stimulation further complicates investigation of behavioral or clinical outcomes associated with deep brain targeting, because adjacent and superficial areas are included in the treatment field and thus may confound observed associations. For example, one study applied continuous TMS to the right anterior insular cortex and control regions (occipital and somatosensory cortices) in healthy volunteers using a superficial (i.e., planar figure-of-eight) coil (Pollatos et al., 2016). Their results suggested that inhibiting the right anterior insula was associated with a significant decrease in cardiac and respiratory interoceptive accuracy (measured by a heartbeat counting task) as well as decreased perceptual confidence. There is debate about the targetability of the anterior insula using conventional superficial coils (Figure 4-4), however, with investigators noting that by using this approach the anterior insula receives about 2 5% of the maximum cortical energy deposition and greatest deposition in the overlying frontal and temporal opercula (Coll etal., 2017; Pollatos and Kammer, 2017).
Recently, a family of coil designs called Hesed (H) coils have been developed to achieve deep brain neuromodulation at the expense of a wide, relatively non-focal treatment field.
74


Figure 4-4. Finite element model results of the predicted electromagnetic field produced by a standard superficial planar transcranial magnetic stimulation 70mm figure-of-eight coil targeting the right anterior insula. Predicted current flux density (left) and normalized absolute value of the electric field (right) illustrating the pattern of energy deposition. Maximum energy is deposited in the scalp, superficial soft tissues, and cerebrospinal fluid due to high tissue conductivities. This illustrates the difficulty in targeting deep structures, such as the insula or anterior cingulate cortex. This is consistent with model results of insular targeting reported by Pollatos and colleagues (Pollatos and Kammer, 2017). Created using SimNIBS version 2.1.1 with right anterior insular target [36,10, -6] in MNI space and default parameters (Thielscher etal., 2015).
These coils provide near-complete stimulation of the frontal lobes. Dinur-Klein et al (Dinur-Klein et al., 2014) randomized a large sample of 115 heavy smokers to 13 daily treatments of high-frequency, low-frequency, or sham TMS using an H-coil designed to target the bilateral ventrolateral prefrontal cortex, dorsolateral prefrontal cortex, and insula. Smoking was measured by participants’ self-report and urine cotinine levels. High-frequency TMS, but not low-frequency or sham TMS, during the presentation of visual smoking cues resulted in a 44% reduction in smoking at 3 months and 33% reduction at 6 months. Counter-intuitively, there was no significant difference in self-reported craving, suggesting that these effects may reflect
75


enhanced cognitive control rather than reduced incentive salience or reduced sensation of cigarette craving. This suggests that their findings of decreased cigarette consumption after high-frequency stimulation may reflect enhanced dorsolateral prefrontal cortex activity and top-down cognitive control. Malik S et al (Malik et al., 2018) applied excitatory and inhibitory TMS to the bilateral insula and surrounding cortical opercula using the H-coil in eight healthy participants in a within-subject, crossover, blinded, sham-controlled, proof-of-concept study. Synaptic effects were measured using PET with a dopamine agonist tracer. They demonstrated that inhibitory (1 Hz) TMS compared to sham and excitatory (10 Hz) TMS significantly decreased dopamine concentrations in the substantia nigra and sensorimotor striatum, with a trend towards significance in the associative striatum. In both these studies claiming to modulate the insula, the investigators could not definitively confirm that the insular cortex is indeed being stimulated, although future studies using fMRI could address this question.
Neuromodulatory methods not only include TMS, but also include transcranial direct current stimulation (tDCS) and deep brain stimulation (DBS). tDCS involves applying an electrical current to the brain between two electrodes, which affects neural tissues within the path of least electrical resistance. tDCS has been used to target the dorsolateral prefrontal cortex in smokers with reduction in cue-induced cravings; however, the non-focal nature of this method limits its utility in targeted, neuroanatomically-driven neuromodulation (Sailing and Martinez, 2016). DBS, on the other hand, involves surgically implanting a stimulating electrode into target brain tissue. DBS targeting the ventral striatum in smokers has been reported in only one study, which reported higher rates of successful cessation compared to unaided smoking cessation in the general population (Kuhn etal., 2009). However, surgically-placed deep brain stimulation is invasive and practically limited by significant ethical considerations.
In summary, applications of non-invasive methods of brain stimulation in nicotine addiction are currently limited by their lack of spatial specificity and depth of targetability.
76


While dorsolateral prefrontal cortex neuromodulation has been shown to improve abstinence rates and mitigate cravings, other cortical areas such as the anterior insula have broader empirical support and may result in stronger treatment responses. Application of TMS to nicotine use disorder is promising but future studies are needed to define optimal targets, paradigms, and patient population. There is a clear clinical need for better smoking cessation treatments, and growing evidence specifically implicates the insula as a rational neuroanatomical target for investigational modulation therapies.
4.5. Conclusion
The insula is functionally heterogeneous, with distinct patterns of connectivity with large-scale brain networks associated with numerous functions and behaviors. Animal models and human lesion studies suggest that the insula is necessary for the maintenance of nicotine-seeking behaviors and nicotine-taking behaviors, likely through nicotine craving. Given the limited efficacy of standard-of-care treatments for nicotine use disorder, neuromodulation of this region may contribute to the next generation of cessation treatments by offering what henceforth has not been available: a targeted, neuroanatomically-driven approach to smoking cessation therapy. Moreover, because substance use disorders in general - including nicotine use disorder - are thought to be initiated and reinforced by maladaptive alterations in the dopamine reward system and associated corticolimbic and cognitive control circuits, such a targeted neuroanatomically-driven approach may advance treatments for other drug addictions as well. There is a clear clinical need for better smoking cessation treatments. Evidence strongly implicates the insula as a rational neuroanatomical target for modulation therapies. Neuromodulation of insula function has significant potential to improve smoking cessation rates amongst smokers, but continued technical developments and research are needed to overcome challenges in depth and specificity of targeting.
77


4.6. Acknowledgements
Several co-authors contributed to this chapter: Jason Tregellas PhD, Benzi Kluger MD MS, Korey Wylie MD, Joshua Gowin PhD, and Jody Tanabe MD. This work was funded by NIH F32 DA041011 (MFR), RSNA RR-1620 (MFR), and CCTSI M-15-81 (MFR).
78


CHAPTER V
INSULAR INHIBITORY NEUROMODULATION IN SMOKERS DECREASES CIGARETTE CRAVINGS AND BRAIN RESPONSES TO CIGARETTE CUES: A RANDOMIZED CONTROLLED
TRIAL
5.1. Abstract
Cigarette addiction is a leading preventable cause of mortality, morbidity, and healthcare costs. Several lines of evidence suggest that the insula contributes to urges to smoke, and that inhibiting the insula can disrupt cigarette addiction. However, no studies to date have attempted therapeutic insular modulation in people. We hypothesized low-frequency repetitive transcranial magnetic stimulation (LF-rTMS) targeting the right anterior insula would decrease cigarette cravings and brain responses to cigarette cues. We conducted a randomized, singled-blinded, sham-controlled, phase I clinical trial in which active smokers (n=40) interested in quitting received a single session of either right anterior insula deep LF-rTMS (n=20) or sham treatment (n=20). Primary outcomes included craving measures and cigarette craving cue task fMRI (3T) brain activity responses, measured before and after treatment. Finite element model simulations of energy deposition intracranially informed a priori selection of regions of interest used to corroborate whole-brain results. Compared to sham treatment, right insula LF-rTMS reduced self-reported cigarette craving (p=0.033). Right insula LF-rTMS also decreased brain activity responses to visual cigarette cues at the whole-brain level in primary sensorimotor cortices, supplementary motor area, and right anterior insula (p < 0.001, pvoxel < 0.005, k >
464 voxels). There were no brain regions in which LF-rTMS caused increased activity response to cigarette cues. A single session of right anterior insula deep LF-rTMS reduced cigarette
79


cravings and brain activity in response to cigarette cues. These findings provide proof-of-concept of a potential neuroanatomical target for smoking cessation therapy.
5.2. Introduction
Cigarette addiction is a leading preventable cause of premature death, morbidity, and healthcare costs (Health and Services, 2014). Each year, approximately 480,000 people in the U.S. prematurely die from smoking-related illnesses. Overall, 1 in 5 U.S. deaths are caused by smoking (Health and Services, 2014). It is estimated that 46 million (~1 in 5) Americans smoke. About half of these smokers attempt to quit each year, with 70% of smokers wanting to quit (Babb, 2017). Unfortunately, standard-of-care smoking cessation treatments are largely ineffective. Approximately 80% of patients attempting to quit relapse within six months, despite combined pharmacologic and behavioral therapies (Tobacco, 2008). Craving is a defining feature of nicotine use disorders and predicts relapse (Potvin et al., 2015; Saunders and Robinson, 2013). For these reasons, exploring novel treatments to reduce craving remains an important objective.
Evidence has implicated the insula in the maintenance of cigarette craving and smoking behaviors. Naqvi and colleagues (Naqvi et al., 2007) reported that a significant proportion of patients with damage to the insula compared to other brain areas were able to stop smoking easily without cravings or relapse. Subsequent prospective studies confirmed that insular lesions in smokers strongly predicted spontaneous continuous abstinence, fewer nicotine withdrawal symptoms, and reduced cravings (Abdolahi et al., 2015a, b, 2017; Suner-Soler et al., 2012). This pattern was also observed in animal experiments using pharmacologic and electrical lesioning of the insula (Forget etal., 2010; Kutlu etal., 2013; Pushparaj et al., 2013). For example, insular lesioning via intracranial GABA-agonist microinfusion in nicotine-dependent rats significantly reduced nicotine motivation-, nicotine seeking-, and nicotine taking-behaviors, with no effect on food behaviors (Forget et al., 2010). However, to date no
80


studies have attempted to modulate smokers’ insular function in vivo to investigate possible therapeutic benefits.
One promising approach to modulate insular function is through transcranial magnetic stimulation (TMS). TMS is a non-invasive neuromodulation technique that has recently demonstrated early-phase success in promoting smoking cessation. Studies of the corticospinal motor tract have shown that low-frequency repetitive TMS (< 1 Hz, LF-rTMS) is inhibitory and high-frequency repetitive TMS (> 5 Hz, HF-rTMS) is facilitatory, with aftereffects closely paralleling long-term depression and long-term potentiation mechanisms of neuroplasticity, respectively (Hoogendam etal., 2010).
While no studies to date have investigated insular TMS in smokers, several studies have investigated effects of LF- and HF-rTMS targeting the dorsolateral prefrontal cortex (DLPFC) on cigarette craving and consumption (for reviews, see Kedzior et al. (2018); Makani et al. (2017); Sailing and Martinez (2016); Song et al. (2018)). Multiple studies demonstrate that excitatory HF-rTMS to the DLPFC reduces cigarette craving. The mechanism is thought to involve augmenting "top-down” executive control, thereby enhancing cognitive suppression of "bottom-up” craving urges. There are currently no human studies, however, that investigate disrupting this system in smokers by inhibiting "bottom-up” craving urges from the presumed cortical source of craving itself. Similar bottom-up approach has been attempted in cocaine users by targeting cortical reward centers in the orbitofrontal cortex (Hanlon et al., 2017). Hanlon et al. reported that left frontopolar TMS delivered in a single day in cocaine addicts significantly decreased TMS-evoked BOLD signal in the orbitofrontal cortex and insula. Only one study attempted insular modulation in smokers - by targeting the bilateral DLPFC using an H-coil (Dinur-Klein et al., 2014), but the large treatment field precluded distinguishing effects of insular from overlying DLPFC stimulation. A small number of studies have attempted to modulate posterior-superior insular function in healthy participants, reporting changes in
81


thermal/pain sensation (Ciampi de Andrade et al., 2012; Lenoir et al., 2018) and interoceptive tasks such as a heartbeat counting (Pollatos et al., 2016), although the targetability of the insula is disputed (Coll et al., 2017). Spagnolo et al. targeted the bilateral prefrontal cortices and insulae using the H-coil in healthy individuals and reported no measurable effects on either a blink suppression task or a forced-choice risk-taking task (Spagnolo et al., 2018). No studies to date have selectively targeted the insula in smokers. We sought to address this gap in the literature with the current study.
We conducted a randomized, singled-blinded, sham-controlled, parallel-group phase I clinical trial (www.ClinicalTrials.gov identifier: NCT 02590640) to answer the question: does right insular inhibition in smokers reduce cigarette craving and brain responses to cigarette cues? This proof-of-concept study hypothesized that a single session of LF-rTMS (inhibitory) targeting the right anterior insula in active smokers would acutely decrease cigarette craving and alter brain activation in response to cigarette smoking cues.
5.3. Materials and Methods
This study was reviewed and approved by the Colorado Multiple Institutional Review Board in accordance with the Declaration of Helsinki. All participants provided written informed consent Recruitment, enrollment, data collection, treatments, and MRI examinations were conducted at the University of Colorado Anschutz Medical Campus. All investigators except the physician administering treatments (MFR) were blinded to group assignment until completion of blinded analysis of primary outcomes, defined as self-reported craving and brain activity responses to cigarette cues.
5.3.1. Study Design
This was a randomized, singled-blinded, sham-controlled, parallel-group, clinical trial in which active smokers interested in quitting received either right anterior insula deep LF-rTMS
82


or sham treatment Randomization was performed using a computerized random number
generator f http: / /www.random.orgl in blocks of two. A physician trained in TMS and seizure management administered all treatments. Participants were blinded to group assignment and asked if they believed they received real or sham treatment, with their responses recorded.
This study was subject to a protocol deviation. In four participants originally randomized to LF-rTMS, there was insufficient time to measure RMT and apply LF-rTMS between scheduled MRI exams. This timing issue was due to random factors unrelated to the participant’s smoking or physiology; it occurred when there was another research MRI scheduled immediately after the current study’s post-treatment MRI. Subsequent logistical changes to the scheduling of participants obviated this issue (i.e., participants were scheduled in evenings such that there were no exams for other research studies scheduled after the current study’s pre-treatment MRI, allowing for greater flexibility in timing).
5.3.2. Sample Population
Participants were recruited through internet advertisements, publicly-posted flyers, and tobacco cessation consultations at the University of Colorado Hospital Emergency Department and Inpatient Services. Ninety-one healthy, right-handed, treatment-seeking, smokers (> 10 cig/d for >1 year) between 18 and 55 years old were recruited; forty completed full study appointments (Figure 5-1). Telephone screening included brief medical, substance use, and psychiatric histories to determine eligibility before scheduling the study appointment Exclusion criteria included [1] use of non-cigarette tobacco products; [2] current use of nicotine replacement therapy, bupropion, or varenicline; [3] major medical disorders; [4] current pregnancy or pregnancy-seeking; [5] active abuse or dependence of illicit substances; [6] MRI or TMS contraindications; and [7] self-reported major psychiatric disorder.
Eligible participants were invited for study appointments. Participants were instructed to maintain their normal smoking habits and not smoke for >3 h prior to their study
83


appointment, confirmed by exhaled carbon monoxide concentration ([CO]Exhaied)- Brief abstinence was imposed to establish a baseline state of craving and promote susceptibility to smoking-related visual cues.
At the beginning of the study appointment, each participant provided medical, surgical, and social histories and underwent a brief neurologic examination by a physician (Figure 5-2).
Figure 5-1. CONSORT enrollment diagram for this phase 1 human trial. Careful attention was paid to maintaining all participants’ blinding to group assignment throughout each study visit.
Subjects’ breath was tested for alcohol content to exclude participants acutely under the influence of alcohol. [COjExhaied was measured with a MicroSmokelyzer (Bedfont Scientific; Kent, United Kingdom) to confirm acute abstinence ([COjExhaied <10 ppm). A urine sample was
84


Full Text

PAGE 1

ADDICTION IS A BRAIN DISEASE: COMPUTATIONAL AND FU NCTIONAL NEUROANATOM Y OF SUBSTANCE USE D ISORDERS USING ADVANCED MRI A ND NEUROSTIMULATION INDUCED BRAIN LESION S by MICHAEL FRANCIS REGNER B.A., University of Wisconsin Madison, 2009 M.D., University of Wisconsin Madison, 2013 A dissertation submitted to the Faculty of the Graduate School of the University of Colorado in partial fulfillment of the requirements for the degree of Doctor of Philosophy Bioengineering 2019

PAGE 2

ii Copyright © 2019 MICHAEL FRANCIS REGNER, M.D. ALL RIGHTS RESERVED. Medicine is an ever changing science. Care has been taken to confirm the accuracy of the information presented and to describe generally accepted practices. However, the authors, editors, and publisher are not responsible for errors or omissions or for any consequences from application of the information in this book and make no warranty, express ed or implied, with respect to the contents of the publication. Some drugs and /or medical devices presented in this publication have Food and Drug Administration ( FDA) clearance for limited use in restricted research settings. It is the responsibility of the health care provider to ascertain the FDA status of each drug or device planned for use in their clinical practice.

PAGE 3

iii This dissertation for the Doctor of Philosophy degree by Michael Francis Regner has been approved for the B ioengineering Program by Kendall Hunter ( Chair ) Jody L. Tanabe ( Co Advisor ) Jason R. Tregellas ( Co Advisor ) Benzi Kluger Vitaly Kheyfets Robin Shandas Approved: May 18, 2019

PAGE 4

iv Regner, Michael Francis (Ph.D., Bioengineering) Addiction is a Brain Disease: Computational and Functional Neuroanatomy of Substance Use Disorders Using Advanced MRI and Neurostimulation Induced Brain Lesions Dissertation directed by Professor Jody Tanabe and Professor Jason Tregellas ABSTRACT Substance use disorder s (SUD) are a group of psychiatric diseases associated with significant mortality, morbidity, economic losses, social and family problems, and per sonal distress. Advances in the ability to image brain structure and function in vivo provide an opportunity to understand addictions . Novel techniques, such as transcranial magnetic stimulation, allow the ability to alter neurocircuitry function for poten tial diagnostic and therapeutic benefit. This dissertation sought to study SUD as diseases of neurocircuitry affecting reward, craving, and goal oriented behavior. We conducted advanced neuroimaging experiments involving two different populations of SUD: l ong term abstinent, severely dependent cocaine and methamphetamine addicts (Chapters 2 and , 3); and, acutely abstinent, moderately dependent cigarette smokers (Chapters 4 and 5 and Appendix 1). Chapters 2 and 3 investigated structural and functional neuroi maging differences, respectively, in long term abstinent cocaine and methamphetamine addicts compared to healthy controls. In Chapter 2, we observed that grey matter volumes in stimulant dependence differed compared to healthy controls (n=127) , and that th ese changes differ ed by sex. W e discuss the important implications with regards to sex differences in the natural history of psychostimulant dependence. In Chapter 3, we investigated brain function as rest by functional connectivity changes amongst large scale brain networks in a subset of the Chapter 2 population (n=100). We found that after long term abstinence, large scale brain networks

PAGE 5

v thought to be involved in the pathogenesis of addiction follow ed a pattern of increased top down cognitive control, which may reflect the necessary increased executive control over habit and reward systems to maintain disease remission. Stimulant dependence also demonstrated greater global efficiency and lower local efficiency amongst large scale brain networks compar ed to healthy controls , suggesting abnormal brain organization despite long term abstinence. Chapters 4, 5, and Appendix 1 introduce and report results on our phase 1, sham controlled, single blinded, randomized clinical trial ( www.ClinicalTrials.gov identifier: NCT 02590640 ) to investigate the effect of inhibitory insular transcranial magnetic stimulation (TMS) on cigarette cravings and brain function in acutely abstinent, moderately dependent smokers . Compared to sham treatment, inhibitory TMS targeting the right insula reduced self reported cigarette craving and decreased brain activity responses to visual cigarette cues in primary sensorimotor cortices, supplementary motor area, and right anterior insula. The se findings provide proof of concept of a potential neuroanatomical target for smoking cessation therapy. While the results should be considered preliminary, they provide hope that TMS could be developed as a treatment strategy to help reduce the burden of cigarette addiction. Overall, t hese neuroimaging studies in two different populations of substance use disorders provide evidence that addiction is a brain disease with endogenous (i.e., increased top down control, Chapter 3) and exogenous (i.e., therapeutically imposed, Chapter 5) mechanisms of remission and treatment. The form and content of this abstract are approved. I recommend its publication. Approved: Jody L. Tanabe Jason R. Tregellas

PAGE 6

vi F or Emilie w ho is my critic when I feel I have nothing left to perfect, and who sings my praises when I feel I have nothing worthy of praise.

PAGE 7

vii ACKNOWLEDGEMENTS Firstly, I would like to thank my dissertation advisor and mentor, Dr. Jody Tanabe. She gave me her tim e and an opportunity when I was a brand new medicine intern with nothing to offer. Then , she challenged me to pursue new directions and pushed me to develop as a scientist as my skills developed . Over the past six years Jody has been a great role model and friend . For this , I am forever grateful. I am particularly grateful to my dissertation advisor, Dr. Jason Tregellas, without whom this work would not have been possible. Jason has been invariably generous with his time, expertise, and resources. I am grat eful to several others without whom this work would not have been possible: Dr. Kendall Hunter, Dr. Benzi Kluger, Dr. Vitaly Khetfets, Dr. Robin Shandas, Dr. Deb Glueck, and Dr. Gerald Dodd III. I thank the Departments of Bioengineering and Radiology for t heir unwavering support. I am particularly grateful to my radiology co residents, who were invariably supportive of the call trades and schedule changes facilitating this research. I thank my first research mentor, Dr. Jack Jiang, and my good friend and medical school classmate/roommate, Dr. Matthew Hoffman. I thank my wife, Dr. Emilie Regner , for her constant support . We have both pursued unusual pathways as physician scientists . I am ever fortunate to tak e this journey with you . I thank my late mother, Claire, who died from stage IV lung cancer (and indirectly , a lifetime of cigarette addiction) in December 2017. It is with a heavy heart that I wish you could have seen the final pro duct of this effort . I thank my family, particularly my father Scott and mother in law Kaye, for their constant support. Lastly , but m ost importantly, I would like to thank our patients and research volunteers. A ddictions are often misguidedly considered a g conferring stigma and guilt that impair s awareness, treatment , and scientific progress . W e must nevertheless persist and continue to shine light on these psychiatric diseases. Our field is indebted to those individuals who volunteered to walk into the light of scientific examinatio n , if only for a few hours .

PAGE 8

viii D ECLARATION OF ORIGINAL WORK I affirm that all work in this Doctoral Dissertation is my original work. Further, I confirm that all writing is my own writing. Work from others has been cited appropriately.

PAGE 9

ix TABLE OF CONTENTS CHAPTER 1. INTRODUCTION ................................ ................................ ................................ ................................ ................. 1 1.1. Neuroimaging in Substance Use Disorders ................................ ................................ ............ 3 1.2. Substance Use Disorders as a Class of Heterogeneous Neurocircuitry Diseases ... 4 1.3. Dissertation Narrative Outline ................................ ................................ ................................ .... 6 2. SEX DIFFERENCES IN G REY MATTER CHANGES A ND BRAIN BEHA VIOR RELATIONSHIPS IN STI MULANT DEPENDENCE ................................ ................................ ................ 9 2.1. Abstract ................................ ................................ ................................ ................................ ................. 9 2.2. Introduction ................................ ................................ ................................ ................................ ..... 10 2.3. Materials and M ethods ................................ ................................ ................................ ................ 12 2.4. Results ................................ ................................ ................................ ................................ ................ 17 2.5. Discussion ................................ ................................ ................................ ................................ ......... 23 2.6. Conclusion ................................ ................................ ................................ ................................ ......... 26 2.7. Acknowledgements ................................ ................................ ................................ ....................... 27 3. TOP DOWN NETW ORK EFFECTIVE CONNEC TIVITY IN ABSTINENT SUBSTANCE DEPENDENT INDIVIDUAL S ................................ ................................ ................................ ....................... 28 3.1. Abstract ................................ ................................ ................................ ................................ .............. 28 3.2. Introduction ................................ ................................ ................................ ................................ ..... 29 3.3. Materials and Methods ................................ ................................ ................................ ................ 33 3.4. Results ................................ ................................ ................................ ................................ ................ 40 3.5. Discussion ................................ ................................ ................................ ................................ ......... 45 3.6. Conclusion ................................ ................................ ................................ ................................ ......... 54 3.7. Acknowledgements ................................ ................................ ................................ ....................... 54

PAGE 10

x 4. THE INSULA IN NICOTINE USE DISORD ER: FUNCTIONAL NEURO IMAGING AND IMPLICATIONS FOR NEU ROMODULATION ................................ ................................ ........................ 55 4.1. Abstract ................................ ................................ ................................ ................................ .............. 55 4.2. Introduction ................................ ................................ ................................ ................................ ..... 56 4.3. Insular Role in Nicotine Use Disorder ................................ ................................ ................... 58 4.4. Implications of Insular Role in Nicotine Use Disorder on Neuromodulatory Therapeutic Development ................................ ................................ ................................ .......... 71 4.5. Conclusion ................................ ................................ ................................ ................................ ......... 77 4.6. Acknowledgements ................................ ................................ ................................ ....................... 78 5. INSULAR INHIBITORY N EUROMODULATION IN SM OKERS DECREASES CIGA RETTE CRAVINGS AND BRAIN R ESPONSES TO CIGARETT E CUES: A RANDOMIZED CONTROLLED TRIAL ................................ ................................ ................................ ................................ .... 79 5.1. Abstract ................................ ................................ ................................ ................................ .............. 79 5.2. Introduction ................................ ................................ ................................ ................................ ..... 80 5.3. Materials and Methods ................................ ................................ ................................ ................ 82 5.4. Results ................................ ................................ ................................ ................................ ................ 95 5.5. Discussion ................................ ................................ ................................ ................................ ...... 103 5.6. Conclusion ................................ ................................ ................................ ................................ ...... 108 5.7. Acknowledgements ................................ ................................ ................................ .................... 108 5.8. Supplement: Justification of Analysis Approach for Primary and Secondary Outcomes ................................ ................................ ................................ ................................ ........ 109 6. CONCLUSION ................................ ................................ ................................ ................................ ................ 114 6.1. Specific Knowledge Gaps Addressed ................................ ................................ .................. 114 6.2. Limitations ................................ ................................ ................................ ................................ ..... 116 6.3. Future Work ................................ ................................ ................................ ................................ .. 117

PAGE 11

xi 6.4. Concluding Remarks ................................ ................................ ................................ .................. 118 REFERENCES ................................ ................................ ................................ ................................ ................................ 119 APPENDIX ................................ ................................ ................................ ................................ ................................ ...... 138

PAGE 12

xii TABLE OF FIGURES FIGURE 1 1. Simplified scheme illustrating how common drugs of abuse modulate the reward circuit. et al., 2015). ... 2 1 2. Simplified illustration of the natural history of substance use disorders. Active substance dependence (red) is characterized by sequential stages of binge/intoxication, withdrawal/negative affect, and craving/preoccupation (Koob and Volkow, 2010) that repeat a variable number of cycles, represented by m . Active disease episodes may be interrupted by periods of short or long term abstinence, represented by n . Relapse is indicated in blue. Note that at every stage of potential or active disease (top, in white) there is a possible avenue towards disease remission (bottom, in grey). Used w ith permission from Regner et al. (2016). .... 5 2 1. Sample population inclusion and exclusion criteria. ................................ ................................ .......................... 12 2 2. T value map illustrating statistically significant greater grey matter volumes in healthy control women compared to stimulant dependent wom en, after controlling for age and brain size ( p <0.001). ................................ ................................ ................................ ................................ ................................ ............... 18 2 3. T value map illustrating statistically significant greater regional grey matter volumes in control women compared to control men, after controlling for age and brain size ( p <0.001). ...................... 19 2 4. Left , T value map of statistically significant negative correlations found on whole brain level between stimulant dependence symptom count and grey matter volume within the bilateral nucleus accumbens (the so Right , negative correlation between the substance dependence symptom count and nucleus accumbens total volume as defined by ROI. ................................ ................................ ................................ ................................ ................................ ............................ 20 2 5. Sex by group by behavior interactions on GMV. Left , scatterplots illustrating the sex by group interaction on the correlation between approach and GMV in the bilateral middle frontal gyri (dorsolateral prefrontal cortex). Right , scatterplots illustrating the sex by group interaction on the correlation between impulsivity and GMV in the left superior temporal gyrus an d left insula.

PAGE 13

xiii Middle , clusters of whole brain significance demonstrating sex by group by impulsivity (red) and sex by group by approach (green) interactions. ................................ ................................ ................................ ... 21 2 6. Boxplots illustrating the total grey matter volume in each subpopulation for each r egion of interest. The top and bottom edges of the boxes indicate the third and first interquartiles, respectively. Lines in the middle of the boxes indicate the medians. Whiskers above and below the boxes indicate the 90 th and 10 th percentiles, respect ively. Points above and below the whiskers indicate the 95 th and 5 th percentiles, respectively. Significance was tested using two way ANCOVA. C = healthy control individuals. SDI = stimulant dependent individuals. .................. 22 3 1. Simplified illustration of the natural history of substance use disorders. Active substance dependence is characterized by sequential stages of binge/intoxication, withdrawal/negative affect, and craving/preoccupation (Koob and Volkow, 2010)th at repeat a variable number of cycles, represented by m . Active disease episodes may be interrupted by periods of short or long term abstinence, represented by n . ................................ ................................ ................................ .................. 31 3 2. Diagram of the fMRI BOLD signal pre processing, network of interest definition , and effective connectivity analysis pipelines. ................................ ................................ ................................ ................................ .... 36 3 3. Effective connectivity network density graphs of SDI and controls. Thickness of each line corresponds to the Granger causal strength and color corresponds to the efferent network. SDI network den sity was significantly greater than healthy controls (p<0.001). ................................ ......... 42 3 4. Directed GC matrices for SDI (left) and controls (right). Colorbar corresponds to logarithm of F values. ................................ ................................ ................................ ................................ ................................ ...................... 43 3 5. Effective connectivity matrix illustrating the FDR corrected group differences between SDI and healthy controls. Colorbar corresponds to p values. White arrows indicate Granger causal direction. ................................ ................................ ................................ ................................ ................................ ................. 43 3 6. Left, NOI dem onstrating significantly greater Granger causal relationships in SDI compared to controls. Illustrated are the NOI relationships that differed between groups; colors correspond to NOI labels on right. Printed numbers in upper left corner of each slice corr espond to Z axis coordinates in MNI space. Right, Granger causal relationships demonstrating greater directed information flow in SDI compared to controls (FDR corrected). RECN, right executive control

PAGE 14

xiv network; dDMN, dorsal default mode network; BGN, basal ganglia network. Values indicate Granger causal p values. Middle, whole brain illustrations of NOI identified. Arrows reflect Granger causal influence. ................................ ................................ ................................ ................................ ................ 44 3 7. Correlations between mean beta value within the RECN and dDMN with impulsivity, approach, and negative affect metrics. Solid black lines indicate the linear regression, solid colored lines indicate the 95% confidence interval, and colored shaded regions indicate the prediction in a given network, and their score on the given behavioral metric. ................................ ................................ ................................ ........................ 45 3 8. Global efficiency (left) and local efficiency (right) in SDI and controls as a function of network cost. ................................ ................................ ................................ ................................ ................................ ........................... 46 4 1. Connectivity base d signal flow diagram of anterior insular control of bottom up versus top down mechanisms of salience. The right dorsal anterior insula is involved in processing salience of externally oriented stimuli and it is correlated with the executive control networ k (an externally directed system). The right ventral anterior insula is involved in processing salience of internally oriented stimuli and it is correlated with the default mode network (an internally directed system). ................................ ................................ ................................ ................................ ................................ . 57 4 2. Diagrammatic illustrati on of the natural history of nicotine use disorder. We attempt to synthesize the neuroimaging literature of nicotine use disorder into four pharmacologically and behaviorally informed stages of nicotine use disorder. Acute nicotine exposure reflects the a cute pharmacologic action on neural circuitry. Chronic nicotine exposure reflects pharmacologic dependency and results from repeated acute nicotine exposure; it manifests as maladaptive changes in reward, salience, and executive control circuitry. Acute ab stinence provides a model for understanding the neural basis of the nicotine withdrawal syndrome and craving. Long term abstinence serves as a model of neuroplastic recovery from nicotine use disorder. Relapse (bottom arrows) is the mechanism by which acti ve disease is maintained. Associated neuroimaging findings are summarized in Table I. ................................ ................................ ............................. 60 4 3. Differential resting state connectivity of the dorsal (left) and ventral (right) right anterior insula, using the Human Connectome Project Connectome Workben ch (n = 1206), uncorrected. Dorsal

PAGE 15

xv right anterior insula is strongly connected with frontoparietal regions involved in executive control; in contrast, ventral right anterior insula connectivity is strongly connected with default mode network regions involv ed in internal feelings and self referential processing. Note that the externally directed system (dorsal anterior insula and executive control network) and internally directed system (ventral anterior insula and default mode network) are inversely correl ated, consistent with mutual inhibition (see Figure 4 1). ................................ ................................ .... 70 4 4. Finite element model results of the predicted electromagnetic field produced by a standard superficial planar transcranial magnetic stimulation 70mm figure of eight coil targeting the r ight anterior insula. Predicted current flux density (left) and normalized absolute value of the electric field (right) illustrating the pattern of energy deposition. Maximum energy is deposited in the scalp, superficial soft tissues, and cerebrospinal flu id due to high tissue conductivities. This illustrates the difficulty in targeting deep structures, such as the insula or anterior cingulate cortex. This is consistent with model results of insular targeting reported by Pollatos and colleagues (Pollatos an d Kammer, 2017). Created using SimNIBS version 2.1.1 with right anterior insular target [36, 10, 6] in MNI space and default parameters (Thielscher et al., 2015). ............... 75 5 1. CONSORT enrollment diagram for this phase 1 human trial. Careful attention was paid to ................ 84 5 2. rigorous sham or deep LF rTMS targeting the right anterior insula. The time delay between the end of treatment and behavioral survey was <5 minutes. The time delay between the end of treatment and post treatment MRI was 11.5 ± 1.9 minutes (mean ± SD). ................................ ............... 85 5 3. Self reported cigarette craving (QSU Brief) by grou p and time. LF rTMS causally reduced self reported craving compared to sham by per protocol two way repeated measures ANOVA of group × time (p = 0.033). No statistically significant difference was observed in craving after sham treatment, although there is a trend towards placebo effect (p = 0.084). ................................ .... 99 5 4. Pre treatment main effects of cigarette cue exposure compared to neutral cues, collapsed across groups. Brain activity significantly increased in the bilateral dorsolateral prefrontal cortex, primary visual cortex, and higher level visual cortex during cigarette cues compared to neutral

PAGE 16

xvi cues (red color bar). Brain activity significantly decreased in the bilateral posterior cingulate gyrus and precuneus during cigarette cues compared to neutral cues (b lue, p < 0.001). Results were corrected for multiple comparisons using familywise p < 0.05, voxelwise p < 0.005, cluster ................................ ................................ ................................ ................................ ..................... 100 5 5. Whole brain interaction effect using a per protocol 2x2 mixed factorial interactio n effect of time (within subjects: post treatment pre treatment) × group (between subjects: LF rTMS sham). Significantly decreased cigarette cue brain response after LF rTMS compared to sham was observed in the right primary sensorimotor cortex, bilate ral supplementary motor cortex (premotor), and right dorsal anterior insula (blue, p < 0.001). No significant increased cigarette cue brain response was observed after LF rTMS compared to sham. Results were corrected for multiple comparisons using familywi voxels. ................................ ................................ ................................ ................................ ................................ ................... 101 5 6. FEM informed ROI analysis of brain activity responses using a per protocol analysis of covariance. ROI beta values and significance levels were extracted using the MarsBar tool box after accounting for nuisance covariates, using the SPM12 design used for whole brain analysis. ................................ ................................ ................................ ................................ ................................ ................................ .. 102 5 7. Correlation between absolute change in QSU Brief self reported craving (post treatment > pre treatment) and change in brain activity responses to cigarette cues [(post treatment > pre treatment) × (cigarette cues > neutral cues)] within the LF rTMS group defined per protocol. Results were corrected for multiple comparisons using familywise p < 0.05, voxelwise p < 0.005, els. ................................ ................................ ................................ ................................ ..... 103

PAGE 17

1 CHAPTER I 1. INTRODUCTION Human use of psychoactive drugs is described in some of the earliest human historical records. P athologic use of ingested substances was described in early Greek and Roman literatur e; f or example, the Greek scientist Aristotle (384 BC 322 BC) provided the first detailed description of alcohol withdrawal, while the Roman physician Celsus ( 25 BC 50 AD ) precociously described alcohol dependence as a systemic disease (Crocq, 2007) . Historically, h uman use of psychoactive s ubstances can be categorized into four major categories: 1. S piritual or religious ceremonial use, such as imbibing the wine during Catholic mass; 2. M edi c al or therapeutic use, such as modern prescription opioids for pain control; 3. S ocially acceptable use for p leasure and/or socialization , such as cigarette smoking amongst construction workers; 4. Abuse or dependence of a substance, in a manner discordant with pro social norms and/or despite . This dissertation focuses on the four th category of use, addiction, which by the Diagnostic Statistical Manuel 5th edition (DSM 5) is categorized under S ubstance use disorder s (SUD) in the Diagnostic Statistical Manuel 5th edition (DSM 5) are a heterogeneous set of psychiatric diseases characterized by physical dependence, diminished control over substance use despite negative consequences, and craving s to use the substance (American Psychiatric Association, 2013) . SUD are characterized in severity by the number of sym ptoms, with 2 or 3 symptoms representing mild and greater than 6 symptoms representing severe SUD (American Psychiatric Association, 2013) .

