Structural and Functional Neural Targets of Addiction Treatment in Adolescents and Young Adults: A Systematic Review and Meta-Analysis

Abstract

Objective:

Addictive disorders start during adolescence for most individuals, and developmental differences in brain maturation and response to treatments are present. Recent studies in adults have identified associations between addiction treatment response and regional and circuit specific brain dysfunction, suggesting candidate neural treatment targets. The purpose of this systematic review and meta-analysis was to qualitatively and quantitatively summarize findings from structural and functional neuroimaging studies that examine neural correlates of treatment response in adolescents and young adults with addictive disorders.

Methods:

A systematic review and meta-analysis of peer-reviewed studies was conducted following PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) guidelines. Studies were selected if they included individuals aged 13–26 with a DSM-IV or DSM-5 (Diagnostic and Statistical Manual, Fourth and Fifth Edition) addictive disorder diagnosis, used neuroimaging, administered a treatment/intervention, and reported within- or between-subject contrasts in brain structure or activity across treatment/intervention and a control condition or brain-behavior correlations with treatment-outcome variables. Quantitative meta-analyses used an activation-likelihood estimation (ALE) approach.

Results:

Out of 3177 citations, 27 studies were included in the qualitative analysis. Qualitative analyses revealed anatomical, connectivity, and functional brain-behavior associations with response to addiction interventions across a broad array of cortical and subcortical brain regions and associated networks. Eighteen functional magnetic resonance imaging studies involving 354 participants and 88 brain foci were included in the ALE meta-analysis. Despite significant heterogeneity in study design and methods, six ALE activation clusters localized to the anterior cingulate cortex, inferior frontal gyrus, supramarginal gyrus, middle temporal gyrus, precuneus, and putamen showed consistent brain-behavior associations with treatment-outcome variables.

Conclusions:

Cortical and subcortical brain regions involved in cognition, emotion regulation, decision-making, reward, and self-reference are associated with treatment response in addicted youth. These results are consistent with findings in the adult literature and suggest overlapping neural treatment targets across developmental stages.

Introduction

Addictive disorders, including alcohol and substance use disorders, as well as nonsubstance “behavioral” addictions, such as gambling disorder, represent a significant public health problem estimated to cost the United States in excess of $400 billion annually (Uhl and Grow 2004). Most individuals who develop addictive disorders in their lifetime report first using and experiencing problems with alcohol or other drugs during adolescence (Healthday 2011). In fact, alcohol and drug initiation, progression to chronic use patterns, and initial onset of addictive disorders peak during adolescence and young adulthood (Johnston et al. 2015). Similar to substance use, other appetitive behaviors, such as video gaming or Internet use, and problems related to compulsive engagement in these behaviors (e.g., Internet gaming disorder) are also elevated in youth (McGue and Iacono 2005) With growing evidence for negative long-term consequences of youth substance use on cognition, mental health, risk-taking behaviors, and academic/job achievement much scientific focus is on early interventions and targeted treatment of addictive disorders in youth (Eaton et al. 2008).

A number of therapeutic interventions have been developed and tested in youth with addictive disorders. They can be broadly categorized into behavioral and pharmacological interventions. Behavioral (i.e., “psychotherapeutic”) interventions, including cognitive-behavioral therapy (CBT), motivational interviewing, family-based therapies, and integrative behavioral therapies, have the strongest evidence base and represent the primary treatment modality for addicted youth (Hogue et al. 2014). While these interventions generally show modest efficacy, there is considerable response variation, and nearly two-thirds of youth relapse to their additive behaviors within 6 months of treatment (Cornelius et al. 2003; Dennis et al. 2004). The field does not have a clear sense of who responds to which treatment, why, and, if treatment is related to changes in brain morphology or function.

A fundamental barrier to progress in treatment science has been the lack of understanding about the basic mechanisms of treatment-related behavioral change and the mediators of treatment response (NIH 2009). In particular, the neurobiological mechanisms underlying individual response to addiction treatment are poorly understood (Nestor et al. 2011; Boyce and Lynne-Landsman 2013; Chung et al. 2016).

Imaging studies in adults with addictive disorders who are in long-term abstinence point to treatment- and abstinence-related improvements in prefrontal cortical brain function that may underscore the importance of cognitive control processes to avoid relapse (Nestor et al. 2011; Garavan et al. 2013). A neuroimaging meta-analysis by Konova and colleagues identified a number of candidate neural treatment targets in adults with addiction problems that included the ventral striatum, anterior cingulate cortex (ACC), inferior frontal gyrus (IFG), middle frontal gyrus (MFG), orbitofrontal cortex (OFC), and precuneus (Konova et al. 2013). These brain regions undergo significant maturation during adolescence (Casey et al. 2010; Hammond et al. 2014). The extent to which these and other brain regions and circuits are related to treatment in youth and represent neural treatment target candidates is unknown.

