Research Article

Common and distinct drug cue reactivity patterns associated with cocaine
and heroin: An fMRI meta-­analysis
Jordan M. Dejoiea, Nicole Seniaa, Anna B. Konovab, David V. Smithc, Dominic S. Fareria

Gordon F. Derner School of Psychology, Adelphi University, Garden City, NY, United States
a

b
Department of Psychiatry, Rutgers University, New Brunswick, NJ, United States
Department of Psychology and Neuroscience, Temple University, Philadelphia, PA, United States
c

Corresponding Author: Dominic S. Fareri ([email protected])

ABSTRACT
Substance use and substance use disorders represent ongoing major public health crises. Specifically, the use of
substances such as cocaine and heroin are responsible for over 50,000 drug-­related deaths combined annually. We
used a comparative meta-­analysis procedure to contrast activation patterns associated with cocaine and heroin cue
reactivity, which may reflect substance use risk for these substances. PubMed and Google Scholar were searched for
studies with within-­subject whole brain analyses comparing drug to neutral cues for users of cocaine and heroin pub-
lished between 1995 and 2022. A total of 18 studies were included, 9 in each subgroup. Voxel-­based meta-­analyses
were performed using seed-­based d mapping with permuted subject images (SDM-­PSI) for subgroup mean analyses
and a contrast meta-­regression comparing the two substances. Results from our mean analysis indicated that users
of heroin showed more widespread activation in the nucleus accumbens, right inferior and left middle temporal gyrus,
right thalamus, and right cerebellum. Cocaine use was associated with recruitment of dorsolateral prefrontal cortex
during cue reactivity. Direct comparison of cue reactivity studies in heroin relative to cocaine users revealed greater
activation in dopaminergic targets for users of heroin compared to users of cocaine. Differential activation patterns
between substances may underlie differences in the clinical characteristics observed in users of cocaine and heroin,
including seeking emotional blunting in users of heroin. More consistent research methodology is needed to provide
adequate studies for stringent meta-­analyses examining common and distinct neural activation patterns across sub-
stances and moderation by clinically relevant factors.

Keywords: fMRI, cocaine, heroin, whole-­brain, cue reactivity, drug

1. INTRODUCTION ­ t al., ­2013). In fact, opioids and stimulants, account for

e
almost 81,000 and 15,000 deaths annually, together repre-
The growing prevalence of substance use disorders
senting approximately 80% of all substance-­ related
among the US population, particularly related to illicit sub-
stances lacking regulation (i.e., opiates, psychostimu- deaths (­Center ­for ­Disease ­Control ­and ­Prevention, ­2023;
lants), thus represents a significant threat to public health. ­National ­Institute ­on ­Drug ­Abuse, ­2022). Given the high
Though the variety of illicit drugs available for use has rates of mortality associated with cocaine and heroin use
increased over time, stimulants (e.g., cocaine) and opioids relative to other substances (­Center ­for ­Disease ­Control
(e.g., heroin, fentanyl, prescription medications) remain ­and ­Prevention, ­2023; ­Darke ­et al., ­2023; ­Lewer ­et al.,
two of the most lethal classes of illicit drugs (­Degenhardt ­2022; ­National ­Institute ­on ­Drug ­Abuse, ­2022; ­Peacock

Received: 18 October 2023 Revision: 21 May 2024 Accepted: 28 May 2024 Available Online: 14 June 2024

© 2024 The Authors. Published under a Creative Commons Imaging Neuroscience, Volume 2, 2024
Attribution 4.0 International (CC BY 4.0) license. https://doi.org/10.1162/imag_a_00211
J.M. Dejoie, N. Senia, A.B. Konova et al. Imaging Neuroscience, Volume 2, 2024

­ t al., ­2021; ­Schneider ­et al., ­2021), we selected these

e ­& ­Kober, ­2022). Characterizing the ways in which the
substances as targets for the present study. These sub- brain differentially responds to craving in people who use
stances are also representative of the broader opioid and cocaine and heroin may provide an important way to
stimulant drug classes, whose use (and co-­use) has surged assess shared and distinct vulnerability mechanisms and
nationally in what has been labeled as the ‘twin opioid-­ may have implications for developing more targeted
stimulant epidemics’. Lastly, although within the broader interventions to prevent their use. Cue reactivity para-
class of stimulant and opioid drugs, others have received digms are reliably related to subjective experiences of
considerable recent attention (e.g., fentanyl), cocaine and drug cravings and are thus widely used, well–­supported
heroin remain the most well studied in fMRI studies of laboratory assessments in samples of individuals using
cue reactivity. substances (­Carter ­& ­Tiffany, ­1999; ­Drummond, ­2000),
Nevertheless, differences in usage patterns between and have been used previously to reveal similarities and
these classes of drugs have also been well documented. differences between individuals who use stimulants, opi-
For example, heroin use is often associated with a more oids, or both (e.g., ­Hochheimer ­et al., ­2023). In fact, a
consistent, chronic, and high rate of use whereas cocaine recent study found that when comparing cue induced
tends to involve more episodic, intermittent, and moderate and tonic craving between opioids and stimulant users in
use (­Hser, ­Evans, ­et al., ­2008; ­Hser, ­Huang, ­et al., ­2008; treatment, opiate cues consistently resulted in greater
­Sartor ­et al., ­2014). Users of heroin also tend to begin use reports of craving (­Hochheimer ­et al., ­2023). Typically,
at a younger age and as a result have a longer duration of such paradigms involve exposure to a drug cue or set of
lifetime use than people who use cocaine (­Hser, ­Evans, drug cues (e.g., images of a drug or associated parapher-
­et al., ­2008; ­Hser, ­Huang, ­et al., ­2008). Onset of heroin, nalia, auditory drug cues via stories or sounds) and the
and opioid use often occurs in conjunction with traumatic subsequent measurement of self-­ reported craving or
events, and with a desire to numb both physical and emo- physiological responses to these cues (e.g., psychophys-
tional pain (­Purdue ­et al., ­2024; ­Shand ­et al., ­2011), while iological signals, patterns of neural activation; ­Carter ­&
users of cocaine may demonstrate higher levels of impul- ­Tiffany, ­1999; ­Courtney ­et al., ­2016; ­Drummond, ­2000).
sivity and impaired inhibition, resulting in higher rates of Task-­based fMRI studies implementing cue reactivity
risk-­taking and reward-­seeking behaviors such as sexual paradigms have been extensively used to characterize the
promiscuity compared to people who use heroin (­Ahn ­& neural circuitry of substance use disorders, including for
­Vassileva, ­2016; ­Bjork ­et al., ­2022; ­Bornovalova ­et al., identifying common and specific patterns associated with
­2005; ­Lejuez ­et al., ­2005; ­Mitchell ­& ­Potenza, ­2014; use of licit and illicit substances (e.g., ­Chase ­et al., ­2011;
­Vassileva ­et al., ­2014). As these behavioral differences do Kuhn & Gallinat, 2011; Zilberman et al., 2019). However, to
not appear to depend on acute drug effects, it is possible our knowledge, there have been no systematic compari-
that they may reflect differences in susceptibility between sons that focus specifically on cocaine versus heroin drug
those who use cocaine versus heroin. cue reactivity, although such comparisons would be highly
Interestingly, whether one is in a social or non-­social relevant given ongoing public health concerns associated
environment appears to impact drug choice (Caprioli et al., with these substances and clinically relevant differences
2009; ­De ­Pirro ­et al., ­2018). In both animal and human in their usage. Studies have revealed that targets of dopa-
models, heroin use tends to be more common within non-­ minergic input such as the striatum, prefrontal cortex, and
social contexts, whereas cocaine use is often preferred in thalamus are recruited during cocaine use (­Bartra ­et al.,
social environments (­Badiani ­et al., ­2011, ­2019; Caprioli ­2013; ­Haber ­& ­Knuston, ­2010; ­Huang ­et al., ­2018; ­Wang
et al., 2009). Such contextual preferences also amplify the ­et al., ­2016) and play a key role in craving and sustaining
respective effects of each substance (­De ­Pirro ­et al., ­2018). the cycle of addiction, consistent with rodent models
In sum, these differences suggest that in addition to shared (­Everitt ­& ­Robbins, ­2005; ­Goldstein ­& ­Volkow, ­2002;
mechanisms, there may also be divergent mechanisms ­Huang ­et al., ­2018; ­Jentsch ­& ­Taylor, ­1999). Similarly, her-
that are important to identify to better understand patterns oin recruits dopaminergic targets such as the dlPFC,
of usage of cocaine and heroin. Identifying divergent insula, and orbitofrontal cortex (OFC; Liu et al., 2021); it
mechanisms (i.e., pinpointing behavioral and neurobiolog- also seems to target other systems including the glutama-
ical risk-­factors for the development of a particular SUD, tergic system (Gheraldini et al., 2015) which is conceptu-
developing drug-­specific treatment interventions) is imper- alized as the basis for heroin’s analgesic effect.
ative given the rapidly increasing rates of use and associ- Further, prior meta-­ analyses studying users of these
ated harm caused by these illicit substances. substances have largely utilized broad inclusion criteria, or
A common component of substance use disorders examined outcomes other than functional activation (e.g.,
that is closely tied to reuse and relapse is drug craving gray matter, structural differences; ­Hill-­Bowen ­et al., ­2022;
(i.e., the strong wanting or desire for a substance; ­Vafale ­Klugah-­Brown ­et al., ­2021; ­Long ­et al., ­2022; ­Pando-­Nuande