PAGE 18

2 All drugs of abuse cause the release of dopamine into the nucleus accumbens (Di Chiara and Imperato, 1988; Koob, 1992) , producing a that positively reinforc es consumption ( Figure 1 1 ) . The mesocorticolimbic reward circuit consists of dopaminergic neurons with cell bodies in the ventral tegmental area of the midbrain and terminal projections in the ventral striatum (nucleus accumb ens ), prefrontal c ortex, and limbic /paralimbic regions. When this circuit is stimulated, d opamine is released into synaptic cleft s in these terminal regions. Dopamine binds to postsynaptic receptors and is subsequently rapidly re sequestered within the presynaptic axon al bo utons by the dopamine transporter (DAT), terminating the reward signal. Cocaine (studied in Chapter s 2 and 3), for example, artificially enhances dopamine signaling by inhibiting the dopamine transporter , thus prolonging dopaminergic activity. Nicotine (studied in Chapters 4 and 5 and Appendix 1 ) acts directly on dopaminergic Figure 1 1 . Simplified scheme illustrating how common drugs of abuse modulate the reward circuit. Modified with permission (Nestler et al., 2015) .

PAGE 19

3 reward neurons in the VTA as well as modulating interneurons. Over time, drug use and drug related stimuli become increasingly associated with the hedonic effects of drug induced dopamine transmission , and th e se stimuli serve as incentivized cues that trigger craving, anticipatory euphoria, or even symptoms of withdrawal (O'Brien et al., 1992) . Drug dependence causes neurocircuitry adaptations that develop i n response to repeated administration, as the after the development of addiction, in turn, unmasks withdrawal syndromes because the forces that keep these neuroc ircuitry alterations balanced are no longer present. 1.1. Neuroimaging in Substance Use Disorders Neuroimaging has fostered a paradigm shift in the understanding of SUD over the past two decades uroimaging has demonstrated that SUD are brain diseases with predictable alterations in neurocircuitry underlying reward, craving, learning, and cognitive control (Tanabe et al., 2019) . Neuroimaging provides both structural and functional observations of altered neuroanatom y. For example, quantitative structural neuroimaging using voxel based morphology (see Chapter 2) has demonstrated lower grey matter tissue volumes compared to healthy adults in the anterior cingulate, dorsolateral prefrontal, medial prefrontal, and parietal cortex in people dependent upon cocaine (Connolly et al., 2013; Regner et al., 2015) , amphetamine (Daumann et al., 2011; Mackey and Paulus, 2013; Tanabe et al., 2009b) , and nicotine (Brody et al., 2004; Pan et al., 2013) . In addition to brain structural changes , neuroimaging has furthered our understanding of the function al mechanisms in the brain that promote and maintain SUD. For instance , intravenous ly administered nicotine (Stein et al., 1998) and cocaine ( Breiter et al., 1997) both acutely increase blood oxygen level dependent (BOLD) fMRI signal in the striatum, amygdala , and prefrontal cort ex. This pattern of reward circuitry activation after acute drug

PAGE 20

4 administration has been shown to be altered even in the absence of drug administration. A large meta analysis investigated the brain changes associated with monetary reward anticipation and r eward outcome in 643 individuals with addictive behaviors (including all common drugs of abuse, gambling disorder, and gaming disorder) compared to 609 healthy control individuals (Luijten et al., 2017) . They observed that during reward anticipation, people with substance, gambling, and gaming addictions exhibited ventral striatum hyperactivation. During reward outcome, they exhibited decreased dorsal striatum hypoactivati on. This suggests that brain processing of reward signals is perturbed in addiction, even for rewards unrelated to the addiction. Neuroimaging has also demonstrated significant long term shifts in the cognitive function in SUD in the absence of drug administration . A parametric m eta analysis of fMRI cue reactivity studies discovered that nicotine, alcohol, and cocaine individuals have abnormally increased brain responsivity to drug cues compared to neutral cues in the ventral s triatum, anterior cingulate cortex, and amygdala regions involved in reward and emotional regulation (Kühn and Gallinat, 2011) . Although there are compelling common mechanisms of SUD across different substances and across individua ls, significant heterogeneity exists. 1.2. Substance Use Disorders as a Class of Heterogeneous Neurocircuitry Diseases Substance use disorders are understood to be chronic relapsing remitting diseases that vary by drug, stage, and individual factors. First, SUD demonstrate brain and behavioral difference s by drug and mode of use. For example, Hanlon et al. studied individual s dependent upon cocaine (n=55 ) , alcohol (n=53), and nicotine (n=48) (Hanlon et al., 2018) , and reported that each substance use disorder demonstrated overlapping yet ana tomically cue induced craving fMRI brain response maps. This suggests that neuromodulatory efforts to therapeutically alter craving in SUD may need to be drug specific. Second, there is significant evidence that each stage of addiction ( Figure 1 2 ) is associated with distinct patterns of brain functional changes observed through neuroimaging compared to healthy controls (Koob and

PAGE 21

5 Volkow, 2010) . Third, and lastly, there are numerous individu al factors that contribute to and modulate the pathology of SUD. These include genetic predisposition s , socio demographics, problems of inhibitory control, co morbid diagnoses (e.g., conduct disorder), sibling use of substances, family history of addiction (i.e., upbringing independent of genetic inheritance) , neighborhood poverty and disorganization, and even locoregional laws and norms (American Psychiatric Association, 2013; Beyers et al., 2004; Conger, 1997; Hawkins et al., 1992; Kendler et al., 2013; Samhsa, 2018; Scaramella and Keyes, 2001; Stone et al., 2012; von Sydow et al., 2002) . Figure 1 2 . Simplified illustration of the natural history of substance use disorders. Active substance dependence (red) is characterized by sequential stages of binge/intoxication, withdrawal/negative affect , and craving/preoccupation (Koob and Volkow, 2010) that repeat a varia ble number of cycles, represented by m . Active disease episodes may be interrupted by periods of short or long term abstinence, represented by n . Relapse is indicated in blue. Note that at every stage of potential or active disease (top, in white) there i s a possible avenue towards disease remission (bottom, in grey). Used with permission from Regner et al. (2016) .

PAGE 22

6 1.3. Dissertation Narrative Outline The overarching technical theme heavily influencing each chapter is the signal processing technique for structural and function brain MRI, including statistical parametric mapping to make hypothesis based inferences. For each analysis, T1 weighted images a re segmented into tissue probability maps , including maps of grey matter, white matter, cerebrospinal fluid, bone, soft tissue, and air. These are non linearly normalized into a standard space . Non linear w vector fi elds across the field of view ) are computed from native to standard space. Time varying functional images are slice tim e and motion corrected, then forward warped into standard space. Ultimately , generalized linear models are used to compute individual an d group level contrast maps (T statistic or F statistic maps) for task based fMRI, or alternatively connectivity maps for resting state fMRI (either through spatiotemporal independent component analysis or seed based whole brain correlations). The goal of this dissertation w as to use advanced MRI based neuroimaging and signal processing methods to address gaps in the literature regarding two different populations of SUD at different points along the spectrum of disease severity: long term abstinent, severe ly dependent cocaine and methamphetamine addicts (Chapters 2 and 3); and, acutely abstinent, moderately dependent cigarette smokers (Chapter 4 and 5 and Appendix 1 ). While these two populations differ in severity, they also differ in primary drug of abuse (psychostimulants versus nicotine ) and stage of disease (chronic remission versus acute withdrawal). Chapters 2 and 3 investigate structural and functional neuroimaging differences, respectively, in long term abstinent psychostimulant ( cocaine and methamphetamine ) addicts compared to healthy controls. In Chapter 2, we report one of the large st sample s (n=127) demonstrating that the differences in grey matter volumes in stimulant dependence compared to healthy controls after long term abstinence differ by sex. Although limited by the cross -

PAGE 23

7 sectional nature of the study, we discuss the important implications with regards to sex differe nces in the natural history of psychostimulant dependence. In Chapter 3, we investigated resting state functional connectivity changes in stimulant dependence compared to healthy controls in a subset of the population reported in Chapter 2 (n=100). We sho w that large scale brain networks defined by independent component analysis follow a pattern of increased top down cognitive control in long term abstinent psychostimulant addicts compared to healthy controls , which may reflect increased executive control over habit and reward systems promoting remission . Stimulant dependence also demonstrated greater global efficiency and lower local efficiency amongst large scale brain networks, suggesting abnormal brain organization despite long term abstinence. Chapter s 4 and 5 and Appendix 1 report functional neuroimaging changes in acutely abstinent smokers after a phase 1, sham controlled, single blinded, randomized clinical trial ( www.ClinicalTrials.gov identifier: NCT 0 2590640 ) designed to test the hypothesis that inhibit ing the right anterior insula will decrease cigarette cravings and alter brain activity and connectivity. Chapter 4 serves i s a review introduc ing and motivating our trial of inhibitory TMS in acutely ab stinent, moderate smokers . We specifically review compelling neuroimaging and neuroscience evidence that the insula may serve as an important therapeutic target for promoting smoking cessation , and we discuss possible mechanisms to pursue such neuromodulat ion . We introduce a useful method of non invasively modulating brain activity: transcranial magnetic stimulation (TMS). Chapter 5 reports the main results of the clinical trial : inhibiting the right anterior insula in smokers using low frequency deep TMS results in decreased cigarette craving and decreased brain fMRI responses to cigarette cues compared to sham . Appendix 1 reports some but not all exploratory data collected during the trial, including changes in resting state c onnectivity after treatment compared to sham.

PAGE 24

8 Each individual dissertation chapter address es a specific yet important gap in the literature on the structural and functional computational neuroanatomy of SUD . Together, however, these neuroimaging studies at differ ent points along the spectrum of substance use disorders provide support for a conclusion still controversial in the scientific literature (Lewis, 2018) : addiction is a brain disease. We further extend this conclusion by investigating possible endogenous (i.e., increased top down control, Chapter 3) and exogenous (i.e., therapeutically imposed, C hapter 5) mechanisms of disease remission and treatment. While these results span different subtypes of substance use disorders, they provide hope that both endogenous and exogenous treatment strategies are possible to help reduce the burden of addiction.

PAGE 25

9 CHAPTER II 2. SEX DIFFERENCES IN G REY MATTER CHANGES AND BRAIN B EHAVIOR RELATIONSHIP S IN STIMULANT DEPENDENCE 2.1. Abstract This first chapter investigated whether sex modulates the effects of stimulant dependence on grey matter volumes (GMV) in long term abstinence. We further sought to characterize how sex modulates brain behavior relationships between GMV and specific behavioral measures, su ch as drug symptom count, behavioral approach, and impulsivity. In this prospective case control study, 127 age and sex matched participants (68 controls [28F/40M] and 59 SDI [28F/31M]) underwent T1 weighted SPGR IR brain MRIs on a 3T system. Images wer e segmented using voxel based morphometry MATLAB based software. After adjusting for age, education, and head size, sex by group interactions and main effects were analyzed over the whole brain using ANCOVA, thresholded at p <0.05, corrected for multiple c omparisons with family wise cluster correction. Drug symptom count and behavioral measurements were correlated with whole brain GMV and five a priori regions of interest based on extant literature. Sex by group interactions on GMV were significant in num erous regions ( p <0.001). Compared to female controls, female SDI had significantly less GMV in widespread brain regions ( p <0.001). There were no significant GMV differences in male controls versus male SDI ( p =0.625). Drug symptom count negatively correl ated with nucleus accumbens GMV in women (left: r p =0.047; right: r p =0.031) but not men (left: r p =0.737; right: r p =0.349). Behavioral approach (p=0.002) and impulsivity ( p =0.013) correlated negatively with frontal and te mporal GMV changes in female SDI but not in other groups, demonstrating a sex by group interaction. Vast GMV changes in SDI were

PAGE 26

10 observed in women but not men after prolonged abstinence. Sexual dimorphism in drug related neuroanatomical changes and brain behavior relationships may be a mechanism underlying the different clinical profiles of addiction in women compared to men. Future structural neuroimaging and clinical studies on substance use disorders should account for the modulatory effects of sex. Th is chapter has been peer reviewed and published: Regner MF, Dalwani M, Sakai JT, Yamamoto D, Perry RI, Honce J, Tanabe J. Sex modulates grey matter changes and brain behavior relationships in substance dependence. Radiology . 2015; 277(3):801 812. PMID: 26 133201 . 2.2. Introduction Substance use disorders are common, with lifetime prevalence estimated to be 10.3% in the United States (Miller and Hendrie, 2009) . Understanding the neurobiology of substance dependence is requisite to advancing treatment s. Neuroanatomical changes in drug addiction have been studied extensively using voxel based morphometry (Ersche et al., 2013) . Structural changes have been observed in the orbitofrontal cor tex (OFC), medial frontal gyrus (MedFG), anterior cingulate gyrus (ACG), insula, and nucleus accumbens in i ndividuals who abuse stimulants (Ersche et al., 2013; Koob and Volkow, 2010) . In the largest meta analysis of stimulant dependence to date, Ersche et al. (Ersche et al., 2013) reported significant decreases in gray matter (GM) in the insula, ventromedial prefrontal cortex, inferior frontal gyrus, ACG, and anterior thalamus. GM changes have also been studied adult sibling pairs, of whom one sibling was dependent on stimulants and the other had no stimulant dependence history with an age and sex matched control group (Ersche et al., 2012) . That study revealed changes in limbic and sensory areas in both me mbers of the sibling pair compared to controls, suggesting that GM volume changes may predate addiction and could potentially be an endophenotype for substance use disorder.

PAGE 27

11 Few previous studies have investigated the role of sex on changes in brain structure in stimulant dependence. This is surprising considering the well characterized sex differences in clinical presentation and natural history of stimulant addiction (Becker et al., 2012; Perry et al., 2013) . Women exhibit a telescoping clinical course compared to men in that they begin cocaine or amphetamine use at earlier ages (Becker et al., 2012; Griffin e t al., 1989; Mendelson et al., 1991) , show accelerated escalation of drug use (Brady and Randall, 1999; Greenfield et al., 2010; Lynch, 2006) , report more difficulty quitting (Back et al., 2005; Lynch, 2006) , and upon seeking treatment report using larger quantities of these drugs compared to men (Becker et al., 2012; Kosten et al ., 1993) . Neuroendocrine factors have been hypothesized to underlie an accelerated clinical cou rse (Becker et al., 2012) . Another hypothesis is that compared to men, women respond differently to stress which i nfluences drug related behavior (Potenza et al., 2012) . However, scant evidence exists for a neuroanatomical correlate of these clinical differences. Many studies recruit primarily men to exclude confounding sex effects (Barros Loscertales et al., 2011; Fein et al., 2002; Franklin et al., 2002) , while other studies do not include sex as a factor in their analyses of GM in stimulant dependence (Ersche et al., 2013; Sim et al., 2007; Tanabe et al., 2009a) . In fact, only two studies have described structural differences betwee n sexes in stimulant dependence (Rando et al., 2013; Tanabe et al., 2013) . Rando et al. (Rando et al., 2013) reported lower GMV in the left inferior frontal gyrus, insula, superior temporal gyrus, and hippocampus in female SDI compared to female controls while male SDI exhibited less GMV in the precentral gyrus and mid cingulate gyrus compared to male controls. However, this study was significantly limited by the potential effects of recent alcohol use (mean 87 drinks in the month prior to scanning) and lack of long term abstinence (mean 3 weeks of abstinence prior to scanning), allowing acute effect of substances to skew resu lts. Tanabe et al. (Tanabe et al., 2013) reported differential effects of sex on insular volumes in SDI: female SDI had smaller insulae, whereas male SDI had larger insulae. This study was limited by

PAGE 28

12 the small sample size (28 SDI) and r udimentary methodology (using FreeSurfer to estimate GMV). The paucity of large, prospective, well controlled studies to investigate long term sex differences associated with abstinent stimulant dependence is addressed by this study. This study investigated whether sex modulates the effects of stimulant dependenc e on grey matter volumes (GMV) in long term abstinence. We further sought to characterize how sex modulates brain behavior relationships between GMV and specific behavioral measures, such as drug symptom count, behavioral approach, and impulsivity. 2.3. Materi als and Methods 2.3.1. Subjects One hundred twenty seven individuals including 68 controls (28F/40M) and 59 (28F/31M) SDI were prospectively recruited ( Figure 2 1 ). Controls were similar to SDI on age and sex. Demographic information is reported in Table 2 1 . SDI were recruited from a residential treatment program at the University of Colorado School of Medicine Addiction Research and Treatment Services. Inclusion criteria for SDI: lifetime dependence on stimulants Figure 2 1 . Sample population inclusion and exclusion criteria.

PAGE 29

13 (methamphetamine, cocaine, or amphetamine class substances) (DSM IV). Control subjects were recruited from the community and excluded if dependent on alcohol or other drugs of abuse excluding tobacco. Exclusion criteria for all subjects: depression within the last two months, psychosis, neurological illness, prior head trauma resulting in greater than fifteen minutes loss of consciousness, prior neurosurgery, positive HIV status, diabetes, hepatitis C, bipolar disorder, other major medical illness, inabilit y to tolerate MRI, IQ < 80, positive urine screen (AccuTest TM ), or positive saliva screen (AlcoScreen TM ). All participants provided written informed consent approved by the Colorado Multiple Institutional Review Board. Table 2 1 . Demographic description of the sample population. Data are presented as mean ± SD where appropriate. Healthy Controls SDI p value Men Women Men Women Group Sex G x S Sample Size * 40 28 31 28 NS NS NS Age (yr) 33.2 ± 9.9 31.9 ± 7.6 36.7 ± 7.6 33.7 ± 7.9 NS NS NS Right handedness * 92.5% 89.3% 96.7% 89.3% NS NS NS Education (yr) 14.0 ± 1.6 15.4 ± 1.1 12.6 ± 1.5 12.2 ± 1.7 <0.001 NS 0.001 Years of Use (yr) 16.1 ± 7.0 15.2 ± 8.0 NS Abstinence (mo) 17.5 ± 15.6 11.2 ± 7.9 NS Drug symptom count 33.6 ± 14.3 25.6 ± 10.2 0.015 Negative affect 16.2 ± 6.7 14.3 ± 3.5 20.8 ± 9.0 22.1 ± 7.1 <0.001 NS NS Behavioral approach 39.3 ± 4.4 35.6 ± 8.3 42.1 ± 5.0 44.3 ± 4.7 <0.001 NS 0.005 Impulsivity 59.1 ± 8.7 55.2 ± 7.1 69.6 ± 10.7 73.7 ± 11.3 <0.001 NS 0.022 Percent Satisfying Dependence Criteria * Stimulants 100% 100% NS Cocaine 67% 41% NS Amphetamine 83% 89% NS Nicotine 73% 74% NS Alcohol 67% 67% NS Cannabis 50% 37% NS Opiates 37% 19% NS Club drugs 10% 7% NS Sedatives 10% 0% NS Hallucinogens 10% 4% NS

PAGE 30

14 2.3.2. Structured Interviews for Inclusion s and Exclusions Composite International Diagnostic Interview Substance Abuse Module (CIDI SAM) is a computerized structured interview that assesses substance dependence diagnoses and symptoms for 11 different drugs of abuse (Cottler et al., 1989) . All subjects were administered the CIDI SAM to verify sti mulant dependence in the SDI group and to exclude controls with abuse or dependence on substances other than tobacco. Diagnostic Interview Schedule version IV (CDIS IV) is a computerized structured interview used to screen for psychiatric disorders (Robins et al., 1995a) . All subjects were administered the CDIS IV to exclude those with lifetime psychoses, lifetime bipolar disorder, or major depressive disorder in the last two months. The CIDI SAM interview (Gelhorn et al., 2008) computes four abuse (i.e., legal problems due to drug) and s even dependence (i.e., uncontrolled substance use escalation) symptoms for each drug class. Drug use severity was calculated by adding abuse and dependence symptom counts. This approach is consistent with the single set of clinically relevant criteria in DSM V, which was released after data collecti on for the current study. The Barratt Impulsiveness Scale is a 30 item self reported questionnaire used to quantify impulsiveness (Patton and Stanford, 1995) . Participants rate whet her phrases and words describing aspects of impulsivity are self descriptive. The Behavioral Activation System (BAS) scale is a 13 item self reported questionnaire used to measure responsiveness of motivational systems (Campbell Sills et al., 2004; Carver and White, 1994) ; this quantifies positive affective and approach response tendencies to appetitive stimuli. 2.3.3. MRI Exami nation Brain MRI was performed using a 3T MR scanner (General Electric, Milwaukee, Wisconsin) and standard quadrature head coil. High resolution T1 weighted SPGR IR sequences were acquired for each subject using the following parameters: TR=45ms, TE=20ms,

PAGE 31

15 flip angle=45°, 256x256 matrix, 240x240mm 2 field of view (0.9x0.9mm 2 in plane), 1.7mm slice thickness, and coronal plane acquisition. All images were evaluated by a board certified neuroradiologist for structural abnormalities. No examinations were exclu ded on this basis. 2.3.4. Image Processing T1 weighted brain MR images were processed using the VBM8 toolbox (http://dbm.neuro.uni jena.de/vbm8/) and SPM8 (http://www.fil.ion.ucl.ac.uk/spm/) software. Images were segmented into GM, WM, and CSF probability maps a natomically co registered using Diffeomorphic Anatomical Registration Through Exponentiated Lie Algebra (DARTEL) (Ashburner, 2007) . Custom DARTEL templates were created for our sample population and u sed to register the images. Segmented images were non linearly modulated after registration to preserve relative regional volume, after correcting for different brain sizes. Segmented GM probability maps of each subject were visually inspected for qualit y control by a radiologist ; no images were excluded. Normalized, modulated images were smoothed with an 8mm 3 full width at half maximum Gaussian kernel. 2.3.5. Whole brain analyses Morphometric analysis Whole brain analyses on grey matter volumes (GMV) were perf ormed using a two way ANCOVA testing for sex by group interactions and main effects of group and sex. All analyses were adjusted for age, education, and head size measured by total intracranial volume. Age did not differ between groups but is an important covariate as it directly affects global GMV (Good et al., 2001) . Education differed by group ( Table 2 1 ). Significance levels were set at p <0.05 corrected for multiple comparisons with family wise error (FWE) using AlphaSim Monte Carlo simulations (10,000 simulations) and voxel wise threshold p< 0 .005. Cluster threshold corresponded to 1202 voxels (each voxel 3.375 mm 3 ) or 4056.8 mm 3 . Whole brain analysis

PAGE 32

16 interpretation was restricted to the supratentorial space given reported methodological difficulties of infratentorial space segmentation (Diedrichsen et al., 2009) . Behavioral morphometric analysis Within SDI, whole brain regression analyses examined GMV association with drug use severity. In exploratory analyses, behavior by sex by group interactions were regressed against whole morphometric methods). Significance was determined using the aforementioned clus ter based FWE of p <0.05 as well as threshold free cluster enhancement (TFCE) with correction for multiple comparisons using FWE of p<0.05 (Smith and Nichols, 2009) . TFCE was used to evaluate small structures less than 4.05cm 3 such as the nucleus accumbens, which would oth erwise be mathematically precl uded from reaching significance (Radua et al., 2014; Smith and Nichols, 2009) . 2.3.6. Regions of Interest (ROI) W hile whole brain analyses offers statistical robustness, cross validation with pre defined ROIs using prior knowledge improves classification performance (Chu et al., 2012; Kerr et al., 2014; Nieto Castanon et al., 2003) . Whole brain voxel level analysis is data driven without regard to specific anatomy, while pre defined ROI analysis is hypothesis driven for specific neuroanatomical structures based on prior knowledg e. To confirm results from whole brain analyses, five a priori neuroanatomical structures were hypothesized to differ in SDI compared to controls based on their involvement in reward, learning, executive control, and affective proce ssing, which are altere d in SDI (Ersche et al., 2013; Koob and Volkow, 2010; Li and Sinha, 2008) : OFC, MedFG, ACG, insula, and nucleus accumbe ns. Masks for these structures were created using the Automated Anatomic Labeling (AAL) atlas toolbox (Tzourio Mazoyer et al., 2002) . Total GMV of each structure was calculated by summing voxels in the ROI masked

PAGE 33

17 structural GM map multiplied voxel wise by the voxel modulation. GMV in each ROI was analyzed using a two way A NCOVA on group, sex, and sex by group interactions, after adjusting for age, education, and head size. To correct for multiple comparisons, results were considered significant at FWE p<0.05 with Bonferroni correction for five ROIs (pairwise comparison p <0. 01). In exploratory analyses, GMV in ROI were regressed against drug symptom count. Behavior by sex by group interactions were regressed against ROIs and considered significant with a Bonferroni correction for FWE across all structures ( p< 0.05; pairwise comparison p <0.01). 2.4. Results 2.4.1. Demographics, drug severity, and behavioral comparisons SDI and controls were similar in age and sex ( Table 2 1 ). There was no sex by group interaction on age. There was a sex by group interaction on education ( F 1,127 =10.936, p <0.001) and main effects of group ( F 1,127 =74.914, p <0.001). Female SDI had f ewest years followed by male SDI, male controls, and female controls with the most years of education. SDI had mean 2.2 fewer years of education than control subjects. Within SDI, there were sex differences in drug use severity ( p =0.015) with men having greater drug symptom count. There were no sex differences in drug exposure, abstinence duration, or years of drug abuse. There were significant sex by group interactions in behavioral approach and impulsivity: female SDI had highest approach and impulsivi ty, followed by male SDI, male controls, and female controls with the least approach and impulsivity.

PAGE 34

18 2.4.2. Whole Brain Analysis Sex by group interaction A significant sex by group interaction on GMV was found in widespread areas of cerebra l cortex, thalamus, and basal ganglia. To further characterize the interaction, the effect of group was investigated in each sex separately. No GMV differences were observed between control and SDI men. In contrast, large, widespread differences in GMV w ere observed between control women and SDI women ( Figure 2 2 ). Compared to control women, SDI women had Figure 2 2 . T value map illustrating statistically significant greater grey matter volumes in healthy control women compared to stimulant dependent women, after controlling for age and brain size ( p <0.001) .

PAGE 35

19 significantly less GMV in frontal lobe (OFC, MedFG, superior frontal gyrus), limbic regions (insula, amygdala, cingulate gyrus), temporal lobe (tempo ral pole, uncus, parahippocampal gyrus, hippocampus, occipitotemporal gyri, superior temporal gyrus, middle temporal gyrus), a nd inferior parietal lobule. Greater GMV observed in control women compared to SDI women showed anatomical congruence to the sex by group interaction effect. Main effect of sex A significant main effect of sex was found ( Figure 2 3 ) , with women exhibiting comparatively greater regional GMV than men widely thro ughout cerebral cortex, thalamus, Figure 2 3 . T value map illustrating statistically significant greater regional grey matter volumes in control women compared to control men, after controlli ng for age and brain size ( p <0.001).

PAGE 36

20 and basal ganglia ( p <0.001). Subgroup analysis demonstrated anatomically similar significant differences between healthy control men and women. Main effect of group A significant main effect of group was found throughout frontal, temporal, insular, and parietal regions ( p <0.001). Controls had significantly greater GMV than SDI. 2.4.3. Whole Brain Correlations with Drug Use Characteristics Significant negative correlations were found between drug use severity and bilateral nucleus accumbens GMV ( Figure 2 4 ; left, 290 voxels, MNI coordinates [ 9, 4, 2], p(FWE corr) <0.0001; right, 271 voxels, MNI coordinates [8, 10, 6], p(FWE corr) <0.0001). Years of substance use did not correlate with any GMV. Abstinence posi tively correlated with a small area of left superior frontal gyrus (171 voxels, MNI coordinates [ 18, 38, 33], p(FWE Figure 2 4 . Left , T value map of statistically significant negative correlations found on whole brain level between stimulant dependence symptom count and grey matter volume within the bilateral nucleus accumbens (the so in SDI. Right , negative correlation be tween the substance dependence symptom count and nucleus accumbens total volume as defined by ROI.

PAGE 37

21 c orr) =0.006). Neither years of substance use, abstinence, or drug use severity were found to have significant sex interactions on GMV. 2.4.4. Ex ploratory Whole Brain Group by Sex by Behavior Interactions Significant three way interactions between behavioral approach, group, and sex were seen in the bilateral middle frontal gyri ( Figure 2 5 , green; Table 2 2 ) . In these areas, in female SDI behavioral approach correlated negatively with GMV, where as female controls and both groups of men had positive correlation coefficients. Significant three way interactions between impulsivity, group, and sex were seen in bilateral superior and middle temporal gyri, right insula, right superior temporal sulcus, and right inferior temporal gyrus ( Figure 2 5 , red ; Table 2 2 ). In these areas in female SDI, impulsivity correlated negatively with GMV, whereas female controls and both groups of men had positive correlation coefficients. Table 2 2 . Significant sex by group interactions on correlations between behavior and GMV. Figure 2 5 . Sex by group by behavior interactions on GMV. Left , scatterplots illustrating the sex by group interaction on the correla tion between approach and GMV in the bilateral middle frontal gyri (dorsolateral prefrontal cortex). Right , scatterplots illustrating the sex by group interaction on the correlation between impulsivity and GMV in the left superior temporal gyrus and left insula. Middle , clusters of whole brain significance demonstrating sex by group by impulsivity (red) and sex by group by approach (green) interactions.

PAGE 38

22 x y z k p value (FWE cor) Structure Behavioral activation x Group x Sex, negative correlation 39 26 45 371 <0.001 Left middle frontal gyrus 46 18 46 48 12 40 353 <0.001 Right middle frontal gyrus 42 20 34 Impulsivity x Group x Sex, negative correlation 56 39 6 4475 <0.001 Right superior temporal gyrus, right insula 57 6 20 51 5 12 68 24 9 312 0.001 Left superior temporal gyrus 27 84 9 1112 0.005 Right superior temporal gyrus 39 57 14 32 72 12 46 42 7 176 0.006 Left middle temporal gyrus 15 91 8 79 0.016 Left inferior frontal gyrus 9 81 2 6 96 5 2.4.5. ROI Analysis Two way ANCOVA revealed statistically significant sex by group interactions and main effects of group for all structures except nucleus accumbens ( Figure 2 6 ). Post hoc pairwise Figure 2 6 . Boxplots illustrating the total grey matter volume in each subpopulation for each region of interest. The top and bottom edges of the boxes indicate the third and first interquartiles, respectively. Lines in the middle of the boxes indicate the medians. Whiskers above and below the boxes indicate the 90 th and 1 0 th percentiles, respectively. Points above and below the whiskers indicate the 95 th and 5 th percentiles, respectively. Significance was tested using two way ANCOVA. C = healthy control individuals. SDI = stimulant dependent individuals.