Understanding the neurobiological mechanisms of addiction treatment response in youth may facilitate early identification of teenagers who are at risk for poor treatment outcomes and their associated health consequences and promote development of effective treatment and prevention strategies. In addition, understanding the neurobiologic basis of addiction and recovery from addiction during adolescence may inform policy and public health initiatives relevant to this developmental period. Identifying neural targets of addiction treatment in youth may hold the key to refining existing treatments and informing the future development of diagnostic and prognostic biomarkers and brain circuit-targeted interventions (Dalley et al. 2011).

In the present study we qualitatively and quantitatively summarize structural and functional neuroimaging studies that examine neural correlates of treatment response in adolescents and young adults with addictive disorders using an activation-likelihood estimation (ALE) meta-analysis (Eickhoff et al. 2012). Coordinate-based meta-analyses, including ALE, allow for the aggregation of neuroimaging data to reliably identify localization of anatomical and activation patterns showing contrasts that are convergent across studies. Predicting that the number of eligible studies would be small, limiting our ability to perform appropriately powered subgroup contrasts, we primarily sought to identify common brain regions and circuits showing significant neural correlates with addiction treatment response across intervention and types of addictive disorders.

This was done by examining brain-behavior associations with treatment outcome variables and activation or anatomical contrasts for intervention versus nonintervention comparison conditions looking for concordance across studies.

We used exploratory analyses contrasting subgroups of studies. The main subgroup analyses examined studies using single time point magnetic resonance imaging (MRI) scans that reported pretreatment neural associations with response to the intervention and studies using multiple MRI scans with a pre- to post-treatment design or intervention versus control condition contrast. The aim of these subgroup analyses was to determine if the brain regions and circuits that predict treatment outcome overlap with or are distinct from the brain regions and circuits that show intervention-related change in function. Additional exploratory analyses contrasted subgroups of studies based upon participant age and type of substance use or addictive disorder treated.

Methods

A systematic review of peer-reviewed studies was conducted following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines and methods (Moher et al. 2009). A subset of studies from the review that included coordinate-level data was used in the ALE meta-analyses.

Search strategy

We searched for studies indexed in the online databases PubMed/Medline, Embase, PsycINFO, Web of Science, and Cochrane Library from January 1990 through February 2018. The following search terms were used: “transitional age youth,” “young adult,” “adolescent,” “teen,” “treatment,” “intervention,” “treatment response,” “treatment outcome,” “predictor,” “mechanism,” “substance use disorder,” “alcohol use disorder,” “cannabis use disorder,” “marijuana use disorder,” “opioid use disorder,” “tobacco use disorder,” “cocaine use disorder,” “benzodiazepine use disorder,” “addictive disorder,” “gambling disorder,” “video game addiction,” “internet addiction,” “neuroimaging,” “functional magnetic resonance imaging (fMRI),” “electrophysiology (EEG),” “positron emission tomography (PET),” “neurophysiology,” “brain activity,” “functional connectivity,” “brain circuit,” or “neural.” Additional studies were identified by manually scanning the references of included studies and recent review articles.

Study selection

Studies were selected if they met the following criteria: (1) included >10 participants and participants were between the ages 13 and 26; (2) used diagnostic criteria for substance use or addictive disorders as specified by the Diagnostic and Statistical Manual, Fourth and Fifth Edition (DSM-IV and DSM-5) or described chronic (>weekly) impairing substance use; (3) used any neuroimaging modality (e.g., MRI, PET, and EEG); (4) administered a treatment/intervention; (5) reported within- or between-subject contrasts in brain structure or activity across treatment/intervention and a control condition or brain-behavior correlations with treatment-outcome variables; and (6) provided information about the inclusion/exclusion criteria from individual studies, clinical characteristics, and demographics of the study sample. Of the studies identified with the above criteria, those that reported peak activation coordinates in Talairach or Montreal Neurologic Institute (MNI) space were used in the ALE meta-analysis.

Data extraction

Articles were extracted, organized, and reviewed using Covidence software. Initial independent abstract evaluations were done to identify potential articles of interest by two authors (A.A. and N.R.). Data extraction accuracy of 25 randomly selected studies showed correspondence greater than 99%. Abstract evaluation was followed by an independent full-text review of articles. Group discussion was used to resolve uncertainties about inclusion criteria and finalize the list of articles included in the qualitative review and the ALE meta-analysis.

To facilitate exploration and interpretation of results, we stratified the studies based upon neuroimaging technique, analytic approach, and tissue class into four categories as follows: structural MRI, diffusion tensor imaging (DTI), functional connectivity MRI, and fMRI. As only one structural MRI study was identified, the ALE meta-analyses focused on fMRI experiments only. Studies examining gray matter volume (n = 1), white matter integrity (i.e., fractional anisotropy) (n = 2), and brain activation using functional connectivity fMRI approaches (n = 7) contained insufficient number of experiments for quantitative analysis, but were included in the qualitative analysis.