2
J.M. Dejoie, N. Senia, A.B. Konova et al. Imaging Neuroscience, Volume 2, 2024

­ t al., ­2021). Although studies have begun to identify com-

e versus neutral condition whole-­brain contrast within the
mon neural circuits implicated in cocaine and heroin use, substance using group; (5) the study utilized a cue reac-
there is no existing meta-­analytic synthesis of the extant tivity paradigm; and (6) reported coordinates in either
literature to quantitatively evaluate similarities and differ- Montreal Neurological Institute (MNI) or Talairach and
ences in the functional neural underpinnings of cocaine Tournoux space. Studies were excluded if (1) they lacked
versus heroin use. These methodological differences a substance versus neutral whole-­brain contrast analysis;
impact the ability to characterize similarities and differ- (2) included PET data; (3) there was insufficient reporting
ences in cue reactivity between users of each substance. of fMRI data or findings; (4) they did not report coordi-
Additionally, an important factor to consider is whether nates from whole-­brain analyses; (5) neither people who
similarities/differences in neural recruitment during cue use cocaine nor heroin were included in the study; and (6)
reactivity across users of cocaine and heroin vary as a analyses were not separated according to drug type.
function of individual differences in length of drug use/chro- Studies were selected based on the above inclusion
nicity. Thus, a more stringent synthesis regarding the neural and exclusion criteria and were reviewed by multiple
circuits implicated in cocaine and heroin usage is needed. authors (J.D., N.S., D.F.). The corresponding author made
To address these gaps, we conducted a meta-­analysis final decisions regarding inclusion/exclusion conflicts.
of fMRI studies aimed at comparing activation patterns Study quality was assessed using a recently proposed
during cue reactivity tasks for drug relative to neutral cues checklist for cue reactivity studies (­Ekhtiari ­et al., ­2022).
in people who use cocaine and people who use heroin. We Given the recency of this publication compared to the
applied strict inclusion criteria for our meta-­analysis such dates of publication for our studies, we set our threshold
as utilizing studies implementing within subjects, whole-­ for quality at 50% of criterion met within a category for at
brain analyses of drug versus neutral cues. We expected least 5 of the 10 proposed categories. All studies included
that studies of both people who use cocaine and people in the current study exceeded this threshold meeting crite-
who use heroin will show significant engagement of reward-­ rion for at least 7 of the 10 proposed categories. Figure 1
related and dopaminergic circuits during cue reactivity depicts the process of literature search and study selec-
tasks, given the well-­documented recruitment of these cir- tion. 1025 studies were first identified via the literature
cuits by both substances. We also expected that people search; any duplicates from searches were removed. Next,
who use cocaine will exhibit greater activation in corticostri- 949 studies were excluded based on title/abstract review.
atal circuits than people who use heroin, given possible dif- The remaining 76 full text articles were assessed for eligi-
ferences in levels of impulsivity, risk taking, and impaired bility. Exclusion of studies occurred for a variety of rea-
inhibition observed in cocaine relative to heroin. We also sons, including: (1) participant groups were composed of
explored the impact of length of use on neural correlates of people who use opiates, in general, as opposed to solely
cue reactivity, with the expectation that length of use would people who use heroin; (2) only between-­group analyses
be negatively associated with activation in dopaminergic were conducted comparing people who use substances
regions, including the prefrontal cortex and regions associ- to healthy controls—no within-­subjects analyses examin-
ated with reward processing such as the striatum. ing responses to drug versus neutral cues; (3) lack of
reporting of whole brain analyses, a focus on only connec-
tivity or solely ROI analyses; and (4) lastly, an implementa-
2. METHOD tion of experimental tasks other than a drug versus neutral
cue task (i.e., stroop word task, drug versus food cues,
2.1. Literature search and study selection
etc.). Of the 76 full-­text studies, 58 studies were excluded
A literature search was conducted via PubMed and Goo- based on the above criteria, leaving 18 studies remaining
gle Scholar (up to April 2022) to identify applicable stud- (total N = 447). Each subgroup—­cocaine and heroin—­was
ies. Keywords used were (‘cocaine’, ‘heroin’, ‘substance composed of 9 studies (Ncocaine = 273; Nheroin = 174) that
use’, ‘cue-­reactivity’, ‘fMRI’). Search strings used were were included in this meta-­analysis (see Table 1). Group
(‘cocaine fMRI cue-­reactivity', ‘heroin fMRI cue-­reactivity') differences in sample characteristics, compared using
along with other combinations of the keywords. All refer- independent samples t-­tests, across included studies are
ence lists from identified studies were checked manually presented in Table 3.
for additional applicable studies. In the current meta-­
analysis, studies were included if (1) analyses were con-
2.2. Seed based d mapping
ducted via fMRI; (2) participants were medically healthy
aside from their substance use adults over the age of 18; Seed-­based d mapping (SDM) SDM-­ PSI v. 6.22 (for-
(3) participants were either abstaining from or currently merly Signed Differential Mapping) is a specific statisti-
engaging in drug use; (4) the study included a substance cal method used for meta-­analyzing structural and/or

3
J.M. Dejoie, N. Senia, A.B. Konova et al. Imaging Neuroscience, Volume 2, 2024

Fig. 1. Flowchart PRISMA diagram outlining the process of literature search and study selection.

Table 1. Characteristics of 18 fMRI studies included in the contrast meta-­analysis, sub-­group mean analyses,
and the 15 included in exploratory meta-­regression.
Included in
Author Year N Task Drug Space meta-­regression
Duncan, E., Boshoven, W. Harenski, K., et al. ­2007 10 Script guided C MNI Y
mental imagery
Elton, A., Smitherman, S., Young, J., et al. ­2015 20 Script guided C MNI Y
mental imagery
Goldstein, R., Tomasi, D., Alla-­Klein, N., et al. ­2009 15 Word-­cue task C TAL Y
Hanlon, C. A., Dowdle, L. T., Gibson, N. B., et al. ­2018 55 Image-­cue task C MNI N
Hassani-­Abharian, P., Ganjgahi, H., ­2015 20 CRET task H MNI Y
Tabatabaei-­Jafari, H., et al
Kaag, A. M., Reneman, L., Homberg, J., et al. ­2018 59 Image-­cue task C MNI Y
Li, Q., Wang, Y., Zhang, Y., et al. ­2012 24 Image-­cue task H MNI Y
Li, Q., Wang, Y., Zhang, Y., et al. ­2013 19 Image-­cue task H MNI Y
Mei, W., Zhang, J. X., & Xiao, Z ­2010 12 Image-­cue task H MNI Y
Prisciandaro, J. J., Myrick, H., Henderson, S., ­2013 25 Image-­cue task C MNI Y
McRae-­Clark, A. L., Santa Ana, E. J., et al.
Prisciandaro, J. J., Myrick, H., Henderson, S., ­2013 30 Image-­cue task C MNI N
McRae-­Clarck, A. L., & Brady, K. T.
Prisciandaro, J. J., McRae-­Clark, A. L., ­2014 38 Image-­cue task C MNI Y
Myrick, H., et al.
Walter, M., Denier, N., Gerber, H., et al. ­2015 27 Image-­cue task H MNI Y
Wang, W., Li, Q., Wang, Y., et al. ­2011 14 Image-­cue task H MNI Y
Wang, A.-­L., Elman, I., Lowen, S. B., et al. ­2015 32 Event related and H TAL N
image cue tasks
Xiao, Z., Lee, T., Zhang, J. X., et al. ­2006 14 Image-­cue task H TAL Y
Zhang, S., Zhornitsky, S., Wang, W., et al. ­2020 52 Image-­cue task C MNI Y
Zijlstra, F., Veltman, D. J., Booji, J., et al. ­2009 12 Image-­cue task H MNI Y

functional differences in the brain across studies using a from individual studies are entered. Next, during pre-­
variety of different neuroimaging techniques, including processing, the upper and lower bounds of possible
fMRI, VBM, DTI, or PET (­Albajes-­Eizagirre, ­Solanes, effect sizes (Hedges g) and variances are calculated for
­Vieta, ­et al., ­2019). The SDM method consists of a series each study. A mean-­analysis is then performed where
of steps. To begin, coordinates of identified cluster peaks maximum-­ likelihood estimation (MLE) and MetaNSUE

4
J.M. Dejoie, N. Senia, A.B. Konova et al. Imaging Neuroscience, Volume 2, 2024

are implemented to estimate the most likely effect size inter-­study heterogeneity of each cluster surviving cor-
for each included study and its standard error (­Albajes- rection was measured by the I2 index, which represents
­Eizagirre, ­Solanes, ­Vieta, ­et al., ­2019). MetaNSUE is a the proportion of total variation due to study heterogene-
meta-­analytic method that decreases bias and permits ity (Higgins & Thompson, 2002). I2 > 50% commonly indi-
inclusion of non-­ statistically significant unreported cates considerable heterogeneity. The maps resulting
effects (NSUEs) (­Albajes-­Eizagirre, ­Solanes, ­Radua, from the contrast analyses were also corrected to a level
­2019). MetaNSUE involves multiple imputations that are of p < 0.005 TFCE-­FWER, with a cluster extent threshold
subsequently created by adding noise to the estimations of ≥ 10 voxels. Subsequent bias tests (e.g., funnel plots,
within the previously calculated bounds (­Albajes-­Eizagirre, I2 index, metabias test, and excess significance tests)
­Solanes, ­Radua, ­2019). Each of the imputed data sets is were conducted for this analysis using the SDM software.
then meta-­ analyzed, and Rubin’s rules (a pooling of The I2 value is generated during the extraction of the
parameter estimates in an imputed data set to provide peaks and represents the percentage of heterogeneity
confidence intervals and p-­values) are applied to com- across the data and studies included in the meta-­analysis
bine the imputed meta-­ analyzed datasets. Lastly, a (­Albajes-­Eizagirre, ­Solanes, ­Fullana, ­et al., ­2019). The
subject-­based permutation test is conducted in which funnel plot test identified if there was asymmetry in the
the maximum statistics, (i.e., the largest z-­value reflect- plot (i.e., whether there were larger effect sizes in smaller
ing the peak of activation across the whole brain in each studies). Last, the excess significance test indicates if
study), from the combined meta-­analysis images, are there was publication bias (i.e., whether studies are only
saved (­Albajes-­Eizagirre, ­Solanes, ­Radua ­et al., ­2019). A published if the results are significant; ­Albajes-­Eizagirre,
family-­ wise error rate correction (FWER) for multiple ­Solanes, ­Fullana, ­et al., ­2019).
comparisons is then performed using the distribution of
the maximum statistic.
2.5. Exploratory meta-­regression

2.3. Meta-­analysis procedure We also tested whether length of use moderated activa-
tion by incorporating length of use and a substance type
Coordinates of peak activation and additional pertinent x length of use interaction into the model. Due to limited
information (e.g., statistical values [p-­ value, t-­value, z-­ data in the included studies regarding demographic and
value], statistical thresholds, and sample sizes) for each clinical features, no other features were included as
within-­group drug cue > neutral contrast from the included meta-­regressors. The inter-­study heterogeneity of each
studies were collected and converted into text files. Any cluster surviving correction was measured by the I2 index,
p-­values or z-­values reported in the studies were con- which represents the proportion of total variation due to
verted to t-­values using the SDM online converter. We then study heterogeneity (Higgins & Thompson, 2002). I2 >
conducted a voxel-­based meta-­analysis using SDM-­PSI v. 50% commonly indicates considerable heterogeneity.
6.22 for all studies included, irrespective of drug type, and The maps resulting from the contrast analyses were also
for each subgroup. We also conducted a subsequent lin- corrected with p < 0.005 TFCE-­ FWER, with a cluster
ear model. The same parameters were utilized for all anal- extent threshold of ≥ 10 voxels. Nilearn was used to visu-
yses and are outlined as follows. Pre-­preprocessing for alize results from all analyses (­Abraham ­et al., ­2014).
each group was conducted within a gray matter mask
using the default settings within SDM-­PSI: a 20-­mm aniso-
tropic full width half maximum (FWHM) kernel and 2-­mm 3. RESULTS
voxel size (­Albajes-­Eizagirre, ­Solanes, ­Vieta, ­et al., ­2019).
3.1. Clinical and demographic differences
The mean analysis and permutation tests were then per-
formed within each group of studies separately (cocaine, Assessment of group differences in age and length of use
heroin), with the number of imputations set to 50 and the revealed a significant difference in age with cocaine par-
number of permutations set to 5000. The results of the ticipants being significantly older but no difference in
meta-­ analyses of the subgroups were then corrected length of use (see Table 3 for full description of study sam-
using the FWER correction with 5000 permutations and ple characteristics). Further, all studies excluded partici-
subsequently thresholded at p < 0.05. pants with a co-­occurring SUD outside the drug of interest
in the study; however, most studies permitted recreational
tobacco and alcohol use. Four studies permitted inclusion
2.4. Contrast analysis
of participants with mood disorders but did not specify
We conducted a linear regression to compare the two whether any participants met criteria. Abstinence was
subgroups, Heroin > Cocaine and Cocaine > Heroin. The reported variably across studies, making it difficult to

5
J.M. Dejoie, N. Senia, A.B. Konova et al. Imaging Neuroscience, Volume 2, 2024

Table 2. Clinical characteristics of heroin and cocaine subgroups.