PAGE 39

23 compar isons revealed greater GMV in control women compared to SDI women in total volumes of each significant structure ( p <0.001) but not between control men and SDI men, consistent with whole brain results. 2.4.6. ROI Correlations with Behavior Drug use severity correlated negatively with nucleus accumbens GMV ( Figure 2 4 , right). This correlation was driven by SDI women; a steeper and significant negative correlation wa s seen in women (left: r p =0.047; right: r p =0.031) compared to men (left: r p =0.737; right: r p =0.349) in which the correlations were not significant. No other correlations between drug characteristics or behavioral metric s were significant. No behavior by sex by group or behavior by group interactions were significant in ROI structures. 2.5. Discussion The current finding of significantly lower GMV in abstinent stimulant dependent women compared to healthy control women is str iking for two reasons : (1) no group differences were observed in men, and (2) the involved regions are anatomically vast and overlap substantially with pathways implicated in reward, learning, executive control, and affective processing (Koob and Volkow, 2010) . The widespread anatomical extent to which men and women differ in relation to abstinent substance dependence has not been reported. These differences in SDI women compared to men could reflect a greater neuroanatomical endophenotype that predisposes them to stimulant dependence or a vulnerability to morphologic changes that result from stimulant dependence more so in women than men. Decreased GMV in female SDI compared to female controls was most striking in limbic regions, particularly the insula, further suggesting a functional role of these structures in mediating t he clinical phenotype.

PAGE 40

24 We expanded upon these structural results with brain behavioral correlations. Nucleus accumbens volume was negatively correlated with drug use severity, consistent with its role in reward and salience. Previous studies have shown t hat ventral striatal activity, including nucleus accumbens, correlates with th e intensity of received rewards (Sescousse et al., 2013) . Persons abusing and dependent upon stimulants undergo pathologic overstimulation of this nucleus and may exhibit compen satory down regulation of neuronal synapses with reduced dendritic branching, number of axonal boutons, and degree of axon mye lination leading to reduced GMV (Draganski and Kherif, 2013; Fields, 2013) . The negative relationship observed in this study between drug use severity and nucleus accumbens volume was significant in women but not in men, despite men exhibiting greater drug use severity. This suggests women may demonstrate greater susceptibility to drug use severity changes, possibly through neuroendocrine mechanisms which will be discussed below. Behavioral approach and impulsivity both interacted with group and sex to signific antly correlate with GMV ( Figure 2 4 , Table 2 2 ). Higher behavioral approach and imp ulsivity were associated with lower GMV in female SDI. Behavioral approach characterizes level of arousal and response to cues toward favorable outcomes and positive affective states; higher approach motivates behavior. Impulsivity describes decreased in hibitory control over potential actions leading to reward. Previous studies have reported significant sex differences between approach and imp ulsivity characteristics in SDI (Perry et al., 2013) ; however, this is the first study to report structural neuroanatomical correlates of these findings. Higher approach in female SDI was correlated with lower GMV in the bilateral DLPFC and may reflect a deficit in top down control over approach behaviors toward drug cues. The current structural and brain behavioral relationship differences by s ex may result from neuroendocrine factors. For example, sex and ovarian hormones affect the number, density, and firing rate of dopaminergic neurons, with women

PAGE 41

25 showing enhanced dopaminergic system engagement during initial drug exposure and exacerbated n egative affective state during drug withdrawal (Becker et al., 2012) . Few studies have investigated sexual dimorphism GMV in stimulant dependence; in fact, only two studies report structural sex differences in SDI: Rando et al. (Rando et al., 2013) and Tanabe et al. (Tanabe et al., 2013) . Consistent with our findings, Rando et al. (Rando et al., 2013) observed greater GMV in healthy compared to cocaine dependent women in the left inferior frontal gyrus, left insula, left superior temporal gyrus, right temporo occipital cortex, and left hippocampus. Tanabe et al. (Tanabe et al., 2013) observed a differential effect of sex on small regions of the insula. Consistent with our results, they reported that SDI women exhibited smaller insulae compared to controls. However, we report that the differences span nearly the entire insulae bilaterally. Both of these studies had modest patient sample sizes, 36 in Rando and 28 in Tanabe. A meta analysis performed by Ersche et al. (Ersche et al., 2013) studied 494 stimulant dependent subjects (79% men) and 428 healthy control subjects (69% men) and reported smaller GMV in SDI compared to controls in the insulae, inferior frontal gyrus, ACG, and anterior thalamus; however, this study did not comment on any sex effects. Other studies of drug effects on brain morph ometry exclude women altogether (Barros Loscertales et al., 2011; Fein et al., 2002; Franklin et al., 2002) , pointing to the need for prospective studies to investigate effects of sex. Here we report significant neuroanatomical sexual dimorphism in the largest prospective sex by group sample of long term abstinent stimulant dependence to date. The lack of group differences in men was unexpected. Unlike our study, Rando et al. found small differences in men, with lower GMV in a small portion of the precentral and mid cingulate gyrus in cocaine dependent compared to healthy men. There are several possible explanations for this difference, such as recent large alcohol intake (mean 87 drinks in prior month), short le ngth of abstinence (mean 3 weeks), and significantly older SDI than controls in

PAGE 42

26 the Rando population. Our sample had much longer abstinence, mean 13.5 months. It has been (Connolly et al., 2013) . For example, in the Ersche et al. (Ersche et al., 2013) meta analysis of 494 stimulant dependent subjects, only four of the 13 studies included subjects abstinent for more than one month, with the majority of studies investigating active users. Thus, most studies examined acute drug ef fects. Because our study included subjects abstinent for at least 60 days, there may have been (Connolly et al., 2013) who found that in men, GMV positively cor related with early abstinence but tapered at 35 weeks to become equivalent to those of drug naïve controls. Given the average 13.5 months abstinence in our study, GMV recovery may have already reached a steady state in men by the time of recruitment. One limitation of this study is the polysubstance use characteristics of the SDI population. While this precludes us from relating structural changes to a single drug, our sample has biological and ecological validity as it reflects an important, real world, clinical population of SDI. Epidemiological data demonstrate that stimulant dependence does not often occur in isolation; instead most stimulant dependence individuals meet dependenc e criteria for other substances (Sara et al., 2012; Stinson et al., 2005) . Importantly, the G MV differences observed here were not due to differences in drug exposure or symptom severity. Another limitation is that our sample was referred from the justice system and we cannot exclude the possibility that antisocial personality traits contributed to the findings. Third, SDI and controls differed in years of education. Although we statistically covaried for this confounding variable in all analyses it is possible that education could influence the observed differences. 2.6. Conclusion Vast neuroanatomic al changes observed in abstinent SDI were present in women but not in men. In particular, structures involved in reward, learning, executive control, and

PAGE 43

27 affective processing pathways were affected: insula, OFC, ACC, MedFG, and nucleus accumbens. These c hanges correlate with drug use and behavioral measures and may help to explain differences in the clinical course of stimulant dependence in women compared to men. 2.7. Acknowledgements Several co authors contributed to this chapter: Manish Dalwani MS, Dorot hy Yamamoto PhD, Robert Perry MD, Joseph Sakai MD, Justin Honce MD, and Jody Tanabe MD. This study was supported by the National Institute of Drug Abuse (NIDA) grants DA024104 (JT) and DA 027748 (JT).

PAGE 44

28 CHAPTER III 3. TOP DOWN NETWORK EFF ECTIVE CONNECTIVITY I N ABSTINENT SUBSTANC E DEPENDENT INDIVIDUAL S 3.1. Abstract This chapter reports resting state large scale brain network connectivity changes in a subsample of the population presented in Chapter 2. We hypothesized that compared to healthy controls, long term abs tinent substance dependent individuals (SDI) will differ in their effective connectivity between large scale brain networks and demonstrate increased directional information from executive control to interoception , reward , and habit related networks. In addition, using graph theory to compare network efficiencies we predicted decreased small worldness in SDI compared to controls. 50 SDI and 50 controls of similar sex and age completed psychological surveys and resting state fMRI. fMRI results were analyze d using group independent component analysis; 14 networks of interest (NOI) were selected using template matching to a canonical set of resting state networks. The number, direction, and strength of connections between NOI were analyzed with Granger Causal ity. Within group thresholds were p<0.005 using a bootstrap permutation. Between group thresholds were p<0.05, FDR corrected for multiple comparisons. NOI were correlated with behavioral measures, and group level graph theory measures were compared. Compar ed to controls, SDI showed significantly greater Granger causal connectivity from right executive control network (RECN) to dorsal default mode network (dDMN) and from dDMN to basal ganglia network (BGN). RECN was negatively correlated with impulsivity, be havioral approach, and negative affect; dDMN was positively correlated with impulsivity. Among the 14 NOI, SDI showed greater bidirectional connectivity; controls showed more unidirectional connectivity. SDI demonstrated greater global efficiency

PAGE 45

29 and lower local efficiency. Increased effective connectivity in long term abstinent drug users may reflect improved cognitive control over habit and reward processes. Higher global and lower local efficiency across all networks in SDI compared to controls may refle ct connectivity changes associated with drug dependence or remission and requires future, longitudinal studies to confirm. This chapter has been peer reviewed and published: Regner MF, Saenz N, Maharajh K, Yamamoto D, Mohl B, Wylie KP, Tregellas J, Tanabe J. Top Down Network Effective Connectivity in Abstinent Substance Dependent Individuals. PLOS ONE . 2016 Oct 24; 11(10): e0164818. PMID: 27776135 . 3.2. Introduction Substance dependence is a significant public health problem with an estimated 10.3% lifetime prev alence in the United States (Miller and Hendrie, 2009) . Across substances of abuse, a generalizable pattern develops beginning with an initial stage of rewarding effects from occasional use and developing into a pathologic stage of loss of cont rol, escalated use, compulsive drug seeking, and significant negative consequences (Wise and Koob, 2014) . Individuals with substance dependence have been shown to exhibit higher levels of impulsivity, behavioral approach, and negative affect (Perry et al., 2013) , and these differences have been associated with structural (Regner et al., 2015) and functional (Bell et al., 2014; Hyatt et al., 2012; Krmpotic h et al., 2013; Wisner et al., 2013) brain d ifferences compared to healthy controls. While task based studies using fMRI and PET have contributed significantly to our understanding of functional brain changes in specific neuroanatomical areas (Jasinska et al., 2014) , resting state fMR I (rsfMRI) provides opportunity to explore large scale networks and network interactions independent of task specific neuropsychological constructs (Fox and Greicius, 2010) . Advantages of rsfMRI include less confounding by differences in task paradigms, correlation of resting state networks (RSN) to specific tasks and neuropsychiatric

PAGE 46

30 constructs (Smith et al., 2009) , and reproducibility due to simplified e xperimental design and data acquisition (Chen et al., 2008) . Stimulant dependence is cha racterized by complex behaviors and, like other neuropsychiatric diseases, is thought to reflect pathology at the circuit level rather than a single brain structure (Koob and Volkow, 2010) . Moreover, activity and connectivity differences in stimulant dependence have been demonstrated using rsfMRI across disease stages and may explain the progressive behavioral phenotype changes across the natural history of the disorder (Sutherland et al., 2012) . For example, active drug addiction stages inc lude (I) binge/intoxication, (II) withdrawal/negative affect, and (III) preoccupation/anticipation (Koob and Volkow, 2010) ; involved circuits at these stages include (I) ventral tegmental area and striatum; (II) amygdala, bed nucleus of the stria terminalis, and ventral striatum; and (III) prefrontal cortex, hippocampus, baso lateral amygdala, cingulate, and insula. Sequential cycling through these active disease stages is hypothesized to result in the neuroadaptive changes that give rise to compulsive drug seeking and drug taking ( Figure 3 1 ). Brain activity and connectivity at different disease stages have been correlated with individual differences in executive function, interoception, reward, and habit formation. For example, Gu et al. (Gu et al., 2010) observed decreased rsfMRI connectivity between nodes within the mesocorticolimbic reward pathway in active cocaine users compared to healthy controls. These findings are consistent with animal models, in which rats dependent upon and self administering cocaine demonstrated decreased connectivity compared to control rats (Lu et al., 2014) ; affected pathways in this sample of rats included connections between the dorsolateral prefrontal cortex (PFC) and ventral striatum, as well as between the p relimbic cortex (homologous to anterior cingulate gyrus in humans) and entopeduncular nucleus (homologous to globus pallidus interna in humans) (Lu et al., 2014) . These active disease findings stand in contrast to findings in disease remission. In short term abstinent cocaine

PAGE 47

31 dependence ( 3 days), Wilcox et al. (Wilcox et al., 2011) observed increased rsfMRI connectivity between the ventral striatum and ventromedial PFC. Camch ong and colleagues (Camchong et al., 2014) measured resting state functional connectivity amongst reward processing regions in a cohort of stimulant dependent individuals at two time points, 5 weeks abstinence and 13 weeks abstinence, with comparison to a matched healthy contr ol group. Abstinent stimulant dependent patients demonstrated increased functional connectivity compared to controls, consistent with prior studies in patients of 1.4 years of abstinence (Krmpotich et al., 2013) and 5.7 ye ars of abstinence (Camchong et al., 2013) . Although Camchong et al. (Camchong et al., 2014) found abstinent stimulant dependent patients demonstrated increased functional connectivity compared to controls, patients who relapsed between time points demonstrated decreased conne ctivity compared to patients who Figure 3 1 . Simplified illustration of the natural history of substance use disorders. Active substance dependence is characterized by sequential stages of binge/intoxication, withdrawal/negative affect, and craving/preoccupation (Koob and Volkow, 2010) that repeat a variable number of cycles, represented by m . Active disease episodes may be interrupted by periods of short or long term abstinence, represented by n .

PAGE 48

32 maintained abstinence. The authors speculated that this reduction in functional connectivity from 5 to 13 weeks in relapsers compared to abstinent patients may be associated with these ence. These studies suggest that group differences in connectivity may be related to different stages of dependence/remission, possibly representing a transition from hypoconnectivity in limbic and s ubcortical regions during active dependence to increased top down executive control in sustained abstinence. Understanding differences in large scale brain connectivity depends upon characterizing the relative activity within networks as well as between them. Two modes of functional interactions between brain re gions include functional connectivity and effective connectivity. Functional connectivity is the simultaneous and temporally coherent activation of separate brain regions. Effective connectivity characterizes the directional flow of information. One method of characterizing effective connectivity is Granger causality (Seth, 2010) , which is methodologically straightforward but requires careful application and interpretation. T o date, no study has investigated the effective connectivity differences in stimulant dependence. This is important because understanding the direction of information flow in large scale brain networks may further elucidate mechanisms of abstinence and exp lain previously reported changes. To improve substance dependence treatments, a better understanding of the connectivity characteristics associated with long term remission are needed and may help to predict successful abstinence, evaluate treatment effica cy, and develop novel treatments. This study investigated the effective connectivity and graph theory characteristics of large scale networks in the resting brain in long term abstinent SDI compared to healthy controls to provide holistic, organ level meas ures of brain connectivity and organization for comparisons between groups. We hypothesize d that compared to healthy controls, long term abstinent SDI will demonstrate altered effective connectivity between large -

PAGE 49

33 scale brain networks and increased directio nal information from executive control to interoception , reward , and habit related networks. 3.3. Materials and Methods 3.3.1. Sample Population Fifty substance dependent individuals (SDI) and 50 healthy controls matched in age and sex were prospectively recruited between October 2010 and June 2013. Demographic information is reported in Table 3 1 . SDI were recruited from a residential treatment program at t he University of Colorado Denver Addiction Research Treatment Services. Inclusion criteria for SDI were lifetime DSM IV psychostimulant dependence (methamphetamine, cocaine, or amphetamine class substances) and abstinence from all drugs of abuse for a min imum 60 days, verified through close supervision and random urine screens. Participants were permitted to have previously met dependence criteria for substances other than psychostimulants due to the high prevalence of polysubstance use in people dependent upon psychostimulants. Average abstinence duration was 12.8 ± 12.4 months. Healthy controls were recruited from the community and excluded if dependent on alcohol or other drugs of abuse except tobacco. Exclusion criteria for all participants included maj or depression within the last two months, psychosis, neurological illness, prior head trauma with loss of consciousness exceeding 15 minutes , prior neurosurgery, HIV, bipolar disorder, other major medical illness, inability to tolerate MRI, positive urine or saliva screen ( AccuTest TM , AlcoScreen TM ), and IQ < 80. All participants provided written informed consent approved by the Colorado Multiple Institutional Review Board.

PAGE 50

34 3.3.2. Structured Interviews and Questionnaires All participants received structured interviews and behavioral measures. Drug dependence was assessed using the computerized Composite International Diagnostic Interview Substance Abuse Module (CIDI SAM) (Cottler et al., 1989) . IQ was estimated with matri x and verbal reasoning Wechsler Abbreviated Scale of Intelligence subtests (WASI, Table 3 1 . Demographic, drug use, and behavioral characteristics of the sample population. BAS, Behavioral activation scale; BIS, Behavioral inhibition scale; BIS 11, Barratt impulsiveness scale version 11; PANAS X, Positive and Negative Affect Schedule Expanded Fo rm. SDI Control t value p value Demographics N 50 (22F/28M) 50 (25F/25M) 0.869 Age 34.18±7.63 31.6±8.57 1.589 0.115 Education 12.48±1.42 14.66±1.47 7.560 <0.001 Abstinence (mo) 12.8 ± 12.4 Drug Dependence Stimulants 50 Nicotine 35 8 Alcohol 35 Opioids 16 Cannabis 20 Other 9 Behavioral Metrics BIS 20.64±3.26 20.10±3.56 0.792 0.430 BAS 76.74±6.54 69.40±9.27 4.575 <0.001 Drive 12.80±2.17 10.34±2.26 5.551 <0.001 Fun Seeking 13.16±1.845 11.38±2.18 4.414 <0.001 Reward 17.40±1.94 17.10±1.75 0.812 0.419 BIS 11 72.16±11.54 57.30±6.71 7.871 <0.001 Motor 27.30±4.72 22.78±3.02 5.707 <0.001 Non Planning 27.08±4.96 20.48±3.33 7.809 <0.001 Attentional 17.78±3.81 14.04±3.14 5.361 <0.001 PANAS X Positive Affect 35.32±6.09 35.92±6.09 0.488 0.626 Negative Affect 21.54±7.98 14.72±3.48 5.540 <0.001

PAGE 51

35 Psychological Corporation, 1999) and was recorded to exclude subjects with low scores (IQ < 80) . The Diagnostic Interview Schedule version IV is a computerized structured int erview used to screen for psychiatric disorders. Participants completed this interview to exclude those with a history of psychiatric disorders as described above. Substance dependence severity was operationalized as the number of total substance dependenc e and abuse symptoms, quantified by the Diagnostic Interview Schedule version IV (Gelhorn et al., 2008; Robins et al. , 1995b) . The Behavioral Inhibition and Activation Scale is a 20 item self reported questionnaire used to measure responsiveness of motivational syste ms (Campbell Sills et al., 2004; Carver and White, 1994) . Behavioral approach and inhibition were operationalized as the total Behavioral Activation and Inhibition Scales, respectively. The Barratt Impulsiveness Scale (BIS 11) is a 30 item self reported questionnaire used to measure impulsivity; participants rated whether phrases and words describing aspects of impulsivity were self descriptive (Patton et al., 1995) . Impulsiveness was operationalized as the total Barratt score. Positive and Negative Affect Schedule Expanded Form (PANAS X) quantifies a on a scale of self description (Crawford and Henry, 2004) . Positive and negative affect were operationalized as the total PANAS X score for positive and negative attributes. 3.3.3. MRI Examination and Image Analysis MRI Acquisition Brain MRI was performed using a 3T MR scanner (General Electric, Milwaukee, Wisconsin) and standard quadrature head coil. Head motion was minimized using a VacFix head conforming vacuum cushion (Par Scientific A/S, Odense, Denmark). mm of head motion were excluded. High resolution T1 weighted SPGR IR sequences (TR=45ms, TE=20ms, flip angle=70°, 256 × 256 matrix, 240 × 240mm 2 field of view (0.9 × 0.9mm 2 in

PAGE 52

36 plane), 1.7mm slice thickness, and coronal plane acquisition ) and r esting state functional scans ( TR=2000ms, TE=30ms, flip angle=30°, axial acquisition, 64 × 64 matrix, 3.4 mm × 3.4 mm in plane voxel size, 3mm slice thicknes s, 1mm gap , 150 volumes) were acquired . During fMRI acquisition, p articipants were instructed to close their eyes, not think of anything in particular, and not fall asleep. Image Preprocessing Resting fMRI images were processed using the SPM8 toolbox in MATLAB. The first four volumes of each examination were excluded to avoid saturation effects ( Figure 3 2 ). Standard pre processing steps included slice timing correction, rigid realignment and motion correction (motion >1 voxel/TR was censored), spatial normalization, and de noising. Motion parameters (three rotation and three translation parameters) for censor ship were calculated for each time point using corresponding SPM realignment pre processing values. Anatomical volumes were segmented into gray matter, white matter, and CSF tissue maps, and the resulting binary masks were eroded (1 isotropic voxel) to mit igate partial volume effects. CSF and white matter time Figure 3 2 . Diagram of the fMRI BOLD signal pre processing, network of interest definition, and effective connectivity analysis pipelines.

PAGE 53

37 series were obtained using the mean signals from voxels based on eroded CSF and white matter SPM template masks. Mask erosion and time series extraction were performed using functions contained in the CONN toolbox (Whitfield Gabrieli and Nieto Castanon, 2012) . After linear trends were removed, time series of the motion parameters, WM s ignal, and CSF signal were removed from the resting state BOLD data using linear regression, and the resultant residual BOLD time series were band pass filtered (0.008 Hz < f < 0.15 Hz) (Braun et al., 2012) . The resultant filtered time series were spatially smoothed with a 6mm full width at half maximum Gaussian kernel. Networks of Int erest (NOI) Definition and Behavioral Correlations Group independent component analysis (ICA) was performed using the GIFT toolbox as previously reported in the literature (Krmpotich et al., 2013; Tanabe et al., 2011; Tregellas et al., 2011a; Tregellas et al., 2011b) in order to define the networks of interest (NOI). For the purposes of this study, the term resting state networks (RSN) refers to the canonical spatial maps used to define the NOI. The term NOI refers to the independent components identified in our sample population and labeled by their corresponding RSN. The dimensionality of the data from each subject was first reduced to 100 components using principal component analysis. Subsequent group level ICA yielded 34 components, the number of which was determined using the minimum description length (MDL) algorithm (Li et al., 2007) . Fourteen canonical RSN templates ( Tab le 3 2 ) were provided by Stanford's Functional Imaging in Neuropsychiatric Disor ders (FIND) Laboratory (Shirer et al., 2012) . At the group level, the 34 identified components were spatially correlated with the canonical RSN templates. Components with the highest spatial correlation to the canonical template were labeled with the corresponding standard RSN label. These labelled components formed the set of NOI for subsequent graph analysis. All components were visually inspected by a neuroradiology fellow (N.S.) and radiology resident (M.R.) independently to confirm accuracy with the canonical RSN templates .

PAGE 54

38 Concordance between inspectors was 100%. Subject specific spatial maps and time courses were estimated using the GICA back reconstruction function in GIFT . For each subject, the strength (or coherence) of each NOI was operationalized as the mean beta value across spatial dimensions for that component in the mixing matrix. These values were regressed against impulsivity, approach, inhibition, and negative affect. the neuroimaging findings within the context of measurable behavioral characteristics. Effective Connectivity Analysis For each individual, the time courses corresponding to the NOI were obtained from the back reconstruction function in group ICA. These NOI signals were linear trend removed, normalized to zero mean and unit variance, and band pass filtered at 0.008 0.15 Hz . The Tab le 3 2 . Canonical RSN included in the analysis and their corresponding symbol abbreviations. Spatial maps were provided by Stanford's Functional Imaging in Neuropsychiatric Disor ders (FIND) Laboratory. Resti ng State Network Symbol Auditory Network AN Anterior Salience Network aSN Basal Ganglia Network BGN Dorsal Default Mode Network dDMN High Visual Network HVN Left Executive Control Network LECN Language Network LN Precuneus Network PCN Posterior Salience Network pSN Primary Visual Network PVN Right Executive Control Network RECN Sensorimotor Network SMN Ventral Default Mode Network vDMN Visuospatial Network VSN

PAGE 55

39 resultant NOI time courses were temporally concatenated across individuals into SDI and control groups (Deshpande et al., 2010a; Ding and Lee, 2013) . Effective connectivity betwe en all 14 NOI time courses in the SDI and control graphs were calculated using Granger causality (GC) analysis implemented in the Granger Causal Connectivity Analysis (GCCA) MATLAB toolbox (Seth, 2010) . Significance was estimated by compar ing observed group difference to a randomized null hypothesis distribution, and the test statistic was determined by the percentile position of the observed difference (i.e., the proportion of randomizations with values greater than or equal to the observe group labels were randomized and GC connectivity differences estimated for each randomization permutation until the aggregate randomization distribution achieved statistical stability. Signifi q < 0.05 to correct for multiple comparisons. Global Network Measures To provide global network measures, graph theory measures were used to describe the topology of the graphs of NOI. The purpose w as to provide holistic, organ level measures of brain connectivity for comparisons between groups. These measures included total weighted network density, local efficiency (derived from clustering coefficients), and global efficiency (derived from path len previously described in detail (Rubinov and Sporns, 2010) . To compute graph theory metrics the GC connectivity matrices were converted to a binary directed adjacency matrix, w here connectivity strengths above and below a certain threshold cost level are set to 1 and 0, respectively (Ginestet et al., 2011) . The cost threshold, , was calculated by first sorting all elements other than the auto correlating diagonals (identity axis) of the connectivity matrix in descending rank and keeping only the top x% (sub -

PAGE 56

40 graph, ). Cost was then computed as the fraction of highest strength edges above the given threshold divided by the total number o f edges: where represents an edge saturated network with the same and the function represents the cardinality of . Therefore, a low value reflects a sparse network. The Brain Connectivity Toolbox (Rubinov and S porns, 2010) was used to calculate global and local efficiency at each cost level. Since there was no a priori reason to select a particular network cost threshold, connectivity metrics as a function of cost were computed and integrated across the cost domain [0, 1] (Ginestet et al., 2011) : This approach abides by prior methodological recommendations to separate network cost from network topology (Ginestet et al., 2011) . To test for signifi cant differences between groups, a non randomized and graph theory measures were calculated for each randomization permutation until the randomization distribution demonstrat ed statistical stability. These distributions formed the null hypothesis distributions representing no group differences for total weighted network density, integrated global efficiency, and integrated local efficiency. Significance of groups differences f or each metric was determined by the percentile position within the null hypothesis distribution. 3.4. Results 3.4.1. Sample Population Demographics There were no significant differences in age ( p = 0.12) or sex ( p = 0.87 ) between groups ( Table 3 1 ) . Years of education ( p <0.01) and IQ ( p < 0.01) differed between groups. Controls had

PAGE 57

41 higher IQ and more years of education than SDI . All SDI met DSM IV depe ndence criteria for stimulants. Drug use characteristics are also summarized in Table 3 1 . Eight controls met dependence criteria for tobacco. No controls met dependence criteri a for drugs or alcohol. 3.4.2. Behavioral Metrics Behavioral characteristics are summarized in Table 3 1 . No group difference in BIS inhibition was observed ( p =0.43). A significant group difference in behavioral approach was observed, with SDI exhibiting higher total BAS scores than controls ( p <0. 001). Further analysis p ( p p =0.42). As expected, SDI reported higher impulsivity than controls ( p <0.001) as well as significan t differences in the motor, non planning, and attentional subscales ( p <0.001). No significant difference in positive affect was observed ( p =0.626); however, a large difference in negative affect was observed ( p <0.001), with SDI demonstrating greater negati ve affect scores than controls.

PAGE 58

42 3.4.3. Network Analysis Directed Connectivity Analysis SDI network density was significantly greater compared to controls ( p <0.001, Figure 3 3 ). This measure reflects increased overall mean GC causal connectivity strength between all 14 NOI compared to controls. Specifically, GC analysis results show that among 182 possible betwee n network pairs ( Figure 3 4 ), only three pairs differed significantly across group in the FDR corrected data ( Figure 3 5 ). Compared to controls, SDI showed stronger effective connectivity from the RECN to the dDMN and from the dDMN to the BGN ( Figure 3 6 ). In addition, SDI showed stronger effective connectivity from the SMN to the VSN. SDI showed Figure 3 3 . Effective connectivity network density graphs of SDI and controls. Thickness of each line corresponds to the Granger causal strength and color corresponds to the efferent network. SDI network density was significantly greater than healthy controls (p<0.001).

PAGE 59

43 greater bidirectional connectivity (reciprocal GC connections) whereas controls showed more unidirectional connectivity among the 14 network components. Figure 3 4 . Direct ed GC matrices for SDI (left) and controls (right). Colorbar corresponds to logarithm of F values. Figure 3 5 . Effective connectivity matrix illustrating the FDR corrected group differences between SDI and healthy controls. Colorbar corresponds to p values. White arrows indicate Granger causal direction.

PAGE 60

44 Network Behavioral Correlations The strength of the RECN correlated negatively with impulsivity (p<0.001), behaviora l approach (p<0.001), and negative affect across the population (p=0.006) ( Figure 3 7 ). In contrast, the strength of the dDMN correlated positively with impulsivity (p<0.001), but not behavioral approach (p=0.034) or negative affect (p=0.030) after correcting for multiple compar isons ( Figure 3 7 ). No NOI correlated with positive affect or educational attainment in years. Global Graph Measures Group comparison results for local and global efficien cy across the domain of co st functions are illustrated in Figure 3 8 . Global efficiency was significantly hig her in SDI than in Figure 3 6 . Left, NOI demonstrating sign ificantly greater Granger causal relationships in SDI compared to controls. Illustrated are the NOI relationships that differed between groups; colors correspond to NOI labels on right. Printed numbers in upper left corner of each slice correspond to Z axi s coordinates in MNI space. Right, Granger causal relationships demonstrating greater directed information flow in SDI compared to controls (FDR corrected). RECN, right executive control network; dDMN, dorsal default mode network; BGN, basal ganglia networ k. Values indicate Granger causal p values. Middle, whole brain illustrations of NOI identified. Arrows reflect Granger causal influence.

PAGE 61

45 controls (p< 0.01) , suggesting greater global integration. Local efficiency was higher in controls (p<0.05) , suggesting greater local specialization. These findings in conjunction suggest reduced small worldness in SDI compared to healthy controls. 3.5. Discussion This study revealed greater effective connectivity ne twork patterns in abstinent substance dependent individuals compared to healthy controls. Specifically, in drug users who have been abstinent for on average over one year, effective connectivity analysis revealed increased information flow from the RECN to dDMN and dDMN to BGN compared to controls. Figure 3 7 . Correlations between mean beta value within the RECN and dDMN with impulsivity, approach, and negative affect metrics. Solid black lines indicate the linear regression, solid colored lines indicate the 95% confidence interval, and colo red shaded regions indicate the and their score on the given behavioral metric.

PAGE 62

46 The areas of increased effective connectivi ty observed in our study correspond to regions involved in executive control (RECN), interoception (dDMN), reward (BGN), and habit (BGN). The mean strength of the RECN component correlated negatively with impulsivity, behavioral approach, and negative affe ct. In contrast, the mean strength of the dDMN component correlated positively with impulsivity and trended towards positive correlations with behavioral approach and negative affect. Given the prolonged abstinence of our SDI sample population, these findi ngs are consistent with the hypothesis that successful long term abstinence is associated with increased top down cognitive control. 3.5.1. Increased Effective Connectivity from RECN to dDMN The pattern of increased effective connectivity from RECN to dDMN is con sistent with increased top down executive control in long term abstinence. Previous work has demonstrated both task related and resting state hyperactivity within executive control and default mode cortices in abstinent stimulant dependence associated with heightened behavioral Figure 3 8 . Global efficiency (left) and local efficiency (right) in SDI and controls as a function of network cost.