Data analysis

ALE meta-analysis procedures

Spatial concordance among the reported voxel-based activation foci was computed using a modified ALE algorithm (Eickhoff et al. 2012). ALE treats each voxel-based activation focus as a Gaussian probability distribution (Eickhoff et al. 2012; Turkeltaub et al. 2012). Gaussian distributions are pooled voxel-wise within experimental contrasts and across contrasts within a group to create a whole-brain ALE map. Within this whole-brain ALE map, each voxel is assigned a unique ALE value that represents the likelihood of experimental effects (e.g., BOLD fMRI signal contrast between intervention and comparison condition) in that voxel. ALE maps are tested against a null distribution that reflects the random spatial association between different experiments and thresholded to a specific statistical value. Voxels that pass this statistical threshold are reported as ALE clusters of significant meta-analytic convergence.

ALE meta-analyses were carried out in GingerALE 2.3.7 (brainma-p.org/ale). In preparation for the ALE Map creation, coordinates and cluster sizes associated with significant activation or deactivation were first converted to Talairach space using the MNI to Talairach conversion tool provided by GingerALE toolbox. For each meta-analysis, we used the nonadditive algorithm (Turkeltaub et al. 2012) to minimize within-experiment effects. Inference was made at cluster level (p < 0.05, 1000 permutations) with an uncorrected voxel-wise p-value of 0.001. Cluster-level inference has been shown to provide a better balance between sensitivity and specificity compared with other methods to correct for multiple comparisons currently available in GingerALE (Eickhoff et al. 2012). The α levels are in line with those used in previous ALE meta-analyses (Fusar-Poli et al. 2009; Barron et al. 2012).

Exploratory subgroup analyses

fMRI studies selected for ALE meta-analysis inclusion were divided into two groups as follows: (1) studies reporting pretreatment brain activation that was associated with treatment outcome or abstinence variables (i.e., neural correlates of response to addiction intervention) and (2) studies showing treatment versus control condition or baseline/pretreatment versus. post-treatment brain activation contrasts (i.e., regions showing intervention-related change in brain function). We had also planned to perform subgroup analyses stratified by participant age/developmental stage—young adult (19–26 years of age) versus adolescent (13–18 years of age) and by the type of addictive disorder (Nicotine, Marijuana, Alcohol, Stimulants, and Internet Gaming Disorder), but there were insufficient experiments across these categories for appropriately powered ALE meta-analyses.

Results

Systematic review

Figure 1 shows the PRISMA flow diagram of the search process. The initial search identified 3177 citations, of which 28 studies were included in the qualitative analysis. The most common reasons for exclusion of citations were: the use of adult instead of adolescent population, the absence of a treatment or intervention, and the absence of at least two comparison points in time (baseline vs. post-treatment).

FIG. 1.

Flowchart outlining selection procedure of studies.

These 28 studies represent neuroimaging data from 572 participants with 94 identified activation and structural foci not including resting state networks (summary shown in Table 1). Eighteen (64%) studies were fMRI studies, seven were functional connectivity fMRI studies, two were diffusion tensor studies, and one was structural MRI study. In studies included in the qualitative analysis, there was heterogeneity across study designs and across the types of intervention participants received to target their addictive behaviors (Supplementary Table S1). Studies used randomized controlled designs applying evidence-based (e.g., behavioral, pharmacologic, and integrated interventions) and experimental interventions (e.g., virtual reality cue exposure), along with naturalistic study designs examining outcomes from youth enrolled in community-based outpatient substance use treatment programs.

Table 1.

Study Characteristics and Qualitative Analysis

Imaging modality	No. of studies	No. of subjects	No. of foci	% Age (No. of foci)	% Type addiction (No. of studies, foci)	Task/stimulus	Intervention(s)	Brain areas and circuits showing treatment outcome contrasts
Structural MRI/VBM	1	20	2	100% young adult	100% cannabis (1 study, 2 foci)		Integrated cognitive behavioral interventions (CM, CBT, or CM+CBT)	Reduced L,R putamen GM volume^a
Diffusion tensor imaging	2	62	4	100% adolescent	100% alcohol (2 studies, 4 foci/tracts)		Community-based outpatient SUD treatment	Reduced FA in prefrontal, orbitofrontal, and temporal WMTs; increased Insula WMV^a
Functional connectivity MRI	7	136		29% adolescent, 71% young adult	29% cannabis, 43% IGD, 14% alcohol, 14% stimulant, 14% mixed	100% seed-based analyses; 86% resting-state fMRI (6 studies); 14% task-based fMRI with PPI (1 study); MI sustain vs change talk paradigm	Community-based OP SUD treatment (2 studies), craving-focused CBT (3 studies), MET (1 study), monitored abstinence (1 study)	Insula resting-state FC; within-network DMN and ECN resting-state FC; OFC connectivity during MI change talk
Functional MRI	18	354	88	41% adolescent (55); 58% young adult (78)	33% cannabis (6 studies, 27 foci), 22% alcohol (4 studies, 25 foci), 27% tobacco (5 studies, 19 foci), 22% IGD (4 studies, 6 foci), 18% mixed	Cue reactivity task, MID, reward-related antisaccade task, BART, amygdala reactivity task, Stroop Color Word Task, Go/No-Go task	CBT, MET, CM, family-based therapy, integrated EBTs (CBT+CM; MET/CBT); pharmacotherapy+integrated EBTs; cue exposure therapy; monitored abstinence; community-based OP, IOP, residential/inpatient SUD treatment	ACC, DLPFC, OFC, IFG, SFG, frontal polar cortex, insula, ventral striatum, NAC, putamen, caudate, amygdala, SMG, STG, MTG, Precuneus, posterior cingulate, lingual gyrus

In VBM and DTI studies brain area/circuits are described in relation to poorer treatment outcome (i.e., reduced L,R putamen volume associated with nonabstinence in treated cannabis users).