Abstinent Mean age Length
Study Drug (Y/N) Co-­use Sex (years) of use (mos)
Duncan, E., Boshoven, W., C Y No Male 43.6 15.9
Harenski, K., et al.
Elton, A., Smitherman, S., Young, J., C N No Male 40 176.4
et al.
Goldstein, R., Tomasi, D., Alla-­Klein, C N Alcohol, tobacco, 12 M/ 3 F 43.6 188.4
N., et al. marijuana
Hanlon, C. A., Dowdle, L. T., C Y Nicotine 38 M/17 F 42.7 NA
Gibson, N. B, et al.
Hassani-­Abharian, P., Ganjgahi, H., H N No Male 31.9 36.24
Tabatabaei-­Jafari, H., et al
Kaag, A. M., Reneman, L., C N No other opiates Male 31.4 72
Homberg, J., et al.
Li, Q., Wang, Y., Zhang, Y., et al. H Y No other opiates Male 32.8 78.6
Li, Q., Wang, Y., Zhang, Y., et al. H Y No Male 32.2 80.5
Mei, W., Zhang, J. X., & Xiao, Z. H Y No 12 M/1 F 33.5 54
Prisciandaro, J. J., Myrick, H., C Y Tobacco, Male 48.8 215.6
Henderson, S. McRae-­Clark, nicotine
A. L., Santa Ana, E. J., et al.
Prisciandaro, J. J., Myrick, H., C Y Marijuana, 24 M/6 F 48.4/31.2 NA
Henderson, S., McRae-­Clarck, alcohol, nicotine
A. L., & Brady, K. T.
Prisciandaro, J. J., McRae-­Clark, C Y No 33 M/5 F 47.7/43.96 213.5
A. L., Myrick, H., et al.
Walter, M., Denier, N., Gerber, H., et al. H Y No 19 M/8 F 41.1 253
Wang, W., Li, Q., Wang, Y., et al. H Y No Male 36.1 94.8
Wang, A.-­L., Elman, I., Lowen, H Y No 17 M/15 F 29.19 NA
S. B., et al.
Xiao, Z., Lee, T., Zhang, J. X., et al. H Y No Male 33.2 85.2
Zhang, S., Zhornitsky, S., Wang, W., C Y No 42 M/10 F 46.7/41.0 189.4
et al.
Zijlstra, F., Veltman, D. J., Booji, J., H Y No Male 42.8 192
et al.
Abstinence reflects whether participants were free from substances at the time of data collection.

directly compare and determine participants’ length of lowing the general meta-­analysis in SDM (see Table 4 and
abstinence at the time of participation. When evaluating Fig. 2). Regions within these clusters (peak x, y, z coordi-
abstinence based on whether participants were abstinent nates reported in MNI space) include the (1) right poste-
at data collection (yes/no), participants across all studies rior cingulate (6, -­44, 30), (2) right inferior frontal gyrus
were largely abstinent. Overall, participants were largely (48, 12, 32), (3) left inferior frontal gyrus (-­50, 10, 28), (4)
biological males. In fact, 10 of the 18 studies were an all-­ left superior frontal gyrus (-­4, 54, 16), and (5) left inferior
male sample. Of the remaining 8 mixed-­sex studies, 7 of temporal gyrus (-­46, -­68, -­8). One peak, the left inferior
them were still composed of a majority of male partici- temporal gyrus, had an I2 value that exceeded 40%
pants (see Tables 2 & 3). In general, cocaine studies (I2 = 52%), suggesting that the between-­study variance
appeared to have more mixed-­sex samples, though still was low with most variance attributed to within-­study
largely male dominant. Treatment history was evaluated sampling error, and therefore, the metabias test for this
based on whether participants were undergoing peak was not interpreted. The remaining peaks did not
medication-­assisted treatment at the time of data collec- have I2 values that exceeded 25% and therefore, meta-
tion. Across all studies, participants were not undergoing bias tests were interpreted. Metabias tests on the remain-
treatment except those in two of the heroin studies. ing peaks revealed an additional cluster, the left inferior
frontal gyrus that demonstrated bias/heterogeneity
(p > 0.05). The T2 value was reviewed for this peak and
3.2. Aggregate meta-­analysis
revealed low dispersion (T2 = 0.01). All other findings
Five clusters/peaks were identified as statistically signifi- were supported as Metabias test results were non-­
cant across all studies for drug cues > neutral cues fol- significant (all p’s > 0.2).

6
J.M. Dejoie, N. Senia, A.B. Konova et al. Imaging Neuroscience, Volume 2, 2024

Table 3. Group differences in age and length of use across subgroups.

Heroin Cocaine t p-­value
Age (m, SD, range) 34.75, 4.48, 29.2-­42.8 42.18, 4.88, 31-­48.8 3.36 p < 0.01
Length of use (m, SD, range) 109.29, 74.11, 36.2-­253 154.31, 78.3, 15.9-­215.6 1.14 p > 0.05
Abstinent at time of study 3/9 1/9 —­ —­
Co-­occurring SUDS 0/9 0/9 —­ —­
Treatment (medication 2/9 0/9 —­ —­
assisted at time of study)
Biological sex (% male) 86% (174 total) 87% (304 total) —­ —­

Table 4. Location of neural activations for aggregate meta-­analysis of Heroin and cocaine studies for drug > neutral cues.
ID Region x y z Voxels p SDM -­Z
1 Right posterior cingulate gyrus, BA 23 6 -­44 30 1817 <0.001 5.915
2 Right inferior frontal gyrus, opercular part, BA 44 48 12 32 620 <0.001 7.222
3 Left inferior frontal gyrus, opercular part, BA 44 -­50 10 28 185 0.002 6.058
4 Left superior frontal gyrus, medial, BA 10 -­4 54 16 223 0.01 4.970
5 Left inferior temporal gyrus. BA 37 -­46 -­68 -­8 75 0.02 4.932
Peaks organized by decreasing voxel size.

Fig. 2. Activated peaks associated with the aggregate mean analysis of 18 cocaine and heroin cue reactivity studies for
drug > neutral cues. Threshold was set at p < 0.005 TFCE-­FWER and cluster size ≥ 10 voxels. Coordinates were reported
in MNI space. See Table 4 for a full list of activated peaks.

3.3. Meta-­analysis of subgroups amus (6, 48, 42), (4) left amygdala (-­28, -­2, -­22), (5) left
inferior temporal gyrus (-­48, -­66, -­6), and (6) right cere-
Fourteen clusters/peaks were identified as statistically
significant in the Heroin subgroup for drug cues > neu- bellum (24, -­74, -­44). For a full list of significantly acti-
tral cues following the general meta-­analysis in SDM vated regions and their respective coordinates, see
(see Fig. 3). Regions within these clusters (peak x, y, z Table 5. Only one cluster/peak—­right precentral gyrus/
coordinates reported in MNI space) include the (1) left inferior frontal gyrus (46, 4, 30)—­was identified as sta-
medial cingulate (-­2, 30, 32), (2) right superior (6, 48, 42) tistically significant in the cocaine subgroup for drug
and inferior (48, 14, 32) frontal gyri, (3) right anterior thal- cues > neutral cues following the general meta-­analysis

7
J.M. Dejoie, N. Senia, A.B. Konova et al. Imaging Neuroscience, Volume 2, 2024

Fig. 3. Activated peaks associated with the mean analysis of 9 heroin cue reactivity studies where drug > neutral cue.
Threshold was set at p < 0.005 TFCE-­FWER and cluster size ≥ 10 voxels. Coordinates were reported in MNI space. See
Table 4 for a full list of activated peaks.

Table 5. Location of neural activations for mean analysis heroin and cocaine subgroups for drug > neutral cues.
ID Region x y z Voxels p SDM -­Z
Heroin subgroup
1 Left median cingulate/paracingulate gyri, BA 23 -­2 -­30 32 3194 <0.001 6.521
2 Right superior frontal gyrus, medial, BA 9 6 48 42 2866 <0.001 4.885
3 Right anterior thalamic projections 4 -­16 -­2 1402 <0.001 5.766
4 Right inferior frontal gyrus, opercular part, BA 44 48 14 32 1096 <0.001 5.554
5 Left amygdala, BA 36 -­28 -­2 -­22 815 <0.001 5.5
6 Left inferior temporal gyrus, BA 37 -­48 -­66 -­6 815 <0.001 5.830
7 Right cerebellum, hemispheric lobule, VIIB 24 -­74 -­44 781 <0.001 5.082
8 Right amygdala, BA 36 28 -­2 -­24 653 <0.001 5.186
9 Left inferior frontal gyrus, opercular part, BA 44 -­50 8 28 663 <0.001 5.65
10 Right angular gyrus, BA 7 32 -­60 52 664 <0.001 5.354
11 Right inferior temporal gyrus, BA 37 48 -­70 -­4 670 <0.001 5.179
12 Left inferior parietal (excluding supramarginal and angular) -­50 -­32 46 590 0.005 4.263
gyri, BA 2
13 Left inferior parietal (excluding supramarginal and angular) -­36 -­62 48 387 0.005 4.507
gyri, BA 7
14 Left caudate nucleus -­8 10 10 249 0.003 4.629
Cocaine subgroup
1 Right precentral gyrus/inferior frontal gyrus, BA 44 46 4 30 144 0.008 5.394
Peaks organized by decreasing voxel size.

(see Fig. 4). Considering the small number of studies 3.4. Contrast analysis
(less than 10) that fell within each subgroup, bias test
Our primary interest was in examining differential recruit-
results were conducted but not interpreted (SDM-­PSI v. ment of neural circuits at the whole-­brain level between
6.22). However, based on the results of the aggregate the cocaine and heroin subgroups in response to drug rel-
meta-­ analysis results, the left inferior temporal gyrus ative to neutral cues. We performed a contrast analysis to
and left inferior frontal gyrus peaks revealed in the investigate this question. We observed increased activa-
heroin-­only meta-­analysis may be biased. tion in the Heroin > Cocaine contrast in several regions

8
J.M. Dejoie, N. Senia, A.B. Konova et al. Imaging Neuroscience, Volume 2, 2024

Fig. 4. Activated peaks associated with the mean analysis of 9 cocaine cue reactivity studies where drug cue < neutral
cue. Threshold was set at p < 0.005 TFCE-­FWER and cluster size ≥ 10 voxels. Coordinates were reported in MNI space.
Image shows the precentral gyrus.

Fig. 5. Peak activations associated with the contrast analysis of heroin > cocaine subgroups for drug > neutral cues.
Threshold was set at p < 0.005 TFCE-­FWER and cluster size ≥ 10 voxels. Coordinates were reported in MNI space.

(see Fig. 5), including: (1) right inferior (48, -­74, -­4) and heterogeneity was found for each significant peak
superior (30, -­68, 50) parietal gyrus; (2) thalamus (6, -­14, (I2 = 0.74-­32.2%). Funnel plots were symmetric, suggest-
2); (3) caudate nucleus (-­10, 10, 10); (4) right superior fron- ing that none of the results were a function of a small sub-
tal gyrus (6, 48, 40); and (5) right cerebellum (20, -­80,-­40). set of included studies or by studies with a small sample
For a full list of significantly activated regions and their size. Metabias tests supported these findings as results
respective coordinates, see Table 6. Low between-­study were all non-­ significant (all p’s > 0.58). No significant

9
J.M. Dejoie, N. Senia, A.B. Konova et al. Imaging Neuroscience, Volume 2, 2024

Table 6. Location of neural activations for heroin > cocaine contrast of drug > neutral cue studies.
Bias test
ID Region x y z Voxels p SDM -­Z (p-­value) I2
1 Right inferior temporal gyrus, BA 19 48 -­74 -­4 628 <0.001 4.313 0.603 32.199
2 Right superior parietal gyrus, BA 7 30 -­68 50 473 <0.001 4.387 0.598 11.118
3 Right anterior thalamic projections 6 -­14 2 297 <0.001 4.345 0.641 7.040
4 Left caudate nucleus -­10 10 10 94 0.01 3.376 0.716 6.025
5 Right superior frontal gyrus, BA 9 6 48 40 81 0.01 3.318 0.703 16.626
6 Right cerebellum, crus II 20 -­80 -­40 67 0.03 2.655 0.771 0.739
7 Left middle temporal gyrus, BA 37 -­54 -­62 2 54 0.02 3.169 0.712 4.254
8 Left inferior parietal (excluding supramarginal -­50 -­32 46 26 0.03 2.821 0.720 8.739
and angular) gyri, BA 2
Peaks arranged by decreasing voxel size.