PAGE 63

47 monitoring (Connolly et al., 2012; Mayer et al., 2013; Wilcox et al., 2011) ; however, this is t he first study to suggest that these neural signals follow a directional flow of information from RECN to dDMN. Connolly et al. (Connolly et al., 2012) conducted a cognitive control task based fMRI study of short (2.4 ± 1.34 weeks) and long term (69 ± 17.49 weeks) abstinent co caine addicts. Abstinent cocaine users demonstrated increased activity in PFC, cingulate, and inferior frontal gyri compared to healthy controls. Moreover, short term abstinent individuals showed right dorsolateral PFC (corresponding to RECN in our study) hyperactivity positively correlating with inhibitory control. Long term abstinent individuals showed the same finding as well as anterior and mid cingulate (corresponding to part of the dDMN in our stu dy) hyperactivity positively correlating with cognitive errors and heightened behavioral monitoring in abstinence. The present study showed that the RECN strength was negatively correlated with rrelated with impulsiveness. Our results advance our understanding of neural network changes during substance use disorder remission: as abstinence progresses, cortices within the RECN and dDMN may become hyperactive to exert top down executive control in a directed fashion; this neuroadaptive change may be associated with decreases in impulsivity and increases in inhibitory control. However, the top down cognitive control hypothesis is not straightforward because in addition to executive function and cogn itive control, affect plays an important role. Albein Urios et al. (Albein Urios et al., 2014 ) showed that short term abstinent (2.5 ± 5.5 months) cocaine dependent individuals had increased right dorsolateral PFC and bi lateral temporoparietal cortex activation during negative emotion experiences without a concomitant increase in the subjective negative experience itself, suggesting an exaggerated neural response in these regions is required to produce normal levels of em otional salience. The regions reported closely resemble by visual comparison the RECN identified by our analysis. Albein Urios et al. posited that these areas demonstrate increased sensitization toward negative emotions in SDI. If

PAGE 64

48 increased RECN top down c ontrol is a durable feature of long term abstinence, the literature thus far suggests that its manifestations in human behavior are complex and not reducible to a single neuropsychological construct. Our finding that RECN strength is negatively correlated to negative affect while dDMN strength trended towards positive correlation with negative affect provides further evidence that the top down control model may involve affective components as well, possibly through reciprocal connectivity with limbic areas. Right sided lateralization of our ECN findings is not unexpected given the asymmetric functional specialization of cerebral hemispheres in healthy humans. Cocaine dependent patients exhibit reduced resting state interhemispheric connectivity compared to healthy controls in prefrontal and parietal cortices (Meunier et al., 2012) , suggesting increased lateralization of function. Connolly et al. (Connolly et al., 2012) reported that hyperactivity in the inferior frontal gyrus correlated with inhibitory control was greater in the left hemisphere in short term abstinent individuals and greater in the right hemisphere in long term abstinent individuals. They hypothesized that a shift from left to right inferior frontal gyrus for inhibitory control may reflect a transition from short term to long term abstinence. Our results of increased RECN effecti ve connectivity in long term abstinent stimulant dependence are consistent with this hypothesis, although future longitudinal studies are required for substantiation. 3.5.2. Increased Effective Connectivity from dDMN to BGN Although the DMN is incompletely understood, growing evidence demonstrates its roles in internally directed tasks such as spontaneous cognition (Mantini and Va nduffel, 2013) , self referential (Vessel et al., 2013) and autobiographical thought (Buckner et al., 2008) , and social understanding of others (Li et al., 2014) . We demonstrated increased effective connectivity from the dD MN to BGN; however, this finding must be interpreted in the context of the structures within the NOI identified as BGN ( Figure 3 4 ). This NOI included basal ganglia,

PAGE 65

49 thalamus, amygdala, hippocampus, hypothalamus, midbrain, and pons. Thus, BGN included several key regions of the bottom up mesocorticolimbic circuit including the ventral tegmental area, nucleus accumbens, amygdala, and striatum. Prior studies have demonstrated hypoactivity in stimulant users in the dDMN and in BGN as well as decreased connectivity between these networks. In active cocaine users, Tomasi et al . (Tomasi et al., 2015) showed that cocaine cues disengaged fMRI activity in the ventral striatum, hypothalamus, and DMN in proportion to density of striatal dopamine receptors by PET. DMN activation has been shown to predict p erformance errors, is diminished in active cocaine dependence, and the extent of altered error preceding activation has been reported to correlate with years of cocaine use (Bednarski et al., 2011) . Gu et al. (Gu et al., 2010) used a seed based fMRI paradigm in active cocaine users and found significantly decreased functional connectivity between multiple regions of the DMN and BGN. McHugh et al. (McHugh et al., 2014) showed that individuals successfully abstinent 30 days after detoxification had stronger functional conne ctivity between the amygdala, ventromedial PFC, and anterior cingulate cortex compared to those who had relapsed. By visual comparison, these regions correspond to structures within the NOI identified as dDMN and BGN in our study. Connolly et al. (Connolly et al., 2012) demonst rated increased anterior and mid cingulate activity in long term abstinent compared to short term abstinent individuals, activity which correlated with heightened behavioral monitoring. Together, these prior studies suggest that DMN activity may change dur ing abstinence. Initial hypoactivation during early abstinence may transition to hyperactivation and increased connectivity with long term abstinen ce . One interpretation is that findings of increased effective connectivity from dDMN to BGN in long term abs tinence may be a compensatory mechanism related to behavioral monitoring not seen in active users. However, longitudinal studies are needed to demonstrate this.

PAGE 66

50 3.5.3. Increased Global and Decreased Local Integration Our findings of increased bidirectional connectivity, increased global efficiency, and decreased local efficiency in long term abstinent SDI compared to healthy controls suggests pathologically greater global integration and lower local integration in SDI; that is, a connectomic de crease in small worldness. Similar findings in humans have only been reported using EEG data in 1 3 week abstinent methamphetamine dependent persons. Ahmadlou et al. (Ahmadlou et al., 2013) showed that these patients demonstrated a deviation from small wo rldness and increased global hypersynchronization in the gamma frequency band, the EEG band most reactive to cognitive information processing. In contrast, active cocaine users demonstrated less global connectivity compared to healthy controls during a Str oop task; however, after adjusting for individual connectivity, cocaine dependent individuals showed greater intrinsic connectivity in the ventral striatum, putamen, inferior frontal gyrus, anterior insula, thalamus and substantia nigra (Mitchell et al., 2013) . Several animal studies provide important context for the interpretation of our findings. Schwarz et al. (Schwarz et al., 2012) used a pharmacological challenge des ign which revealed that rats under the acute effects of amphetamine compared to a saline vehicle exhibited less clustering (small worldness) and increased connectedness within somatosensory, motor, cingulate, prefrontal, and insular cortices. In the rhesus monkey model, active cocaine self administration was associated with decreased global functional connectivity that selectively affected top down prefrontal circuits and control behavior while sparing limbic and striatal areas (Murnane et al., 2015) . Interestingly, impaired connectivity between prefrontal and striatal areas during abstinence predicted cocaine intake when these monkeys were again provided access to cocaine (i.e., prediction of relapse), consistent with th e connectivity pattern associated with relapse in humans as reported by Camchong et al. (Camchong et al., 2014) .

PAGE 67

51 Together these findings suggest t here is globally decreased connectivity in active users and short term abstinent with a transition to globally increased connectivity in long term abstinent users. These findings may improve clinical management if global connectivity patterns can be used t o predict abstinence success or trajectory in humans. Future longitudinal studies comparing global connectivity in active, short term, and long term abstinent drug users must be performed to address this question. Another approach could involve correlating abstinence duration with global connectivity across individuals, an approach we could not implement due to the group level nature of our statistical design. 3.5.4. Limitations Controversy Surrounding Granger Causality While our study provides several important n ovel findings, it has limitations. Influences between specialized neural systems exist on a spectrum of temporal lag. Functional connectivity using temporal correlation reflect influences with causal latencies that are below the temporal resolution of the repetition time. These influences are not truly contemporaneous in vivo , but appear so by fMRI as a result of low temporal sampling and temporal blurring induced by the hemodynamic response function. Time lag based measures such as Granger causality reflec t slower influences with greater causal latencies that occur on the order of hundreds of milliseconds, which may provide greater power in predicting cause effect relationships at the timescale of conscious thought (Tononi et al., 2016) . Neural signals between two nodes may have significantly different physiologic func tions depending upon the directionality. As a result, segregating neural influences according to their directionality is necessary in order to properly examine brain function. Methods of examining effective connectivity using fMRI data include structural e quation modelling (Buchel and Friston, 1997) and dynamic causal modelling (Friston et al., 2003) . These methods require a priori hypotheses describing the theoretical connectivity structure and are limited to models

PAGE 68

52 consisting of a small number of nodes. We used an alternative method, Granger causality, which is based on time lag regressions and is more data driven. Granger causality is increasingly used in fMRI based neuroscience (Chiong et al., 2013; Cohen Kadosh et al., 2016; Feng et al., 2016; Wen et al., 2013; Zhang et al., 2017) and has been pre viously applied specifically to independent component analysis as in our study (Demirci et al., 2009; Diez et al., 2015; Ding and Lee, 2013; Stevens et al., 2009; Zhong et al., 2012) . However, criticisms of the application of Granger causality to fMRI data have in cluded (Deshpande et al., 2010b) : (1) lack of evidence that Granger causality in fMRI level time series reflects causality in neuronal level time series, (2) insufficient temporal sampling relative to the timescale of ne uronal events, and (3) the possibility that spurious findings may result from systematic differences in hemodynamic response functions. Several recent developments have provided evidence that fMRI Granger causality reliably reflects neuronal causality (Deshpande and Hu, 2012; Deshpande et al., 2010b; Schippers et al., 2011; Wen et al., 2013) . Seth and colleagues (Seth et al., 2013) demonstrated that Granger causality is reliably invariant to inter regional differences in the hemodynamic response function, including the time to peak. However, they reported si gnificant effects of temporal resolution on their results. Wen and colleagues (Wen et al., 2013) demonstrated that fMRI based Granger causali ty is a monotonic function of neural Granger causality. Importantly, they showed that this relationship can be reliably detected using conventional fMRI temporal resolution and noise levels as was used here. However, they cautioned that differences in the hemodynamic response could lead to spurious results. The impact of hemodynamic response variability is currently debated. Schippers and colleagues (Schippers et al., 2011) demonstrated that hemodynamic response variability was minimized by multisubject group inference. Statistica lly, this is intuitive because population averaging will augment systematic differences (e.g., true neuronal differences) while

PAGE 69

53 suppressing random or pseudorandom differences (e.g., hemodynamic response variability). Some authors speculate that HRF variabi lity could be systematic (Smith et al., 2012) , and indeed this is a confound that by design exists in the majority of between group fMRI studi es using independent samples (D'Esposito et al., 2003; Hillman, 2014; Murphy et al., 201 3) . Accordingly, we cannot exclude that systematic differences in the neurovascular response to neural activity between g roups may have contributed to our findings. Other Limitations Network resolution was limited by the manner in which independent component analysis identifies temporally coherent signals across the brain. For example, the network component identified as BG N included several non basal ganglia structures, such as the thalamus, amygdala, hippocampus, and midbrain. Additionally, concatenation across individuals precluded correlation of individual psychological measures to resting state network Granger causality ; as such, correlations between the strength of each NOI and behavioral metrics were used to provide psychological context for the findings. Lastly, polysubstance use and low educational attainment among psychostimulant users may be viewed as potential con founds or representation of real world clinical features. There is significant literature describing the correlation between drug use and low educational attainment; it is debated whether low educational attainment is the cause or result of drug use disord ers (Fergusson et al., 2003; Swaim et al., 1997; Yamada et al., 1996) . More recently, however, authors have reported that this correlation is due in part to shared genetic factors (Bergen et al., 2008) while others report that it is due to shared environmental or non genetic familial risk factors (Grant et al., 2012; Verweij et al., 2013) . These studie s suggest that low educational attainment is a behavioral component of the pathology of substance use disorders. With regard to polysubstance use among SDI, while this limitation prevents our findings from being attributed to a single drug, it strengthens our results by providing biological and ecological validity.

PAGE 70

54 Epidemiologic studies have demonstrated that psychostimulant dependence does not naturally occur in isolation; rather, most patients meet dependence criteria for other drugs of abuse (Sara et al., 2012; Stinson et al., 2005) . Our sample population thus reflects the real world, clinical population of patients with stimulant dependence. 3.6. Conclusion Increased effective connectivity in long term abstinent drug users may reflect improved cognitive control and behavioral mo nitoring (ECN) over self referential thought (DMN), habit (BGN), and reward (BGN) processes in long term abstinent drug users. Higher global and lower local efficiency across all networks in SDI compared to healthy controls may reflect connectivity changes associated wi th drug dependence or remission. F uture, longitudinal studies are necessary to definitively characterize connectomic changes across the natural history of substance use disorders. 3.7. Acknowledgements Several co authors contributed to this chap ter: Naomi Saenz MD, Keeran Maharajh PhD, Dorothy Yamamoto PhD, Brianne Sutton (nee Mohl) PhD, Korey Wylie, Jason Tregellas PhD, and Jody Tanabe MD. This work was supported by the National Institute of Drug Abuse (NIDA) grants DA024104 (JT), DA 027748 (JT ), and DA 041011 (MFR). The authors report no biomedical financial interests or potential conflicts of interest.

PAGE 71

55 CHAPTER IV 4. THE INSULA IN NICOTI NE USE DISORDER: FUN CTIONAL NEUROIMAGING AND IMPLICATIONS FOR NEU ROMODULATION 4.1. Abstract In this chapter, we revi ew the literature and provide motivation for our randomized clinical trial of inhibitory TMS targeting the insula in smoker, presented in Chapter 5. Animal and human literature suggest that the insula is necessary for nicotine use disorders. Yet, much rema ins unknown about how insular function drives nicotine use. Insular subdivisions show distinct patterns of connectivity with large scale brain networks, and each subdivision is During acute withdrawal, the insula arbitrates bottom up versus top down saliency mechanisms to guide behavior either to pursue smoking or to avoi d relapse and this arbitration is associated with craving and the nicotine withdrawal syndrome. The purpose of nicotine use disorder and present evidence for sui tability as a neuromodulation target to promote cessation. Given the limited efficacy of standard of care treatments for nicotine use disorder, insular neuromodulation may contribute to the next generation of cessation treatments by offering what hencefort h has not been available: a minimally invasive, anatomically driven approach to smoking cessation therapy. This chapter is currently under review for publication in Neuroscience & Biobehavioral Reviews , pending revisions.

PAGE 72

56 4.2. Introduction Given the serious health problems caused by smoking, it is not surprising that the majority of smokers, nearly 7 in 10 in 2015 , want to quit (Babb, 2017) . Unfortunately, however, smoking cessation treatments are largely ineffective, with m ost abstinence attempts fail ing within the first 24 hours . A pproximately 80% of patients relaps e by six months , d espite combine d pharmac ologic and behavioral therapies (Tobacco, 2008) . Whereas only 1 patient out of 10 receiving behavioral therapy alone remains abstinent, only 2 patients out of 10 receiving pharmacological (e.g. nicotine replacement) and behavioral therapy remain abstinent after six months (Stead and Lancaster, 2012) . These low success rates underscore the need for more effective interventions and improved prognostication of individuals most likely to benefit from a given intervention. To achieve these goals, a better understanding of the neura l underpinnings of compulsive nicotine use is required. A significant advance in our understanding of the neurobiology of smoking behavior was made in 2007, when Naqvi and colleagues observed that lesions to the insula disrupt cigarette smoking (Naqvi et al., 2007) . These findings and subsequent studies that will be described in greater detail below convincingly demonstrated that the insula plays a critical role in smoking maintenance and cravings, and that the region may be a therapeutic target for smoking cessation . (Abdolahi et al., 2015a, b, 2017; Forget et al., 2010; Naqvi et al., 2007; Pushparaj et al., 2013; Suner Soler et al., 2012) . The insula is involved in a wide array behaviors and functions, including salience, interoception, aw areness, affect, anticipation, uncertainty, self recognition, prediction error, perception, attention, and cognitive processing (Nieuwenhuys, 2012) . Recent meta analys es of functional MRI studies suggest that the insula contains two to seven functionally distinct regions (Cauda et al., 2012; Chang et al., 2013; Deen et al., 2011; Kelly et al., 2012; Kurth et al., 2010) . Current insula m odels suggest that the region plays a role in three broad categories of

PAGE 73

57 function. (Deen et al., 2011; Uddin et al., 2014) . The dorsal anterior insula is associ ated with cognitive control functions , such as attention, inhibitory control, and goal directed cognitive tasks (Dosenbach et al., 2007) . The ventral anterior insula is involved in with emotional limbic fun ctions , including pe ripheral physiological responses to emotional experiences, as measured by heart rate or galvanic skin response (Mutschler et al., 2009) . Finally, the p osterior insula mediates sensorimo tor interoceptive functions , and receives rich afferents from spinothalamocortical pathway carrying nociceptive, thermal, and other interoceptive information (Craig, 2002) . Despite this functional diversity within insular subunits, several closely related theories about the functional role of the insula in craving and smo king behaviors have emerged. While differing somewhat in their models and interpretations, the theories all sensations versu Figure 4 1 ). In this context, the Figure 4 1 . Connectivity based signal flow diagram of anterior insular control of bottom up versus top down mechanisms of salience. The right dorsal anterior insula is involved in processing salience of externally oriented stimuli and it is correlated with the exec utive control network (an externally directed system). The right ventral anterior insula is involved in processing salience of internally oriented stimuli and it is correlated with the default mode network (an internally directed system).

PAGE 74

58 insula may serve as a sensory signal bottleneck, such that insular lesions impair cognitive processing of craving and nicotine withdrawal sensations. The following sectio ns of this review describe and discuss: (1) evidence of insular involvement in nicotine use disorder pathophysiology and (2) possible therapeutic strategies to target the region for smoking cessation using neuromodulation, a non invasive technique capable of altering the function of specific brain regions and networks. Given the limited efficacy of standard of care treatments for nicotine use disorder, insular neuromodulation may contribute to cessation treatments by offering a non invasive, anatomically dr iven approach to smoking cessation therapy. 4.3. Insular Role in Nicotine Use Disorder 4.3.1. Introduction: Insular Lesions Disrupt Smoking Behaviors Converging evidence strongly implicates the insula in the maintenance of smoking behaviors and cigarette craving. Naqv i and colleagues (Naqvi et al., 2007) reported that smokers with cerebrovascular damage to the right insula were able to stop smoking easily without cravings or relapse, supporting a role of the insula in a ddi c tion. A subsequent, large , p rospective study over a one year period also found that insular lesions in smokers were strongly associated with becoming a non smoker (Suner Soler et al ., 2012) . Abdolahi and colleagues (Abdolahi et al., 2015b) conducted a prospective cohort study with three month follow up in 15 6 smokers hospitalized for acute ischemic stroke , of which 38 were insular strokes. They reported insular damage was associated with increased odds of three month continuous abstinence as well as cessation from all nicotine products at three months. Insula r damage in the same cohort was also associated with fewer nicotine withdrawal symptoms and cravings compared to those with non insular strokes (Abdolahi et al., 2015a, 2017) . These findings have been corroborated in animal models of nicotine dependence . For example, insular inactivation in rat models significantly reduced nicotine motivation, nicotine seeking , and

PAGE 75

59 nicotine taking behaviors, with no effect on food behaviors (Forget et al., 2010; Pushparaj et al., 2013) . These findings will be discussed in the context of animal model neuromodulation in Section 4.4 : Implications of Insular Role in Nicotine Use Disorder on Neuromodulatory Therapeutic Development (page 71 ). T hese human and animal studies together demonstrate that insular lesions disrupt smoking behaviors and underscore the need to understand insular function in smokers in vivo . 4.3.2. Neuroimaging the Insula in Nicotine Use Disorder Nicotine exposure causes two distinct sets of effects based on timing: acute states and chronic effects. Acute states refer to the short term behavioral changes and altered brain function independent of dependency. Chronic e ffects refer to pharmacologic dependence and pharmacodynamics are distinguished from the durable neural changes caused by chronic use; that is, nicotine use disord er reflects chronic effects resulting from repeated acute exposure to nicotine. We synthesize the neuroimaging literature into four distinct stages of nicotine use disorder and recovery ( Figure 4 2 ). First, we review studies of the acute effects of nicotine on neurobiology (page 60 ) . Second, we review studies comparing chronic smokers t o controls to understand dependence ( page 62 ). Third, we review studies of nicotine dependent individuals during acute abstinence to understand the mechanisms of the nicotine withdrawal syndrome and craving ( page 65 ). Fourth, we review long term abstinence as a model of neuroplastic recovery from nicotine use disorder ( page 67 ). Long term abstinence provides neuroimaging biomarkers of recovery may serve as useful indicators of treatment efficacy. Finally, we attempt to synthesize the findings from these four stages of the disease into a single model ( Section 4.3.3 : Putting it All Togethe r: Unified Models of the Role of the Insula in Nicotine Use Disorder Pathogenesis , page 68 ). Our findings are su mmarized in Table 4 1 .

PAGE 76

60 Stage 1: Acute Nicotine Exposure ( Neural Pharmacodynamics) dopamine release from ventral tegmental area neurons into the nucleus accumbens and prefrontal cortex (Volkow et al., 2012) . Nicotine acts as a ligand at nicotinic acetylcholine receptors (nAChRs), a family of ligand gated ion channels involved in th ree major circuits: (1) synapse on dopaminergic neurons in the ventral tegmental area, (2) widely projecting cholinergic neurons from the basal forebrain (nucleus b asalis of Meynert) involved in attention, and (3) fast acting excitatory post synaptic potentials in autonomic ganglia associated with Figure 4 2 . Diagrammatic illustration of the natural history of nicotine use disorder. We attempt to synthesize the neuroimaging literature of nicotine use disorder into four pharmacologically and behaviorally informed stages of nicotine use disorder. Acute nicotine exposure reflects the acute pharmacologic action on neural circuitry. Chronic nicotine exposure reflects pharmacologic dependency and results from repeated acute nicotine exposure; it manifests as maladaptive changes in reward, salience, and executive cont rol circuitry. Acute abstinence provides a model for understanding the neural basis of the nicotine withdrawal syndrome and craving. Long term abstinence serves as a model of neuroplastic recovery from nicotine use disorder. Relapse (bottom arrows) is the mechanism by which active disease is maintained. Associated neuroimaging findings are summarized in Table I. Stage 1: Stage 2: Stage 3: Stage 4:

PAGE 77

61 autonomic and visceral sensations (Nestler et al., 2015) . Through these molecular mechanisms, acute nicotine exposure affects circuits involved in arousal, reward, at tention, and autonomic regulation. Neuroimaging allows scientists to image in vivo the downstream effects of nicotine induced changes in these circuits. Table 4 1 . Summary of large scale brain network neuroimaging findings and associated role of the insula across different stages of nicotine use disorder. The neuropharmacologic mechanisms associated with each disease stage are listed in italics. Disease Stage Drug Induced Mechanism Neuroimaging Findings Role of the Insula Acute Nicotine Exposure Neural Pharmacodynamics Default mode network activity (Tanabe 2011, Sutherland 2015) ECN activity (Sutherland 2015) SN DMN ECN connectivity (Lerman 2014) Anterior insula activity (Sutherland 2015) Mediating a higher order representation of positive somatosensory, viscerosensory, and interoceptive sensations associated with drug reward Chronic Nicotine Exposure Pharmacologic Dependency Insulo cingulate connectivity Positively associated with: FTND, successful future abstinence Negatively associated with: incongruent errors on Stroop task, lifetime nicotine consumption (pack years), future relapse (Janes et al., 2010; Lin et al., 2017) Associated striatal adaptations, progressively ventral to dorsal reflecting habit formation New homeostatic set point for visceral sensations associated with drug reward Cigarette related memory retrieval (Janes et al., 2015b) Acute Abstinence Nicotine Withdrawal Syndrome SN DMN ECN connectivity (Lerman 2014) Granger causality from insula to other brain regions (Ding and Lee, 2013) Right Anterior Insula DMN connectivity associated with craving mag nitude (Moran Santa Maria et al., 2015) Representing the negative somatosensory, viscerosensory, and interoceptive sensations associated with cravings (Abdolahi et al., 2015a, b, 2017) Chronic Abstinence Neuroplastic Recovery Right anterior insular activation in response to cue exposure, associated with lifetime nicotine consump tion (Nestor et al., 2011; Zanchi et al., 2015) Salience Network Coherence (Zanchi et al., 2015) Representing the relative hyper saliency of drug cues for further monitoring and decision making Hyper saliency of drugs cues is associated with insular activity and is durable up to 1 year

PAGE 78

62 Acute nicotine exposure alters neural activation and connectivity patterns observed by fMRI in both healthy adults and individuals with nicotine use disorder. For example, nicotine administration relative to placebo in non smokers l ed to decreased default mode network activity (Hahn et al., 2007; Tanabe et al., 2011) and significantly increased local efficiency o f connectivity by whole brain graph theory analysis, particularly in right sided limbic and paralimbic areas (Wylie et al., 2012) . Sutherland and colleagues (Sutherland et al., 2015) conducted a large Activation Likelihood Estimation (ALE) meta analysis of acute effects of nicotinic agonists on brain activity changes in smokers as measured by fMRI or PET. The sample population included 796 participa nts spanning 77 different contrasts and experiments, including the resting state. Compared to placebo, nicotinic agonist administration was associated with decreased activity in the bilateral anterior insulae but mixed effects in the left middle insula. Ni cotinic agonists also resulted in significantly decreased activity within default mode network regions and increased activity in executive control network regions. When comparing vehicles of nicotine administration (pure nicotine pharmacologic administrati on versus cigarette smoking), decreased left middle insula activity was common by conjunction analysis to both manipulations, while decreased right anterior insula activity was specific to cigarette smoking compared to pharmacologic administration. Importa ntly, this study did not evaluate or control for effects of satiation versus abstinence and heterogeneity of tasks (including resting state), which confound interpretation. Relative to placebo, however, acute nicotine exposure overall consistently leads to decreased anterior insular activity, decreased default mode network activity, and increased executive control network activity. Stage 2: Chronic Nicotine Exposure in the Cigarette Sated State (Pharmacologic Dependency) Several studies have examined indiv iduals with a nicotine use disorder during a resting state fMRI scan. One study found that r educed circuit strength between the insula and d orsal

PAGE 79

63 anterior cingulate cortex , the two principal nodes of the salience network, was associated with increased addi ction severity (Moran et al., 2012) . These associati ons were observed when participants were scanned both after smoking or after acute abstinence , suggesting that decreased salience network coherence reflect s a chronic effect of nicotine use disorder rather than an acute pharmacologic effect. Salience netwo rk coherence has been consistently associated with severity of nicotine use disorder across studies (Bi et al., 2017; Lin et al., 2017; Moran et al., 2012; Wilcox et al., 2017; Zhou et al., 2017) . For example, Zhou and colleagues (Zhou et al., 2017) reported that reduced connectivity between the insula and anterior cingulate cortex was associated with increased nicotine use disorder severity . Li and colleagues (Lin et al., 2017) extended these findings , showing that reduced circuit strength between right insula and anterior cingulate cortex was associated with higher number of incongruent errors during a cognitive co ntrol task, implicating this circuit in top down cognitive control of saliency. More importantly, diminished circuit strength between the se regions was associated with greater lifetime nicotine consumption. Overall, these studies provide converging evidenc e that reduced salience network coherence at rest is a marker of chronic nicotine use and reflects addiction severity. Insular connectivity may also have prognostic importance related to vulnerability to relapse during future cessation attempts . For examp le, d ecreased connectivity between the insula and brain regions involved in cognitive control, including the dorsal anterior cingulate and dorsolateral prefrontal cortices , was associated with greater risk of future relapse after attempted cessation (Janes et al., 2010) , possibly reflecting a mechanism of reduced top down control. In another study , cir cuit strength between the insula and dorsal anterior cingulate cortex was significantly associated with enhanced smoking cue reactivity in areas involved in attention and motor planning , such as the right ventrolateral prefrontal cortex and dorsal striatum (Janes et al., 2015a) . Interestingly, the authors reported that insular anterior cingulate

PAGE 80

64 connectivity in smokers was durable over a one hour period and not associated with subjective craving or exhaled carbon monoxide, suggesting that increased salience network coherence may represent a chronic effect (i.e., a neural signature of hypersensitive cue reactivity in nicotine use disorder) . Recent, larger studies have corroborat ed these findings that salience coherence is important in mediating chronic effects of nicotine use disorder . Wilcox and colleagues studied 144 individuals with nicotine use disorder during the resting state and report ed that decreased circuit strength bet ween the insula and dorsal anterior cingulate was significantly correlated with higher cigarette consumption (Wilcox et al., 2017) . After controlling for addiction severity , increased circuit strength between these r egions was associated with greater likelihood of successful abstinence. Similarly, a 10 week longitudinal study (Addicott et al., 2015) found that increased insular conne ctivity to executive control and striatal regions was seen in non relapsers (i.e., successful abstainers) compared to relapsers . This suggests lower insular connectivity may be associated with relapse vulnerability. Together, these studies suggest that cir cuit strength between the insula and both (1) anterior cingulate, and (2) regions involved in cognitive control are not only markers of nicotine use disorder, but are also meaningful for prognosis, since it is associated with ability to quit smoking. Insul a activation during various tasks is also a potentially useful biomarker of nicotine use disorder. For example, Janes and colleagues . (Janes et al., 2017) stu died 23 smokers during a cessation attempt, 10 of whom remained abstinent during a two week follow up. Relative to successful abstainers, s mokers who relapsed demonstrated increased right insular activation in response to cigarette cues , suggesting that th is activation predicted likelihood of future use . In another study using a subsample of smokers from the Human Connectome Project dataset, individuals who smoked more cigarettes had greater right anterior insular activation in response to viewing faces exp ressing negative emotions such as anger (Dias et al., 2016) . These studies suggest that greater insula activation in response to both smoking cues and emotional

PAGE 81

65 cues may indicate a higher propensity for smoking and relapse. Neuroimaging smokers in nicotine withdrawal and experiencing cigarette cravings provides a p ossible mechanism for these observations. Stage 3: Acute Abstinence (Nicotine Withdrawal Syndrome) Acute abstinence in heavy smokers invariably causes the nicotine withdrawal syndrome , characterized by cigarette craving, hedonic dysregulation, cognitive di fficulties, and increased negative affect (Jackson et al., 2015) . Craving is a negatively reinforcing aspect of n icotine use disorder and is important for conferring relapse vulnerability (Ferguson and Shiffman, 2009) . In a longitudinal study of smokers during abstinence , the strength of urges to smoke showed an exponential decline over 12 months of abstinence (Ussher et al., 2013) . S ix months after cessation, 13% of ex smokers still reported strong urges but after 12 months, no ex smokers reported strong urges, although 34% reported some urges. Since the nicotine withdrawal syndrome and craving in particular are remarkably durable over time , the effects of acute abstinence on brain activity and connectivity cou ld provide insights into nicotine use disorder and its refractoriness to treatment . Several studies have examined how brain circuits are altered during acute withdrawal, suggesting altered large scale brain network dynamics between salience, executive cont rol, and default mode networks . For example, a within subject study of the effect of 24 hour abstinence compared to satiation on resting state connectivity in smokers demonstrated that abstinen ce compared to satiety was associated with weaker mutual inhibi tion between the default mode and salience networks (Lerman et al., 2014) . Weaker between network coupling predicted abstinence induced cravings to smoke and less suppression of default mode network activity during a working memory task (Lerman et al., 2014) . T he insula specifically may be involved in causing the altered network connectivity observed in withdrawal . To investigate this, Ding and colleagues (Ding and Lee, 2013) studied 21 heavy smokers in cigarette sated and abstinent

PAGE 82

66 conditions . A fter smoking replenishment , directed connectivity from salience network to default mode network was significantly reduced and directed connectivity f rom both executive control and default mode networks to the salience network was enhanced. Moreover, the insula showed significantly increased directed connectivity with salience, default mode, and executive control regions in cigarette abstinence compared to satiation. This suggests that directed information flow from the insula to other brain regions is increased in abstinent compared to sated heavy smokers , possibly reflecting increased signaling of withdrawal symptoms and craving . Moran Santa Maria and colleagues (Moran Santa Maria et al., 2015) studied acutely abstinent smokers using a n fMRI visual craving cue task. Psychophysiologic interaction with a seed in the right anterior insula was used to infer directed connectivity. Results demonstrated significantly greater effective connectivity from the right anterior insula to the bilateral precuneus , a key node of the default mode network, during smoking compared to neutral cues. Insula to precuneus effective connectivity showed a significant positive correlation with craving magnitude, providing further evidence that this circuit between salience and default mode networks play s an important role in cue induced craving. Causal effects of cigarette cues on brain function during acute withdrawal were also investigated by Claus and colleagues (Claus et al., 2013) . The investigators studied neural responses to cigarette cues in interaction centered on a left dorsal anterior insular seed. Results suggested that smoking cues compared to neutral cues caused stronger connectivity between the left insula and multiple nodes , including right insula (anterior and posterior), amygdala, somatosensory cortex, orbitofrontal cortex , and striatum. In addition, during smoking video exposure significant positive correlations were observed between insula activity and dependence severity ; and again, salience network coherence was associated with dependence severity . The authors speculate d that the anterior insula may contribute to the initial evaluation of cigarette cue

PAGE 83

67 value, interoceptive processing of withdrawal symptoms, and engagement of motor circuits in preparation for drug seeking behavior. Drug expectancy , or prior belief s about impending acute nicotine administration , has been shown to be a factor that modulates the effects of acute withdrawal. Gu and colleagues (Gu et al., 2016) studied 24 overnight abstinent smokers who performed a sequential reward learning task immediately after a cigarette smoking intervention. S mokers received either a 0.6 mg nicotine cigarette or a de nicotinized cigarette and were either told that the cigarette Only w hen smokers received a cigarette with nicotine and were t old that it contained nicotine, significant activation in the ventral anterior insula was observed during a reward learning task , which was positively correlated with craving magnitude . This suggests that the anterior insula is not only involved in interoc eptive processing, but that anterior insula processing of craving and reinforcement learning is modulated by drug expectancy, presumably through top down cognitive influences. Stage 4: Chronic Abstinence (Neuroplastic Recovery) It is unclear whether neura l function returns to healthy levels following long term abstinence, or if the differences associated with nicotine use disorder are durable even after years of abstinence. Few studies have investigated the neuroimaging correlates of long term abstinence i n nicotine use disorder . This is unfortunate, because although chronic nicotine exposure results in upregulation of nicotinic acetylcholine receptors throughout the brain (Breese et al., 1997; Gentry and Lukas, 2002) , former smokers exhibit nicotinic acetylcholine receptors concentrations similar to non smokers (Breese et al., 1997) , suggesting that pathologic u pregulation is reversible. Similar evidence of neuroplastic recovery is suggested by behavioral changes during chronic abstinence. Measures of impulsivity have been shown to be abnormally elevated in active smokers, but former smokers show levels similar t o never -

PAGE 84

68 smokers (Bickel et al., 1999) . Studies of ex smokers thus may provide important insights into the successful mainten ance of smoking cessation. Despite limited literature, several small studies have examined insula connectivity dynamics in chronic abstinence. Zanchi and colleagues (Zanchi et al., 2015) studied non smokers, active smokers, and ex smokers during a craving cue task fMRI scan and reported several findings supporting insular role in recovery . First, ex smokers with greater right anterior insular activity in response to cigarette cue s also had higher life time nicotine consumption. Second, ex smokers demonstrated decreased circuit strength between the right anterior insula and anterior cingulate compared to non smokers, but no significant difference was observed between ex smokers and active smokers. This s uggests that insular function may not completely recover in long term abstinence, possibly reflecting a mechanism of persistent craving. Another fMRI study (Nestor et al., 2011) of smokers, ex smokers, and healthy controls us ed a n attentional bias paradigm with neutral cues, emotionally evocative c ues, and smoking cues . Across all cue conditions, ex smokers exhibited significantly greater activation in the right anterior insula compared to active smokers and controls. In a separate experiment employing a go/no go paradigm to investigat e motor respon se inhibition and cognitive error monitoring, the ex smokers had significantly greater error related activation than both controls and smokers in the left insula. Taken together, these results suggest that heightened insular monitoring of cues and errors c ontribute to the successful maintenance of abstinence. Higher right anterior insular activity in ex smokers compared to healthy controls may reflect a hypervigilance against smoking cues necessary for successful long term abstinence. 4.3.3. Putting it All Togethe r: Unified Models of the Role of the Insula in Nicotine Use Disorder Pathogenesis important role in higher cognitive function in normal, non addicted persons. One of t he primary

PAGE 85

69 functions of the anterior insula is salience detection (Seeley et al., 2007) , such as identifying stimulus features that stand out or are of instinctual or learned importance . Saliency involves the selection of stimuli from a continuous stream of internal and external sensory inputs for additional processing. Another theory is that the anterior insula serves as the apex of a predict ive cortical hierarchy that spans all sensory systems (Barrett and Simmons, 2015; Chanes and Barrett, 2016) , selecting goal relevant stimuli for attention and cognitive processing. The insula is unique amongst cortical areas in that it contains sequential yet overlapping maps from all exter oceptive and interoceptive senses (Craig, 2009, 2011) . These higher order maps are successively re represented from posterior insula to middle insula to anterior insula, progressively acquiring additional sensory input maps, interoceptive signals, and rewa rd signals along the way. The anterior insula then provides a single cortical representation of how an individual is feeling at a given time: (Craig, 2009) (Craig, 2011) . Despite these slightly differing models and interp retations, the evidence suggests broad involvement of the anterior insula in sensitivity control mechanisms of salience. Normal insular physiology thus provides a framewor k for understanding insular pathophysiology in nicotine use disorders. disorder. One hypothesis centers on the salience network, comprised of the insula and anterior cingula te cortex. The insula is believed to serve as a toggle, directing brain function towards internal or external stimuli, in order to maintain homeostasis of cognitive resources and guide goal directed behavior (Sutherland et al., 2012) . Internal focus is reflected by greater default mode network (endogenous oriented) activity, whereas external focus is reflected by greater

PAGE 86

70 exe cutive control network (exogenous oriented) activity ( Figure 4 3 toggling between these two networks is hypothesized to be usurped in nicotine use disorder. The concept that the insula directs attention towards the most homeostatically relevan t stimuli internal or external provides a neurobiological model to explain both cognitive changes and functional connectivity findings of acute nicotine ingestion, nicotine satiety in dependence, and nicotine withdrawal syndrome. This review of the evidence suggests that the critical role of the insula in maintaining nicotine use disorder is related to its function in providing conscious awareness of craving and withdrawal symptoms. Lerman and colleagues (Lerman et al., 2014) provided supporting evidence for this model, reporting decreased Figure 4 3 . Differential resting state connectivity of the dorsal (left) and ventral (right) right anterior insula, using the Human Connectome Project Connectome Workbench (n = 1206), uncorrected. Dorsal right anterior insula is strong ly connected with frontoparietal regions involved in executive control; in contrast, ventral right anterior insula connectivity is strongly connected with default mode network regions involved in internal feelings and self referential processing. Note tha t the externally directed system (dorsal anterior insula and executive control network) and internally directed system (ventral anterior insula and default mode network) are inversely correlated, consistent with mutual inhibition (see Figure 4 1 ).

