Continuous theta burst stimulation of the medial prefrontal cortex reduces drug cue reactivity with a potential for improving outpatient treatment outcomes in cocaine use disorder

Over the past decade, the US has been affected by a re-emerging stimulant use public health crisis with 1.4 million people meeting criteria for crack/cocaine use disorders (CUD). Together, crack/cocaine and other stimulants were involved in tens of thousands of overdose deaths, rising at striking rates (>300% from 2015). In addition to the devastating death toll, crack/cocaine use and CUD cause escalating mental, physical and social health consequences, with high relapse rates following treatment (as high as 75% two months post-treatment). Yet, there are no FDA approved pharmacological treatments for CUD, together showcasing the urgent need for the development and testing of evidence-based treatments to support recovery in this population.

Several models of addiction posit that recovery would be supported by enhancing control over craving, the intense subjective urge to use drugs, a criterion in the diagnosis of substance use disorder (SUD) and a predictor of clinical outcomes such as relapse. An important correlate is drug cue-reactivity whereby drug-associated cues elicit strong conditioned responses, inclusive of increased activation of the prefrontal cortex (PFC) and limbic subcortical regions (e.g., the striatum), as documented by decades of preclinical and human studies. Similarly to craving, this automatic hyper-reactivity to drug cues contributes to drug use and compulsive drug-seeking and is predictive of the development of disease and of relapse after treatment. For example, in a meta-analysis of neuroimaging human SUD studies, one unit increase in cue-reactivity was associated with more than double the odds of future drug use/relapse (1). Importantly, altering this neural pattern is achievable both with abstinence and by therapeutic interventions (2).

Repeated transcranial magnetic stimulation (rTMS), a noninvasive tool that leverages focal electromagnetic pulses to induce neural plasticity in preclinical models and selectively modulate cortical excitability in humans, has been recognized by the FDA for depression, obsessive compulsive disorder and smoking cessation treatment. Evidence from the last decade has also suggested that rTMS of the PFC reduces craving and drug intake in people with CUD. Instead of targeting the dorsolateral PFC, the main target in the prior rTMS studies in CUD, here the authors relied on convergent evidence that suggested that the medial PFC (mPFC) is a viable, transdiagnostically relevant, target for reducing drug cue reactivity in addiction. Data from their own lab showed that single TMS pulses to the left frontal pole induced BOLD fMRI alterations in the mPFC and striatum in healthy individuals and those with cocaine and alcohol dependence. Of most relevance, the authors demonstrated that continuous theta burst stimulation (cTBS), a patterned form of rTMS, to the frontal pole decreased brain reactivity to alcohol cues and alcohol drinking, increasing retention in an intensive outpatient treatment program, effects seen for up to three months in 50 individuals with alcohol use disorder. Building on this prior work, here this team describes results of a similar protocol in CUD, the first clinical trial evaluating the frontal pole of the mPFC as a potential TMS target for this population.

In the current randomized, sham-controlled, double-blind trial, the authors first reported robustness of the sham control and tolerability and feasibility of 10 sessions of cTBS targeting the left frontal pole of the mPFC in 33 treatment-engaged outpatients with CUD (3). They then reported lower fMRI drug cue reactivity in the mPFC (inclusive of the superior and middle orbitofrontal cortex) with real cTBS at the 2^nd and 3^rd months after intervention completion (as compared to a pre-treatment baseline and the 1^st month after treatment). A treatment by time interaction was significant in the anterior cingulate cortex and insula, where decreased drug cue reactivity was most pronounced three months after treatment. At reduced thresholds, whole brain analyses were promising (suggesting reduced orbitofrontal cortical cue reactivity and relative attenuation of functional connectivity between the stimulation site and striato-limbic regions in the real vs. sham cTBS condition). In post-hoc analyses, baseline cocaine craving was a significant covariate, such that greater craving scores were associated with a greater decrease in cocaine cue reactivity. Importantly, although not reaching statistical significance, a greater portion of individuals who received real cTBS remained abstinent relative to sham (1 month: 92.0%; 2 month: 100.0%; 3 month: 85.0% vs. 66.6% at these same time intervals in the sham condition with no group differences in study retention). This difference was notable because these outpatients also received program-sponsored housing that was contingent on total cocaine abstinence; in the context of this already-effective technique, cTBS conferred additional benefit to treatment outcomes.