Table 7. Location of neural activations for exploratory meta-­regression of interaction of drug (heroin and cocaine) for
drug > neutral cues and average length of use.
ID Region x y z Voxels p SDM -­Z I2
1 Left superior frontal gyrus, medial, BA 10 -­2 62 20 271 <0.001 -­3.017 1.48
2 Left superior frontal gyrus, dorsolateral, BA 9 -­18 38 46 193 <0.001 -­3.003 3.15
Peaks arranged by decreasing voxel size.

clusters/peaks survived correction for the Cocaine > Her- 4. DISCUSSION

oin contrast. No study had an I2 value that exceeded 32%,
In the present study, we sought to investigate the differen-
suggesting that the between-­study variance was low with
tial neural underpinnings of these two major drug classes
most variance attributed to within-­study sampling error.
with respect to drug cue reactivity, a clinically significant
process associated with recurrent use. We specifically
3.5. Exploratory meta-­regression focused on studies implementing cue reactivity para-
digms, a gold-­standard approach to the clinical neurobiol-
An important consideration when examining activation
ogy of craving, with a focus on responses to drug cues
differences across groups is whether they are moderated
relative to neutral cues. Our results demonstrate that there
by the lifetime length of use of the substance, as resulting
are previously underappreciated differences in activation
activation patterns to drug cues may vary in accordance
between people who use heroin and people who use
with acute and chronic usage. To address this question,
cocaine. Specifically, increased (rather than decreased)
we conducted an exploratory meta-­regression examining
activation in dopaminergic targets such as the striatum
a whole-­brain interaction between drug type and average
length of use (in months). Fifteen of the 18 original stud- and mPFC, as well as the temporal and parietal gyri and
ies were used for this analysis, as 3 studies did not pro- cerebellum were found in people who use heroin that are
vide average length of use data (see Table 1 for full list of not readily attributable to demographic or clinical differ-
included studies). This exploratory analysis revealed a ences across groups (see Tables 2 & 3).
negative interaction in clusters, including the left superior
frontal gyrus (both the medial (-­2, 62, 20) and dorsolateral
4.1. Aggregate and substance specific effects
areas (-­18, 38, 46)), which suggests that activation here
decreased as length of use increased in people who use Aggregate mean analysis results revealed significant acti-
heroin (see Table 7 and Fig. 6). Low between-­study het- vation in the posterior cingulate gyrus, a region within the
erogeneity was found for each significant peak (I2 = 1.48-­ default mode network (DMN). The posterior cingulate, as
3.14%). Funnel plots were symmetric, suggesting that well as the DMN at large, has been (e.g., right posterior
none of the results were a function of a small subset of cingulate gyrus) widely implicated in cue reactivity, as
included studies or by studies with a small sample size. well as other paradigms, across substance use disorders
The potential publication bias tests were all non-­ (­Devoto ­et al., ­2020; ­Englemann ­et al., ­2012; ­Janes ­et al.,
significant (p = 0.998). No significant clusters of activa- ­2016; ­Wang ­et al., ­2020). Further, our results also revealed
tion survived whole-­brain correction for people who use significant activation in the superior and inferior frontal
cocaine. No study had an I2 value that exceeded 32%. gyrus, which converge with recent meta-­analyses, though

10
J.M. Dejoie, N. Senia, A.B. Konova et al. Imaging Neuroscience, Volume 2, 2024

Fig. 6. Peak activations across 15 studies of the interaction between length of use and drug type for drug > neutral cues.
Threshold was set at p < 0.005 TFCE-­FWER and cluster size ≥ 10 voxels. Coordinates reported in MNI space.

not all specific to cue reactivity (­Chase ­et al., ­2011; ­Yang ­et al. ­(2009) conducted a study examining neural
­Hill-­Bowen ­et al., ­2021; ­Klugah-­Brown ­et al., ­2020). In responses to cue reactivity in people who use heroin,
sum, our results are in line with previous findings as well reporting activation in the amygdala and cerebellum.
as behavioral models of addiction, suggesting dysfunc- Studies have also implicated increased recruitment of
tion in habit development, modulation of behavior, and frontal regions and reward circuity, albeit with some vari-
cognitive control (­Buhle ­et al., ­2014; Everitt & Robbins, ation in the specific peak locations when compared with
2016; ­Klugah-­Brown ­et al., ­2020; Koob & Volkow, 2016; our findings (­Pollard ­et al., ­2023; ­Yang ­et al., ­2009). How-
­Wiers ­et al., ­2015; ­Zilverstand ­et al., ­2018). ever, support for this result is variable, with other studies
Results from the mean analysis of heroin studies finding decreased activation in the left precentral gyrus
revealed significant activation in the left cingulate, the (­Moeller ­et al., ­2010). It is difficult to compare these stud-
right superior and inferior frontal gyrus, the right anterior ies as methodology and task design varies greatly, how-
thalamic projections, the right and left amygdala, the left ever, indicating a need for replication of these findings
caudate nucleus, and the right cerebellum. These find- using consistent task designs.
ings are consistent with rodent work indicating that acute
heroin administration within the VTA enhanced calcium
4.2. Between-­substance effects
signaling in dopaminergic neurons as well as in seroto-
nergic neurons within the dorsal raphe nucleus (DRN) Interestingly, and contrary to our hypotheses, our direct
(­Wei ­et al., ­2018), which, in turn, was associated with comparison of cue reactivity between substances
increased locomotor function. We note that similar pat- revealed increased recruitment of dopaminergic regions
terns of results were reported in this study for administra- and other areas in people who use heroin relative to peo-
tion of cocaine into the VTA and DRN (­Wei ­et al., ­2018), ple who use cocaine. Given the robust literature on the
though we did not see consistent recruitment of dopami- involvement of reward processing regions such as the
nergic regions in our mean analysis of cocaine studies. striatum and PFC across cocaine, heroin, and other sub-
Instead, results for the cocaine mean analyses revealed stances (Cooper et al., 2019; ­Hobkirk ­et al., ­2019;
significant peaks of activation in the right precentral ­Klugah-­Brown ­et al., ­2020; ­Motzkin ­et al., ­2014; ­Patel
gyrus. However, the influence of heterogeneity on cocaine ­et al., ­2013; ­Sell ­et al., ­2001; ­Volkow ­et al., ­2010; ­Yan
results could not be rigorously tested due to a lack of ­et al., ­2023), we did expect to find at least some regions
data of clinical features. Our findings converge with some of overlap, but our meta-­analysis may have been under-
prior fMRI studies examining cue reactivity in people who powered to detect such effects. However, this could be
use heroin and people who use cocaine. For example, explained by differential effectiveness of drugs eliciting

11
J.M. Dejoie, N. Senia, A.B. Konova et al. Imaging Neuroscience, Volume 2, 2024

cue reactivity; in other words, heroin cues were more may parallel these findings, as both cocaine and heroin
effective at eliciting cue reactivity. A recent study found users show general activation in regions associated with
that when comparing cue induced and tonic craving attention and cognitive control (e.g., prefrontal cortex,
between opioid and stimulant users in treatment, opioid parietal regions); however, users of heroin seem to show
cues consistently resulted in greater reports of craving more significant activation in regions associated with
(­Hochheimer ­et al., ­2023). Such results may, in turn, sup- cognitive control compared to users of cocaine, a finding
port the increased number of regions recruited in our which may be related to the differences in attentional dis-
sample of heroin cue reactivity studies. Alternatively, it is ruptions found in behavioral tasks (­Bjork ­et al., ­2022;
also possible that users of cocaine in the studies in our ­Psederska ­& ­Vassileva, ­2023).
sample may be better able to engage regulatory mecha- A recent study comparing whole-­brain white matter
nisms during their responses to cue-­induced craving as abnormalities in users of both substances found that
suggested by dlPFC activation, a region consistently while both people who use heroin and people who use
implicated in regulation of craving (­Brevers ­et al., ­2023; cocaine showed abnormalities in white matter, people
­George ­& ­Koob, ­2013; ­Hayashi ­et al., ­2013; ­Kober ­et al., who use heroin showed significantly decreased fractional
­2010; Li et al., 2017). anisotropy overall in the brain compared with people who
However, given the known neuropharmacological and use cocaine (­Gaudreault ­et al., ­2022), suggestive of
clinical differences between these substances, the diver- decreased white matter integrity. Our results may dove-
gence in neural activation that emerged in our study is tail with these in that we find studies of people who use
plausible. Support for this possibility comes from animal heroin showing more diffuse activation than people who
models that have demonstrated distinct recruitment of use cocaine in dopaminergic regions, as well as the right
subpopulations of neurons within the striatum and PFC inferior and left middle temporal gyrus, the right thala-
when comparing neural responses to cocaine and heroin mus, and the right cerebellum. Additional evidence from
(­Badiani ­et al., ­2011; ­Chang ­et al., ­1998; ­Ettenberg ­et al., studies examining volumetric impacts of cocaine and
­1982; ­Pettit ­et al., ­1984). Further, in spite of differing loca- heroin use have also converged on the involvement of the
tions of activation associated with heroin and cocaine nucleus accumbens, which has shown significant gray
cues, there is evidence from other domains to suggest that matter reduction, in people who use both substances
differing locations of activation can be mapped back to the when compared to healthy controls; however, we note
same network, ultimately implicating an overall common that this pattern was more robust in people who use her-
pathway (­Jousta ­et al., ­2022; Padmanabhan et al., 2019; oin (­Carlezon ­& ­Thomas, ­2009; ­Ceceli ­et al., ­2022; ­Seifert
­Stubbs ­et al., ­2023). Such a pattern suggests that viewing ­et al., ­2015). Similar findings were reported in the vmPFC
all substances as having the same neurobiological sub- (­Ceceli ­et al., ­2022). These observed volumetric differ-
strates because of common activation of regions such as ences may suggest biomarkers of susceptibility to sub-
the striatum and PFC, to some degree, may be an over- stance (ab)use or consequences of prolonged use, which
generalization. It is, therefore, plausible that the activa- may, in turn, be related to alterations in functional activa-
tion differences we observed reflect preliminary neural tion (­Ceceli ­et al., ­2022; ­Seifert ­et al., ­2015). The incorpo-
underpinnings for the cognitive, behavioral, and clinical ration of both functional and volumetric measurements,
differences between heroin and cocaine use, including while controlling for contextual factors (e.g., duration of
frequency of use and impulsive behavior, but replication use, recency of use, age of onset) when examining the
and further research is needed (­Bornovalova ­et al., ­2005; implications of substance use in future studies, may help
­Hser, ­Evans, ­et al., ­2008; ­Hser, ­Huang, ­et al., ­2008; ­Lejuez address these questions (­Oakes ­et al., ­2007).
­et al., ­2005).
For example, cocaine and heroin have both been
4.3. Length of use moderates heroin cue reactivity
implicated in attentional dysfunction though the way in
activation
which each drug disrupts attention varies, such that
individuals who use heroin display delayed attentional Our exploratory meta-­regression demonstrated that acti-
responses, whereas users of cocaine demonstrate vation in the left superior frontal gyrus, both the medial and
increased interference with attention (­Bjork ­et al., ­2022). dorsolateral areas, decreased as length of use increased in
Opioid use has also been linked to impacts on decision people who use heroin, controlling for age. However, no
making, including decreased response inhibition and significant relationship between cocaine and length of use
delayed gratification (­Psederska ­& ­Vassileva, ­2023). Opi- was found. Literature examining the effect of duration of
ate use may affect other aspects of decision making, use on brain activation is limited; however, our findings fit
such as risky choices, differentially depending on length within literature, suggesting that drug use is known to
of abstinence (­Psederska ­& ­Vassileva, ­2023). Our results impact the prefrontal cortex (­Goldstein ­& ­Volkow, ­2012). A