PAGE 87

71 between network coherence amongst salience, default mode, and executive control networks in abstinence compared to satiety. They reported that weaker between network coupling predicted abstinence induced cravings and less suppression of default mode activity during performance of a subsequent working memory task, possibly reflecting a mechanism of cognitive and attentive impairments commonly observed during the nicotine withdrawal syndrome. In summary, the evidence suggests that insular function is disrupted compared to healthy controls across all stages of nicotine use disorder ( Table 4 1 ). Salience network coherence between insular and anterior cingula te nodes is particularly important for craving induced behaviors, reflects disease severity, and has prognostic value. However, large scale longitudinal studies are needed to understand the altered connectivity profiles of the insula with salience, default mode, and executive control regions at different disease stages. Although the evidence is still largely comprised of single site, small population studies, it provides a compelling neurobiological argument for future work. In the following section, we exp lore in detail the possibility of targeting the insula with neuromodulation as a therapeutic strategy for promoting abstinence. 4.4. Implications of Insular Role in Nicotine Use Disorder on Neuromodulatory Therapeutic Development While pharmacotherapy can be u sed with some efficacy to diffusely modulate dysregulated brain circuits with the goal of promoting abstinence, new treatment strategies clearly are needed. One possibility is a targeted circuit node approach, aligned with current understanding of the und erlying pathology. Such a neurocircuit based approach may improve successful cessation by intervening upon or modulating specific neuroanatomical structures that serve as key nodes within behaviorally relevant circuits, such as the insula in the case of ni cotine use disorder. A promising candidate for this approach is neuromodulation.

PAGE 88

72 4.4.1. Therapeutic Neuromodulation in Animal Models of Nicotine Use Disorder Aside from systemic pharmacologic approaches, animal models have shown significant benefit of targeted ne uromodulation . For example, one study (Forget et al., 2010) reported that insular inactivation via intracranial GABA agonist microinfusion in nicotine dependent rats significantly reduced nicotine motivation, nicotine seeking , and nicotine taking behaviors, with no effect on food behaviors. These findings were further confirmed using an alternative lesion ing method , bilateral insular deep brain stimulation , in a rat model of nicotine dependence (Pushparaj et al., 2013) . Kutlu an d colleagues (Kutlu et al., 2013) extended these findings by showing that locally infused D 1 but not D 2 antagonists into the rostral anterior administration acutely by more than 50%, with repea ted D 1 antagonist infusions resulting in continued decreases in consumption without evidence of tolerance. The cause and effect relationship between decreased D 1 activity in the insula and decreased nicotine self administration suggests that mesocorticolim bic dopaminergic afferents onto the anterior insula are critically involved in promoting and maintaining nicotine dependence (Kutlu et al., 2013) . Disrupting this insular mechanism leads to diminished nicotine consumption, possibly throug h diminished interoception of reward (or lack of reward) signals. 4.4.2. Therapeutic Neuromodulation in Humans with Nicotine Use Disorder T ranscranial magnetic stimulation (TMS) is a neuromodulation technique that shows promise as a means to target insula function. TMS is a noninvasive intervention in which extracorporeal current carrying electrical coils are used to induce rapid, transient, focal magnetic fields targeting a specific brain region. The se transient magnetic field fluxes cause electromagnetic induction in underlying neural tissues that alter neural transmembrane potentials and in turn affect neural activity. Applying a sequence of TMS pulses causes long term effects that either facilitate or inhibit neuronal excitability, depending upon multipl e

PAGE 89

73 factors , including pulse parameters and stimulation site. Based on studies of the corticospinal motor tract, low repetitive TMS is inhibitory and high repetitive TMS is faciliatory, with aftereffects closely parallel ing long term depress ion and long term potentiation mechanisms of neuroplasticity (Hoogendam et al., 2010) . Since 2003, several studies in the English language literature have investigated the role of high or low frequency TMS targeting the dorsolateral prefrontal cortex in cigarette craving mitigation (Salling and Martinez, 2016) . The focus in the literature on high frequency TMS of the dorsolateral prefrontal cortex in smokers is likely related to its demonstrated efficacy in major depressive disorder (Brunoni et al., 2017; Milev et al., 2016) and relat ive accessibility of this region as a superficial target site compared to other, deeper structures. Overall , the se studies demonstrate that both single session and repeated sessions of TMS to the dorsolateral prefrontal cortex reduces cigarette craving, although in some studies the cigarette consumption effects were mixed, highlighting the need for objective endpoints and response measures (as opposed to self reported craving measures) in clinical trials. Given the efficacy of dorsolateral prefrontal cortex neuromodulation in promoting smoking cessation, it stands to reason that other cortical areas involved in nicotine use disorder, such as the insula, may be useful therapeutic targets. Enhancing top down control of craving rel ated behaviors has been shown effective. However, this is arguably an indirect method by augmenting top down control through excitation of dorsolateral prefrontal cortex, this approach is thought to enhance suppression of bottom up craving urges. Alterna tively, inhibiting the bottom up craving urges from the cortical source itself may be a more direct and effective treatment. H uman stroke and animal neuromodulation studies reviewed here implicate a crucial cause and effect role of the insula in maintainin g bottom up craving sensations and nicotine consuming behaviors. In chapter 5, we report the main results of a randomized sham controlled clinical trial ( www.ClinicalTrials.gov identifier: NCT 02590640 ) in

PAGE 90

74 acti ve smokers to investigate the efficacy of insular inhibitory neuromodulation in directly reducing cigarette craving at the presumed neural source . One of the major technical limitations in TMS is the limited spatial depth of electromagnetic induction and i ts inverse relationship with focality. Theoretical simulations have demonstrated that commercially available figure of eight coils can penetrate the cortex 1.0 to 3.5 cm within normal safety parameters (Deng et al., 2013) . Across different coil geometries, stimulation of deeper brain targets necessitates greater electrical field spread ( reduced focality ) . This tradeoff between electric field depth and focality pose s important physical challenges to stimulating the anterior insula. Moreover, non target brain stimulation further complicates investigation of behavioral or clinical outcomes associated with deep brain targeting, because adjacent and superficial areas are included in the treatment field and thus may confound observed associations. For example, one study applied continuous TMS to the right anterior insular cortex and control regions (occipital and somatosensory cortices) in healthy volunteers using a superficial (i.e., planar figure of eight ) coil (Pollatos et al., 2016) . Their results suggested that inhibiting the right anterior insula was associated with a significant decrease in cardiac and respiratory interoceptive accuracy (measured by a heartbeat counting task) as well as decreased perceptual confidence. There is debate about the targetability of the anterior insula using conventional superficial coils ( Figure 4 4 ), however, with investigators noting that by using this approach the anterior insula receives about 25% of the maximum cortical energy deposition and greatest deposition in the overlying frontal and temporal opercula (Coll et al., 2017; Pollatos and Kammer, 2017) . Recently, a family of coil designs called Hesed (H) coils have been developed to achieve deep brain neuromodulation at the expense of a wide, relatively non focal treatment field.

PAGE 91

75 These coils provide near complete stimulation of the frontal lobes. Dinur Klein et al (Dinur Klein et al., 2014) r andomized a large sample of 115 heavy smokers to 13 daily treatments of high frequency, low frequency, or sham TMS using an H coil designed to target the bilateral ventrolateral prefrontal cortex, dorsolateral prefrontal cortex, and insula. Smoking was mea report and urine cotinine levels. High frequency TMS, but not low frequency or sham TMS, during the presentation of visual smoking cues resulted in a 44% reduction in smoking at 3 months and 33% reduction at 6 months. Counter in tuitively, there was no significant difference in self reported craving, suggesting that these effects may reflect Figure 4 4 . Finite elemen t model results of the predicted electromagnetic field produced by a standard superficial planar transcranial magnetic stimulation 70mm figure of eight coil targeting the right anterior insula. Predicted current flux density (left) and normalized absolute value of the electric field (right) illustrating the pattern of energy deposition. Maximum energy is deposited in the scalp, superficial soft tissues, and cerebrospinal fluid due to high tissue conductivities. This illustrates the difficulty in targeting d eep structures, such as the insula or anterior cingulate cortex. This is consistent with model results of insular targeting reported by Pollatos and colleagues (Pollatos and Kammer, 2017) . Created using SimNIBS version 2.1.1 with right anterior insular target [36, 10, 6] in MNI space and default parameters (Thielscher et al., 2015) .

PAGE 92

76 enhanced cognitive control rather than reduced incentive salience or reduced sensation of cigarette craving. This suggests that their finding s of decreased cigarette consumption after high frequency stimulation may reflect enhanced dorsolateral prefrontal cortex activity and top down cognitive control. Malik S et al (Malik et al., 2018) appl ied excitatory and inhibitory TMS to the bilateral insula and surrounding cortical opercula using the H coil in eight healthy participants in a within subject, crossover, blinded, sham controlled, proof of concept study. Synaptic effects were measured usin g PET with a dopamine agonist tracer. They demonstrated that inhibitory (1 Hz) TMS compared to sham and excitatory (10 Hz) TMS significantly decreased dopamine concentrations in the substantia nigra and sensorimotor striatum, with a trend towards significa nce in the associative striatum. In both these studies claiming to modulate the insula, the investigators could not definitively confirm that the insular cortex is indeed being stimulated, although future studies using fMRI could address this question. Neu romodulatory methods not only include TMS, but also include transcranial direct current stimulation (tDCS) and deep brain stimulation (DBS). tDCS involves applying an electrical current to the brain between two electrodes, which affects neural tissues with in the path of least electrical resistance. tDCS has been used to target the dorsolateral prefrontal cortex in smokers with reduction in cue induced cravings; however, the non focal nature of this method limits its utility in targeted, neuroanatomically dr iven neuromodulation (Salling and Martinez, 2016) . DBS, on the other hand, involves surgically implanting a stimulating electrode into target brain tissue. DBS targeting the ventral striatum in smokers has been reported in only one study, which re ported higher rates of successful cessation compared to unaided smoking cessation in the general population (Kuhn et al., 2009) . However, surgically placed deep brain stimulation is invasive and practically limited by significant ethical considerations. In summary, applications of non invasive methods of brain stimulation in nicotine addiction are currently limited by their lack of spatial specificity and depth of targetability.

PAGE 93

77 While dorsolateral prefrontal cortex neuromodulation has been shown to improv e abstinence rates and mitigate cravings, other cortical areas such as the anterior insula have broader empirical support and may result in stronger treatment responses. Application of TMS to nicotine use disorder is promising but future studies are needed to define optimal targets, paradigms, and patient population. There is a clear clinical need for better smoking cessation treatments, and growing evidence specifically implicates the insula as a rational neuroanatomical target for investigational modulati on therapies. 4.5. Conclusion The insula is functionally heterogeneous, with distinct patterns of connectivity with large scale brain networks associated with numerous functions and behaviors. Animal models and human lesion studies suggest that the insula is necessary for the maintenance of nicotine seeking behaviors and nicotine taking behaviors, likely through nicotine craving. Given the limited efficacy of standard of care treatments for nicotine use disorder, neuromodulation of this region may contribute to the next generation of cessation treatments by offering what henceforth has not been available: a targeted, neuroanatomically driven approach to smoking cessation therapy. Moreover, because substance use disorders in general including nicotine use dis order are thought to be initiated and reinforced by maladaptive alterations in the dopamine reward system and associated corticolimbic and cognitive control circuits, such a targeted neuroanatomically driven approach may advance treatments for other drug addictions as well. There is a clear clinical need for better smoking cessation treatments. Evidence strongly implicates the insula as a rational neuroanatomical target for modulation therapies. Neuromodulation of insula function has significant potential to improve smoking cessation rates amongst smokers, but continued technical developments and research are needed to overcome challenges in depth and specificity of targeting.

PAGE 94

78 4.6. Acknowledgements Several co authors contributed to this chapter: Jason Tregella s PhD, Benzi Kluger MD MS, Korey Wylie MD, Joshua Gowin PhD, and Jody Tanabe MD. This work was funded by NIH F32 DA041011 (MFR) , RSNA RR 1620 (MFR) , and CCTSI M 15 81 (MFR).

PAGE 95

79 CHAPTER V 5. INSULAR INHIBITORY N EUROMODULATION IN SM OKERS DECREASES CIGARETTE CRAVINGS AND BRAIN R ESPONSES TO CIGARETT E CUES: A RANDOMIZED CONTROLLED TRIAL 5.1. Abstract Cigarette addiction is a leading preventable cause of mortality, morbidity, and healthcare costs. Several lines of evidence suggest that the insula contributes to urges to smoke, and that inhibiting the insula can disrupt cigarette addiction. However, no studies to date have attempted therapeutic insular modulation in people. We hypothesized low frequency repetitive transcranial magnetic stimulation (LF rTMS) ta rgeting the right anterior insula would decrease cigarette cravings and brain responses to cigarette cues. We conducted a randomized, singled blinded, sham controlled, phase I clinical trial in which active smokers (n=40) interested in quitting received a single session of either right anterior insula deep LF rTMS (n=20) or sham treatment (n=20). Primary outcomes included craving measures and cigarette craving cue task fMRI (3T) brain activity responses, measured before and after treatment. Finite element m odel simulations of energy deposition intracranially informed a priori selection of regions of interest used to corroborate whole brain results. Compared to sham treatment, right insula LF rTMS reduced self reported cigarette craving (p=0.033). Right insul a LF rTMS also decreased brain activity responses to visual cigarette cues at the whole brain level in primary sensorimotor cortices, supplementary motor area, and right anterior insula (p < 0.001, pvoxel < 0.005, k > 464 voxels). There were no brain regio ns in which LF rTMS caused increased activity response to cigarette cues. A single session of right anterior insula deep LF rTMS reduced cigarette

PAGE 96

80 cravings and brain activity in response to cigarette cues. These findings provide proof of concept of a poten tial neuroanatomical target for smoking cessation therapy. 5.2. Introduction Cigarette addiction is a leading preventable cause of premature death, morbidity, and healthcare costs (Health and Services, 2014) . Each year, approximately 480,000 people in the U.S. prematurely die from smoking related illnesses. Overall, 1 in 5 U.S. deaths are caused by smoking (Health and Services, 2014) . It is estimated that 46 million (~1 in 5) Americans smoke. About half of these smokers at tempt to quit each year, with 70% of smokers wanting to quit (Babb, 2017) . Unfortunately, standard of care smoking cessation treatments are largely ineffective. A pproximately 80% of patients attempting to quit relaps e within six months , d espite combined pharmac ologic and behavioral therapies (Tobacco, 2008) . Craving is a defining feature of nicotine use disorders and predicts relapse (Potvin et al., 2015; Saunders and Robinson, 2013) . For these reasons, exploring novel treatments to reduce craving remains an important objective. Evidence has implicated the insula in the maintenance of cigarette craving and smoking behaviors. Naqvi and colleagues (Naqvi et al., 2007) reported that a significant proportion of patients with damage to the insula compared to other brain areas were able to stop smoking easily without cravings or relapse. Subsequent prospective studies confirmed that insular lesions in smokers strongly predicted spontaneous continuous abstinence, fe wer nicotine withdrawal symptoms, and reduced cravings (Abdolahi et al., 2015 a, b, 2017; Suner Soler et al., 2012) . This pattern was also observed in animal experiments using pharmacologic and electrical lesioning of the insula (Forget et al., 2010; Kutlu et al., 2013; Pushparaj et al., 2013) . For example, insular lesioning via intracranial GABA agonist microinfusion in nicotine dependent rats significantly reduced nicotine motivation , nicotine seeking , and nicotine taking behaviors, with no effect on food behaviors (Forget et al., 2010) . However, to date no

PAGE 97

81 in vivo to investigate possible therapeutic benefits. One promising approach to modulate insular function is through t ransc ranial magnetic stimulation (TMS). TMS is a non invasive neuromodulation technique that has recently demonstrated early phase success in promoting smoking cessation. S tudies o f the corticospinal motor tract have shown that low frequency repetitive TMS ( Hz , LF rTMS) is inhibitory and high frequency repetitive TMS ( , HF rTMS) is facili t atory, with aftereffects closely paralleling long term depress ion and long term potentiation mechanisms of neuroplasticity, respectively (Hoogendam et al., 2010) . While no studies to date have investigated insular TMS in smokers, several studies have investigated effects of LF and HF rTMS targeting the dorsolateral prefrontal cortex (DLPFC) o n cigarette craving and consumption (for reviews, see Kedzior et al. (2018); Makani et al. (2017); Salling and Martinez (2016); Song et al. (2018) ). Multiple studies demonstrate that excitatory HF rTMS to the DLPFC reduce s cigarette craving . The mechanism is thought to involve are currently no human studies , however, that investigate disrupting this s source of craving itself . Similar bottom up approach has been attempted in cocaine users by targeting cortical reward centers in the orbitofrontal cortex (Hanlon et al., 2017) . Hanlon et al. reported that left frontopolar TMS delivered in a sin gle day in cocaine addicts significantly decrease d TMS evoked BOLD signal in the orbitofrontal cortex and insula. Only one study attempted insular modulation in smokers by targeting the bilateral DLPFC using an H coil (Dinur Klein e t al., 2014) , but the large treatment field precluded distinguishing effects of insular from overlying DLPFC stimulation. A small number of studies have attempted to modulate posterior superior insular function in healthy participants, reporting changes in

PAGE 98

82 thermal/pai n sensation (Ciampi de Andrade et al., 2012; Lenoir et al., 2018) and interoceptive tasks such as a heartbeat counting (Pollatos et al., 2016) , although t he targetability of the insula is disputed (Coll et al., 2017) . Spagnolo et al. targeted the bilateral prefrontal cortices and insulae using the H coil in healthy individuals and reported no measurable effects on either a blink suppression task or a forced choice risk taking task (Spagnolo et al., 2018) . No studies to date have selectively targeted the insula in smokers. We sought to address this gap in the literature with the current study. We conducted a randomized, singled blinded, sham controlled, parallel group phase I clinical trial ( www.ClinicalTrials.gov identifier: NCT 02590640 ) to answer the question: does right insular inhibition in smokers reduce cigarette craving and brain response s to cigarette cues ? This proof of concept study hypothesized that a single session of LF rTMS (inhibitory) targeting the right anterior insula in active smokers would acutely decrease cigarette craving and alter brain activation in response to cigarette smoking cues. 5.3. Materials and Methods This study was reviewed and approved by the Colorado Multiple Institutional Review Board in accordance with the Declaration of Helsinki. All participants provided written informed consent. Recruitment, enrollment, data collection, treatments, and MRI examinations were c onducted at the University of Colorado Anschutz Medical Campus. All investigators except the physician administering treatments (MFR) were blinded to group assignment until completion of blinded analysis of primary outcomes, defined as self reported cravin g and brain activity responses to cigarette cues. 5.3.1. Study Design This was a randomized, singled blinded, sham controlled, parallel group, clinical trial in which active smokers interested in quitting received either right anterior insula deep LF rTMS

PAGE 99

83 or sham treatment. Randomization was performed using a computerized random number generator ( http://www.random.org ) in blocks of two. A physician trained in TMS and seizure management administered all treatments. Participants were blinded to group assignment and asked if they believed they received real or sham treatment, with their responses recorded. This study was subject to a protocol deviation. In four participants originally randomized to LF rTMS, there was insufficient time to measure RMT and apply LF rTMS between scheduled MRI exams. This timing issue was due to random factors unrelated to the scheduled immediately after the current stu treatment MRI. Subsequent logistical changes to the scheduling of participants obviated this issue (i.e., participants were scheduled in evenings such that there were no exams for other research studies scheduled after the tre atment MRI, allowing for greater flexibility in timing). 5.3.2. Sample Population Participants were recruited through internet advertisements, publicly posted flyers, and tobacco cessation consultations at the University of Colorado Hospital Emergency Department and Inpatient Services. Ninety one hea lthy, right handed, treatment seeking, full study appointments ( Figure 5 1 ). Telephone screening included brief medical, substance use, and psychiatric histories to determine eligibility before scheduling the study appointment. Exclusion criteria included [1] use of non cigarette tobacco product s; [2] current use of nicotine replacement therapy, bupropion, or varenicline; [3] major medical disorders; [4] current pregnancy or pregnancy seeking; [5] active abuse or dependence of illicit substances; [6] MRI or TMS contraindications; and [7] self rep orted major psychiatric disorder. Eligible participants were invited for study appointments. Participants were instructed

PAGE 100

84 appointment, confirmed by exhaled carbon monoxide concentration ([CO] Exhaled ). Brief abstinence was imposed to establish a baseline state of craving and promote susceptibility to smoking related visual cues. At the beginning of the study appointment, each participant provided medical, surgical, and social histories and un derwent a brief neurologic examination by a physician ( Figure 5 2 ) . influence of alcohol. [CO] Exhaled was measured with a MicroSmokelyzer (Bedfont Scientific; Kent, Unit ed Kingdom) to confirm acute abstinence ([CO] Exhaled Figure 5 1 . CONSORT enrollment diagram for this phase 1 human trial. Careful attention was .

PAGE 101

85 collected for semi quantitative urine cotinine testing and urine toxicology screening for common drugs of abuse. 5.3.3. Behavioral Measures Each participant completed standardized surveys including the Fagerström Test for Nicotine Dependence (FTND), Wisconsi n Inventory of Smoking Dependence Motivations (WISDM), Barratt Impulsiveness Scale, and Behavioral Approach/Inhibition Scales (BAS/BIS). All subjects completed craving assessments using the Questionnaire of Smoking Urges Brief (QSU B, 10 items) immediately before the pre and post treatment MRI examinations. 5.3.4. MRI Examination MRI exams were acquired before and after treatment at the University of Colorado Brain Imaging Center using a Siemens 3 Tesla Magnetom Skyra scanner (Siemens AG; Munich, Germany) and 20 channel neurovascular coil . Structural images included a T1 weighted 3D magnetization prepared rapid gradient multi echo sequence (MPRAGE; sagittal plane acquisition; repetition time [TR] = 2300 ms; echo time [TE] = 2.24 ms; inversion time [TI] = 900 Figure 5 2 . Timeline of research appointment. Treatment included either rigorous sham or deep LF rTMS targeting the right anterior insula. The time delay between the end of treatment and behavioral survey wa s <5 minutes. The time delay between the end of treatment and post treatment MRI was 11.5 ± 1.9 minutes (mean ± SD).

PAGE 102

86 ms; e cho train length [ETL] = 250 ms; flip angle = 8°; 1 mm slice thickness, 176 slices; FOV = 220 mm with 256 × 256 matrix; time = 5:21 min). Functional images were acquired with a T2 weighted echo planar gradient echo sequence (GE EPI FID; axial oblique plan e acquisition; echo time [TE] = 28 ms; repetition time [TR] = 2000 ms; flip angle = 70°; slice thickness = 3mm with 1mm gap, 32 slices; FOV = 220 mm with 64 × 64 matrix; acquisition time = 6:04 min). Head vacuum cushion (Par Scientific A/S, Odense, Denmark). Pre and post treatment MRI exams also included resting state fMRI, ASL, and MR spectroscopy sequences, reported separately ( Figure 5 2 ) . 5.3.5. Cigarette Craving Cue fMRI Task fMRIs were acquired while participants viewed visual stimuli using fiberoptic binocular ach run to achieve spin history field homogeneity. Stimuli included pseudorandomized , temporally jittered blocks of smoking related and neutral related cues, separated by cross hair fixation to improve fitting of generalized linear models. S moking cue image s included the heads and mouths of people smoking, people lighting a cigarette, hands holding a cigarette , and cigarettes in ashtrays or in pack s . Neutral images were matched to smoking images with regards to color, complexity, and form; presence and numbe r of faces and body parts; presence and type of places; and presence and number of tools (cigarettes and cigarette packs were considered tools). This matching controlled for activation of specialized cortical areas involved in cognitive processing of faces , body parts, places, tools, and numbers/counting. Each block lasted 18 seconds and consisted of four stimuli shown for 4.5 seconds each. Participants were asked to lie quietly while viewing images and imagine themselves in the specific situations portraye d.

PAGE 103

87 5.3.6. Determining Resting Motor Threshold Resting motor threshold (RMT) was determined by single pulse TMS of the left paracentral lobule (lower extremity motor cortex). The left paracentral lobule was chosen as the motor target because of its greater depth c ompared to other primary motor cortex targets, comparable to the insula. The coil was positioned over the left paracentral lobule and adjusted the target location until each single pulse TMS triggered muscular contraction. TMS power was then decreased to d efine resting motor threshold (RMT) defined as the minimal amplitude required to generate at least 4 of 8 stimulations with motor activity with right tibialis anterior electromyography response amplitude greater than 0.5 mV. All participants, including tho se in the sham TMS treatment arm, underwent RMT determination. Four subjects in the treatment group and four subjects in the sham group did not have a reliably measured RMT due to insufficient time between scheduled MRI examinations. This timing issue was occurred only when there was another research MRI exam scheduled immediately after the p hysiology also contributed (e.g., participant using the restroom or phone after the first MRI, delay in T1W image transfer). Subsequent scheduling changes described above obviated this issue. 5.3.7. Right Insular Deep Low Frequency Repetitive Transcranial Magneti c Stimulation weighted images were loaded into the BrainSight computer system (Rogue Resolutions; Cardiff, United Kingdom) allowing for precise cortical targeting using real time infrared based stereotactic intracranial navigation. Scalp and pial surface digital reconstructions were created for each participant. Pointer and coil positions were visualized continuously in real time relative to surface reconstructions and raw multiplanar T1 weighted images throughout the treatment (displayed on a monitor visible to participant and physician).

PAGE 104

88 Participants and investigators, except the physician who applied treatments, were blinded to the treatment arm. Right insula LF rTMS was administered to the right anterior insular second gyrus brevus usin g a custom angulated, 80 mm double figure of eight coil manufactured by Magstim Ltd (Whitland, Camarthenshire, UK). Targets were selected based on visual weighted images by a radiologist, and individual target coordinates were recorded in MNI space. Target variability during the treatment was permitted up to ± 3 mm in any direction. Treatment field trajectories were orthogonal relative to surface. Due to uncomfortable jaw clenching in three subjects, the trajectory angle was manipulated up to 30° relative to the scalp surface orthogonal to mitigate treatment field engagement of the temporalis muscle. LF rTMS included a single 25 min train of 1500 total 1 Hz pulses at 90% of RMT u sing the described target and trajectory. All participants completed a post treatment written survey to [1] document whether they believed they received real or sham treatment, and [2] obtain a post treatment QSU Brief craving assessment. 5.3.8. Sham Treatment Sh am TMS used a custom built sham coil designed to look, sound, and feel identical to real treatment. To mimic LF rTMS, cutaneous electrodes were placed on participants and connected to an electrical amplifier controller. Cutaneous current was administered f or 25 min at 1 Hz to trigger facial muscle contractions and skin sensations. All participants, including those in the LF rTMS treatment arm, had skin electrodes placed on the scalp to ensure uniformity of participant experience; however, only those in the sham group received current. All participants completed post treatment written surveys. 5.3.9. Blinded Analysis Upon completion of study enrollment, blinded analysis was performed of craving and fMRI task results. Because only single blinding could be applied dur ing data collection (i.e., the

PAGE 105

89 physician applying LF rTMS versus sham was unblinded by necessity), a second blind was applied at the analysis stage to insulate results from potential investigator bias. Blinding of participant labels was applied by a biosta tistician uninvolved in the study to this point (AJ); the blinding label key was stored on a secure server to which none of the co investigators had access. Blinded analyses included: [1] t wo way repeated measure s analysis of variance ( [LF rTMS > sham] × [ post treatment > pre treatment]) performed on QSU Brief craving measures, [2] visual inspection of motion and nuisance variance correction of fMRI on a subject by subject basis, [3] main effect of cigarette cues on fMRI activation across groups (cigarette cue > neutral cue contrasts, pre treatment scans only, one sample t test), and [4] interaction effect of group by time on fMRI activity task contrasts ([cigarette cue > neutral cue] × [post treatment > pre treatment] × [blinded group A > blinded group B]) . The results of these blinded analyses were reviewed with senior co authors and documented (MFR, JRT, JLT) before unblinding group labels and analyzing all data. 5.3.10. General Data Analysis After unblinding, all non imaging data collected on participants was digitally stored in a secure database. Demographic, psychometric, and behavioral comparisons using Satterthwaite two sample t tests , Cochran Mantel Haenszel tests were calculated using SAS based JMP Pro 14.1.0. A t wo way repeated measure s analysis of variances (sham versus LF rTMS × pre versus post treatment) w as performed on QSU Brief craving regarding group assignment with actual group assignment. Because this proof of concept clinical trial was intended to evaluate efficacy rather than effectiveness, we performed per protocol or as treated analysis (i.e., answering t he question unless explicitly stated otherwise. Secondary analyses included both intention to treat and

PAGE 106

90 at is the effect of assignment to a given Detailed justification of this approach is provided in Section 5.8 Supplement: Justification of Analysis Approach for Primary and Secondary Outcomes , page 109 . 5.3.11. Neuroimaging S ignal Pre Processing Structural and functional images were pre processed using MATLAB 2017 and SPM12 software (Wellcome Trust Centre for Neuroimaging ; London, UK) . T1 weighted images were segmented and normalized to MNI space. BOLD pre processing included: [1] removal of first four TRs (at acquisition), [2] slice timing correction, [3] rigid realignment of BOLD images to the first TR, [4] motion scrubbing/censoring (Power et al., 2014) , [5] non neural noise correction using aCompCor (Behzadi et al., 2007; Musc helli et al., 2014) , and [6] non linear deformat censored; binary censoring indicators and rigid realignment parameters were included as first level covariates. Both total number of valid TRs (i.e., non censored) and mean framewi se displacement across each run were included as second level nuisance covariates. Participants Brain activation responses induced by cue blocks were modeled at the first level with a bo x car function convolved with a double gamma hemodynamic response function. BOLD data were visually inspected by three investigators (MFR, JRT, JLT) on a subject by subject basis during blinded analysis to document agreement on quality control prior to an alysis (Power et al., 2014) . For both pre treatment and post treatment scans, this included: [1] line plots of realignment parameters, [2] line plots indicating censored TRs, and [3] BOLD signal greyplots before and after nuisance variance correction. Motion and nuisance correction appeared appropriate in all subjects; none were exclude d on this basis.

PAGE 107

91 5.3.12. Neuroimaging Data Analysis The main effect of cigarette cues (cigarette cue > neutral cue) at the pre treatment timepoint across groups (one sample t test) was computed to inspect effects of visual cue exposure on brain activity. Effect of treatment group was evaluated using a 2x2 mixed independence factorial design. Time was the within subjects factor (pre versus post treatment); group was the between subjects factor (LF rTMS versus sham). First level SPM contrasts were construct ed for each participant using paired t tests (post treatment [cigarette > neutral cues] > pre treatment [cigarette > neutral cues]). Second level SPM contrasts were constructed using a two sample t test comparing first level contrasts by group. Images were masked at the second level using a grey matter mask with the cerebellum excluded. Whole brain significance threshold was set at familywise p < 0.05 (AlphaSim corrected assuming spatial auto correlation; voxelwise p < 3712 mm 3 3 ) (Cox et al., 2017; Eklund et al., 2016) . 5.3.13. Finite Element Method (FEM) Simulation and ROI Selection ROI selection was based on a finite element method (FEM) head model of induced current density created using SimNIBS v2.1 ( www.SimNIBS.org (Bicalho Saturnino et al., 2019; Thielscher et al., 2015; Wang and Eisenberg, 1994; Windhoff et al., 2013) ). Finite element method (FEM) is a numerical method for solving differential equations (DEs), which is useful when analytic solutions to these problems are impractical or impossible. Our model treats the application of LF rTMS to the head as a boundary value problem governed by DEs describing the electromagnetic potential throughout a space with varying tissue conductivity . Several strengths relevant to neurostimulation simulation include realistic representation of complex geometries, tissue conductivity/anisotropy properties, and estimation of local effects (Opitz et al., 2011) .