ACC, anterior cingulate cortex; BART, balloon analog risk task; CBT, cognitive-behavioral therapy; CM, contingency management; DLPFC, dorsolateral prefrontal cortex; DMN, default mode network; DTI, diffusion tensor imaging; EBT, evidence based treatment; ECN, executive control network; FC, functional connectivity; fMRI, functional magnetic resonance imaging; GM, gray matter; IFG, inferior frontal gyrus; IGD, Internet gaming disorder; MET/MI, motivational interviewing; MID, Monetary incentive delay task; MTG, middle temporal gyrus; NAC, nucleus accumbens; OFC, orbitofrontal cortex; SFG, superior frontal gyrus; SMG, supramarginal gyrus; STG, superior temporal gyrus; VBM, voxel based morphometry; WM, white matter; WMT, white matter tract; WMV, white matter volume.

The ways that the studies measured treatment outcome were also heterogeneous, with some studies using biochemical assessment and others using self-report (Supplementary Table S1). Most studies included in the qualitative and quantitative analyses stratified youth who received the intervention into treatment responders, defined as youth who achieved abstinence (self-reported or biologically verified) by the end of treatment, compared to nonresponders on brain measures. Of the studies that reported on treatment efficacy in the samples, many showed significant individual differences in clinical response to the substance use interventions consistent with prior treatment studies in addictive disorders. Across studies, brain regions and associated networks were associated with treatment response in the qualitative analyses.

Structural neuroimaging studies

The structural studies, which included a single voxel based morphometry experiment and two DTI experiments, reported that reduced frontal and striatal gray and white matter volumes were associated with poorer treatment response.

Functional connectivity fMRI studies

Functional connectivity fMRI studies reported connectivity strength associations with treatment outcome across resting state networks, including the executive control network default mode network, and insula, a core hub of the salience network. Due to heterogeneity of the functional connectivity studies, no clear pattern of brain activity changes due to treatment emerged; however, in most studies statistically significant functional connectivity differences subsided after treatment regardless of sample characteristics or intervention.

ALE meta-analysis

Eighteen fMRI studies involving 354 participants and 88 brain foci were included in the ALE meta-analyses (summary shown in Table 2). Six activation foci localized to the ACC, IFG, supramarginal gyrus (SMG), middle temporal gyrus (MTG), precuneus, and putamen showed significant brain-behavior associations with treatment-outcome variables or abstinence across all studies.

Table 2.

Activation-Likelihood Estimation Analysis Summary

Cluster No.	Label	BA	Volume (mm³)	Weighted center			Extrema value	x	y	z
Cluster No.	Label	BA	Volume (mm³)	x	y	z	Extrema value	x	y	z
All studies
1	Anterior cingulate cortex	24	728	0.3	32.6	3.1	0.013989376	−2	32	4
2	Inferior frontal gyrus	47	448	−41	25	1	0.015782826	−42	24	0
3	Supramarginal gyrus	40	448	−57	−49	23	0.015782129	−56	−50	22
4	Precuneus	7	344	2.7	−69	43.7	0.013044049	4	−70	44
5	Middle temporal gyrus	21	272	55	−36	−3	0.011279688	55	−36	−3
Studies showing pretreatment brain activation associated with treatment outcome or abstinence variables
1	Anterior cingulate cortex	24	776	−2.1	29.2	9.7	0.012838	−4	28	10
2	Precuneus	7	608	2	74.6	37.8	0.01306	2	−74	38
Studies showing intervention-related changes in brain activation
1	Anterior cingulate cortex	24	736	0.3	32.6	3.1	0.013989376	−2	32	4
2	Inferior gyrus cortex	47	448	−41	25	1	0.015782828	−42	24	0
3	Supramarginal gyrus	40	448	−57	−49	23	0.015782129	−58	−48	24
4	Middle temporal gyrus	21	272	55	−36	−3	0.011279688	55	−36	−3

A series of exploratory ALE meta-analyses were carried out in GingerALE 2.3.7 on fMRI studies reporting Talairach or MNI coordinates with threshold set at uncorrected p-value <0.001.

ALE, activation-likelihood estimation; fMRI, functional magnetic resonance imaging; MNI, Montreal Neurologic Institute.