12
J.M. Dejoie, N. Senia, A.B. Konova et al. Imaging Neuroscience, Volume 2, 2024

model put forth by ­Volkow ­et al. ­(2010) posits that one of ing a standardized demographic and clinical data collec-
the effects of chronic drug use is decreased reward sensi- tion procedure that includes relevant contextual factors
tivity or hypoactivation in the prefrontal cortex. It is unclear (e.g., time since last use, length of use, amount used) to
whether this is a direct reflection of the effects of chronic enhance external validity and better understand the rela-
drug use on the brain or perhaps better explained by the tionship between these factors and neural responses to
tolerance hypothesis—­that is, that over time, one needs drug cues (Ekhatiari et al., 2022). Further, future research
an increasing amount of a drug to produce the same effect should examine how the nuanced activation differences
(­Peper, ­2004). Additionally, Ferenzci et al. (2016) also found identified here may relate to differences identified in other
a similar pattern of deactivation in the PFC with length of areas of substance use, such as responsiveness to treat-
substance use, though they also demonstrated a subse- ment (­Lichenstein ­et al., ­2021). Lastly, longitudinal studies
quent increase in BOLD activation in the striatum. While may help answer whether and how such neural activation
our findings did not show an increased BOLD response in patterns during cue reactivity precede or succeed prob-
the striatum, this relationship warrants more exploration as lematic substance use.
our lack of findings in this area may be a function of dura- Taken together, our findings show that while there is
tion of use for study participants and our small sample some overlap in neural activation across heroin and
size. This may require comparison across casual versus cocaine use, there are distinct neural patterns uniquely
chronic use and examining developmental samples in associated with each substance. Additionally, these differ-
future studies. Further, larger samples could evaluate both ential activation patterns in people who use heroin may be
length of use and amount of use over time to explore this a function of length of use. Therefore, it is imperative that
relationship further. Additionally, future studies should fur- research continue to evaluate both the convergence and
ther examine the relationship between length of use and divergence of activation patterns across substances, as
heroin as our study may not have incorporated enough well as individual differences in length of use among other
variation in length of cocaine use to fully capture whether variables in order to relate such findings to behavioral dif-
a relationship was meaningful. ferences as well as individualized treatments.

4.4. Limitations DATA AND CODE AVAILABILITY

Limitations of this study include a small sample size due to All text files used in this meta-­analysis are accessible via
limited availability of literature for within-­group cue reactiv- the Open Science Framework at https://osf​.­io​/­gmu5v.
ity studies examining only people who use heroin or
cocaine. The small sample size impacts the generalizabil-
AUTHOR CONTRIBUTIONS
ity of these results to widespread populations of people
who use cocaine and heroin. Another limitation of the cur- J.M.D. (Data Acquisition, Methodology, Writing – o­ riginal
rent study is the lack of or inconsistencies in demographic draft, and Visualization), N.S. (Conceptualization, Data
and clinical data collected across studies, further limiting Curation), D.V.S. (Conceptualization, Methodology,
full consideration of these variables as moderators. Import- Writing – r­eview & editing, and Visualization), A.B.K.
ant co-­variates such as additional/co-­occurring diagno- (Conceptualization, Writing – review & editing, and Visu-
ses, amount of use, frequency of use, and so on need to alization), and D.S.F. (Supervision, Formal Analysis,
be collected across all studies to enhance generalizability Writing – review & editing, and Visualization)
and provide opportunities for answering more complex
and nuanced questions related to drug addiction. Further,
FUNDING
there was variability across studies in the duration of use
and time since recent use, preventing the ability to control This manuscript is supported in part by funding from the
for these factors in the analysis. National Institute on Mental Health (R15MH122927 to
D.S.F.), the National Institute on Drug Abuse (R01DA053282,
R01DA054201 to A.B.K.), and the National Institute on
4.5. Future considerations
Aging (RF1-­AG067011 to D.V.S).
Given the clinical and behavioral findings that social con-
text is related to drug choice, future studies should con-
DECLARATION OF COMPETING INTEREST
sider incorporating a social context into cue reactivity
tasks. This would expand the clinical and behavioral find- The authors hereby declare that there are no competing
ings to include the effects of social context on functional financial or personal interests that could have appeared
activation. Future studies should also consider implement- to influence the work reported in this paper.

13
J.M. Dejoie, N. Senia, A.B. Konova et al. Imaging Neuroscience, Volume 2, 2024

ETHICS STATEMENT (2014). Cognitive reappraisal of emotion: A meta-­analysis

of human neuroimaging studies. Cerebral Cortex, 24(11),
Not applicable due to the meta-­analytic study. 2981–2990. https://doi​.­org​/­10​.­1093​/­cercor​/­bht154
Caprioli, D., Celentano, M., Dubla, A., Luciantonio, F.,
Nencini, P., & Badiani, A. (2009). Ambience and drug
REFERENCES choice: Cocaine- and heroin-taking as a function
Abraham, A., Pedregosa, F., Eickenberg, M., Gervais, P., of environmental context in humans. Biological
Mueller, A., Kossaifi, J., Gramfort, A., Thirion, B., & Psychiatry, 65(10), 893–899. https://doi.org/10.1016/j
Varoquaux, G. (2014). Machine learning for neuroimaging .biopsych.2008.12.009
with scikit-­learn. Fronteirs in Neuroinformatics, 8, 14. Carlezon, W. A., & Thomas, M. J. (2009). Biological
https://doi​.­org​/­10​.­3389​/­fninf​.­2014​.­00014 substrates of reward and aversion: A nucleus
Ahn, W.-­Y., & Vassileva, J. (2016). Machine-­learning accumbens activity hypothesis. Neuropharmacology,
identified substance-­specific behavioral markers 56(Suppl 1), 122–132. https://doi​.­org​/­10​.­1016​/­j​
for opiate and stimulant dependence. Drug Alcohol .­neuropharm​.­2008​.­06​.­075
Dependence, 161, 247–257. https://doi​.­org​/­10​.­1016​/­j​ Carter, B. L., & Tiffany, S. T. (1999). Meta-­analysis of
.­drugalcdep​.­2016​.­02​.­008 cue-­reactivity in addiction research. Addiction, 93(3),
Albajes-­Eizagirre, A., Solanes, A., Fullana, M. A., 327–340. https://doi​.­org​/­10​.­1046​/­j​.­1360​-­0443​.­1999​
Ioannidis, J. P. A., Fusar-­Poli, P., Torrent, C., Solé, B., .­9433273​.­x
Bonnín, C. M., Vieta, E., Mataix-­Cols, D., & Radua, J. Ceceli, A. O., Huang, Y., Krongberg, G., Malaker, P., Miller, P.,
(2019). Meta-­analysis of voxel-­based neuroimaging King, S., Gaudreault, P.-­O., McClain, N., Gabay, L., Vasa,
studies using seed-­based d mapping with permutation D., Ekin, D., Alia-­Klein, N., & Goldstein, R. Z. (2022).
of subject images (SDM-­PSI). Journal of Visualized Common and distinct cortico-­striatal volumetric changes
Experiments (153). https://doi​.­org​/­10​.­3791​/­59841 in cocaine and heroin use disorder. Brain, 146(4),1662–
Albajes-­Eizagirre, A., Solanes, A., & Radua, J. (2019). 1671. https://doi​.­org​/­10​.­1093​/­brain​/­awac366
Meta-­analysis of non-­statistically significant unreported Center for Disease Control and Prevention. (2023). Drug
effects. Statistical Methods in Medical Research, 28(12), overdose. https://www​.­cdc​.­gov​/­drugoverdose​/­deaths​
3741–3754. https://doi​.­org​/­10​.­1177​/­0962280218811349 /­other​-­drugs​.­html
Albajes-­Eizagirre, A., Solanes, A., Vieta, E., & Radua, J. Chang, J.-­Y., Janak, P. H., & Woodward, D. J. (1998).
(2019). Voxel-­based meta-­analysis via permutation of Comparison of mesocorticolimbic neuronal responses
subject images (PSI): Theory and implementation for during cocaine and heroin self-­administration in freely
SDM. Neuroimage, 186, 174–184. https://doi​.­org​/­10​ moving rats. The Journal of Neuroscience, 18(8),
.­1016​/­j​.­neuroimage​.­2018​.­10​.­077 3098–3115. https://doi​.­org​/­10​.­1523​/­JNEUROSCI​.­18​-­08​
Badiani, A., Belin, D., Epstein, D., Calu, D., & Shaham, Y. -­03098​.­1998
(2011). Opiate versus psychostimulant addiction: The Chase, H. W., Eickhoff, S. B., Laird, A. R., & Hogarth, L.
differences do matter. Nature Reviews Neuroscience, (2011). The neural basis of drug stimulus processing
12(11), 685–700. https://doi​.­org​/­10​.­1038​/­nrn3104 and craving: An activation likelihood estimation meta-­
Badiani, A., Caprioli, D., & De Pirro, S. (2019). Opposite analysis. Biological Psychiatry, 70(8), 785–793. https://
environmental gating of the experienced utility (‘liking’) doi​.­org​/­10​.­1016​/­j​.­biopsych​.­2011​.­05​.­025
and decision utility (‘wanting’) or heroin versus cocaine Cooper, S., Robison, A. J., & Mazei-­Robison, M. S. (2017).
in animals and humans: Implications for computational Reward circuitry in addiction. Neurotherapeutics, 14(3),
neuroscience. Psychopharmacology, 236, 2451–2471. 687–697. https://doi​.­org​/­10​.­1007​/­s13311​-­017​-­0525​-­z
https://doi​.­org​/­10​.­1007​/­s00213​-­019​-­05318​-­9 Courtney, K. E., Schacht, J. P., Hutchinson, K.,
Bartra, O., McGuire, J. T., & Kable, J. W. (2013). The Roche, D. J. O., & Ray, L. A. (2016). Neural substrates of
valuation system: A coordinate-­based meta-­analysis of cue-­reactivity: Association with treatment outcomes and
BOLD fMRI experiments examining neural correlates of relapse. Addiction Biology, 21(1), 3–22. https://doi​.­org​/­10​
subjective value. Neuroimage, 76, 412–427. https://doi​ .­1111​/­adb​.­12314
.­org​/­10​.­1016​/­j​.­neuroimage​.­2013​.­02​.­063 Darke, S., Duflou, J., Peacock, A., Chrzanowska, A.,
Bjork, J. M., Keyser-­Marcus, L., Vassileva, J., Ramey, T., Farrel, M., & Lappin, J. (2023). Rates, characteristics
Houghton, D. C., & Moeller, F. G. (2022). Attentional and toxicology of cocaine-­related deaths in Australia,
function and inhibitory control in different substance use 2000–­2021. Addiction, 111(2), 297–306. https://doi​.­org​
disorders. Psychiatry Research, 313, 114591. https://doi​ /­10​.­1111​/­add​.­16055
.­org​/­10​.­1016​/­j​.­psychres​.­2022​.­114591 De Pirro, S., Galati, G., Pizzamiglio, L., & Badiani, A. (2018).
Bornovalova, M. A., Daughters, S. B., Hernandez, G. D., The affective and neural correlates of heroin vs cocaine
Richards, J. B., & Lejuez, C. W. (2005). Impulsivity and use in addiction are influenced by environmental setting
risk-­taking propensity between primary users of crack but in opposite directions. The Journal of Neuroscience,
cocaine and primary users of heroin in a residential 38(22), 5182–5195. https://doi​.­org​/­10​.­1523​/­jneurosci​
substance-­use program. Experimental and Clinical .­0019​-­18​.­2018
Psychopharmacology, 13(4), 311–318. https://doi​.­org​/­10​ Degenhardt, L., Whiteford, H. A., Baxter, A. J., Charlson, F. J.,
.­1037​/­1064​-­1297​.­13​.­4​.­311 Hall, W. D., Freedman, G., Burstein, R., Johns, N., Engell,
Brevers, D., Baeken, C., De Smet, S., Catoira, B., R. E., Flaxman, A., Murray, C. L., & Vos, T. (2013). Global
De Witte, S., He, Q., Maurage, P., Schulze-­Steinen, L., burden of disease attributable to illicit drug use and
Sescousse, G., Vila Verde, C., Vögele, C., & Billieux, J. dependence: Findings from the global burden of disease
(2023). Stimulation of the dorsolateral prefrontal cortex study 2010. Lancet, 382(9904), 1564–1574. https://doi​.­org​
modulates brain cue reactivity ot reward (un)availability. /­10​.­1016​/­s0140​-­6736(13)61530​-­5
Cortex, 164, 51–62. https://doi​.­org​/­10​.­1016​/­j​.­cortex​.­2023​ Devoto, F., Zapparoli, L., Spinelli, G., Scotti, G., &
.­03​.­008 Paulesu, E. (2020). How the harm of drugs and their
Buhle, J. T., Silvers, J. A., Wagner, T. D., Lopez, R., availability affect brain reactions to drug cues: A
Onyemekwu, C., Kober, H., Weber, J., & Ochsner, K. N. meta-­analysis of 64 neuroimaging activation studies.

