Guest Editorial – Methylphenidate in Cocaine Use Disorder (CUD): Rethinking Trial Design Around CUD Comorbid with ADHD

Our recent systematic review and meta-analysis concluded that methylphenidate (MPH) does not significantly reduce craving or cocaine use in patients with cocaine use disorder (CUD).¹ Taken at face value, this is a negative result, adding to a long list of pharmacotherapies that have not shown efficacy for CUD, a condition with no FDA-approved medication to date.² We argue, however, that the pooled negative finding may have hidden a more specific and testable signal: across the studies in our review, evidence of benefit consistently clustered in patients with CUD and comorbid attention deficit hyperactivity disorder (ADHD). As a consequence, we believe clinical studies should stop evaluating MPH in unselected populations with CUD but instead enroll specifically for CUD with comorbid ADHD.

The Subgroup Signal

Our review included searching PubMed, Web of Science, and Scopus for randomized and nonrandomized studies of MPH in CUD, with or without comorbid ADHD. Of the total, 20 studies met inclusion criteria and 6 were suitable for a formal meta-analysis, showing that MPH did not have any significant effect on craving or cocaine use compared to placebo. However, this negative pooled estimate is an average across studies that were not designed or powered to detect a subgroup-specific effect. When we examined the pattern across the full set of 20 studies, a more specific relationship emerged: craving decreased significantly in roughly half of the studies, and in nearly all of those participants who also carried an ADHD diagnosis. Studies that administered MPH without considering the ADHD comorbidity, or intravenously, showed no benefit or even worsening of craving. On the contrary, studies that gave oral MPH at sustained doses to patients with diagnosed ADHD showed the clearest signal of benefit.¹

This pattern is consistent with a plausible mechanistic explanation. Specifically, MPH is a dopamine and noradrenaline transporter inhibitor with substantial pharmacodynamics that overlaps with cocaine itself. The stimulant replacement therapy, similar to the agonist treatments for opioid and tobacco use disorders, should be especially effective where dopaminergic dysregulation is the primary driver of the substance use, such as in patients whose stimulant use may partly reflect an underlying, undertreated ADHD. It is also consistent with known prevalence data: adult ADHD has been reported in roughly 15%-25% of treatment-seeking individuals with CUD, several-fold higher than in the general population, yet many of the trials in our review did not systematically screen for or report ADHD status.³ A drug evaluated in a population where the responsive subgroup is neither reliably identified nor consistently reported is likely to appear ineffective regardless of its efficacy in that subgroup.

The subgroup hypothesis is not new, but it has been under-tested with MPH specifically, and it has rarely been tested by design. Only 2 of the randomized trials in our own meta-analysis enrolled specifically for comorbid adult ADHD and CUD. Schubiner and colleagues found no difference in craving between MPH and placebo, though the trial was small and its craving measure could not be pooled quantitatively.⁴ By contrast, Levin and colleagues, in a larger double-blind comparison of methylphenidate and placebo among cocaine-dependent treatment seekers with adult ADHD, reported a reduced probability of a positive cocaine screen with MPH,⁵ a signal that did not reach statistical significance on its own but is consistent with the pattern we identified across the broader dataset. Neither trial was adequately powered for its comorbid subgroup, however both are consistent with the logic of agonist replacement therapy trials in substance use disorders.

Real-World Evidence

If the subgroup hypothesis is correct, it should be detectable in real-world prescribing data, where clinicians already exercise judgment about likely responders. In a retrospective cohort of adult patients with a dual diagnosis of ADHD and CUD, Manni and colleagues reported a marked reduction in cocaine use among patients treated with medications for ADHD,⁶ though as an uncontrolled retrospective design it carries some bias. However, a large propensity-score-matched cohort study published earlier this year is consistent with this prediction. Magagnoli and colleagues examined 109,608 veterans with a documented positive cocaine screen, 503 of whom had a record of MPH exposure.⁷ After matching on age, sex, ADHD diagnosis, and psychiatric and substance use comorbidities, MPH exposure was associated with significantly lower odds of a subsequent positive screen, both against unexposed patients (OR=0.71, 95% Cl 0.51-0.98) and against bupropion, an active comparator (OR=0.60, 95% Cl 0.41-0.87). This finding is consistent with the subgroup hypothesis of our study.⁷ This analysis does not establish causality, but the 2 cohorts represent independent and converging lines of naturalistic evidence pointing in the same direction similarly to our meta-analysis.

Clinical and Research Implications

For clinicians, this evidence supports a specific and actionable practice recommendation. In patients with CUD and a well-documented and independently confirmed ADHD diagnosis, not inferred from stimulant-seeking behavior alone, there is a preliminary but genuine rationale for considering MPH as an off-label option. This should be accompanied by informed consent that accurately reflects the evidence: MPH is not an approved treatment for CUD, the trial data supporting its use in this subgroup remain limited, and the real-world association may partly reflect selection into treatment rather than treatment effect. Cardiovascular monitoring is also warranted at the doses studied (up to 90 mg/day orally), and treatment response should be assessed with an objective measure such as toxicology-confirmed use rather than relying on self-report alone, given how consistently self-reported outcomes failed to show an effect across the trials in our review. In practice, this means establishing the ADHD diagnosis with a validated instrument or specialist evaluation prior to prescribing, and then carefully monitoring the patient through time.

For researchers, the priority is to design adequately powered randomized trials that enroll specifically for confirmed comorbid ADHD and CUD; use toxicology-confirmed cocaine use and validated ADHD symptom measures as co-primary outcomes; and apply robust, individually titrated dosing. MPH already has an extensive existing literature on this question and a substantial real-world evidence base across various candidate agents, yet there is a lack of trials of this design conducted for it to date.

Repeating the unselected-population design is unlikely to resolve the existing discordance and will likely reproduce the same diluted negative results. Within the literature captured by our review and the recently published cohort study, the evidence converges on a specific, testable hypothesis: MPH is most likely to benefit patients with CUD and comorbid ADHD specifically, not an unselected population with CUD. We believe that future trials should test this hypothesis directly, enrolling for the subgroup rather than the unselected population.