Post hoc ALE analyses examined whether significant intervention-related activation foci associations with treatment outcome were driven by region-specific patterns of neural activation, deactivation, or both (Supplementary Data; Supplementary Table S2; Supplementary Figs. S1 and S2). Results identified eight activation foci, including the ACC, IFG, superior temporal gyrus, SMG, MTG, and subcortical foci in the thalamus and red nucleus, with significant brain-behavior associations showing intervention-related increases in brain activation related to positive treatment outcomes and no activation foci showing significant associations between intervention-related decreases in brain activation and positive treatment outcomes.

ALE subgroup analyses

Subgroup analyses identified activation foci in the ACC, IFG, SMG, and MTG across studies showing intervention-related treatment contrasts between baseline/pretreatment to post-treatment changes in function and in the precuneus and putamen in studies showing pretreatment brain activation associated with treatment outcome or abstinence variables (Fig. 2).

FIG. 2.

ALE meta-analyses showing concordance of functional brain activation in fMRI studies across (A) all neuroimaging-treatment studies; (B) studies using a single fMRI scan at baseline/pretreatment to predict treatment outcomes; and (C) studies showing regional brain activation contrast between baseline/pretreatment and post-treatment. ALE, activation-likelihood estimation; fMRI, functional magnetic resonance imaging.

Discussion

Addictive disorders in youth are associated with increased morbidity and mortality and alterations in brain function and to a lesser extent structure. How regional brain function and structure are related to addiction treatment response is poorly understood. The purpose of this systematic review and meta-analysis was to identify brain regions and circuits associated with addiction treatment response in youth to guide future targeted analyses and studies. To this end, we performed a PRISMA guided systematic review identifying 28 neuroimaging studies for qualitative review and an ALE meta-analysis using 18 fMRI studies involving 354 participants and 88 foci. Qualitative analyses revealed anatomical, connectivity, and functional brain-behavior associations with response to addiction interventions across a broad array of cortical and subcortical brain regions and associated networks. Across studies, six activation foci localized to the ACC, IFG, SMG, precuneus, and MTG were associated with response to addiction interventions in adolescents and young adults.

Common neural treatment targets across interventions

The primary objective of this study was to identify common addiction treatment targets across interventions in youth. Although the qualitative analyses revealed a large number of anatomical, connectivity, and functional brain-behavior associations, our main ALE results narrowed the findings, revealing six relevant brain regions, the ACC, IFG, SMG, precuneus, MTG, and putamen, that showed consistent brain-behavior associations with addiction treatment outcome variables. These brain regions are important for cognitive and emotional regulation, decision-making, reward processing, and language and are involved in self-awareness and introspection. The ACC is implicated in a number of roles, including conflict monitoring, attention, decision-making, and reward learning (Kennerley et al. 2006, 2011). Increased ACC activation during drug cue reactivity paradigms and decreased ACC activation during cognitive control and attentional tasks have been reported in both adolescents and adults with addictive disorders compared with healthy controls (Peters et al. 2012; Schacht et al. 2013). The precuneus and IFG are implicated in self-awareness, introspection, and self-related mental representation (Cavanna and Trimble 2006; Spreng and Grady 2010; Feldstein Ewing et al. 2011; Krishnan-Sarin et al. 2013), while the MTG and SMG are involved in language processing, semantic meaning, and episodic memory (Acheson and Hagoort 2013). Finally, the putamen is involved in motor action, planning, and development of habitual behaviors. Animal models of addictive disorders describe the transition from impulsive drug use to compulsive and habitual drug use as being associated with neuroplastic changes shifting from dorsal to ventral striatal driven behaviors.

Partially consistent with our findings, prior studies in the adolescent addiction literature describe functional and, to a lesser extent, structural abnormalities in frontal and subcortical brain regions among youth who chronically use alcohol and other drugs (Hammond et al. 2014; Lisdahl et al. 2014; Brumback et al. 2015).

The absence of drug related deficits in MTG and SMG in prior studies suggests that activation in these language-processing regions could be relevant to positive treatment outcomes while at the same time being unrelated to the underlying pathophysiology of addictive disorders. Increased neural activation in MTG and SMG may indicate a preservation of language/semantic skills that can be accessed during therapeutic interventions in treatment responders, leading to better outcomes.

The ventral striatum, ACC, IFG, MFG, OFC, and precuneus all represent candidate neural treatment targets in addicted adults (Konova et al. 2013). Thus, our results are also consistent with findings in adult samples suggesting overlap in the neural treatment targets for addictive disorders across ages and developmental stages. This is pertinent given the significant maturation that these brain regions undergo during adolescence. Our findings taken within a larger context suggest that brain regions involved in the development and maintenance of addictive disorders in youth overlap with the brain regions involved in treatment response and recovery. It then goes to follow that if similar brain regions and circuits are involved in the development, maintenance, and recovery from addictive disorders across the life span, then biologically-based treatments that target these circuits and show efficacy in adults should hypothetically also be effective in youth with addictive disorders.