14
J.M. Dejoie, N. Senia, A.B. Konova et al. Imaging Neuroscience, Volume 2, 2024

Translational Psychiatry, 10(1), 429. https://doi​.­org​/­10​ Volcow, N. A. (2009). Dopaminergic response to

.­1038​/­s41398​-­020​-­01115​-­7 drug words in cocaine addiction. The Journal of
Drummond, D. C. (2000). What does cue-­reactivity Neuroscience, 29(18), 6001–6001. https://doi​.­org​/­10​
have to offer clinical research? Addiction, 95 .­1523​/­jneurosci​.­4247​-­08​.­2009
(Supplement 2), S129–S144. https://doi​.­org​/­10​.­1080​ Goldstein, R. Z., & Volkow, N. D. (2002). Drug addiction
/­09652140050111708 and its underlying neurobiological basis: Neuroimaging
Duncan, E., Boshoven, W., Harenski, K., Fiallos, A., Tracy, evidence for the involvement of the frontal cortex.
H., Jovanovic, T., Hu, X., Drexler, D., & Kilts, C. (2007). An American Journal of Psychiatry, 159(10), 1642–1652.
fMRI study of the interaction of stress and cocaine cues https://doi​.­org​/­10​.­1176​/­appi​.­ajp​.­159​.­10​.­1642
on cocaine craving on cocaine-­dependent men. The Goldstein, R. Z., & Volkow, N. D. (2012). Dysfunction of the
American Journal on Addictions, 16, 174–182. https://doi​ prefrontal cortex in addiction: Neuroimaging findings
.­org​/­10​.­1080​/­10550490701375285 and clinical implications. Nature Reviews Neuroscience,
Ekhtiari, H., Zare-­Bidoky, M., Sangchooli, A., Janes, A. C., 12(11), 652–669. https://doi​.­org​/­10​.­1038​/­nrn3119
Kaufman, M. J., Oliver, J. A., Prisciandaro, J. J., Haber, S. N., & Knuston, B. N. (2010). The reward
Wüstenberg, T., Anton, R. F., Bach, P., Baldacchino, A., circuit: Linking primate anatomy and human imaging.
Beck, A., Bjork, J. M., Brewer, J., Childress, A. R., Neuropsychopharmacology, 35, 4–26. https://doi​.­org​/­10​
Claus, E. D., Courtney, K. E., Ebrahimi, M., Filbey, .­1038​/­npp​.­2009​.­129
F. M., Ghahremani, D. G., … Zilverstand, A. (2022). A Hanlon, C. A., Dowdle, L. T., Gibson, N. B., Li, X.,
methodological checklist for fMRI drug cue reactivity Hamilton, S., Centerberry, M., & Hoffman, M. (2018).
studies: Development and expert consensus. Nature Cortical substrates of cue-­reactivity in multiple substance
Protocols, 17(3), 567–595. https://doi​.­org​/­10​.­1038​ dependent populations: Transdiagnostic relevance of the
/­s41596​-­021​-­00649​-­4 medial prefrontal cortex. Translational Psychiatry, 8, 186.
Elton, A., Smitherman, S., Young, J., & Kilts, C. D. (2015). https://doi​.­org​/­10​.­1038​/­s41398​-­018​-­0220​-­9
Effects of childhood maltreatment on the neural Hassani-­Abharian, P., Ganjgahi, H., Tabatabaei-­Jafari, H.,
correlates of stress-­and drug cue-­induced cocaine Oghabian, M. A., Mokri, A., & Ekhtiari, H. (2015).
craving. Addiction Biology, 20(4), 820–831. https://doi​ Exploring neural correlates of different dimensions in
.­org​/­10​.­1111​/­adb​.­12162 drug craving self-­reports among heroin dependents.
Englemann, J. M., Versace, F., Robinson, J. D., Minnix, J. A., Basic and Clinical Neuroscience, 6(4), 271–284.
Lam C. Y., Brown, V. L., & Cincirpini, P. M. (2012). Neural PMID: 26649165; PMCID: PMC4668874.
substrates of smoking cue reactivity: A meta-­analysis of Hayashi, T., Ko, J. H., Strafella, A. P., & Dagher, A.
fMRI Studies. Neuroimage, 61(1), 252–262. https://doi​.­org​ (2013). Dorsolateral prefrontal and orbitofrontal cortex
/­10​.­1016​/­j​.­neuroimage​.­2011​.­12​.­024 interactions during self-­control of cigarette craving.
Ettenberg, A., Pettit, H. O., Bloom, F. E., & Koob, G. F. Proceedings of the National Academy of Sciences of the
(1982). Heroin and cocaine intravenous self-­administration United States of America, 110(11), 4422–4427. https://
in rats: Mediation by separate neural systems. doi​.­org​/­10​.­1073​/­pnas​.­1212185110
Psychopharmacology, 78, 204–209. https://doi​.­org​/­10​ Hedegaard, H., Minińo, A. M., & Warner, M. (2020). National
.­1007​/­bf00428151 center for health statistics data (NCHS brief no. 394).
Everitt, B. J., & Robbins, T. W. (2005). Neural systems of Centers for Disease Control. https://www​.­cdc​.­gov​/­nchs​
reinforcement for drug addiction: From actions to habits /­products​/­index​.­htm.
to compulsion. Nature Neuroscience, 8(11), Article 11. Higgins, J. P. T., & Thompson, S. G. (2002). Quantifying
https://doi​.­org​/­10​.­1038​/­nn1579 heterogeneity in a meta-analysis. Stat Med, 21(11),
Everitt, B. J., & Robbins, T. W. (2016). Drug addiction: 1539–1538. https://doi.­org/10.1002/sim.1186
Updating actions to habits to compulsion ten years on. Hill-­Bowen, D., Riedel, M. C., Salo, T., Flannery, J. S., Poudel,
Annual Review of Psychology, 67, 23−50. https://doi R., Laird, A. R., & Sutherland, M. T. (2022). Convergent gray
.org/10.1146/annurev-psych-122414-033457 matter alterations across drugs of abuse and network-­level
Ferenczi, E. A., Zalocusky, K. A., Liston, C., Grosenick, L., implications: A meta-­analysis of structural MRI studies.
Warden, M. R., Amatya, D., Katovich, K., Mehta, H., Drug and Alcohol Dependence, 240, 109625. https://doi​
Patenaude, B., Ramakrishnan, C., Kalanithi, P., Etkin, .­org​/­10​.­1016​/­j​.­drugalcdep​.­2022​.­109625
A., Knuston, B., Glover, G. H., & Deisseroth, K. (2016). Hill-­Bowen, L. D., Riedel, M. C., Poudel, R., Salo, T.,
Prefrontal cortical regulation of brainwide circuit Flannery, J. S., Camillerio, J. A., Eickhoff, S. B.,
dynamics and reward-related behavior. Science, 351, Laird, A. R., & Sutherland, M. T. (2021). The cue-­reactivity
asc9698. https://doi.org/10.1126.science.aac9698 paradigm: An ensemble of networks driving attention and
Gaudreault, P.-­O., King, S. G., Malaker, P., Alia-­Klein, N., cognition when viewing drug and natural reward-­related
& Goldstein, R. Z. (2022). Whole-­brain white matter stimuli. Neuroscience & Biobehavioral Reviews, 130,
abnormalities in human cocaine and heroin use 201–203. https://doi​.­org​/­10​.­1016​/­j​.­neubiorev​.­2021​.­08​.­010
disorders: Association with craving, recency, and Hobkirk, A. L., Bell, R. P., Utevsky, A. V., Huettel, S., &
cumulative use. Molecular Psychiatry, 28(2), 780–791. Meade, C. S. (2019). Reward and executive control
https://doi​.­org​/­10​.­1038​/­s41380​-­022​-­01833​-­y network resting-­state functional connectivity is
George, O., & Koob, G. F. (2013). Control of craving by the associated with impulsivity during reward-­based
prefrontal cortex. Proceedings of the National Academy decision making for cocaine users. Drug and Alcohol
of Sciences of the United States of America, 110(11), Dependence, 194(1), 32–39. https://doi​.­org​/­10​.­1016​/­j​
4165–4166. https://doi​.­org​/­10​.­1073​/­pnas​.­1301245110 .­drugalcdep​.­2018​.­09​.­013
Gheraldini, C., Di Cesare Mannelli, L., & Bianchi, E. (2015). Hochheimer, M., Strickland, J. C., Rabinowitz, J. A.,
The pharmacological basis of opioids. Clinical Cases in Ellis, J. D., Bergeria, C. L., Hobelmann, J. G., & Huhn,
Mineral and Bone Metabolism, 12(3), 219−221. https:// A. S. (2023). The impact of opioid-­stimulant co-­use on
doi.org/10.11138/ccmbm/2015.12.3.219 tonic and cue-­induced craving. Journal of Psychiatric
Goldstein, R. Z., Tomasi, D., Alia-­Klein, N., Carrillo, J. H., Research, 164, 15–22. https://doi​.­org​/­10​.­1016​/­j​
Maloney, T., Woicik, P. A., Wang, R., Telang, F., & .­jpsychires​.­2023​.­05​.­018