PAGE 108

92 The mechanism of action of TMS upon n eural function is electromagnetic induction loop induces a magnetic field orthogonal to the flow of current according to the right hand rule. ined as: where is the magnetic field induced by the applied current ( ), is an infinitesimal differential vector of the curve of applied current, is the closed line integral around the current curve, and is the permeability of free space. What this means is that if you take any imaginary closed loop path surrounding a current carrying wire (right side of above equation) and sum the magnetic field at all points along that path, that sum is directly pro portional to the current flowing through the wire (left side of above equation) , with the constant of proportionality being the permeability of free space . The complex head and coil geometries, differing tissue conductivities, and voxel wise differences in anisotropy make an analytic prediction of induced intracranial currents impossible. Thus, we used FEM to numerically estimate expected brain parenchymal cu rrents induced by the TMS coil, and in turn, to estimate an expected spatial dose response map. Briefly, the standard five compartment head model (WM, GM, CSF, skull, and skin) was used using a canonical T1W image in MNI space. Tissue conductivities were skin = 0.25 S/m (average between outer skin and fat as given in Truong et al. (2013) ) skull = 0.01 S/m (Dannhauer et al., 2011) CSF GM WM = 0.126 S/m (Thielscher et al., 2011) . Tissue anisotropy was estimated using a canonical DTI image in MNI space with 32 directionality basis. Final FEM mes hes contained about 5 × 10 5 nodes and 3 × 10 6 tetrahedral elements; this discretization has shown satisfactory spatial convergence (Bicalho Saturnino et al., 2019) . Eq. 1

PAGE 109

93 The governing DEs for the electric field used in the FEM are (Bicalho Saturnino et al., 2019; Wang and Eisenberg, 1994) : where is the magnetic vector potential of the TMS coil (the curl of which results in the magnetic field proper: ) and is the electric potential. Magnetic vector potential ( ) can be conceptualized as momentum per unit charge (in units of V · s · m 1 ) or potential energy per unit element of current (in units of J · A 1 · m 1 ). This is analogous to electric potential ( ), conceptualized as potential energy per unit charge (in units of volts, joules per coulomb, or electronvolts per elementary charge). Standard vector calculus operators of dot product ( ) , gradient ( ), curl ( ), and divergence apply. This model assumes a quasi static function of time and the elect ric field as a function of space can be separated into their own homogenous equations of time and space, respectively (Bicalho Saturnino et al., 2019) . This model also assumes homogenous Neumann boundary conditions, meaning that there is no current flow to the outside of the head. The vector potential of the standard 70mm figure of eight planar coil was pre calculated as in (Thielscher and Kammer, 2004) and previously empirically validated (Bungert et al., 2017) . FEM simulation results of the predicted electromagnetic field produced by a standard planar transcranial magnetic stimulation 70mm figure of eight coil targeting the right anterior insula are presented below. Note that this energy field deposition estimate is considered conservative, as in our study we used a custom built angulated coil which theoretically provides a deeper energy deposition field (Deng et al., 2013) . Eq. 2 Eq. 3

PAGE 110

94 The above figure shows the results of the predicted magnetic potential (left, presented as time derivative) and normalized absolute value of the electric field (right), illustrating the pattern of energy deposition. Maximum energy is deposited in the scalp, superficial soft tissues, and cerebrospinal fluid due to high tissue conductivities. This illustrates the difficulty in targeting deep structures, such as the insula or anterior cingulate cortex, using conventional coil designs. This is consistent with model results of insular targeting reported by Pollatos and colleagues (Pollatos and Kammer, 2017) . In any numerical method used to solve a system of partial differential equations, two critical questions must be answere correct observations) (Kheyfets, 2016) . First, t he FEM method used here has demonstrated good convergence with 3% error or less using the free parameters herein specified (Bicalho Saturnino et al., 2019) . Second, t his method as employed here has also shown good comparison with exp erimental observation; it has been recently validated against hand electrophysiologic responses in vivo (Bungert et al., 2017) .

PAGE 111

95 This model is subject to several limitations. First, we estimated TMS targeting the right anterior insula in a standardized neurological space using population averaged T1W images. Inter individual differences in gyral anatomy would be predicted to influence the tissue energy deposition (Bicalho Saturnino et al., 2019) , and thus subject specific estimates of current induction wou ld be more accurate. Second, we modelled the treatment using a planar coil with relatively shallow tissue energy deposition, whereas the coil used in this study was angulated and predicted (in spherical head models) to exhibit a deeper treatment field (Deng et al., 2013) . Lastly, the FEM model is a numerical estimate. Although it has been validated against hand electrophysiologic responses in vivo (Bungert et al., 2017) and converges at or less than ~3% error with free parameters specified here (Bicalho Saturnino et al., 2019) , it has not been validated in a rigorous fashion, such as through using three dimensi onal printed head models with tissues of known conductivities or ex vivo in cadaveric brains with neurosurgically placed depth electrodes. 5.4. Results 5.4.1. Participant Characteristics Recruitment and randomization of the sample population are illustrated in Figure 5 1 . Participant characteristics are provided in Table 5 1 . Participant psychometric survey results are provided in Table 5 2 . Four (10% of the sample) participants originall y assigned to LF rTMS were administratively re assigned (after randomization) due to study appointment timing issues described above. However, participant blinding was maintained. These four subjects reported treatment survey. There was no significant group difference in the number of subjects who believed that they received real TMS (p=0.2 0 5) and the majority believed they received treatment ( Table 5 3 ).

PAGE 112

96 5.4.2. Self Reported Cigarette Craving Insular LF rTMS ( T 19 = 5.323, p < 0.001 ) but not sham ( T 19 = 1.820, p = 0.084 ) caused a significant decrease in cigarette craving. Compared to sham treatment, LF rTMS of the right insula reduced self reported cigarette craving ( F 1,38 = 4.921, p = 0.033, Figure 5 3 ). 5.4.3. Neural Activity Responses to Cigarette Cues: Whole Brain Analysis Cigarette compared to neutral cue exposure caused significant, bidirectional changes in brain activity during the pre treatment fMRI ( Figure 5 4 ). Brain activity significantly increased in the bilateral dorsolateral prefrontal cortex, bilateral primary visual cortex, and bilateral higher level visual cortex during cigarette cues compared to neutral cues. At an uncorrected level, brain activity in the bilateral insula was observed, consistent with prior studies (Claus et al., 2013) . Brain activity significantly decreased in the bilateral posterior cingulate gyrus and precuneus during cigarette cues compared to neutral cues ( Figure 5 4 ) . Compared to sham, LF rTMS of the right insula decreased brain responses to visual cigarette cues at the whole brain level ( Figure 5 5 ). LF rTMS compared to sham reduced brain

PAGE 113

97 Table 5 1 . Sample population demographic and smoking characteristics. LF rTMS Sham Whole Sample Group comparison p value Demographics Age (years) 37.5 ± 11.2 41.0 ± 8.7 39.3 ± 10.1 0.278 Sex by Birth (M/F) 13/7 11/9 24/16 0.518 Education (years) 13.3 ± 1.9 13.3 ± 1.8 13.3 ± 1.8 0.932 Ethnicity (self reported) 0.394 White / Caucasian 17 14 31 Other (Multiple Groups) 3 6 8 Handedness (R/L) 18/2 18/2 36/4 1.000 EtOH (stand. drinks / week) 0.43 ± 1.21 1.25 ± 2.90 0.84 ± 2.23 0.251 Cannabis User (Y/N) 10/10 11/9 21/19 0.751 Cigarette Use Characteristics and History Last Cigarette (hours ) 5.18 ± 3.42 5.1 ± 3.18 5.14 ± 3.26 0.943 Age of First Cig (years) 14.75 ± 3.35 14.05 ± 4.19 14.40 ± 3.76 0.563 Age of Daily Cig Use (years) 17.10 ± 2.73 16.68 4.40 16.89 ± 3.62 0.716 Onset Enjoyment 6.13 ± 3.11 6.50 ± 3.36 6.31 ± 3.20 0.716 Dependence Duration (years) 19.50 ± 12.19 24.38 ± 10.28 21.94 ± 11.40 0.180 Current avg cig consumption (#cigs/day) 19.10 ± 6.69 18.70 ± 4.38 18.90 ± 5.58 0.824 Maximum cig consumption (#cigs/day) 36.60 ± 15.56 31.00 ± 9.68 33.80 ± 13.10 0.181 # Lifetime Quit Attempts 2.50 ± 1.57 2.80 ± 1.91 2.65 ± 1.73 0.591 Longest Abstinence (days) 287.35 ± 467.13 161.50 ± 321.18 224.43 ± 400.78 0.328 # Cigs Smoked in Last 24h 14.85 ± 7.77 13.00 ± 6.89 13.93 ± 7.31 0.431 # Cigs Smoked in Last Week 122.95 ± 39.50 127.40 ± 29.31 125.18 ± 34.41 0.688 FTND 1 5.15 ± 1.76 5.55 ± 2.26 5.35 ± 2.01 0.538 [CO] Exhaled 2 (ppm) 5.00 ± 1.49 5.90 ± 1.41 5.14 ± 1.50 0.057 [Cotinine] Urine 0.485 30 100 ng/mL 1 2 3 100 200 ng/mL 5 8 13 200 500 ng/mL 8 6 14 500 1000 ng/mL 2 3 5 >1000 ng/mL 4 1 5 Is cigarette consumptions decreasing in past week? (Y/N) 13/7 14/6 27/13 0.736 Consumption reduction in past week (#cig/day) 2.60 ± 4.03 1.90 ± 3.34 2.25 ± 3.67 0.553

PAGE 114

98 Table 5 2 . Sample population psychometric survey results. LF rTMS Sham Whole Sample Group comparison p value Mean SD Mean SD Mean SD Wisconsin Inventory of Smoking Dependence Motives (WISDM) Affiliative Attachment 4.04 1.06 4.37 1.03 4.21 1.04 0.323 Automaticity 4.27 0.96 4.15 1.23 4.21 1.09 0.733 Loss of Control 4.21 1.20 4.20 1.18 4.21 1.17 0.973 Behavioral Choice Melioration 3.80 0.96 4.22 1.01 4.01 1.00 0.184 Cognitive Enhancement 3.94 1.20 4.03 1.03 3.99 1.10 0.800 Craving 4.34 1.22 4.30 1.28 4.32 1.24 0.925 Cue Exposure and Associative Processes 4.29 1.08 4.55 0.86 4.42 0.97 0.397 Negative Reinforcement 4.38 1.13 4.22 0.82 4.30 0.98 0.615 Positive Reinforcement 3.93 1.01 4.48 1.02 4.21 1.04 0.047 Social Environmental Goals 4.14 1.08 4.26 1.04 4.20 1.05 0.711 Taste Sensory Processes 4.47 1.05 4.33 1.05 4.40 1.04 0.672 Tolerance 4.17 0.78 4.06 1.21 4.12 1.01 0.735 Weight Control 4.12 0.95 4.29 1.20 4.21 1.08 0.624 Total Scaled Score 54.09 9.10 55.46 8.72 54.77 8.82 0.629 Behavioral A pproach and Behavioral Inhibition Scales (BAS /BIS ) BAS Drive 10.50 2.56 9.70 2.39 10.10 2.48 0.316 BAS Fun Seeking 9.50 2.54 10.15 2.62 9.83 2.57 0.431 BAS Reward Responsiveness 15.20 1.91 16.00 3.03 15.60 2.53 0.324 BAS Total 35.20 5.18 35.85 4.50 35.53 4.80 0.674 BIS Total 17.95 3.25 19.20 2.48 18.58 2.93 0.180 Barratt Impulsiveness Scale Attentional ( 2nd Order ) 18.55 2.98 19.80 3.55 19.18 3.30 0.235 Motor ( 2nd Order ) 27.10 4.61 28.15 4.85 27.63 4.70 0.487 Non planning ( 2nd Order ) 28.90 3.73 28.50 5.01 28.70 4.36 0.776 Total 74.55 7.03 76.45 8.13 75.50 7.57 0.434

PAGE 115

99 Table 5 3 . Sample population treatment characteristics, craving results, and beliefs regarding treatment. LF rTMS Sham Whole Sample Group diff p value Treatment Characteristics T x Power (% MagStim) 69.95 ± 6.38 N/A N/A N/A Subject Belief Regarding T reatment 0.205 Received real Tx 18 15 33 Received sham Tx 2 5 7 Craving Scores (QSU Brief) and Changes Pre Treatment 57.70 ± 15.61 59.25 ± 9.18 58.48 ± 12.66 0.704 Post Treatment 34.05 ± 13.82 50.30 ± 20.75 42.18 ± 19.25 0.006 Absolute Change ( Post Pre) 23.65 ± 19.87 8.95 ± 21.99 16.30 ± 21.98 0.033 Relative Change ([Post Pre]/Pre) 37.24 ± 29.18 13.37 ± 38.45 25.31 ± 35.79 0.034 Figure 5 3 . Self reported cigarette craving (QSU Brief) by group and time. LF rTMS causally reduced self reported craving compared to sham by per protocol two way repeated measures ANOVA of group × time (p = 0.033). No statistically significant difference was observed in craving after sham treatment, although there is a trend towards pla cebo effect (p = 0.084).

PAGE 116

100 Figure 5 4 . Pre treatment main effects of cigarette cue exposure compared to neutral cues, collapsed across groups. Brain activity significantly increased in the bilateral dorsolateral prefrontal cortex, primary visual cortex, and higher level visual cortex during cigarette cues compared to neutral cues (red color bar). Brain activity significantly decreased in the bilateral posterior cingulate gyrus and precuneus during cigarette cues compared to neutral cues (blue, p < 0.001). Results were corrected for multiple comparisons using familywise p < 0.05,

PAGE 117

101 Figure 5 5 . Whole brain interaction effect using a per protocol 2x2 mixed factorial interaction effect of time (within subjects: post treatment pre treatment) × group (between subjects: LF rTMS sham). Significantly decreased cigarette cue brain response after LF rTMS compared to sham was observed in the right primary sensorimotor cortex, bilateral supplementary motor cortex (premotor), and right dorsal anterior insula (blue, p < 0.001). No significant increased cigarette cue brain response was observed after LF r TMS compared to sham. Results were corrected for multiple comparisons using familywise p < 0.05, voxelwise p < 0.005, cluster

PAGE 118

102 activity in primary sensorimotor cortices, premotor cortices, and right anterior insula. There were no b rain areas in which LF rTMS compared to sham increased cigarette cue brain activity. 5.4.4. Neural Activity Responses to Cigarette Cues: FEM Based ROI Analysis Compared to sham, LF rTMS of the right insula reduced brain activity responses to visual cigarette cues within the target stimulation ROI. No significant effects were observed in non target stimulation and non stimulated control ROIs ( Figure 5 6 ). Brain activity responses to cigarette cues within the non target stimulation ROI (right inferior frontal gyrus pars triangularis) did not statistically differ by group. Figure 5 6 . FEM informed ROI analysis of brain acti vity responses using a per protocol analysis of covariance. ROI beta values and significance levels were extracted using the MarsBar toolbox after accounting for nuisance covariates, using the SPM12 design used for whole brain analysis .

PAGE 119

103 5.4.5. Brain Behavioral Relationshi ps At the whole brain level, a significant association was found between change in craving (QSU Brief) and b rain activity responses to cigarette cues within the LF rTMS group ( Figure 5 7 ). Two clusters demonstrated significance: a large cluster spanning the bilateral periroland ic cortex and supplementary motor areas ( T peak = 8.39, p FWE < 0.001) and another in the right anterior insula ( T peak = 6.01, p FWE < 0.001). There was no correlation between change in craving and change in brain activity responses within ROIs. 5.5. Discussion In this study, deep low frequency (LF) rTMS to the right anterior insula in smokers resulted in: [1] decreased cigarette craving, [2] decreased brain responsivity to cigarette cues, Figure 5 7 . Correlation between absolute change in QSU Brief self reported craving (post treatment > pre treatment) and change in brain activity responses to cigarette cues [(post treatment > pre treatment) × (cigarette cues > neutral cues)] within the LF rTMS group defined per protocol. Results were corrected for multiple comparisons using familywise p < 0.05, voxelwise p <

PAGE 120

104 [3] neuroimaging evidence of target engagement consistent with predicted spa tial dose response across the brain, and [4] evidence that the decreased craving correlated with decreased brain responsivity at a whole brain level. Per protocol analysis found that a single session of deep LF rTMS targeting the right anterior insula sign ificantly decreased cigarette cravings in the immediate post treatment time period compared to sham. Evidence suggests that smoking relapse may be associated with decreased cognitive control over insular functions (Janes et al., 2010) , and that the insula promotes smoking through increased cravings and attentional redirection towards cigarette seeking behaviors (Naqvi et al., 2014; Noel et al., 2013) . Our in terpretation is that the observed results are in contrast to and extend prior studies which augmented urges in the dorsolateral prefrontal co rtex (Amiaz et al., 2009; D ieler et al., 2014; Dinur Klein et al., 2014; Eichhammer et al., 2003; Huang et al., 2016; Johann et al., 2003; Kozak et al., 2018; Li et al., 2013; Li et al., 2017; Pripfl et al., 2014; Trojak et al., 2015; Wing et al., 2012) . It is important to note that LF rTMS effects on brain activity have primarily been studied from a mechanistic standpoint in the motor system; our findings provide support for the inhibitory aftereffects hypothesis of LF rTMS action in cortical areas outside the motor system (i.e., associative cortex in addition to primary sensorimotor) (Chen et al., 1997; Miniussi et al., 2013) . TMS mechanisms of action, however, remain largely speculative and multifactorial (Cirillo et al., 2017) . Most subjects in both treatment groups believed they received real treatment. This belief pattern did not differ between groups, strengthening our interpretation that the intervention decreased cravings above and beyond placebo effects. While most prior studies in treatment beliefs to confirm efficacy of the blinding.

PAGE 121

105 We observed that LF rTMS centered upon the right anterior insula in smokers reduced brain activity responses in the target stimulation region during the craving cue fMRI task. This to craving, although it is just possible that LF rTMS to insula affected other withdrawal sensations involving interoceptive perception (Craig, 2002, 2009, 2011; Nieuwenhuys, 2012) not assessed by this study. While our whole brain and ROI neuroimaging results provide supporting evidence for corticotopic specificity, the study remains confounded by this issue. Few TMS studies in smokers use fMRI or advanced neuroimaging to measure effects of neuromodulation on smoking related brain activity. Here, fMRI changes using whole brain and ROI approac hes in conjunction with FEM predictions of the spatial dose response curve provide evidence of a possible physical mechanism explaining the observed changes in craving. Cigarette cues resulted in brain activity in bilateral dorsolateral prefrontal cortex, primary visual cortex, and higher order visual cortex . Our results are consistent with a meta analysis of craving cue fMRI studies that reported brain activity in the extended visual system in smokers, (Engelmann et al., 2012) , a pat tern thought to represent excessive incentive salience or increased allocation of attentional resources toward processing of appealing cue stimuli (Engelmann et al., 2012; Robinson and Berridge, 2008) . This increased activation in response to cigar ette cues was larger by meta analysis in deprived smokers compared to sated possibly suggesting effects of craving. Similar to our results, increased brain activity responses to cigarette cues have been reported in the prefrontal cortex, which may reflec t a preparation to initiate drug seeking behaviors (Claus et al., 2013; Engelmann et al., 2012) . We also observed decreased main effects of cue in the bilateral posterior cingulate gyrus and precuneus , which appears in contrast with reported studies. We cannot confidently explain this resu lt, although one possibility is that most previous studies compared cigarette cues to matched food cues,

PAGE 122

106 which are presumably hedonically relevant, while we compared to matched neutral cues that contained no clear hedonic relevance. Our observation of decr eased brain activity responses in the right primary sensorimotor cortex and right premotor cortex after LF rTMS was unexpected. Since these areas are anatomically distant from the stimulation field, it suggests a possible confounding stimulation effect rel ated to facial sensations/muscular twitches or a corticocortical effect in the sensorimotor system secondary to insular changes. Participants undergoing LF rTMS and sham and involuntary right jaw clenching. However, we expect that if these sensations resulted in confounding brain activity, it would be represented in the contralateral cerebral hemisphere. Another possibility is that the right sensorimotor and premotor findi ngs represent indirect corticocortical association fiber effects from the ipsilateral posterior insula, which is known to be highly connected to sensorimotor areas (Craig, 2002, 2009, 2011; Nieuwenhuys, 2012) . Alternatively, the finding may represent a decreased cognitive preparedness (supplem entary motor area) or representation (primary sensorimotor) to act upon craving cues. This study addresses a gap in the literature on the feasibility and efficacy of insular TMS targeting in smokers. This gap is surprising, given substantial evidence from animal (Forget et al., 2010; Kutlu et al., 2013; Pushparaj et al., 2013) and human (Abdolahi et al., 2015a, b, 2017; Naqvi et al., 2007; Suner Soler et al., 2012) studies suggesting that the insula plays a major role in nicotine use disorder. One reason for this gap is the biophysical limitations of current TMS coil technology. We address this using a novel angled coil and calibrated the RMT to the paracentral lobule. We then transiently lesioned the anterior insula in smokers to reproduce the previously reported insula effects for possible therap eutic benefit. This study has several strengths. There are several prospective, blinded, randomized controlled trials of DLPFC targeting in smokers. This RCT design, applied to a novel target,

PAGE 123

107 theoretically allows for causal inference. Participant blinding was maintained and demonstrably validated. Blinding of primary outcomes analysis by a biostatistician uninvolved in the study provided an additional safeguard against potential bias. Most participants believed significant group differences in this belief, demographics, smoking history, or psychometrics. Nevertheless, in caution and because our primary analysis was per protocol rather than intention to treat, we present our inferences as hypothesis generating ra ther than hypothesis testing. This study is limited in several ways. First, four of 40 total participants (10%) were re assigned post randomization due to study appointment timing issues. This placed the study at risk for performance bias by Cochran crite ria. It is possible that this biased the study, and thus the study must be replicated. For example, if the re assigned participants had systematically higher RMTs, we would expect the treatment effects to be overestimated. If the re assigned participants h ad systematically lower RMTs, we would expect the treatment effects to be underestimated. Second, whether LF rTMS can directly modulate insular activation is controversial (Coll et al., 2017; Pollatos et al., 2016; Pollatos and Kammer, 2017) . We believe that by selecting RMT within the deep paracentral lobule (foot) as opposed to the superficial y using a custom built angled coil designed for deeper targeting, it is reasonable to conclude the insula received subthreshold energy deposition sufficient to affect neural function. Third, in the absence of a comparison cortical stimulation (e.g., dorsol ateral prefrontal cortex or medial frontal cortex), the corticotopic specificity of our findings to the insula is unknown. It could be that decreased craving results from LF rTMS to any brain area without regional specificity. Even if rTMS neural aftereffects are assumed to spatially match predicted energy deposition, we cannot prove that targeting right anterior insula specifically caused our observations. For example, the right inferior frontal gyrus, involved in inhibitory co ntrol, was subjected to greater energy deposition compared to insula

PAGE 124

108 using our treatment paradigm. However, if we assume the aftereffects of LF rTMS are inhibitory, we would expect the effects on a major inhibitory center to cause cognitive disinhibition, increased cigarette craving, and increased brain activity responses. Lastly, the small sample size limited statistical power. While the per protocol effects were significant, the intent to treat effects was not, suggesting that the approach may work but re quires replication in a larger sample. 5.6. Conclusion In conclusion, per protocol analysis showed that direct inhibition of the right anterior insula reduces craving, supporting previous evidence that insular lesions disrupt craving. We further demonstrated th at LF rTMS targeting the insula reduced insular activity responses to cigarette cues. It remains unclear if this would disrupt smoking behaviors and cigarette consumption with repeated treatments, which should be investigated. Future studies using a single day appointment design must carefully consider the timing between pre and post treatment MRI examinations, if the study uses an MRI scanner shared between multiple research groups. This issue could be easily obviated in a longitudinal study, which we hop e to pursue in a future grant funded study. While these results should be considered preliminary, they provide hope that TMS could be developed as a treatment strategy to help reduce the burden of cigarette addiction. 5.7. Acknowledgements This work was funded by NIH F32 DA041011 (MFR) , RSNA RR 1620 (MFR) , CCTSI M 15 81 (MFR), NIH R01 GM12108 (DHG), NIH R25 GM111901(DHG), and NIH R25 GM111901S1 (DHG). The authors would like to thank MRI technologists Deb Singel and Kendra Huber for their assistance with data col lection . The authors would like to thank Dr. Isabelle Buard, Dr. Peter DeWitt, and Dr. Joshua Gowin for helpful discussion and support. Finally, the authors

PAGE 125

109 would like to thank the anonymous study participants, who generously contributed to scientific rese arch. 5.8. Supplement: Justification of Analysis Approach for Primary and Secondary Outcomes The National Institutes of Health Office of Extramural Research defines a clinical trial as est s a new biomedical intervention in a small group of people (e.g . , 20 80) for the first time to determine efficacy and evaluate safety (NIH) Because this proof of concept, phase I clinical trial was intended to evaluate efficacy rather than effectiveness, we performed an as is the effect of receiving (Spieth et al., 2016) . In addition, we sought to provide meaningful inferences regarding the safety, tolerability, and dose response characteristics of LF rTMS targeting the insula in smokers. Secondary analyses included both intentio n to treat of concept, we elected to focus on efficacy (i.e., an explanatory trial), evaluating whether the inte rvention produces the expected results under ideal circumstances. This is in contradistinction to effectiveness (i.e., a pragmatic trial), (Gartlehner et al., 2006) . Primary Analysis (As Treated) Secondary Analysis (Intention to Treat) Secondary Analysis (Sensitivity Analysis) Sample Size (Tx/Sham) n = 40 (20/20) n = 40 (24/16) n = 36 (16/20) Self Reported Cigarette Craving (QSU Brief) F 1,38 = 4.921, p = 0.033 F 1,38 = 0.715, p = 0.403 F 1,38 = 2.837, p = 0.101 Whole brain response to cigarette cues p < 0.001 p = 0.293 p = 0.096 FEM Informed ROI Analysis Right Insula T = 2.45, p = 0.029 T = 1.15, p = 0.387 T = 1.99; p = 0.081 Right IFG PT T = 2.02, p = 0.074 T = 0.83, p = 0.618 T = 1.10, p = 0.363 Left SMG T = 0.15, p = 0.825 T = 0.62, p = 0.809 T = 0.39, p = 0.958

PAGE 126

110 This study was exposed to a potential performance bias in group assignment, which occurs if there is insufficient adherence to the study protocol by either the participant or the inves tigator (Spieth et al., 2016) . Four subjects in the treatment group and four subjects in the sham group did not have a reliably measured RMT due to insufficient time between scheduled smoking or physiology. It occurred only when there was another research MRI exam scheduled immediately afte treatment MRI. Other random factors unrelated to the first MRI, delay in T1W image transfer). Subsequent scheduling changes descr ibed above obviated this issue. Because LF rTMS could not safely be applied to four patients originally randomized to the treatment group prior to their post treatment scan, they were administratively re assigned post randomization to the sham group by jud gement of the study physician. Careful attention was paid to maintaining the subject blinding throughout the entire study appointment, and post treatment provided inter val validity of the blinding. It is possible that this re assignment of four participants biased the study, and thus the study must be replicated. For example, if the re assigned participants had systematically higher RMTs, we would expect the treatment ef fects to be overestimated. If the re assigned participants had systematically lower RMTs, we would expect the treatment effects to be underestimated. Of note, group differences comparison across demographic, smoking use/history, and psychometric characteri stics suggest an absence of overall differences between groups. We observed that in both secondary analyses (sensitivity analysis excluding subjects with post randomization reassignment and intention to treat analysis labelling participant group labels by initial allocation rather than actual treatment) the directions of associations

PAGE 127

111 were unchanged, although the statistical significance was no longer observed. This pattern may s generating as opposed to hypothesis testing. While we believe these results to be of interest to the scientific community, we recommend a larger sample confirmatio n study be undertaken to verify.

PAGE 128

112 5.8.1. Primary Analysis Per Protocol (n=40) Results Significance level for p Family < 0.05 is k E (AlphaSim corrected assuming spatial auto 3712 mm 3 3 ) (Cox et al., 2017; Eklund et al., 2016) . Anatomical Label MNI Coordinates k E p uncorr p FWE X Y Z Right inferior frontal gyrus 42 4 24 103 0.045 0.837 Right precentral gyrus / supplementary motor area 54 14 50 996 <0.001 <0.001 Right anterior insula 38 20 8 502 <0.001 0.005 Left temporal operculum / Left posterior insula 48 10 2 168 0.014 0.421 Right temporal operculum 54 2 4 200 0.008 0.275 Left precentral gyrus 34 16 68 386 0.001 0.022 Right posterior insula 36 12 2 111 0.038 0.787 Left precentral gyrus 18 14 76 117 0.034 0.747

PAGE 129

113 5.8.2. Secondary Analysis, Sensitivity Results (n=36) Significance level for p Family < 0.05 is k E (AlphaSim corrected assuming spatial auto 3712 mm 3 3 ) (Cox et al., 2017; Eklund et al., 2016) . Anatomi cal Label MNI Coordinates k E p uncorr p FWE X Y Z Left posterior insula 48 8 2 102 0.006 0.404 Right anterior insula 38 20 8 206 <0.001 0.025 Right precentral gyrus 54 14 50 161 0.001 0.083 Right temporal operculum, posterior 52 0 8 120 0.004 0.252

PAGE 130

114 CHAPTER VI 6. C ONCLUSION This dissertation sought to study substance use disorders (SUD) as brain diseases of altered neurocircuitry underlying reward, craving, and goal oriented behavior s . We conducted advanced neuroimaging experiments involving two different populations of SUD: long term abstinent, severely dependent cocaine and methamphetamine addicts (Chapters 2 and 3); and, acutely abstinent, moderately dependent cigarette smokers (Chapter 4 and 5 a nd Appendix 1 ). While these two populations differ in overall disease severity, they also differ in primary drug of abuse (psychostimulants versus nicotine) and stage of disease (chronic remission versus acute withdrawal). 6.1. Specific Knowledge Gaps Addressed In Chapter 2, we investigated differences in grey matter volumes and brain behavior relationships in stimulant dependent individuals compared to healthy controls after long term abstinence. We observed significant differences between stimulant dependent women and health control women in regional grey matter volumes, but no similar differences were observed amongst the men. We further observed that the relationships between grey matter volumes and behavioral metrics relevant to SUD differed by sex. This cr oss sectional study was strong with regards to its sample size and signal processing technique; however, the results remain somewhat ambiguous. For instance, i t is unclear to what extent these differences pre dated the pathology (i.e., different endophenot ype s ), resulted from the pathology (i.e., different drug induced neurotoxicity) , or resulted from differences in recovery from the pathology (i.e., different neuroplasticity). Our results are consistent with the literature in that significant

PAGE 131

115 structural ne uroanatomical changes are associated with severe SUD , and highlight sex as an important variable in neuroimaging studies , despite being often ignored (Lind et al., 2017) . In Chapter 3, we investigated the resting state functional connectivity changes in stimulant dependent individuals compared to healthy controls in a subset of the population reported in Chapter 2. We found that even after lo ng term abstinen ce (average 1 2.8 months) , stimulant dependent persons exhibit ed a brain function effective connectivity when compared to healthy controls. Compared to controls, stimulant dependent individuals showed signific antly greater Granger causal connectivity from right executive control network to default mode network and from default mode network to basal ganglia network. Stimulant dependent individuals also demonstrated greater global efficiency and lower local effic iency , suggesting large scale changes in brain connectivity despite long term abstinence . These findings suggest that i ncreased top down effective connectivity in long term abstinent drug users may reflect adaptive changes that foster successful remission, such as improved cognitive control over habit and reward processes. Chapters 4, 5, and Appendix 1 introduce and report data from our phase 1, randomized, sham controlled, single blinded clinical trial investigating the effect of inhibitory insular transcranial magnetic stimulation (TMS) on cigarette cravings, brain responses to cigarette cues, and resting state brai n connectivity in acutely abstinent moderately dependent smokers. Finite element model simulations of energy deposition intracranially allowed for estimation of a spatial dose response across the brain . Compared to sham treatment, right insula inhibitory T MS reduced self reported cigarette craving and decreased brain activity responses to visual cigarette cues in primary sensorimotor cortices, supplementary motor area, and right anterior insula. The fact that right anterior insular responses to cigarette cu es were diminished after inhibitory TMS targeting the insula provides novel neuroimaging evidence of target

PAGE 132

116 engagement; most prior studies targeted the dorsolateral prefrontal cortex. These findings provide proof of concept of a potential neuroanatomical t arget for smoking cessation therapy. 6.2. Limitations This dissertation was subject to several limitations. First, a major limitation in studying long term abstinence is the cross sectional nature of the study designs. This limited our ability to determine if o bserved changes in brain structural and functional reflected endophenotypes that predated the pathology, changes associated with the disease course itself, or differences in neuroplastic recovery from the disease. Second, our randomized controlled trial w as limited by the lack of a crossover and comparison target group. For example, we could not definitively determine if the changes we observed were secondary to stimulation of the right anterior insula or stimulation of the brain in general (i.e., we coul d not prove corticotopic specificity). In order to definitively prove causality, we would need to design the experiment using a within subject counterbalanced crossover with at least three interventions (e.g., target insula, target prefrontal cortex, sham ). Lastly, our studies in both stimulant and cigarette addicts were confounded in part by polysubstance use/abuse and comorbid low educational attainment , albeit less so in the case of cigarette addicts (confounded only by marijuana use). While this precl udes us from relating structural and functional changes to a single drug, our sample s ha ve biological and ecological validity as they reflect important, real world, clinical population s of SDI. Epidemiological data , for example, demonstrate that stimulant dependence does not often occur in isolation; instead most stimulant dependen t individuals meet dependenc e criteria for other substances (Sara et al., 2012; Stinson et al., 2005) . There is significant literature describing the correlation between drug use and low educational attain ment; it is debated whether low educational attainment is the cause or result of drug use disorders (Fergusson et al., 2003; Swaim et al., 1997; Yamada et al., 1996) . More recently , however, authors have reported that this correlation is due in part to

PAGE 133

117 shared genetic factors (Bergen et al., 2008) while others report that it is due to shared environmental or non genetic familial risk factors (Grant et al., 2012; Verweij et al., 2013) . These studies suggest that low educational attainment may be a pre existing behavioral component of the path ology of substance use disorders. 6.3. Future Work The immediate future work is currently underway. With regards to our randomized controlled trial, future work will investigate the changes induced by inhibitory TMS on measures of resting state brain connectivi ty (Appendix 1), cerebral blood flow, and neurotransmitters thought to be involved in the mechanism of action of TMS, including glutamate, glutamine, and GABA. The FEM model stimulation presented will be compared to individual changes in cerebral blood flo w to determine if this energy deposition map reflects a n empirical spatial dose response curve. The results from these analyses, including those presented in Chapter 5, will form the preliminary data for future scientific proposals that will extend and add ress limitations of the current clinical trial, including comparisons of different cortical targets for TMS . This dissertation provides impetus for several strands of future work. First, l ong term future work is needed to explore the effects and durability of repeated applications of TMS in smokers , including alternative/comparison cortical targets . Second, t he mechanism s of TMS are still largely speculative, and future studies could potentially disentangle these mechanisms by optimizing TMS within the MRI scanner and neuroimaging TMS after different changes in target and pulse sequence . Lastly, substance use disorders span a complex disease spectrum and vary by drug, stage, and individual factors. Finding commonalities amongst these myriad presentations o f addiction with regards to long term remission is critically important. Neuroimaging long term abstinent individuals, as well as individuals progressing through the abstinence process, has potential to improve our understanding of the neurocircuitry

PAGE 134

118 mecha nisms of successful abstinence. These three long term areas of future work involving neuroimaging and neuromodulation could help understand addiction as a brain disease and potentially reduce the burden s associated with it . 6.4. Conclu ding Remarks Each individual chapter addresses a specific yet important gap in the literature. Together, however, these neuroimaging studies in two different populations of substance use disorders provide evidence that : (1) addiction is a brain disease , and (2) addiction is a disease of neurocircuitry with endogenous (i.e., increased top down control, Chapter 3) and exogenous (i.e., therapeutically imposed, Chapter 5) mechanisms of remission and treatment . Future work must continue to study mechanisms of and interventions th at promote successful disease remission.