A major unanswered question remains as to which of these structural and functional abnormalities observed in adults and adolescents with addictive disorders predate the onset of alcohol and drug use and which relate to addictive processes or reflect neuroadaptation or neurotoxicity related to recent or long-term drug or alcohol exposure that may or may not be central to addictive processes (Berridge 2007; Casey et al. 2010). Questions about premorbid factors and how trajectories of structural and functional change in developing brains relates to addictive disorders will be addressed with the Adolescent Brain Cognitive Development (ABCD) study, a longitudinal observational study of 10,000 preadolescents who will receive neuroimaging and health related assessments biannually over the next decade (Jernigan et al. 2018). Furthermore, the ABCD study will aid in clarifying the genetic and environmental contributions, including the role of adverse childhood experiences along with resiliency, on teen brain health and neural maturation, helping to identify neural regions and circuits predictive of persistent substance use and negative outcomes, and providing a roadmap for novel interventions and prevention approaches. Still, the ABCD study and longitudinal naturalistic studies like it are limited by the scope of assessments and are unable to answer questions related to treatment effects, due to a lack of randomization and comparison groups. Thus, they cannot answer clinically relevant questions about mechanisms of behavioral change related to interventions for addictive disorders and about individual differences in treatment response to specific evidence-based interventions (e.g., CBT). As such, there will be a future role for treatment mechanism focused studies and for developmental studies that address important questions not examined with the ABCD battery. By incorporating neuroimaging into randomized controlled trials and addressing the complexities of treatment by measuring and controlling for premorbid/pretreatment cognitive function, effects of chronic, recent, and ongoing substance use during treatment, these mechanism focused studies provide a complimentary role to the ABCD study (Chung et al. 2016).

In terms of directional effects, our post hoc analyses identified a number of cortical and subcortical structures where increased intervention-related neural activation was associated with abstinence. In contrast, no regions showed consistent relationships between deactivation and abstinence. These findings should be interpreted cautiously given the small number of studies and foci included in the analyses. Still, it is promising that they are generally consistent with neuroimaging studies in adults (Nestor et al. 2011; Garavan et al. 2013; Fein and Cardenas 2015).

Brain regions showing intervention-related change in function

While associations between intervention and activation in the ACC, IFG, SMG, precuneus, MTG, and putamen were observed across all studies, significant activation foci among studies examining intervention-related changes in brain function were only found in the ACC, IFG, SMG, and MTG. These regions represent promising targets for biologically-based addiction interventions, given their change in BOLD fMRI signal from pre- to postintervention.

Intervention-related changes in function within the ACC, IFG, SMG, and MTG may reflect active learning and cognitive and emotional regulatory processes using symbolic language to update self-awareness and self-meaning during treatment (Feldstein Ewing et al. 2013). Targeting these regions, whether it be with behavioral, pharmacologic, or other interventions, may result in region or circuit specific mechanisms of behavioral change and dissociable pathways to recovery. For example, interventions that enhance circuit strength in executive control and language processing networks could hypothetically reduce engagement in addictive behaviors and promote abstinence by improving “top-down” cognitive control over negative affect and stress reactivity. Conversely, interventions that target self-referential and default mode networks could act through “bottom-up” mechanisms to reduce defensiveness/guardedness, increase introspection, flexibility of thinking, and openness to different views in the role of drug use and one's identity.

Limitations and future directions

This review has a number of important limitations. Results should be viewed as preliminary given that the number of eligible studies for inclusion in the primary meta-analysis was modest and that there was an insufficient number of experiments available to conduct subgroup or conjunction analyses. Too few studies were available to perform quantitative analyses on structural MRI, DTI, functional connectivity experiments or to examine for sex differences in brain response to treatment, an important area of focus. As the field grows and more studies are published using these methods, quantitative analyses of these approaches and conjunction analyses looking for convergent and divergent findings across these approaches will further inform our understanding of neurobiological mechanisms of treatment response.

Only our main hypotheses about relationships between brain function and addiction treatment response in youth could be properly addressed, as the study was underpowered to examine other differences, for example, across age ranges or types of addictive disorders. Neuroimaging meta-analyses are limited by the quality and degree of heterogeneity in the studies selected for inclusion. Our review found that many of the selected studies were of low quality. Common limitations observed across studies included underpowered studies with small sample sizes, lack of control groups or control conditions, and use of simplistic analytic approaches. Most of the studies used region of interest analyses primarily and only applied whole brain analysis as part of exploratory analyses. Significant heterogeneity was observed in study design, paradigms used, interventions administered, and statistical methods applied. Furthermore, there were inconsistencies in how studies defined treatment response. Still, it is promising that, even, despite quality and heterogeneity issues, we identified a number of brain regions showing convergence across studies. Further studies are needed using large samples of addicted youth and assessing standardized treatments compared with appropriate control conditions. It is important to note that none of the studies included in the review used neuroimaging prediction models or machine learning approaches, such as feature selection, cross-validation, and random label permutations (Jollans and Whelan 2016; Moeller and Paulus 2018). These advanced techniques have been shown to improve generalizability and reproducibility of predictions and should be used in future studies to validate the “candidate” neural treatment targets identified here.