15
J.M. Dejoie, N. Senia, A.B. Konova et al. Imaging Neuroscience, Volume 2, 2024

Hser, Y.-­I., Evans, E., Huang, D., Brecht, M.-­L., & Li, Neuroscience, 33(7), 1318–1326. https://doi.org
L. (2008). Comparing the dynamic course of heroin, /10.1111/j.1460-9568.2010.07590.x
cocaine, and methamphetamine use over 10 years. Lejuez, C. W., Bornovalova, M. A., Daughters, S. B., &
Addictive Behaviors, 33, 1581–1589. https://doi​.­org​/­10​ Curtin, J. J. (2005). Differences in impulsivity and sexual
.­1016​/­j​.­addbeh​.­2008​.­07​.­024 risk behavior among inner-­city crack/cocaine users
Hser, Y.-­I., Huang, D., Brecht, M.-­L., Li, L., & Evans, E. and heroin users. Drug and Alcohol Dependence, 77,
(2008). Contrasting trajectories of heroin, cocaine, and 169–175. https://doi​.­org​/­10​.­1016​/­j​.­drugalcdep​.­2004​.­08​
methamphetamine use. Journal of Addictive Diseases, .­013
27(3), 13. https://doi​.­org​/­10​.­1080​/­10550880802122554 Lewer, D., Brothers, T. D., Van Hest, N., & Hickman, M.
Huang, A. S., Mitchell, J. A., Haber, S. N., Alia-­Klein, N., & (2022). Causes of death among people who used illicit
Goldstein, R. Z. (2018). The thalamus in drug addiction: opioids in england, 2001-­18: A matched cohort study.
From rodents to humans. Philosophical Transactions The Lancet, 7(2), 126–135. https://doi​.­org​/­10​.­1016​/­s2468​
of the Royal Society B: Biological Sciences, 373(1742), -­2667(21)00254​-­1
20170028. https://doi​.­org​/­10​.­1098​/­rstb​.­2017​.­0028 Li, Q., Wang, Y., Zhang, Y., Li, W., Yang, W., Zhu, J.,
Janes, A. C., Betts, J., Jenson, J. E., & Lukas, S. E. (2016). Wu, N., Chang, H., Zheng, Y., Qin, W., Zhao, L., Yuan, K.,
Dorsal anterior cingulate glutamate is associated Liu, J., Wang, W., & Tian, J. (2012). Craving correlates
with engagement of the default mode network with mesolimbic responses to heroin-­related cues in
during exposure to smoking cues. Drug and Alcohol short-­term abstinence from heroin: An event-­related fMRI
Dependence, 167(1), 75–81. https://doi​.­org​/­10​.­1016​/­j​ study. Brain Research, 1469, 63–72. https://doi​.­org​/­10​
.­drugalcdep​.­2016​.­07​.­021 .­1016​/­j​.­brainres​.­2012​.­06​.­024
Jentsch, J., & Taylor, J. R. (1999). Impulsivity resulting from Li, Q., Wang, Y., Zhang, Y., Li, W., Zhu, J., Zheng, Y.,
frontostriatal dysfunction in drug abuse: Implications Chen, J., Zhao, L., Zhou, Z., Liu, Y., Wang, W., &
for the control of behavior by reward-­related stimuli. Tian, J. (2013). Assessing cue-­induced brain response
Psychopharmacology, 146, 373–390. https://doi​.­org​/­10​ as a function of abstinence duration in heroin-­dependent
.­1007​/­PL00005483 individuals: An event-­related fMRI study. PLoS One, 8(5),
Jousta, J., Moussawi, K., Siddiqi, S. H., Abdolhai, A., e62911. https://doi​.­org​/­10​.­1371​/­journal​.­pone​.­0062911
Drew, W., Cohen, A. L., Ross, T. J., Desgpande, H. U., Li, Y., Yuan, K., Bi, Y., Guan, Y., Cheng, J., Zhang, Y., Shi, S.,
Wang, H. Z., Bruss, J., Stein, E. A., Volkow, N. D., Lu, X., Yu, D., & Tian, J. (2017). Neural correlates of 12-h
Grafman, J. H., van Wijngaarden, E., Boes, A. D., & abstinence-induced craving in young adult smokers:
Fox, M. D. (2022). Brain lesions disrupting addiction map A resting-state study. Brain Imaging and Behavior, 11,
to a common human brain circuit. Nature Medicine, 28, 677–684. https://doi.org/10.1007/s11682-016-9544-3
1249–1255. https://doi​.­org​/­10​.­1038​/­s41591​-­022​-­01834​-­y Lichenstein, S. D., Scheinost, D., Potenza, M. N.,
Kaag, A. M., Reneman, L., Homberg, J., van den Brink, W., Carroll, K. M., & Yip, S. W. (2021). Dissociable neural
& van Wingen, G. A. (2018). Enhanced amygdala-­striatal substrates of opioid and cocaine use identified via
functional connectivity during the processing of cocaine connectome-­based modelling. Molecular Psychiatry,
cues in male cocaine users with a history of childhood 26(8), 4383–4393. https://doi​.­org​/­10​.­1038​/­S41380​-­019​
trauma. Frontiers in Psychiatry, 9, 70. https://doi​.­org​/­10​ -­0586​-­Y
.­3389​/­fpsyt​.­2018​.­00070 Long, Y., Pan, N., Ji, S., Chen, Y., Zhang, X., He, M.,
Liu, S., Wang, S., Zhang, M., Xu, Y., Liu, J., & Yuan, K. Suo, X., Yu, Y., Wang, S., & Gong, Q. (2022). Distinct
(2021). Brain responses to drug cues predict craving brain structure abnormalities in attention-­deficit/
changes in abstinent heroin users: A preliminary study. hyperactivity disorder and substance use disorder: A
Neuroimage, 237, 118169. https://doi.org/10.1016/j comparative meta-­analysis. Translational Psychiatry, 12,
.neuroimage.2021.118169 368. https://doi​.­org​/­10​.­1038​/­s41398​-­022​-­02130​-­6
Klugah-­Brown, B., Di, X., Zweerings, J., Mathiak, K., Mei, W., Zhang, J. X., & Xiao, Z. (2010). Acute effects
Becker, B., & Biswal, B. (2020). Common and of sublingual buprenorphine on brain responses to
seperable neural alterations in substance use disorder: heroin-­related cues in early-­abstinent heroin addicts: An
A coordinate-­based meta-­analysis of functional uncontrolled trial. Neuroscience, 170, 808–815. https://
neuroimaging studies in humans. Human Brain Mapping, doi​.­org​/­10​.­1016​/­j​.­neuroscience​.­2010​.­07​.­033
41(16), 4459–4477. https://doi​.­org​/­10​.­1002​/­hbm​.­25085 Mitchell, M. R., & Potenza, M. N. (2014). Addictions and
Klugah-­Brown, B., Jiang, C., Agoalikum, E., Zhou, X., personality traits: Impulsivity and related constructs.
Zou, L., Yu, Q., Becker, B., & Biswal, B. (2021). Common Current Behavioral Neuroscience Reports, 1, 1–12.
abnormality of gray matter integrity in substance https://doi​.­org​/­10​.­1007​/­s40473​-­013​-­0001​-­y
use disorder and obsessive-­compulsive disorder: A Moeller, F. G., Steinberg, J. L., Schmitz, J. M., Ma, L.,
comparative voxel-­based meta-­analysis. Human Brain Liu, S., Kjome, K. L., Rathnayaka, N., Kramer, L. A., &
Mapping, 42, 3871–3886. https://doi​.­org​/­10​.­1002​/­hbm​ Narayana, P. A. (2010). Working memory fMRI activation
.­25471 in cocaine-­dependent subjects: Association with
Kober, H., Mende-­Siedlecki, P., Kross, E. F., & Oschner, K .N. treatment response. Psychiatry Research: Neuroimaging,
(2010). Prefrontal-­striatal pathway underlies cognitive 18, 174–182. https://doi​.­org​/­10​.­1016​/­j​.­pscychresns​.­2009​
regulation of craving. Proceedings of the National .­11​.­003
Academy of Sciences of the United States of America, Motzkin, J. C., Baskin-­Sommers, A., Newman, J. P.,
107(33), 14811–14816. https://doi​.­org​/­10​.­1073​/­pnas​ Kiehl, K. A., & Koenigs, M. (2014). Neural correlates
.­1007779107 of substance abuse: Reduced functional connectivity
Koob, G. F., & Volkow, N. D. (2016). Neurobiology of addiction: between areas underlying reward and cognitive control.
A neurocircuitry analysis. Lancet Psychology, 3(8), 760– Human Brain Mapping, 35(9), 4282–4292. https://doi​.­org​
773. https://doi.org/10.1016/S2215-0366(16)00104-8 /­10​.­1002​/­hbm​.­22474
Kuhn, S., & Gallinat, J. (2011). Common biology of craving National Institute on Drug Abuse. (2022). Overdose Death
across legal and illegal drugs − a quantitative meta- Rates. NIDA.https://nida​.­nih​.­gov​/­research​-­topics​/­trends​
analysis of cue-reactivity brain response. Behavioral -­statistics​/­overdose​-­death​-­rates