PAGE 135

119 REFERENCES Abdolahi, A., Williams, G.C., Benesch, C.G., Wang, H.Z., Spitzer, E.M., Scott, B.E., Block, R.C., van Wijngaarden, E., 2015a. Damage to the insula leads to decreased nicotine withdrawa l during abstinence. Addiction 110, 1994 2003. Abdolahi, A., Williams, G.C., Benesch, C.G., Wang, H.Z., Spitzer, E.M., Scott, B.E., Block, R.C., van Wijngaarden, E., 2015b. Smoking cessation behaviors three months following acute insular damage from stroke . Addictive behaviors 51, 24 30. Abdolahi, A., Williams, G.C., Benesch, C.G., Wang, H.Z., Spitzer, E.M., Scott, B.E., Block, R.C., van Wijngaarden, E., 2017. Immediate and Sustained Decrease in Smoking Urges After Acute Insular Cortex Damage. Nicotine & to bacco research : official journal of the Society for Research on Nicotine and Tobacco 19, 756 762. Addicott, M.A., Sweitzer, M.M., Froeliger, B., Rose, J.E., McClernon, F.J., 2015. Increased Functional Connectivity in an Insula Based Network is Associated with Improved Smoking Cessation Outcomes. Neuropsychopharmacology : official publication of the American College of Neuropsychopharmacology 40, 2648 2656. Ahmadlou, M., Ahmadi, K., Rezazade, M., Azad Marzabadi, E., 2013. Global organization of functional b rain connectivity in methamphetamine abusers. Clinical neurophysiology : official journal of the International Federation of Clinical Neurophysiology 124, 1122 1131. Albein Urios, N., Verdejo Roman, J., Asensio, S., Soriano Mas, C., Martinez Gonzalez, J.M. , Verdejo Garcia, A., 2014. Re appraisal of negative emotions in cocaine dependence: dysfunctional corticolimbic activation and connectivity. Addict Biol 19, 415 426. American Psychiatric Association, T., 2013. Diagnostic and Statistical Manual of Mental D isorders (DSM 5®). American Psychiatric Pub. Amiaz, R., Levy, D., Vainiger, D., Grunhaus, L., Zangen, A., 2009. Repeated high frequency transcranial magnetic stimulation over the dorsolateral prefrontal cortex reduces cigarette craving and consumption. Add iction 104, 653 660. Ashburner, J., 2007. A fast diffeomorphic image registration algorithm. NeuroImage 38, 95 113. Babb, S., 2017. Quitting smoking among adults United States, 2000 2015. MMWR. Morbidity and mortality weekly report 65. Back, S.E., Brady, K .T., Jackson, J.L., Salstrom, S., Zinzow, H., 2005. Gender differences in stress reactivity among cocaine dependent individuals. Psychopharmacology 180, 169 176. Barrett, L.F., Simmons, W.K., 2015. Interoceptive predictions in the brain. Nature reviews. Ne uroscience 16, 419 429. Barros Loscertales, A., Garavan, H., Bustamante, J.C., Ventura Campos, N., Llopis, J.J., Belloch, V., Parcet, M.A., Avila, C., 2011. Reduced striatal volume in cocaine dependent patients. NeuroImage 56, 1021 1026.

PAGE 136

120 Becker, J.B., Perry, A.N., Westenbroek, C., 2012. Sex differences in the neural mechanisms mediating addiction: a new synthesis and hypothesis. Biol Sex Differ 3, 14. Bednarski, S.R., Zhang, S., Hong, K.I., Sinha, R., Rounsaville, B.J., Li, C.S., 2011. Deficits in defau lt mode network activity preceding error in cocaine dependent individuals. Drug Alcohol Depend 119, e51 57. Behzadi, Y., Restom, K., Liau, J., Liu, T.T., 2007. A component based noise correction method (CompCor) for BOLD and perfusion based fMRI. NeuroImag e 37, 90 101. Bell, R.P., Garavan, H., Foxe, J.J., 2014. Neural correlates of craving and impulsivity in abstinent former cocaine users: Towards biomarkers of relapse risk. Neuropharmacology 85, 461 470. Bergen, S.E., Gardner, C.O., Aggen, S.H., Kendler, K .S., 2008. Socioeconomic status and social support following illicit drug use: causal pathways or common liability? Twin Research and Human Genetics 11, 266 274. Beyers, J.M., Toumbourou, J.W., Catalano, R.F., Arthur, M.W., Hawkins, J.D., 2004. A cross nat ional comparison of risk and protective factors for adolescent substance use: the United States and Australia. J Adolesc Health 35, 3 16. Bi, Y., Yuan, K., Guan, Y., Cheng, J., Zhang, Y., Li, Y., Yu, D., Qin, W., Tian, J., 2017. Altered resting state funct ional connectivity of anterior insula in young smokers. Brain imaging and behavior 11, 155 165. Bicalho Saturnino, G., Madsen, K.H., Thielscher, A., 2019. Efficient Electric Field Simulations for Transcranial Brain Stimulation. bioRxiv, 541409. Bickel, W.K ., Odum, A.L., Madden, G.J., 1999. Impulsivity and cigarette smoking: delay discounting in current, never, and ex smokers. Psychopharmacology 146, 447 454. Brady, K.T., Randall, C.L., 1999. Gender differences in substance use disorders. Psychiatr Clin Nort h Am 22, 241 252. Braun, U., Plichta, M.M., Esslinger, C., Sauer, C., Haddad, L., Grimm, O., Mier, D., Mohnke, S., Heinz, A., Erk, S., Walter, H., Seiferth, N., Kirsch, P., Meyer Lindenberg, A., 2012. Test retest reliability of resting state connectivity n etwork characteristics using fMRI and graph theoretical measures. NeuroImage 59, 1404 1412. Breese, C.R., Marks, M.J., Logel, J., Adams, C.E., Sullivan, B., Collins, A.C., Leonard, S., 1997. Effect of smoking history on [3H]nicotine binding in human postmo rtem brain. The Journal of pharmacology and experimental therapeutics 282, 7 13. Breiter, H.C., Gollub, R.L., Weisskoff, R.M., Kennedy, D.N., Makris, N., Berke, J.D., Goodman, J.M., Kantor, H.L., Gastfriend, D.R., Riorden, J.P., Mathew, R.T., Rosen, B.R., Hyman, S.E., 1997. Acute effects of cocaine on human brain activity and emotion. Neuron 19, 591 611. Brody, A.L., Mandelkern, M.A., Jarvik, M.E., Lee, G.S., Smith, E.C., Huang, J.C., Bota, R.G., Bartzokis, G., London, E.D., 2004. Differences between smoker s and nonsmokers in regional gray matter volumes and densities. Biological psychiatry 55, 77 84.

PAGE 137

121 Brunoni, A.R., Chaimani, A., Moffa, A.H., Razza, L.B., Gattaz, W.F., Daskalakis, Z.J., Carvalho, A.F., 2017. Repetitive Transcranial Magnetic Stimulation for t he Acute Treatment of Major Depressive Episodes: A Systematic Review With Network Meta analysis. JAMA psychiatry 74, 143 152. Buchel, C., Friston, K.J., 1997. Modulation of connectivity in visual pathways by attention: cortical interactions evaluated with structural equation modelling and fMRI. Cerebral cortex (New York, N.Y. : 1991) 7, 768 778. Buckner, R.L., Andrews Hanna, J.R., Schacter, D.L., 2008. The brain's default network: anatomy, function, and relevance to disease. Ann N Y Acad Sci 1124, 1 38. Bun gert, A., Antunes, A., Espenhahn, S., Thielscher, A., 2017. Where does TMS Stimulate the Motor Cortex? Combining Electrophysiological Measurements and Realistic Field Estimates to Reveal the Affected Cortex Position. Cerebral cortex (New York, N.Y. : 1991) 27, 5083 5094. Camchong, J., Macdonald, A.W., 3rd, Mueller, B.A., Nelson, B., Specker, S., Slaymaker, V., Lim, K.O., 2014. Changes in resting functional connectivity during abstinence in stimulant use disorder: a preliminary comparison of relapsers and ab stainers. Drug Alcohol Depend 139, 145 151. Camchong, J., Stenger, V.A., Fein, G., 2013. Resting state synchrony in long term abstinent alcoholics with versus without comorbid drug dependence. Drug Alcohol Depend 131, 56 65. Campbell Sills, L., Liverant, G .I., Brown, T.A., 2004. Psychometric evaluation of the behavioral inhibition/behavioral activation scales in a large sample of outpatients with anxiety and mood disorders. Psychol Assess 16, 244 254. Carver, C.S., White, T.L., 1994. Behavioral inhibition, behavioral activation, and affective responses to impending reward and punishment: The BIS/BAS Scales. Journal of personality and social psychology 67, 319. Cauda, F., Costa, T., Torta, D.M., Sacco, K., D'Agata, F., Duca, S., Geminiani, G., Fox, P.T., Verc elli, A., 2012. Meta analytic clustering of the insular cortex: characterizing the meta analytic connectivity of the insula when involved in active tasks. NeuroImage 62, 343 355. Chanes, L., Barrett, L.F., 2016. Redefining the Role of Limbic Areas in Corti cal Processing. Trends in cognitive sciences 20, 96 106. Chang, L.J., Yarkoni, T., Khaw, M.W., Sanfey, A.G., 2013. Decoding the role of the insula in human cognition: functional parcellation and large scale reverse inference. Cerebral cortex (New York, N.Y . : 1991) 23, 739 749. Chen, R., Classen, J., Gerloff, C., Celnik, P., Wassermann, E.M., Hallett, M., Cohen, L.G., 1997. Depression of motor cortex excitability by low frequency transcranial magnetic stimulation. Neurology 48, 1398 1403. Chen, S., Ross, T. J., Zhan, W., Myers, C.S., Chuang, K.S., Heishman, S.J., Stein, E.A., Yang, Y., 2008. Group independent component analysis reveals consistent resting state networks across multiple sessions. Brain research 1239, 141 151.

PAGE 138

122 Chiong, W., Wilson, S.M., D'Esposit o, M., Kayser, A.S., Grossman, S.N., Poorzand, P., Seeley, W.W., Miller, B.L., Rankin, K.P., 2013. The salience network causally influences default mode network activity during moral reasoning. Brain 136, 1929 1941. Chu, C., Hsu, A.L., Chou, K.H., Bandettini, P., Lin, C., Alzheimer's Disease Neuroimaging, I., 2012. Does feature selection improve classification accuracy? Impact of sample size and feature selection on classification using anatomical magnetic resonance images. NeuroImage 60, 59 70. Cia mpi de Andrade, D., Galhardoni, R., Pinto, L.F., Lancelotti, R., Rosi, J., Jr., Marcolin, M.A., Teixeira, M.J., 2012. Into the island: a new technique of non invasive cortical stimulation of the insula. Neurophysiologie clinique = Clinical neurophysiology 42, 363 368. Cirillo, G., Di Pino, G., Capone, F., Ranieri, F., Florio, L., Todisco, V., Tedeschi, G., Funke, K., Di Lazzaro, V., 2017. Neurobiological after effects of non invasive brain stimulation. Brain stimulation 10, 1 18. Claus, E.D., Blaine, S.K., Filbey, F.M., Mayer, A.R., Hutchison, K.E., 2013. Association between nicotine dependence severity, BOLD response to smoking cues, and functional connectivity. Neuropsychopharmacology : official publication of the American College of Neuropsychopharmacolog y 38, 2363 2372. Cohen Kadosh, K., Luo, Q., de Burca, C., Sokunbi, M.O., Feng, J., Linden, D.E.J., Lau, J.Y.F., 2016. Using real time fMRI to influence effective connectivity in the developing emotion regulation network. NeuroImage 125, 616 626. Coll, M.P. , Penton, T., Hobson, H., 2017. Important methodological issues regarding the use of transcranial magnetic stimulation to investigate interoceptive processing: a Comment on Pollatos et al. (2016). Philosophical transactions of the Royal Society of London. Series B, Biological sciences 372. Conger, R., 1997. The social context of substance abuse: a developmental perspective, in: Robertson, E.B., Sloboda, Z., Boyd, G.M., Beatty, L., Kozel, N.J. (Eds.), Rural substance abuse: state of knowledge and issues. NID A, Rockville MD, pp. pp6 36. Connolly, C.G., Bell, R.P., Foxe, J.J., Garavan, H., 2013. Dissociated grey matter changes with prolonged addiction and extended abstinence in cocaine users. PLoS One 8, e59645. Connolly, C.G., Foxe, J.J., Nierenberg, J., Shpan er, M., Garavan, H., 2012. The neurobiology of cognitive control in successful cocaine abstinence. Drug Alcohol Depend 121, 45 53. Cottler, L.B., Robins, L.N., Helzer, J.E., 1989. The reliability of the CIDI SAM: a comprehensive substance abuse interview. Br J Addict 84, 801 814. Cox, R.W., Chen, G., Glen, D.R., Reynolds, R.C., Taylor, P.A., 2017. FMRI Clustering in AFNI: False Positive Rates Redux. Brain connectivity 7, 152 171. Craig, A.D., 2002. How do you feel? Interoception: the sense of the physiologi cal condition of the body. Nature reviews. Neuroscience 3, 655 666. Craig, A.D., 2009. How do you feel -now? The anterior insula and human awareness. Nature reviews. Neuroscience 10, 59 70.

PAGE 139

123 Craig, A.D., 2011. Significance of the insula for the evolution of human awareness of feelings from the body. Ann N Y Acad Sci 1225, 72 82. Crawford, J.R., Henry, J.D., 2004. The positive and negative affect schedule (PANAS): construct validity, measurement properties and normative data in a large non clinical sample. Th e British journal of clinical psychology 43, 245 265. Crocq, M.A., 2007. Historical and cultural aspects of man's relationship with addictive drugs. Dialogues in clinical neuroscience 9, 355 361. D'Esposito, M., Deouell, L.Y., Gazzaley, A., 2003. Alteratio ns in the BOLD fMRI signal with ageing and disease: a challenge for neuroimaging. Nature reviews. Neuroscience 4, 863 872. Dannhauer, M., Lanfer, B., Wolters, C.H., Knosche, T.R., 2011. Modeling of the human skull in EEG source analysis. Human brain mappin g 32, 1383 1399. Daumann, J., Koester, P., Becker, B., Wagner, D., Imperati, D., Gouzoulis Mayfrank, E., Tittgemeyer, M., 2011. Medial prefrontal gray matter volume reductions in users of amphetamine type stimulants revealed by combined tract based spatial statistics and voxel based morphometry. Neuroimage 54, 794 801. Deen, B., Pitskel, N.B., Pelphrey, K.A., 2011. Three systems of insular functional connectivity identified with cluster analysis. Cerebral cortex (New York, N.Y. : 1991) 21, 1498 1506. Demirc i, O., Stevens, M.C., Andreasen, N.C., Michael, A., Liu, J., White, T., Pearlson, G.D., Clark, V.P., Calhoun, V.D., 2009. Investigation of relationships between fMRI brain networks in the spectral domain using ICA and Granger causality reveals distinct dif ferences between schizophrenia patients and healthy controls. NeuroImage 46, 419 431. Deng, Z.D., Lisanby, S.H., Peterchev, A.V., 2013. Electric field depth focality tradeoff in transcranial magnetic stimulation: simulation comparison of 50 coil designs. B rain stimulation 6, 1 13. Deshpande, G., Hu, X., 2012. Investigating effective brain connectivity from fMRI data: past findings and current issues with reference to Granger causality analysis. Brain connectivity 2, 235 245. Deshpande, G., Sathian, K., Hu, X., 2010a. Assessing and compensating for zero lag correlation effects in time lagged Granger causality analysis of FMRI. IEEE transactions on bio medical engineering 57, 1446 1456. Deshpande, G., Sathian, K., Hu, X., 2010b. Effect of hemodynamic variabili ty on Granger causality analysis of fMRI. NeuroImage 52, 884 896. Di Chiara, G., Imperato, A., 1988. Drugs abused by humans preferentially increase synaptic dopamine concentrations in the mesolimbic system of freely moving rats. Proceedings of the National Academy of Sciences of the United States of America 85, 5274 5278. Dias, N.R., Peechatka, A.L., Janes, A.C., 2016. Insula reactivity to negative stimuli is associated with daily cigarette use: A preliminary investigation using the Human Connectome Databas e. Drug Alcohol Depend 159, 277 280.

PAGE 140

124 Diedrichsen, J., Balsters, J.H., Flavell, J., Cussans, E., Ramnani, N., 2009. A probabilistic MR atlas of the human cerebellum. NeuroImage 46, 39 46. Dieler, A.C., Dresler, T., Joachim, K., Deckert, J., Herrmann, M.J., Fallgatter, A.J., 2014. Can intermittent theta burst stimulation as add on to psychotherapy improve nicotine abstinence? Results from a pilot study. European addiction research 20, 248 253. Diez, I., Erramuzpe, A., Escudero, I., Mateos, B., Cabrera, A., Marinazzo, D., Sanz Arigita, E.J., Stramaglia, S., Cortes Diaz, J.M., 2015. Information Flow Between Resting State Networks. Brain connectivity 5, 554 564. Ding, X., Lee, S.W., 2013. Changes of functional and effective connectivity in smoking replenishment on deprived heavy smokers: a resting state FMRI study. PloS one 8, e59331. Dinur Klein, L., Dannon, P., Hadar, A., Rosenberg, O., Roth, Y., Kotler, M., Zangen, A., 2014. Smoking cessation induced by deep repetitive transcranial magnetic stim ulation of the prefrontal and insular cortices: a prospective, randomized controlled trial. Biological psychiatry 76, 742 749. Dosenbach, N.U., Fair, D.A., Miezin, F.M., Cohen, A.L., Wenger, K.K., Dosenbach, R.A., Fox, M.D., Snyder, A.Z., Vincent, J.L., Ra ichle, M.E., Schlaggar, B.L., Petersen, S.E., 2007. Distinct brain networks for adaptive and stable task control in humans. Proceedings of the National Academy of Sciences of the United States of America 104, 11073 11078. Draganski, B., Kherif, F., 2013. I n vivo assessment of use dependent brain plasticity -beyond the "one trick pony" imaging strategy. NeuroImage 73, 255 259; discussion 265 257. Eichhammer, P., Johann, M., Kharraz, A., Binder, H., Pittrow, D., Wodarz, N., Hajak, G., 2003. High frequency rep etitive transcranial magnetic stimulation decreases cigarette smoking. The Journal of clinical psychiatry 64, 951 953. Eklund, A., Nichols, T.E., Knutsson, H., 2016. Cluster failure: Why fMRI inferences for spatial extent have inflated false positive rates . Proceedings of the National Academy of Sciences of the United States of America 113, 7900 7905. Engelmann, J.M., Versace, F., Robinson, J.D., Minnix, J.A., Lam, C.Y., Cui, Y., Brown, V.L., Cinciripini, P.M., 2012. Neural substrates of smoking cue reactiv ity: a meta analysis of fMRI studies. NeuroImage 60, 252 262. Ersche, K.D., Jones, P.S., Williams, G.B., Turton, A.J., Robbins, T.W., Bullmore, E.T., 2012. Abnormal brain structure implicated in stimulant drug addiction. Science 335, 601 604. Ersche, K.D., Williams, G.B., Robbins, T.W., Bullmore, E.T., 2013. Meta analysis of structural brain abnormalities associated with stimulant drug dependence and neuroimaging of addiction vulnerability and resilience. Current opinion in neurobiology 23, 615 624. Fein, G ., Di Sclafani, V., Meyerhoff, D.J., 2002. Prefrontal cortical volume reduction associated with frontal cortex function deficit in 6 week abstinent crack cocaine dependent men. Drug Alcohol Depend 68, 87 93.

PAGE 141

125 Feng, C., Deshpande, G., Liu, C., Gu, R., Luo, Y .J., Krueger, F., 2016. Diffusion of responsibility attenuates altruistic punishment: A functional magnetic resonance imaging effective connectivity study. Human brain mapping 37, 663 677. Ferguson, S.G., Shiffman, S., 2009. The relevance and treatment of cue induced cravings in tobacco dependence. Journal of substance abuse treatment 36, 235 243. Fergusson, D.M., Horwood, L.J., Beautrais, A.L., 2003. Cannabis and educational achievement. Addiction 98, 1681 1692. Fields, R.D., 2013. Changes in brain structu re during learning: fact or artifact? Reply to Thomas and Baker. NeuroImage 73, 260 264; discussion 265 267. Forget, B., Pushparaj, A., Le Foll, B., 2010. Granular insular cortex inactivation as a novel therapeutic strategy for nicotine addiction. Biologic al psychiatry 68, 265 271. Fox, M.D., Greicius, M., 2010. Clinical applications of resting state functional connectivity. Frontiers in systems neuroscience 4, 19. Franklin, T.R., Acton, P.D., Maldjian, J.A., Gray, J.D., Croft, J.R., Dackis, C.A., O'Brien, C.P., Childress, A.R., 2002. Decreased gray matter concentration in the insular, orbitofrontal, cingulate, and temporal cortices of cocaine patients. Biological psychiatry 51, 134 142. Friston, K.J., Harrison, L., Penny, W., 2003. Dynamic causal modelling. NeuroImage 19, 1273 1302. Gartlehner, G., Hansen, R.A., Nissman, D., Lohr, K.N., Carey, T.S., 2006. AHRQ Technical Reviews, Criteria for Distinguishing Effectiveness From Efficacy Trials in Systematic Reviews. Agency for Healthcare Research and Quality (U S), Rockville (MD). Gelhorn, H., Hartman, C., Sakai, J., Stallings, M., Young, S., Rhee, S.H., Corley, R., Hewitt, J., Hopfer, C., Crowley, T., 2008. Toward DSM V: an item response theory analysis of the diagnostic process for DSM IV alcohol abuse and depe ndence in adolescents. J Am Acad Child Adolesc Psychiatry 47, 1329 1339. Gentry, C.L., Lukas, R.J., 2002. Regulation of nicotinic acetylcholine receptor numbers and function by chronic nicotine exposure. Current drug targets. CNS and neurological disorders 1, 359 385. Ginestet, C.E., Nichols, T.E., Bullmore, E.T., Simmons, A., 2011. Brain network analysis: separating cost from topology using cost integration. PloS one 6, e21570. Good, C.D., Johnsrude, I.S., Ashburner, J., Henson, R.N., Friston, K.J., Fracko wiak, R.S., 2001. A voxel based morphometric study of ageing in 465 normal adult human brains. NeuroImage 14, 21 36. Grant, J.D., Scherrer, J.F., Lynskey, M.T., Agrawal, A., Duncan, A.E., Haber, J.R., Heath, A.C., Bucholz, K.K., 2012. Associations of alcoh ol, nicotine, cannabis, and drug use/dependence with educational attainment: evidence from cotwin control analyses. Alcoholism, clinical and experimental research 36, 1412 1420. Greenfield, S.F., Back, S.E., Lawson, K., Brady, K.T., 2010. Substance abuse i n women. Psychiatr Clin North Am 33, 339 355.

PAGE 142

126 Griffin, M.L., Weiss, R.D., Mirin, S.M., Lange, U., 1989. A comparison of male and female cocaine abusers. Arch Gen Psychiatry 46, 122 126. Gu, H., Salmeron, B.J., Ross, T.J., Geng, X., Zhan, W., Stein, E.A., Yang, Y., 2010. Mesocorticolimbic circuits are impaired in chronic cocaine users as demonstrated by resting state functional connectivity. NeuroImage 53, 593 601. Gu, X., Lohrenz, T., Salas, R., Baldwin, P.R., Soltani, A., Kirk, U., Cinciripini, P.M. , Montague, P.R., 2016. Belief about Nicotine Modulates Subjective Craving and Insula Activity in Deprived Smokers. Frontiers in psychiatry 7, 126. Hahn, B., Ross, T.J., Yang, Y., Kim, I., Huestis, M.A., Stein, E.A., 2007. Nicotine enhances visuospatial at tention by deactivating areas of the resting brain default network. The Journal of neuroscience : the official journal of the Society for Neuroscience 27, 3477 3489. Hanlon, C.A., Dowdle, L.T., Correia, B., Mithoefer, O., Kearney Ramos, T., Lench, D., Grif fin, M., Anton, R.F., George, M.S., 2017. Left frontal pole theta burst stimulation decreases orbitofrontal and insula activity in cocaine users and alcohol users. Drug Alcohol Depend 178, 310 317. Hanlon, C.A., Dowdle, L.T., Gibson, N.B., Li, X., Hamilton , S., Canterberry, M., Hoffman, M., 2018. Cortical substrates of cue reactivity in multiple substance dependent populations: transdiagnostic relevance of the medial prefrontal cortex. Translational psychiatry 8, 186. Hawkins, J.D., Catalano, R.F., Miller, J.Y., 1992. Risk and protective factors for alcohol and other drug problems in adolescence and early adulthood: implications for substance abuse prevention. Psychol Bull 112, 64 105. Health, U.D.o., Services, H., 2014. The health consequences of smoking 50 years of progress: a report of the Surgeon General. Atlanta, GA: US Department of Health and Human Services, Centers for Disease Hillman, E.M., 2014. Coupling mechanism and significance of the BOLD signal: a status report. Annual review of neuroscience 37, 161 181. Hoogendam, J.M., Ramakers, G.M., Di Lazzaro, V., 2010. Physiology of repetitive transcranial magnetic stimulation of the human brain. Brain stimulation 3, 95 118. Huang, W., Shen, F., Zhang, J., Xing, B., 2016. Effect of Repetitive Transcrani al Magnetic Stimulation on Cigarette Smoking in Patients with Schizophrenia. Shanghai archives of psychiatry 28, 309 317. Hyatt, C.J., Assaf, M., Muska, C.E., Rosen, R.I., Thomas, A.D., Johnson, M.R., Hylton, J.L., Andrews, M.M., Reynolds, B.A., Krystal, J .H., Potenza, M.N., Pearlson, G.D., 2012. Reward related dorsal striatal activity differences between former and current cocaine dependent individuals during an interactive competitive game. PloS one 7, e34917. Jackson, K.J., Muldoon, P.P., De Biasi, M., D amaj, M.I., 2015. New mechanisms and perspectives in nicotine withdrawal. Neuropharmacology 96, 223 234. Janes, A.C., Farmer, S., Peechatka, A.L., Frederick Bde, B., Lukas, S.E., 2015a. Insula Dorsal Anterior Cingulate Cortex Coupling is Associated with En hanced Brain Reactivity to Smoking Cues.

PAGE 143

127 Neuropsychopharmacology : official publication of the American College of Neuropsychopharmacology 40, 1561 1568. Janes, A.C., Gilman, J.M., Radoman, M., Pachas, G., Fava, M., Evins, A.E., 2017. Revisiting the role o f the insula and smoking cue reactivity in relapse: A replication and extension of neuroimaging findings. Drug Alcohol Depend 179, 8 12. Janes, A.C., Pizzagalli, D.A., Richardt, S., de, B.F.B., Chuzi, S., Pachas, G., Culhane, M.A., Holmes, A.J., Fava, M., Evins, A.E., Kaufman, M.J., 2010. Brain reactivity to smoking cues prior to smoking cessation predicts ability to maintain tobacco abstinence. Biological psychiatry 67, 722 729. Janes, A.C., Ross, R.S., Farmer, S., Frederick, B.B., Nickerson, L.D., Lukas, S.E., Stern, C.E., 2015b. Memory retrieval of smoking related images induce greater insula activation as revealed by an fMRI based delayed matching to sample task. Addict Biol 20, 349 356. Jasinska, A.J., Stein, E.A., Kaiser, J., Naumer, M.J., Yalachkov, Y ., 2014. Factors modulating neural reactivity to drug cues in addiction: a survey of human neuroimaging studies. Neurosci Biobehav Rev 38, 1 16. Johann, M., Wiegand, R., Kharraz, A., Bobbe, G., Sommer, G., Hajak, G., Wodarz, N., Eichhammer, P., 2003. [Tran scranial magnetic stimulation for nicotine dependence]. Psychiatrische Praxis 30 Suppl 2, S129 131. Kedzior, K.K., Gerkensmeier, I., Schuchinsky, M., 2018. Can deep transcranial magnetic stimulation (DTMS) be used to treat substance use disorders (SUD)? A systematic review. BMC psychiatry 18, 137. Kelly, C., Toro, R., Di Martino, A., Cox, C.L., Bellec, P., Castellanos, F.X., Milham, M.P., 2012. A convergent functional architecture of the insula emerges across imaging modalities. NeuroImage 61, 1129 1142. Ke ndler, K.S., Ohlsson, H., Sundquist, K., Sundquist, J., 2013. Within family environmental transmission of drug abuse: a Swedish national study. JAMA psychiatry 70, 235 242. Kerr, W.T., Douglas, P.K., Anderson, A., Cohen, M.S., 2014. The utility of data dri ven feature selection: re: Chu et al. 2012. NeuroImage 84, 1107 1110. Kheyfets, V., 2016. University of Colorado Graduate School Course, Bioengineering 5021: Numerical Methods. Koob, G.F., 1992. Drugs of abuse: anatomy, pharmacology and function of reward pathways. Trends Pharmacol. Sci. 13, 177 184. Koob, G.F., Volkow, N.D., 2010. Neurocircuitry of addiction. Neuropsychopharmacology : official publication of the American College of Neuropsychopharmacology 35, 217 238. Kosten, T.A., Gawin, F.H., Kosten, T.R ., Rounsaville, B.J., 1993. Gender differences in cocaine use and treatment response. J Subst Abuse Treat 10, 63 66. Kozak, K., Sharif Razi, M., Morozova, M., Gaudette, E.V., Barr, M.S., Daskalakis, Z.J., Blumberger, D.M., George, T.P., 2018. Effects of sh ort term, high frequency repetitive transcranial magnetic

PAGE 144

128 stimulation to bilateral dorsolateral prefrontal cortex on smoking behavior and cognition in patients with schizophrenia and non psychiatric controls. Schizophrenia research. Krmpotich, T.D., Tregel las, J.R., Thompson, L.L., Banich, M.T., Klenk, A.M., Tanabe, J.L., 2013. Resting state activity in the left executive control network is associated with behavioral approach and is increased in substance dependence. Drug Alcohol Depend 129, 1 7. Kuhn, J., Bauer, R., Pohl, S., Lenartz, D., Huff, W., Kim, E.H., Klosterkoetter, J., Sturm, V., 2009. Observations on unaided smoking cessation after deep brain stimulation of the nucleus accumbens. European addiction research 15, 196 201. Kühn, S., Gallinat, J. , 2011. Common biology of craving across legal and illegal drugs a quantitative meta analysis of cue reactivity brain response. European Journal of Neuroscience 33, 1318 1326. Kurth, F., Zilles, K., Fox, P.T., Laird, A.R., Eickhoff, S.B., 2010. A link be tween the systems: functional differentiation and integration within the human insula revealed by meta analysis. Brain structure & function 214, 519 534. Kutlu, M.G., Burke, D., Slade, S., Hall, B.J., Rose, J.E., Levin, E.D., 2013. Role of insular cortex D (1) and D(2) dopamine receptors in nicotine self administration in rats. Behavioural brain research 256, 273 278. Lenoir, C., Algoet, M., Mouraux, A., 2018. Deep continuous theta burst stimulation of the operculo insular cortex selectively affects Adelta f ibre heat pain. The Journal of physiology 596, 4767 4787. Lerman, C., Gu, H., Loughead, J., Ruparel, K., Yang, Y., Stein, E.A., 2014. Large scale brain network coupling predicts acute nicotine abstinence effects on craving and cognitive function. JAMA psyc hiatry 71, 523 530. Lewis, M., 2018. Brain Change in Addiction as Learning, Not Disease. The New England journal of medicine 379, 1551 1560. Li, C.S., Sinha, R., 2008. Inhibitory control and emotional stress regulation: neuroimaging evidence for frontal li mbic dysfunction in psycho stimulant addiction. Neurosci Biobehav Rev 32, 581 597. Li, W., Mai, X., Liu, C., 2014. The default mode network and social understanding of others: what do brain connectivity studies tell us. Frontiers in human neuroscience 8, 7 4. Li, X., Hartwell, K.J., Owens, M., Lematty, T., Borckardt, J.J., Hanlon, C.A., Brady, K.T., George, M.S., 2013. Repetitive transcranial magnetic stimulation of the dorsolateral prefrontal cortex reduces nicotine cue craving. Biological psychiatry 73, 71 4 720. Li, X., Sahlem, G.L., Badran, B.W., McTeague, L.M., Hanlon, C.A., Hartwell, K.J., Henderson, S., George, M.S., 2017. Transcranial magnetic stimulation of the dorsal lateral prefrontal cortex inhibits medial orbitofrontal activity in smokers. The Ame rican journal on addictions 26, 788 794. Li, Y.O., Adali, T., Calhoun, V.D., 2007. Estimating the number of independent components for functional magnetic resonance imaging data. Human brain mapping 28, 1251 1266.