Conclusions

Cortical and subcortical brain regions involved in cognition, emotion regulation, reward, and self-reference are associated with treatment response in addicted youth. These results are consistent with findings in the adult literature and suggest overlapping neural treatment targets across developmental stages.

Clinical Significance

Understanding the neurobiological mechanisms of addiction treatment response in youth may hold the key to refining existing treatments, developing new ones, and informing the future development of diagnostic and prognostic biomarkers. The present meta-analysis identified six brain regions, the ACC, IFG, SMG, precuneus, MTG, and putamen, that were associated with treatment response across addiction interventions in youth and represent candidate neural treatment targets for further study. These candidate targets include cortical and subcortical brain regions involved in a diverse array of functions, including cognitive and emotion regulation, decision-making, reward, and self-reference.

Footnotes

Acknowledgments

The authors acknowledge and thank R.L. Findling and E.A. Stein for their help in providing edits of an earlier version of this research project presented as a new research poster at the Society of Biological Psychiatry annual scientific conference 2018 and K. Rathod for her help in compiling and organizing the reference list.

Disclosures

Dr. Hammond receives research support from an American Academy of Child & Adolescent Psychiatry and NIDA career development award (K12-DA000357). Ms. Allick, Rahman, and Nanavati have no disclosures or conflicts. This work was supported by the American Academy of Child & Adolescent Psychiatry, the National Institutes of Health [Grant No. K12-DA000357 (Hammond)].

Supplementary Material

Supplementary Data

Supplementary Figure S1

Supplementary Figure S2

Supplementary Table S1

Supplementary Table S2

References

Acheson

, Hagoort

: Stimulating the brain's language network: Syntactic ambiguity resolution after TMS to the inferior frontal gyrus and middle temporal gyrus. J Cogn Neurosci, 25:1664–1677, 2013.

Barron

, Fox

, Laird

, Robinson

, Fox

: Thalamic medial dorsal nucleus atrophy in medial temporal lobe epilepsy: A VBM meta-analysis. Neuroimage Clin, 2:25–32, 2012.

Berridge

: The debate over dopamine's role in reward: The case for incentive salience. Psychopharmacology (Berl), 191:391–431, 2007.

Boyce

, Lynne-landsman

: Integrating translational neuroscience to improve drug abuse treatment for adolescents. Psychol Addict Behav, 27:547–551, 2013.

Brumback

, Squeglia

, Jacobus

, Pulido

, Tapert

, Brown

: Adolescent heavy drinkers' amplified brain responses to alcohol cues decrease over one month of abstinence. Addict Behav, 46:45–52, 2015.

Casey

, Jones

, Levita

, Libby

, Pattwell

, Ruberry

, Soliman

, Somerville

: The storm and stress of adolescence: Insights from human imaging and mouse genetics. Dev Psychobiol, 52:225–235, 2010.

Cavanna AE Trimble

: The precuneus: A review of its functional anatomy and behavioural correlates. Brain, 129:564–583, 2006.

Chung

, Noronha

, Carroll

, Potenza

, Hutchison

, Calhoun

, Gabrieli

, Morgenstern

, Nixon

, Wexler

, Brewer

, Ray

, Filbey

, Strauman

, Kober

, Feldstein Ewing

: Brain mechanisms of change in addictions treatment: Models, methods, and emerging findings. Curr Addict Rep, 3:332–342, 2016.

Cornelius

, Maisto

, Pollock

, Martin

, Salloum

, Lynch

, Clark

: Rapid relapse generally follows treatment for substance use disorders among adolescents. Addict Behav, 28:381–386, 2003.

10.

Dalley

, Everitt

, Robbins

: Impulsivity, compulsivity, and top-down cognitive control. Neuron, 69:680–694, 2011.

11.

Dennis

, Godley

, Diamond

, Tims

, Babor

, Donaldson

, Liddle

, Titus

, Kaminer

, Webb

, Hamilton

, Funk

: The Cannabis Youth Treatment (CYT) Study: Main findings from two randomized trials. J Subst Abuse Treat, 27:197–213, 2004.

12.

Eaton

, Kann

, Kinchen

, Shanklin

, Ross

, Hawkins

, Harris

, Lowry

, Mcmanus

, Chyen

: Youth risk behavior surveillance—United States, 2007. MMWR Surveill Summ, 57:1–131, 2008.

13.

Eickhoff

, Bzdok

, Laird

, Kurth

, Fox

: Activation likelihood estimation meta-analysis revisited. Neuroimage, 59:2349–2361, 2012.

14.

Fein

, Cardenas

: Neuroplasticity in human alcoholism: Studies of extended abstinence with potential treatment implications. Alcohol Res, 37:125–141, 2015.

15.

Feldstein Ewing

, Filbey

, Sabbineni

, Chandler

, Hutchison

: How psychosocial alcohol interventions work: A preliminary look at what FMRI can tell us. Alcohol Clin Exp Res, 35:643–651, 2011.

16.

Feldstein Ewing

, Mceachern

, Yezhuvath

, Bryan

, Hutchison

, Filbey

: Integrating brain and behavior: Evaluating adolescents' response to a cannabis intervention. Psychol Addict Behav, 27:510–525, 2013.