16
J.M. Dejoie, N. Senia, A.B. Konova et al. Imaging Neuroscience, Volume 2, 2024

Oakes, T. R., Fox, A. S., Johnstone, T., Chung, M. K., 132(0), 195–201. https://doi​.­org​/­10​.­1016​/­j​.­drugalcdep​
Kalin, N., & Davidson, R. J. (2007). Integrating VBM .­2013​.­02​.­009
into the general linear model with voxelwise anatomical Psederska, E., & Vassileva, J. (2023). Neurocognitive
covariates. Neuroimage, 34(2), 500–508. https://doi​.­org​ impulsivity in opiate users at difference lengths of
/­10​.­1016​/­j​.­neuroimage​.­2006​.­10​.­007 abstinence. International Journal of Environmental
Padmanabhan, J. L., Cooke, D., Joutsa, J., Siddiqi, S. H., Research and Public Health, 20, 1236. https://doi​.­org​/­10​
Ferguson, M., Darby, R. R., Soussand, L., Horn, A., Kim, .­3390​/­ijerph20021236
N. Y., Voss, J. L., Naidech, A. M., Brodtmann, A., Egorova, Sartor, C. E., Kranzler, H. R., & Gelernter, J. (2014).
N., Gozzi, S., Phan, T. G., Corbetta, M., Grafman, J., & Characteristics and course of dependence in cocaine-­
Fox, M. D. (2019). A human depression circuit derived dependent individuals who never used alcohol or
from focal brain lesions. Biological Psychiatry, 86(10), marijuana or used cocaine first. Journal of Studies on
749–758. https://doi.org/10.1016/j.biopsych.2019.07.023 Alcohol and Drugs, 75(3), 423–427. https://doi​.­org​/­10​
Pando-­Nuade, V., Toxto, S., Fernandez-­Lozano, S., .­15288​/­jsad​.­2014​.­75​.­423
Parsons, C. E., Alcauter, S., & Garza-­Villarreal, E. A. Schneider, S., Richeter, C., Niethammer, R., & Beisel, L.
(2021). Gray and white matter morphology in substance (2021). Fatal and non-­fatal heroin-­related overdose
use disorders: A neuroimaging systematic review and circumstances and patterns. Substance Use & Misuse,
meta-­analysis. Translational Psychiatry, 11, 11–29. 56(13), 1997–2006. https://doi​.­org​/­10​.­1080​/­10826084​
https://doi​.­org​/­10​.­1038​/­s41398​-­020​-­01128​-­2 .­2021​.­1963986
Patel, K. T., Stevens, M. C., Meda, S. A., Muska, C., Thomas, Seifert, C. L., Magon, S., Sprenger, T., Lang, U. E., Huger,
A. D., Potenza, M. N., & Pearlson, G. D. (2013). Robust C. G., Denier, N., Vogel, M., Schmidt, A., Raude, E.-­W.,
changes in reward circuitry during reward loss in current Borgwardt, S., & Walter, M. (2015). Reduced volume of
and former cocaine users during performance of a the nucleus accumbens in heroin addiction. European
monetary incentive delay task. Biological Psychiatry, 74(1), Archives of Psychiatry and Clinical Neuroscience, 265,
529–537. https://doi​.­org​/­10​.­1016​/­j​.­biopsych​.­2013​.­04​.­029 637–645. https://doi​.­org​/­10​.­1007​/­s00406​-­014​-­0564​-­y
Peacock, A., Tran, L. T., Larney, S., Stockings, E., Santo , Jr., Sell, L. A., Morris, J., Bearn, J., Frackowiak, R. S. J.,
T., Jones, H., Santomauro, D., & Degenhardt, L. (2021). Friston, K. J., & Dolan, R. J. (2001). Activation of reward
All-­cause and cause-­specific mortality among people with circuitry in human opiate addicts. European Journal of
regular or problematic cocaine use: A systematic review Neuroscience, 11(3), 1042–1048. https://doi​.­org​/­10​.­1046​
and meta-­analysis. Addiction, 116, 725–742. https://doi​ /­j​.­1460​-­9568​.­1999​.­00522​.­x
.­org​/­10​.­1111​/­add​.­15239 Shand, F. L., Degenhardt, L., Slade, T., & Nelson, E. C.
Peper, A. (2004). A theory of drug tolerance and (2011). Sex differences amongst dependent heroin
dependence I: A conceptual analysis. Journal of users: Histories, clinical characteristics and predictors
Theoretical Biology, 229(4), 477–490. https://doi​.­org​/­10​ of other substance dependence. Addictive Behaviors,
.­1016​/­j​.­jtbi​.­2004​.­04​.­010 36(1–­2), 27–36. https://doi​.­org​/­10​.­1016​/­j​.­addbeh​.­2010​
Perdue, T., Carlson, R., Daniulaityte, R., Silverstein, S. M., .­08​.­008
Bluthenthal, R. N., Valdez, A., & Cepeda, A. (2024). Stubbs, J. L., Taylor, J. J., Siddiqi, S. H., Schapner,
Characterizing prescription opiod, heroin, and fentanyl L. W. V. J. F., Cohen, A. L., Drew, W., Hanlon, C. A.,
initiation trajectories: A qualitative study. Social Science Abdolhai, A., Wang, H. Z., Honer, W. G., Panenka, W. J.,
& Medicine, 340, 116441. https://doi​.­org​/­10​.­1016​/­j​ & Fox, M. D. (2023). Hetergenous neuroimaging findings
.­socscimed​.­2023​.­116441 across substance use disorder localize to a common
Pettit, H. O., Ettenberg, A., Bloom, F. E., & Koob, G. F. brain network. Nature Mental Health, 1, 772–781. https://
(1984). Destruction of dopamine in the nucleus doi​.­org​/­10​.­1038​/­s44220​-­023​-­00128​-­7
accumbens selectively attenuates cocaine but not Vafaie, N., & Kober, H. (2022). Association of drug cues and
self-­administration in rats. Psychopharmacology, 84, craving with drug use relapse: A systemic review and
167–173. https://doi​.­org​/­10​.­1007​/­bf00427441 meta-­analysis. JAMA Psychiatry, 79(7), 641–650. https://
Pollard, A. A., Hauson, A. O., Lackey, N. S., Zhang, E., doi​.­org​/­10​.­1001​/­jamapsychiatry​.­2022​.­1240
Khayat, S., Carson, B., Fortea, L., Radua, J., & Grant, Vassileva, J., Paxton, J., Moeller, F. G., Wilson, M. J.,
I. (2023). Functional neuroanatomy of craving in heroin Bozgunov, K., Martin, E. M., Gonzalez, R., & Vasilev,
use disorder: Voxel-­based meta-­analysis of functional G. (2014). Heroin and amphetamine users display
magnetic resonance imaging (fMRI) drug cue reactivity opposite relationships between trait and neurobehavioral
studies. Am J Drug Alcohol Abuse, 49(4), 418–430. dimensions of impulsivity. Addictive Behaviors, 39,
https://doi​.­org​/­10​.­1080​/­00952990​.­2023​.­2172423 652–659. https://doi​.­org​/­10​.­1016​/­j​.­addbeh​.­2013​.­11​.­020
Prisciandaro, J. J., McRae-­Clark, A. L., Myrick, H., Volkow, N. D., Wang, G.-­J., Fowler, J. S., Tomasi, D.,
Henderson, S., & Brady, K. T. (2014). Brain activation to Telang, F., & Baler, R. (2010). Addiction: Decreased
cocaine cues and motivation/treatment status. Addiction reward sensitivity and increased expectation sensitivity
Biology, 19(2), 240–249. https://doi​.­org​/­10​.­1111​/­j​.­1369​ conspire to overwhelm the brain’s control circuit.
-­1600​.­2012​.­00446​.­x Bioessays, 32(9), 748–755. https://doi​.­org​/­10​.­1002​/­bies​
Prisciandaro, J. J., Myrick, H., Henderson, S., .­201000042
McRae-­Clark, A. L., & Brady, K. T. (2013). Prospective Walter, M., Denier, N., Gerber, H., Schmid, O., Lanz, C.,
associations between brain activation to cocaine and Brenneisen, R., Riecher-­Rössler, A., Wiesbeck, G. A.,
no-­go cues and cocaine relapse. Drug and Alcohol Scheffler, K., Seifritz, E., McGuire, P., Fusar-­Poli, P., &
Dependence, 131(0), 44–49. https://doi​.­org​/­10​.­1016​/­j​ Borgwardt, S. (2015). Orbitofrontal response to drug-­
.­drugalcdep​.­2013​.­04​.­008 related stimuli after heroin administration. Addiction
Prisciandaro, J. J., Myrick, H., Henderson, S., Biology, 20(3), 570–579. https://doi​.­org​/­10​.­1111​/­adb​.­12145
McRae-­Clark, A. L., Santa Ana, E. J., Saladin, M. E., Wang, A.-­L., Elman, I., Lowen, S. B., Blady, S. J., Lynch,
& Brady, K. T. (2013). Impact of DCS-­facilitated cue K. G., Hyatt, J. M., O’Brien, C. P., & Langleben, D. D.
exposure therapy on brain activation to cocaine cues in (2015). Neural correlates of adherence to extended-­
cocaine dependence. Drug and Alcohol Dependence, release naltrexone pharmacotherapy in heroin

17
J.M. Dejoie, N. Senia, A.B. Konova et al. Imaging Neuroscience, Volume 2, 2024

dependence. Translational Psychiatry, 5, e531. https://doi​ 157–162. https://doi​.­org​/­10​.­1016​/­j​.­drugalcdep​.­2005​.­11​

.­org​/­10​.­1038​/­tp​.­2015​.­20 .­012
Wang, K. S., Haiser, R. H., Peechatka, A. L., Frederick, Yan, H., Xiao, S., Fu, S., Gong, J., Qi, Z., Chen, G.,
B. B., & Janes, A. C. (2020). Temporal dynamics of Chen, P., Tang, G., Su, T., Yang, Z., & Wang, Y. (2023).
large-­scale networks predict neural cue reactivity and Functional and structural brain abnormalities in
cue-­induced craving. Biological Psychiatry: Cognitive substance use disorder: A multimodal meta-­analysis of
Neuroscience and Neuroimaging, 5(11), 1011–1018. neuroimaging studies. Acta Psychiatrica Scandinavia,
https://doi​.­org​/­10​.­1016​/­j​.­bpsc​.­2020​.­07​.­006 147, 345–359. https://doi​.­org​/­10​.­1111​/­acps​.­13539
Wang, K. S., Smith, D. V., & Delgado, M. R. (2016). Using Yang, Z., Xie, J., Shao, Y.-­C., Xie, C.-­M., Fu, L.-­P., Li,
fMRI to study reward processing in humans: Past, D.-­J., Fan, M., Ma, L., & Li, S.-­J. (2009). Dynamic
present, and future. Journal of Neurophysiology, 115, neural responses to cue-­reactivity paradigms in heroin-­
1664–1678. https://doi​.­org​/­10​.­1152​/­jn​.­00333​.­2015 dependent users: A fMRI study. Human Brain Mapping,
Wang, W., Li, Q., Wang, Y., Tian, J., Yang, W., Li, W., 30, 766–775. https://doi​.­org​/­10​.­1002​/­hbm​.­20542
Qin, W., Yuan, K., & Liu, J. (2011). Brain fMRI and Zhang, S., Zhornitsky, S., Wang, W., Dhingra, I., Le, T. M., &
craving response to heroin-­related cues in patients on Li, C.-­S.R. (2020). Cue-­elicited functional connectivity of
methadone maintenance. The American Journal of Drug the periaqueductal gray and tonic cocaine craving. Drug
and Alcohol Abuse, 37(2), 123–130. https://doi​.­org​/­10​ and Alcohol Dependence, 216, 108240. https://doi​.­org​
.­3109​/­00952990​.­2010​.­543997 /­10​.­1016​/­j​.­drugalcdep​.­2020​.­108240
Wei, C., Han, X., Weng, D., Feng, Q., Qi, X., Li, J., & Luo, Zijlstra, F., Veltman, D. J., Booij, J., van den Brink, W., &
M. (2018). Response dynamics of midbrain dopamine Franken, I. H. A. (2009). Neurobiological substrates of
neurons and serotonin neurons to heroin, nicotine, cue-­elicited craving and anhedonia in recently abstinent
cocaine, and MDMA. Cell Discovery, 4, 60. https://doi​ opioid-­dependent males. Drug and Alcohol Dependence,
.­org​/­10​.­1038​/­s41421​-­018​-­0060​-­z 99, 183–192. https://doi​.­org​/­10​.­1016​/­j​.­drugalcdep​.­2008​
Wiers, C. E., Gawron, C. K., Gröpper, S., Spengler, S., .­07​.­012
Stuke, H., Lindenmeyer, J., Walter, H., & Bermpohl, F. Zilberman, N., Lavidor, M., Yadid, G., & Rassovsky, Y.
(2015). Decreased gray matter volume in inferior frontal (2019). Qualitative review and quantitative effect size
gyrus is related to stop-­signal task performance in meta-analyses in brain regions identified by cue-
alcohol-­depdendent patients. Psychiatry Research: reactivity addiction studies. Neuropsychology, 33(3),
Neuroimaging, 233(2), 125–130. https://doi​.­org​/­10​.­1016​/­j​ 319–334. https://doi.org/10.1037/neu0000526
.­pscychresns​.­2015​.­05​.­006 Zilverstand, A., Huang, A. S., Alia-­Klein, N., & Goldstein,
Xiao, Z., Lee, T., Zhang, J. X., Wu, Q., Wu, R., Weng, X., R. Z. (2018). Neuroimaging impaired response inhibition
& Hu, X. (2006). Thirsty heroin addicts show different and salience attribution in human drug addiction: A
fMRI activations when exposed to water-­related and systematic review. Neuron, 98(5), 886–903. https://doi​
drug-­related cues. Drug and Alcohol Dependence, 83, .­org​/­10​.­1016​/­j​.­neuron​.­2018​.­03​.­048

18