Depression in Parkinson’s disease frequently coexists with apathy, anxiety, motor fluctuation, and susceptibility to hallucinations. Bupropion has a pharmacological rationale in this setting, yet the clinical literature is too heterogeneous to separate an apathy-specific benefit from general mood change or dopaminergic adverse effects. To determine whether a phenotype-gated, dual-outcome Bayesian trial can identify a joint antidepressant and anti-apathy advantage while constraining psychosis-related harm. A structured evidence matrix was constructed from 23 publications comprising seven clinical investigations, eleven case reports or case series, and five synthesis or guidance records. Exact numerical observations were retained only when the denominator, instrument, and assessment interval were reported. These observations set the eligibility restrictions, outcome domains, follow-up duration, and safety margin for a double-dummy comparison of bupropion XL with sertraline. Monte Carlo experiments evaluated three sequential analyses at total enrolments of 80, 160, and 240. Joint efficacy required a posterior probability greater than 0.975 that both standardized apathy and depression effects favored bupropion; harm stopping required a posterior probability greater than 0.95 that the psychosis-related event excess exceeded five percentage points. The literature contained five abstract-only records (21.7%) and eleven case-based reports (47.8%). Among reports with compatible paired scores, decreases ranged from 22.4% to 76.0% for depression scales, while a gait score decreased by 11.9% without statistical significance. Under no treatment advantage, the simulated probability of an efficacy decision was 0.0024 and mean enrolment was 111.5. Standardized advantages of 0.45 for apathy and 0.35 for depression yielded an efficacy-decision probability of 0.587 and mean enrolment of 198.1. When the bupropion psychosis-related event rate was 15% and the comparator rate was 3%, the harm-stop probability was 0.641 and mean enrolment fell to 175.4. A dual-outcome sequential trial can distinguish joint symptomatic benefit from mood-only improvement and can terminate early when clinically important neuropsychiatric harm emerges. The design answers a narrower and more clinically actionable question than unrestricted antidepressant efficacy: whether bupropion benefits depressed patients with prominent apathy, stable dopaminergic therapy, and low psychosis liability.
Depression in Parkinson’s disease frequently coexists with apathy, anxiety, motor fluctuation, and susceptibility to hallucinations. Bupropion has a pharmacological rationale in this setting, yet the clinical literature is too heterogeneous to separate an apathy-specific benefit from general mood change or dopaminergic adverse effects. To determine whether a phenotype-gated, dual-outcome Bayesian trial can identify a joint antidepressant and anti-apathy advantage while constraining psychosis-related harm. A structured evidence matrix was constructed from 23 publications comprising seven clinical investigations, eleven case reports or case series, and five synthesis or guidance records. Exact numerical observations were retained only when the denominator, instrument, and assessment interval were reported. These observations set the eligibility restrictions, outcome domains, follow-up duration, and safety margin for a double-dummy comparison of bupropion XL with sertraline. Monte Carlo experiments evaluated three sequential analyses at total enrolments of 80, 160, and 240. Joint efficacy required a posterior probability greater than 0.975 that both standardized apathy and depression effects favored bupropion; harm stopping required a posterior probability greater than 0.95 that the psychosis-related event excess exceeded five percentage points. The literature contained five abstract-only records (21.7%) and eleven case-based reports (47.8%). Among reports with compatible paired scores, decreases ranged from 22.4% to 76.0% for depression scales, while a gait score decreased by 11.9% without statistical significance. Under no treatment advantage, the simulated probability of an efficacy decision was 0.0024 and mean enrolment was 111.5. Standardized advantages of 0.45 for apathy and 0.35 for depression yielded an efficacy-decision probability of 0.587 and mean enrolment of 198.1. When the bupropion psychosis-related event rate was 15% and the comparator rate was 3%, the harm-stop probability was 0.641 and mean enrolment fell to 175.4. A dual-outcome sequential trial can distinguish joint symptomatic benefit from mood-only improvement and can terminate early when clinically important neuropsychiatric harm emerges. The design answers a narrower and more clinically actionable question than unrestricted antidepressant efficacy: whether bupropion benefits depressed patients with prominent apathy, stable dopaminergic therapy, and low psychosis liability.