17.

Fusar-poli

, Placentino

, Carletti

, Landi

, Allen

, Surguladze

, Benedetti

, Abbamonte

, Gasparotti

, Barale

, Perez

, Mcguire

, Politi

: Functional atlas of emotional faces processing: A voxel-based meta-analysis of 105 functional magnetic resonance imaging studies. J Psychiatry Neurosci, 34:418–432, 2009.

18.

Garavan

, Brennan

, Hester

, Whelan

: The neurobiology of successful abstinence. Curr Opin Neurobiol, 23:668–674, 2013.

19.

Hammond

, Mayes

, Potenza

: Neurobiology of adolescent substance use and addictive behaviors: Treatment implications. Adolesc Med State Art Rev, 25:15–32, 2014.

20.

Healthday: Addiction Starts Early in American Society, Report Find. U.S News Health Report, 2011.

21.

Hogue

, Henderson

, Ozechowski

, Robbins

: Evidence base on outpatient behavioral treatments for adolescent substance use: Updates and recommendations 2007–2013. J Clin Child Adolesc Psychol, 43:695–720, 2014.

22.

Jernigan

, Brown

, The ABCD Consortium Coordinators: Introduction to the ABCD study. Dev Cogn Neurosci, 2018. [Epub ahead of print]; DOI: 10.1016/j.dcn.2018.02.002.

23.

Johnston

, O'malley

, Miech

, Bachman

, Schulenberg

: Monitoring the Future National Survey Results on Drug Use: 1975–2014: Overview, Key Findings on Adolescent Drug Use. Ann Arbor, MI, Institute for Social Research, The University of Michigan, 2015.

24.

Jollans

, Whelan

: The clinical added value of imaging: A perspective from outcome prediction. Biol Psychiatry Cogn Neurosci Neuroimaging, 1:423–432, 2016.

25.

Kennerley

, Behrens

, Wallis

: Double dissociation of value computations in orbitofrontal and anterior cingulate neurons. Nat Neurosci, 14:1581–1589, 2011.

26.

Kennerley

, Walton

, Behrens

, Buckley

, Rushworth

: Optimal decision making and the anterior cingulate cortex. Nat Neurosci, 9:940–947, 2006.

27.

Konova

, Moeller

, Goldstein

: Common and distinct neural targets of treatment: Changing brain function in substance addiction. Neurosci Biobehav Rev, 37:2806–2817, 2013.

28.

Krishnan-sarin

, Balodis

, Kober

, Worhunsky

, Liss

, Xu

, Potenza

: An exploratory pilot study of the relationship between neural correlates of cognitive control and reduction in cigarette use among treatment-seeking adolescent smokers. Psychol Addict Behav, 27:526–532, 2013.

29.

Lisdahl

, Wright

, Kirchner-Medina

, Maple

, Shollenbarger

: The effects of regular cannabis use on neurocognition in adolescents and young adults. Curr Addict Rep, 1:144–156, 2014.

30.

Mcgue

, Iacono

: The association of early adolescent problem behavior with adult psychopathology. Am J Psychiatry, 162:1118–1124, 2005.

31.

Moeller

, Paulus

: Toward biomarkers of the addicted human brain: Using neuroimaging to predict relapse and sustained abstinence in substance use disorder. Prog Neuropsychopharmacol Biol Psychiatry, 80:143–154, 2018.

32.

Moher

, Liberati

, Tetzlaff

, Altman

: Preferred reporting items for systematic reviews and meta-analyses: The PRISMA statement. Ann Intern Med, 151:264–269, w264, 2009.

33.

Nestor

, Ghahremani

, Monterosso

, London

: Prefrontal hypoactivation during cognitive control in early abstinent methamphetamine-dependent subjects. Psychiatry Res, 194:287–295, 2011.

34.

NIH: Science of Behavior Change. Bethesda, Maryland, National Institute of Health, 2009, pp. 1–39.

35.

Peters

, Budney

, Carroll

: Clinical correlates of co-occurring cannabis and tobacco use: A systematic review. Addiction, 107:1404–1417, 2012.

36.

Schacht

, Anton

, Randall

, Li

, Henderson

, Myrick

: Effects of a GABA-ergic medication combination and initial alcohol withdrawal severity on cue-elicited brain activation among treatment-seeking alcoholics. Psychopharmacology (Berl), 227:627–637, 2013.

37.

Spreng

, Grady

: Patterns of brain activity supporting autobiographicaly memory, prospection, and theory of mind, and their relationship to the default mode network. J Cogn Neurosci, 22:1112–1123, 2010.

38.

Turkeltaub

, Eickhoff

, Laird

, Fox

, Wiener

, Fox

: Minimizing within-experiment and within-group effects in Activation Likelihood Estimation meta-analyses. Hum Brain Mapp, 33:1–13, 2012.

39.

Uhl

, Grow

: The burden of complex genetics in brain disorders. Arch Gen Psychiatry, 61:223–229, 2004.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.03 MB

0.08 MB