PAGE 145

129 Lin, D.J., Wong, T.T., Ciavarra, G.A., Kaz am, J.K., 2017. Adventures and Misadventures in Plastic Surgery and Soft Tissue Implants. Radiographics 37, 2145 2163. Lind, K.E., Gutierrez, E.J., Yamamoto, D.J., Regner, M.F., McKee, S.A., Tanabe, J., 2017. Sex disparities in substance abuse research: Ev aluating 23 years of structural neuroimaging studies. Drug Alcohol Depend 173, 92 98. Lu, H., Zou, Q., Chefer, S., Ross, T.J., Vaupel, D.B., Guillem, K., Rea, W.P., Yang, Y., Peoples, L.L., Stein, E.A., 2014. Abstinence from cocaine and sucrose self admini stration reveals altered mesocorticolimbic circuit connectivity by resting state MRI. Brain connectivity 4, 499 510. Luijten, M., Schellekens, A.F., Kuhn, S., Machielse, M.W., Sescousse, G., 2017. Disruption of Reward Processing in Addiction : An Image Bas ed Meta analysis of Functional Magnetic Resonance Imaging Studies. JAMA psychiatry 74, 387 398. Lynch, W.J., 2006. Sex differences in vulnerability to drug self administration. Exp Clin Psychopharmacol 14, 34 41. Mackey, S., Paulus, M., 2013. Are there vol umetric brain differences associated with the use of cocaine and amphetamine type stimulants? Neuroscience and Biobehavioral Reviews 37, 300 316. Makani, R., Pradhan, B., Shah, U., Parikh, T., 2017. Role of Repetitive Transcranial Magnetic Stimulation (rTM S) in Treatment of Addiction and Related Disorders: A Systematic Review. Current drug abuse reviews 10, 31 43. Malik, S., Jacobs, M., Cho, S.S., Boileau, I., Blumberger, D., Heilig, M., Wilson, A., Daskalakis, Z.J., Strafella, A.P., Zangen, A., Le Foll, B. , 2018. Deep TMS of the insula using the H coil modulates dopamine release: a crossover [(11)C] PHNO PET pilot trial in healthy humans. Brain imaging and behavior 12, 1306 1317. Mantini, D., Vanduffel, W., 2013. Emerging roles of the brain's default networ k. Neuroscientist 19, 76 87. Mayer, A.R., Wilcox, C.E., Teshiba, T.M., Ling, J.M., Yang, Z., 2013. Hyperactivation of the cognitive control network in cocaine use disorders during a multisensory Stroop task. Drug Alcohol Depend 133, 235 241. McHugh, M.J., Demers, C.H., Salmeron, B.J., Devous, M.D., Sr., Stein, E.A., Adinoff, B., 2014. Cortico amygdala coupling as a marker of early relapse risk in cocaine addicted individuals. Frontiers in psychiatry 5, 16. Mendelson, J.H., Weiss, R., Griffin, M., Mirin, S.M ., Teoh, S.K., Mello, N.K., Lex, B.W., 1991. Some special considerations for treatment of drug abuse and dependence in women. NIDA Res Monogr 106, 313 327. Meunier, D., Ersche, K.D., Craig, K.J., Fornito, A., Merlo Pich, E., Fineberg, N.A., Shabbir, S.S., Robbins, T.W., Bullmore, E.T., 2012. Brain functional connectivity in stimulant drug dependence and obsessive compulsive disorder. NeuroImage 59, 1461 1468. Milev, R.V., Giacobbe, P., Kennedy, S.H., Blumberger, D.M., Daskalakis, Z.J., Downar, J., Modirrous ta, M., Patry, S., Vila Rodriguez, F., Lam, R.W., MacQueen, G.M., Parikh, S.V., Ravindran, A.V., 2016.

PAGE 146

130 Canadian Network for Mood and Anxiety Treatments (CANMAT) 2016 Clinical Guidelines for the Management of Adults with Major Depressive Disorder: Section 4 . Neurostimulation Treatments. Canadian journal of psychiatry. Revue canadienne de psychiatrie 61, 561 575. Miller, T.R., Hendrie, D., 2009. Substance abuse prevention dollars and cents: A cost benefit analysis. US Department of Health and Human Services, Substance Abuse and Mental Health Services Administration, Center for Substance Abuse Prevention. Miniussi, C., Harris, J.A., Ruzzoli, M., 2013. Modelling non invasive brain stimulation in cognitive neuroscience. Neurosci Biobehav Rev 37, 1702 1712. Mitchell, M.R., Balodis, I.M., Devito, E.E., Lacadie, C.M., Yeston, J., Scheinost, D., Constable, R.T., Carroll, K.M., Potenza, M.N., 2013. A preliminary investigation of Stroop related intrinsic connectivity in cocaine dependence: associations with treatm ent outcomes. The American journal of drug and alcohol abuse 39, 392 402. Moran Santa Maria, M.M., Hartwell, K.J., Hanlon, C.A., Canterberry, M., Lematty, T., Owens, M., Brady, K.T., George, M.S., 2015. Right anterior insula connectivity is important for c ue induced craving in nicotine dependent smokers. Addict Biol 20, 407 414. Moran, L.V., Sampath, H., Stein, E.A., Hong, L.E., 2012. Insular and anterior cingulate circuits in smokers with schizophrenia. Schizophrenia research 142, 223 229. Murnane, K.S., G opinath, K.S., Maltbie, E., Daunais, J.B., Telesford, Q.K., Howell, L.L., 2015. Functional connectivity in frontal striatal brain networks and cocaine self administration in female rhesus monkeys. Psychopharmacology 232, 745 754. Murphy, K., Birn, R.M., Ba ndettini, P.A., 2013. Resting state fMRI confounds and cleanup. NeuroImage 80, 349 359. Muschelli, J., Nebel, M.B., Caffo, B.S., Barber, A.D., Pekar, J.J., Mostofsky, S.H., 2014. Reduction of motion related artifacts in resting state fMRI using aCompCor. N euroImage 96, 22 35. Mutschler, I., Wieckhorst, B., Kowalevski, S., Derix, J., Wentlandt, J., Schulze Bonhage, A., Ball, T., 2009. Functional organization of the human anterior insular cortex. Neurosci Lett 457, 66 70. Naqvi, N.H., Gaznick, N., Tranel, D., Bechara, A., 2014. The insula: a critical neural substrate for craving and drug seeking under conflict and risk. Ann N Y Acad Sci 1316, 53 70. Naqvi, N.H., Rudrauf, D., Damasio, H., Bechara, A., 2007. Damage to the insula disrupts addiction to cigarette s moking. Science 315, 531 534. Nestler, E., Hyman, S., Holtzman, D., Malenka, R., 2015. Molecular Neuropharmacology: A Foundation for Clinical Neuroscience. McGraw Hill. Nestor, L., McCabe, E., Jones, J., Clancy, L., Garavan, H., 2011. Differences in "botto m up" and "top down" neural activity in current and former cigarette smokers: Evidence for neural substrates which may promote nicotine abstinence through increased cognitive control. NeuroImage 56, 2258 2275.

PAGE 147

131 Nieto Castanon, A., Ghosh, S.S., Tourville, J. A., Guenther, F.H., 2003. Region of interest based analysis of functional imaging data. NeuroImage 19, 1303 1316. Nieuwenhuys, R., 2012. The insular cortex: a review. Progress in brain research 195, 123 163. NIH, NIH OER Definition of Phase I Clinical Tria l. National Institutes of Health Office of Extramural Research. Noel, X., Brevers, D., Bechara, A., 2013. A neurocognitive approach to understanding the neurobiology of addiction. Current opinion in neurobiology 23, 632 638. O'Brien, C.P., Childress, A.R ., McLellan, A.T., Ehrman, R., 1992. Classical conditioning in drug dependent humans. Ann. N. Y. Acad. Sci. 654, 400 415. Opitz, A., Windhoff, M., Heidemann, R.M., Turner, R., Thielscher, A., 2011. How the brain tissue shapes the electric field induced by transcranial magnetic stimulation. NeuroImage 58, 849 859. Pan, P., Shi, H., Zhong, J., Xiao, P., Shen, Y., Wu, L., Song, Y., He, G., 2013. Chronic smoking and brain gray matter changes: evidence from meta analysis of voxel based morphometry studies. Neuro l. Sci. 34, 813 817. Patton, J.H., Stanford, M.S., 1995. Factor structure of the Barratt impulsiveness scale. Journal of clinical psychology 51, 768 774. Patton, J.H., Stanford, M.S., Barratt, E.S., 1995. Factor structure of the Barratt impulsiveness scale . J Clin Psychol 51, 768 774. Perry, R.I., Krmpotich, T., Thompson, L.L., Mikulich Gilbertson, S.K., Banich, M.T., Tanabe, J., 2013. Sex modulates approach systems and impulsivity in substance dependence. Drug Alcohol Depend 133, 222 227. Pollatos, O., Her bert, B.M., Mai, S., Kammer, T., 2016. Changes in interoceptive processes following brain stimulation. Philosophical transactions of the Royal Society of London. Series B, Biological sciences 371. Pollatos, O., Kammer, T., 2017. Reply to Coll et al. 'Impor tant methodological issues regarding the use of transcranial magnetic stimulation to investigate interoceptive processing' (2017). Philosophical transactions of the Royal Society of London. Series B, Biological sciences 372. Potenza, M.N., Hong, K.I., Laca die, C.M., Fulbright, R.K., Tuit, K.L., Sinha, R., 2012. Neural correlates of stress induced and cue induced drug craving: influences of sex and cocaine dependence. The American journal of psychiatry 169, 406 414. Potvin, S., Tikasz, A., Dinh Williams, L.L ., Bourque, J., Mendrek, A., 2015. Cigarette Cravings, Impulsivity, and the Brain. Frontiers in psychiatry 6, 125. Power, J.D., Mitra, A., Laumann, T.O., Snyder, A.Z., Schlaggar, B.L., Petersen, S.E., 2014. Methods to detect, characterize, and remove motio n artifact in resting state fMRI. NeuroImage 84, 320 341.

PAGE 148

132 Pripfl, J., Tomova, L., Riecansky, I., Lamm, C., 2014. Transcranial magnetic stimulation of the left dorsolateral prefrontal cortex decreases cue induced nicotine craving and EEG delta power. Brain stimulation 7, 226 233. Pushparaj, A., Hamani, C., Yu, W., Shin, D.S., Kang, B., Nobrega, J.N., Le Foll, B., 2013. Electrical stimulation of the insular region attenuates nicotine taking and nicotine seeking behaviors. Neuropsychopharmacology : official pu blication of the American College of Neuropsychopharmacology 38, 690 698. Radua, J., Canales Rodriguez, E.J., Pomarol Clotet, E., Salvador, R., 2014. Validity of modulation and optimal settings for advanced voxel based morphometry. NeuroImage 86, 81 90. Rando, K., Tuit, K., Hannestad, J., Guarnaccia, J., Sinha, R., 2013. Sex differences in decreased limbic and cortical grey matter volume in cocaine dependence: a voxel based morphometric study. Addict Biol 18, 147 160. Regner, M.F., Dalwani, M., Yam amoto, D., Perry, R.I., Sakai, J.T., Honce, J.M., Tanabe, J., 2015. Sex Differences in Gray Matter Changes and Brain Behavior Relationships in Patients with Stimulant Dependence. Radiology 277, 801 812. Regner, M.F., Saenz, N., Maharajh, K., Yamamoto, D.J. , Mohl, B., Wylie, K., Tregellas, J., Tanabe, J., 2016. Top Down Network Effective Connectivity in Abstinent Substance Dependent Individuals. PloS one 11, e0164818. Robins, L., Cottler, L., Bucholz, K., Compton, W., 1995a. The diagnostic interview schedule , version IV. St. Louis, MO: Washington University. Robins, L., Cottler, L., Bucholz, K., Compton, W., North, C., Rourke, K., 1995b. Diagnostic interview schedule for DSM IV. St. Louis, MO: Washington University School of Medicine. Robinson, T.E., Berridge , K.C., 2008. Review. The incentive sensitization theory of addiction: some current issues. Philosophical transactions of the Royal Society of London. Series B, Biological sciences 363, 3137 3146. Rubinov, M., Sporns, O., 2010. Complex network measures of brain connectivity: uses and interpretations. NeuroImage 52, 1059 1069. Salling, M.C., Martinez, D., 2016. Brain Stimulation in Addiction. Neuropsychopharmacology : official publication of the American College of Neuropsychopharmacology 41, 2798 2809. Samh sa, 2018. Risk and Protective Factors. Sara, G., Burgess, P., Harris, M., Malhi, G.S., Whiteford, H., Hall, W., 2012. Stimulant use disorders: characteristics and comorbidity in an Australian population sample. The Australian and New Zealand journal of psy chiatry 46, 1173 1181. Saunders, B.T., Robinson, T.E., 2013. Individual variation in resisting temptation: implications for addiction. Neurosci Biobehav Rev 37, 1955 1975. Scaramella, L.V., Keyes, A.W., 2001. The social contextual approach and rural adoles cent substance use: implications for prevention in rural settings. Clin Child Fam Psychol Rev 4, 231 251.

PAGE 149

133 Schippers, M.B., Renken, R., Keysers, C., 2011. The effect of intra and inter subject variability of hemodynamic responses on group level Granger cau sality analyses. NeuroImage 57, 22 36. Schwarz, A.J., Gozzi, A., Chessa, A., Bifone, A., 2012. Voxel scale complex networks of functional connectivity in the rat brain: neurochemical state dependence of global and local topological properties. Computationa l and mathematical methods in medicine 2012, 615709. Seeley, W.W., Menon, V., Schatzberg, A.F., Keller, J., Glover, G.H., Kenna, H., Reiss, A.L., Greicius, M.D., 2007. Dissociable intrinsic connectivity networks for salience processing and executive contro l. The Journal of neuroscience : the official journal of the Society for Neuroscience 27, 2349 2356. Sescousse, G., Caldu, X., Segura, B., Dreher, J.C., 2013. Processing of primary and secondary rewards: a quantitative meta analysis and review of human fun ctional neuroimaging studies. Neurosci Biobehav Rev 37, 681 696. Seth, A.K., 2010. A MATLAB toolbox for Granger causal connectivity analysis. Journal of neuroscience methods 186, 262 273. Seth, A.K., Chorley, P., Barnett, L.C., 2013. Granger causality anal ysis of fMRI BOLD signals is invariant to hemodynamic convolution but not downsampling. NeuroImage 65, 540 555. Sheffer, C.E., Bickel, W.K., Brandon, T.H., Franck, C.T., Deen, D., Panissidi, L., Abdali, S.A., Pittman, J.C., Lunden, S.E., Prashad, N., Malho tra, R., Mantovani, A., 2018. Preventing relapse to smoking with transcranial magnetic stimulation: Feasibility and potential efficacy. Drug Alcohol Depend 182, 8 18. Shirer, W.R., Ryali, S., Rykhlevskaia, E., Menon, V., Greicius, M.D., 2012. Decoding subj ect driven cognitive states with whole brain connectivity patterns. Cerebral cortex (New York, N.Y. : 1991) 22, 158 165. Sim, M.E., Lyoo, I.K., Streeter, C.C., Covell, J., Sarid Segal, O., Ciraulo, D.A., Kim, M.J., Kaufman, M.J., Yurgelun Todd, D.A., Rensh aw, P.F., 2007. Cerebellar gray matter volume correlates with duration of cocaine use in cocaine dependent subjects. Neuropsychopharmacology : official publication of the American College of Neuropsychopharmacology 32, 2229 2237. Smith, S.M., Bandettini, P .A., Miller, K.L., Behrens, T.E., Friston, K.J., David, O., Liu, T., Woolrich, M.W., Nichols, T.E., 2012. The danger of systematic bias in group level FMRI lag based causality estimation. NeuroImage 59, 1228 1229. Smith, S.M., Fox, P.T., Miller, K.L., Glah n, D.C., Fox, P.M., Mackay, C.E., Filippini, N., Watkins, K.E., Toro, R., Laird, A.R., Beckmann, C.F., 2009. Correspondence of the brain's functional architecture during activation and rest. Proceedings of the National Academy of Sciences of the United Sta tes of America 106, 13040 13045. Smith, S.M., Nichols, T.E., 2009. Threshold free cluster enhancement: addressing problems of smoothing, threshold dependence and localisation in cluster inference. NeuroImage 44, 83 98. Song, S., Zilverstand, A., Gui, W., L i, H.J., Zhou, X., 2018. Effects of single session versus multi session non invasive brain stimulation on craving and consumption in individuals with drug addiction, eating disorders or obesity: A meta analysis. Brain stimulation.

PAGE 150

134 Spagnolo, P.A., Wang, H., Srivanitchapoom, P., Schwandt, M., Heilig, M., Hallett, M., 2018. Lack of Target Engagement Following Low Frequency Deep Transcranial Magnetic Stimulation of the Anterior Insula. Neuromodulation : journal of the International Neuromodulation Society. Spie th, P.M., Kubasch, A.S., Penzlin, A.I., Illigens, B.M., Barlinn, K., Siepmann, T., 2016. Randomized controlled trials a matter of design. Neuropsychiatric disease and treatment 12, 1341 1349. Stead, L.F., Lancaster, T., 2012. Combined pharmacotherapy and behavioural interventions for smoking cessation. The Cochrane database of systematic reviews 10, Cd008286. Stein, E.A., Pankiewicz, J., Harsch, H.H., Cho, J.K., Fuller, S.A., Hoffmann, R.G., Hawkins, M., Rao, S.M., Bandettini, P.A., Blo om, A.S., 1998. Nicotine induced limbic cortical activation in the human brain: a functional MRI study. The American journal of psychiatry 155, 1009 1015. Stevens, M.C., Pearlson, G.D., Calhoun, V.D., 2009. Changes in the interaction of resting state neura l networks from adolescence to adulthood. Human brain mapping 30, 2356 2366. Stinson, F.S., Grant, B.F., Dawson, D.A., Ruan, W.J., Huang, B., Saha, T., 2005. Comorbidity between DSM IV alcohol and specific drug use disorders in the United States: results f rom the National Epidemiologic Survey on Alcohol and Related Conditions. Drug Alcohol Depend 80, 105 116. Stone, A.L., Becker, L.G., Huber, A.M., Catalano, R.F., 2012. Review of risk and protective factors of substance use and problem use in emerging adult hood. Addictive behaviors 37, 747 775. Suner Soler, R., Grau, A., Gras, M.E., Font Mayolas, S., Silva, Y., Davalos, A., Cruz, V., Rodrigo, J., Serena, J., 2012. Smoking cessation 1 year poststroke and damage to the insular cortex. Stroke; a journal of cere bral circulation 43, 131 136. Sutherland, M.T., McHugh, M.J., Pariyadath, V., Stein, E.A., 2012. Resting state functional connectivity in addiction: Lessons learned and a road ahead. NeuroImage 62, 2281 2295. Sutherland, M.T., Ray, K.L., Riedel, M.C., Yane s, J.A., Stein, E.A., Laird, A.R., 2015. Neurobiological impact of nicotinic acetylcholine receptor agonists: an activation likelihood estimation meta analysis of pharmacologic neuroimaging studies. Biological psychiatry 78, 711 720. Swaim, R.C., Beauvais, F., Chavez, E.L., Oetting, E.R., 1997. The effect of school dropout rates on estimates of adolescent substance use among three racial/ethnic groups. American journal of public health 87, 51 55. Tanabe, J., Nyberg, E., Martin, L.F., Martin, J., Cordes, D., Kronberg, E., Tregellas, J.R., 2011. Nicotine effects on default mode network during resting state. Psychopharmacology 216, 287 295. Tanabe, J., Tregellas, J.R., Dalwani, M., Thompson, L., Owens, E., Crowley, T., Banich, M., 2009a. Medial orbitofrontal co rtex gray matter is reduced in abstinent substance dependent individuals. Biological psychiatry 65, 160 164. Tanabe, J., Tregellas, J.R., Dalwani, M., Thompson, L., Owens, E., Crowley, T., Banich, M., 2009b. Medial orbitofrontal cortex gray matter is reduc ed in abstinent substance dependent individuals. Biological psychiatry 65, 160 164.

PAGE 151

135 Tanabe, J., York, P., Krmpotich, T., Miller, D., Dalwani, M., Sakai, J.T., Mikulich Gilbertson, S.K., Thompson, L., Claus, E., Banich, M., Rojas, D.C., 2013. Insula and orb itofrontal cortical morphology in substance dependence is modulated by sex. AJNR Am J Neuroradiol 34, 1150 1156. Tanabe, J.T., Regner, M.F., Sakai, J., Martinez, D., Gowin, J., 2019. Neuroimaging reward, craving, learning, and cognitive control in substanc e use disorders: review and implications for treatment. British Journal of Radiology 0, 20180942. Thielscher, A., Antunes, A., Saturnino, G.B., 2015. Field modeling for transcranial magnetic stimulation: A useful tool to understand the physiological effect s of TMS? Conference proceedings : ... Annual International Conference of the IEEE Engineering in Medicine and Biology Society. IEEE Engineering in Medicine and Biology Society. Annual Conference 2015, 222 225. Thielscher, A., Kammer, T., 2004. Electric fi eld properties of two commercial figure 8 coils in TMS: calculation of focality and efficiency. Clinical neurophysiology : official journal of the International Federation of Clinical Neurophysiology 115, 1697 1708. Thielscher, A., Opitz, A., Windhoff, M., 2011. Impact of the gyral geometry on the electric field induced by transcranial magnetic stimulation. NeuroImage 54, 234 243. Tobacco, T.C.P.G.T., 2008. A clinical practice guideline for treating tobacco use and dependence: 2008 update: a US public healt h service report. American journal of preventive medicine 35, 158. Tomasi, D., Wang, G.J., Wang, R., Caparelli, E.C., Logan, J., Volkow, N.D., 2015. Overlapping patterns of brain activation to food and cocaine cues in cocaine abusers: association to striat al D2/D3 receptors. Human brain mapping 36, 120 136. Tononi, G., Boly, M., Massimini, M., Koch, C., 2016. Integrated information theory: from consciousness to its physical substrate. Nature reviews. Neuroscience 17, 450 461. Tregellas, J.R., Tanabe, J., Ro jas, D.C., Shatti, S., Olincy, A., Johnson, L., Martin, L.F., Soti, F., Kem, W.R., Leonard, S., Freedman, R., 2011a. Effects of an alpha 7 nicotinic agonist on default network activity in schizophrenia. Biological psychiatry 69, 7 11. Tregellas, J.R., Wyli e, K.P., Rojas, D.C., Tanabe, J., Martin, J., Kronberg, E., Cordes, D., Cornier, M.A., 2011b. Altered default network activity in obesity. Obesity (Silver Spring, Md.) 19, 2316 2321. Trojak, B., Meille, V., Achab, S., Lalanne, L., Poquet, H., Ponavoy, E., Blaise, E., Bonin, B., Chauvet Gelinier, J.C., 2015. Transcranial Magnetic Stimulation Combined With Nicotine Replacement Therapy for Smoking Cessation: A Randomized Controlled Trial. Brain stimulation 8, 1168 1174. Truong, D.Q., Magerowski, G., Blackburn, G.L., Bikson, M., Alonso Alonso, M., 2013. Computational modeling of transcranial direct current stimulation (tDCS) in obesity: Impact of head fat and dose guidelines. NeuroImage. Clinical 2, 759 766. Tzourio Mazoyer, N., Landeau, B., Papathanassiou, D., Crivello, F., Etard, O., Delcroix, N., Mazoyer, B., Joliot, M., 2002. Automated anatomical labeling of activations in SPM using a macroscopic anatomical parcellation of the MNI MRI single subject brain. NeuroImage 15, 273 289. Uddin, L.Q., Kinnison, J., Pe ssoa, L., Anderson, M.L., 2014. Beyond the tripartite cognition emotion interoception model of the human insular cortex. Journal of cognitive neuroscience 26, 16 27.

PAGE 152

136 Ussher, M., Beard, E., Abikoye, G., Hajek, P., West, R., 2013. Urge to smoke over 52 weeks of abstinence. Psychopharmacology 226, 83 89. Verweij, K.J., Huizink, A.C., Agrawal, A., Martin, N.G., Lynskey, M.T., 2013. Is the relationship between early onset cannabis use and educational attainment causal or due to common liability? Drug Alcohol Dep end 133, 580 586. Vessel, E.A., Starr, G.G., Rubin, N., 2013. Art reaches within: aesthetic experience, the self and the default mode network. Frontiers in neuroscience 7, 258. Volkow, N.D., Wang, G.J., Fowler, J.S., Tomasi, D., 2012. Addiction circuitry in the human brain. Annual review of pharmacology and toxicology 52, 321 336. von Sydow, K., Lieb, R., Pfister, H., Höfler, M., Wittchen, H. U., 2002. What predicts incident use of cannabis and progression to abuse and dependence? A 4 year prospe ctive examination of risk factors in a community sample of adolescents and young adults. Drug and Alcohol Dependence 68, 49 64. Wang, W., Eisenberg, S.R., 1994. A three dimensional finite element method for computing magnetically induced currents in tissue s. IEEE Transactions on Magnetics 30, 5015 5023. Wen, X., Rangarajan, G., Ding, M., 2013. Is Granger causality a viable technique for analyzing fMRI data? PloS one 8, e67428. Whitfield Gabrieli, S., Nieto Castanon, A., 2012. Conn: a functional connectivity toolbox for correlated and anticorrelated brain networks. Brain connectivity 2, 125 141. Wilcox, C.E., Calhoun, V.D., Rachakonda, S., Claus, E.D., Littlewood, R.A., Mickey, J., Arenella, P.B., Hutchison, K.E., 2017. Functional network connectivity predict s treatment outcome during treatment of nicotine use disorder. Psychiatry research 265, 45 53. Wilcox, C.E., Teshiba, T.M., Merideth, F., Ling, J., Mayer, A.R., 2011. Enhanced cue reactivity and fronto striatal functional connectivity in cocaine use disord ers. Drug Alcohol Depend 115, 137 144. Windhoff, M., Opitz, A., Thielscher, A., 2013. Electric field calculations in brain stimulation based on finite elements: an optimized processing pipeline for the generation and usage of accurate individual head model s. Human brain mapping 34, 923 935. Wing, V.C., Bacher, I., Wu, B.S., Daskalakis, Z.J., George, T.P., 2012. High frequency repetitive transcranial magnetic stimulation reduces tobacco craving in schizophrenia. Schizophrenia research 139, 264 266. Wise, R.A ., Koob, G.F., 2014. The development and maintenance of drug addiction. Neuropsychopharmacology : official publication of the American College of Neuropsychopharmacology 39, 254 262. Wisner, K.M., Patzelt, E.H., Lim, K.O., MacDonald, A.W., 3rd, 2013. An in trinsic connectivity network approach to insula derived dysfunctions among cocaine users. The American journal of drug and alcohol abuse 39, 403 413.

PAGE 153

137 Wylie, K.P., Rojas, D.C., Tanabe, J., Martin, L.F., Tregellas, J.R., 2012. Nicotine increases brain functi onal network efficiency. NeuroImage 63, 73 80. Yamada, T., Kendix, M., Yamada, T., 1996. The impact of alcohol consumption and marijuana use on high school graduation. Health economics 5, 77 92. Zanchi, D., Brody, A.L., Montandon, M.L., Kopel, R., Emmert, K., Preti, M.G., Van De Ville, D., Haller, S., 2015. Cigarette smoking leads to persistent and dose dependent alterations of brain activity and connectivity in anterior insula and anterior cingulate. Addict Biol 20, 1033 1041. Zhang, Y., Li, Q., Wen, X., C ai, W., Li, G., Tian, J., Zhang, Y.E., Liu, J., Yuan, K., Zhao, J., Wang, W., Zhou, Z., Ding, M., Gold, M.S., Liu, Y., Wang, G.J., 2017. Granger causality reveals a dominant role of memory circuit in chronic opioid dependence. Addict Biol 22, 1068 1080. Zh ong, C., Bai, L., Dai, R., Xue, T., Wang, H., Feng, Y., Liu, Z., You, Y., Chen, S., Tian, J., 2012. Modulatory effects of acupuncture on resting state networks: a functional MRI study combining independent component analysis and multivariate Granger causal ity analysis. J Magn Reson Imaging 35, 572 581. Zhou, S., Xiao, D., Peng, P., Wang, S.K., Liu, Z., Qin, H.Y., Li, S.S., Wang, C., 2017. Effect of smoking on resting state functional connectivity in smokers: An fMRI study. Respirology (Carlton, Vic.) 22, 11 18 1124.

PAGE 154

138 APPENDIX INSULAR INHIBITORY N EUROMODULATION IN SM OKERS DECREASES CIGA RETTE CRAVINGS AND ALTERS RESTING S TATE BRAIN CONNECTIV ITY 1.1. Materials and Methods The methods of this section are identical to Chapter 5 unless explicitly stated otherwise. 1.1.1. MRI Examination MRI exams were acquired before and after treatment at the University of Colorado Brain Imaging Center using a Siemens 3 Tesla Magnetom Skyra scanner (Siemens AG; Munich, Germany) and 20 channel neurovascular coil . Structural images included a T1 weighted 3D magnetization prepared rapid gradient multi echo sequence (MPRAGE; sagittal plane acquisition; repetition time [TR] = 2300 ms; echo time [TE] = 2.24 ms; inversion time [TI] = 900 ms; echo train length [ETL] = 250 ms; flip angle = 8°; 1 mm slice thickness, 176 slices; FOV = 220 mm with 256 × 256 matrix; time = 5:21 min). Resting state functional images were acquired with a T2 weighted echo planar gradient echo sequence (GE EPI FID; axial oblique plane acquisition; echo time [TE] = 28 ms; r epetition time [TR] = 2000 ms; flip angle = 70°; slice thickness = 3mm with 1mm gap, 32 slices; FOV = 220 mm with 64 × 64 matrix; acquisition time cushion (Par Scientific A/S, Ode nse, Denmark). 1.1.2. MRI Signal Pre Processing Structural and functional images were pre processed using MATLAB 2017 and SPM12 software (Wellcome Trust Centre for Neuroimaging ; London, UK) . T1 weighted images were segmented and normalized to MNI space. BOLD pre processing included: [1] removal of first four TRs (at acquisition), [2] slice timing correction, [3] rigid realignment of BOLD images to the first TR, [4] motion scrubbing/censoring (Power et al., 2014) , [5] non neural noise

PAGE 155

139 correction using aCompCor (Behzadi et al., 2007; Muschelli et al., 2014) , and [6] non linear censored; binary censoring indicators and rigid real ignment parameters were included as first level covariates. Both total number of valid TRs (i.e., non censored) and mean framewise displacement across each run were included as second level nuisance covariates. Participants re excluded; two met this criterion. BOLD data were visually inspected by three investigators (MFR, JRT, JLT) on a subject by subject basis during blinded analysis to document agreement on quality control prior to analysis (Power et al., 2014) . For both pre treatment and post treatment scans, this included: [1] line plots of realignment parameters, [2] line plots indicating censored TRs, and [3] BOLD signal greyplots before and after nuisance variance correction. Motion and nuisance correction appeared appropriate in all subjects; none were excluded on this basis. 1.1.3. Resting State fMRI Anal yses Seed based whole brain connectivity maps were constructed for both the pre and post treatment timepoints. Seeds were selected according to the previously defined method (see Section 5.3.13 , Finite Element Method (FEM) Simulation and ROI Selection , page 91 ). Significance level for p Family < 0.05 is k E (AlphaSim corrected assuming spatial auto 3 ) (Cox et al., 2017; Eklund et al., 2016) . Effect of treatment group was evaluated using a 2x2 mixed independence factorial design. Time was the within subjects factor (pre versus post treatment); group was the between subjects factor (LF rTMS versus sham). First level SPM contrasts were constructed for each participant using paired t tests (post treatment [seed based connectivity] > pre treatment [seed based connectivity]). Second level SPM contrasts were constructed using a two sample t -

PAGE 156

140 test comparing first level cont rasts by group. Images were masked at the second level using a grey matter mask with the cerebellum excluded. 1.2. Resting State Connectivity Results We observed statistically significant bidirectional seed based whole brain connectivity changes after real tre atment compared to sham for both the right inferior frontal gyrus, pars articularis as well as the right anterior insula. No statistically significant changes were observed in seed based connectivity with the posterior middle temporal gyrus after real tre atment compared to sham .

PAGE 157

141