Taken together, results suggest that TMS is an effective tool that can be combined with housing-based recovery programs in CUD consistent with similar results in people with alcohol use disorder. Reductions in neural cocaine cue reactivity suggest a beneficial impact on prospective relapse, potentially most for those who report greater baseline craving. Although results did not survive whole-brain corrections, studying functional connectivity the authors clearly considered second order effects: via monosynaptic afferent projections from cortical regions that are easier to non-invasively stimulate, e.g., the mPFC, it is possible to impact subcortical regions that are difficult to reach, e.g., the striatum or amygdala (4).

Yet, of the initial 33 randomized subjects, 25 completed the study (76% of those randomized), and 24 (73%) were available at the 1-month follow-up. The cue-reactivity results were driven by the 2- and 3-month follow-up data, characterized by further attrition, where only 20 and 15 subjects were available (61% and 45%), respectively. This subject attrition may have contributed to the absence of significant whole-brain results (for both the cue reactivity and functional connectivity effects), calling for a larger sample size also to replicate the reported regional effects (in as much as the mPFC effect did now show a treatment by time interaction, and the interaction in the anterior cingulate cortex and insula was driven by data at the 3-months follow-up, the most impacted by attrition). In general, given that 64 subjects were prescreened, this intervention may remain an option for the select few; recruitment from a Veteran’s Affairs hospital (88% of participants were men) further highlights the need for testing this intervention in women. Another concern was the absence of effect on craving despite recent studies and a meta-analysis that showed a strong effect of non-invasive brain stimulation over the PFC on reducing craving in stimulant use disorder. Future work could consider capturing the temporal dynamics and magnitude of drug craving (5), and directly comparing the impact of stimulation of the mPFC with other targets (e.g., the dorsolateral PFC) on this and other outcome measures.

Future work can also explore the mechanisms underlying the reported drug cue-reactivity reduction. Based on several addiction theories, including the impaired response inhibition and salience attribution (iRISA) (2), craving, cue-reactivity and drug use can be driven by PFC-mediated dysfunction of inhibitory control (and other higher-order executive functions) and the attribution of excessive salience to the drug and drug cues at the expense of non-drug related reinforcers. Using additional fMRI designs, one could therefore explore whether the current results suggest recovery of emotional regulation [e.g., down regulation with drug cue reappraisal (6) vs. up regulation with savoring of alternative non drug rewards or hedonic restructuring (7)], cognitive control (8), and/or motivation for cognitively effortful behavior (9); of note here are the instructions given to the participants in this trial to consider the negative aspects of cocaine use (instead of the more common passive viewing employed during drug cue reactivity tasks), supporting the engagement of higher-order cognitive processes. Of relevance here is a study that stimulated (with TBS) the right inferior frontal gyrus to report improved inhibitory control related functional connectivity, reducing craving and smoking in cigarette smokers (10).

Overall, results of this study are highly promising despite the small sample size. Spotlighting non-invasive stimulation of the mPFC, this trial provides convincing evidence for a new avenue for intervention in patients with CUD. Using a longitudinal sham-controlled design with up to four fMRI scans per subject (in a within-subjects test-retest approach), the current rigorous, creative, neuroscience-informed study meets the challenges of a neuromodulation trial conducted in a clinical population. In highlighting the mPFC as a promising neural target to reduce drug cue reactivity in SUD, this study is to become a crucial cornerstone in a field characterized by a great deal of variability in stimulation protocols and study designs, especially for stimulant addiction where there are currently no FDA approved treatments. Mechanistically, and as warranting further exploration, results are consistent with the use of brain stimulation not only as a tool for enhancing recovery, but also as a manipulator of the underlying neural circuitry. That is, the observed reductions in drug cue reactivity in this study can serve as causal evidence for the role of the mPFC in CUD. Encouraging precision, this study adds to an impressive body of work that lays the foundations for an individually-tailored program, whereby anatomical and brain-state guided approaches could become increasingly adopted and standardized, such that the right brain region/circuit is targeted for stimulation for the right person at the right time.