A medication can affect cigarette smoking without demonstrating benefit for methamphetamine dependence. A historical phase II trial of varenicline tested those outcomes separately. Keeping them separate prevents a positive secondary finding from replacing negative primary findings.
What treatment was tested?
Briones and colleagues randomized 52 treatment-seeking adults with methamphetamine dependence to varenicline, 27 participants, or placebo, 25. The target dose was oral varenicline 1 mg twice daily after titration. Both groups were offered weekly individual cognitive behavioral therapy during the nine-week medication phase.
The protocol also included an inpatient abstinence-initiation period. Participants provided urine at frequent clinic visits. Missing specimens were treated as methamphetamine-positive for the primary analyses. Twenty-six participants completed the study and 26 dropped out.
What were the primary outcomes?
End-of-treatment abstinence required methamphetamine-negative urine specimens during weeks eight and nine. It was achieved by 4/27 varenicline participants, 15%, and 5/25 placebo participants, 20%; p = 0.9.
The treatment effectiveness score counted methamphetamine-negative urine specimens across treatment. Mean scores were 8.6 with varenicline and 8.1 with placebo. Treatment was not a significant predictor of this count: incidence rate ratio 1.01, 95% confidence interval 0.39–2.70, p = 0.9.
Why preserve the missing-specimen rule?
The analysis did not treat an absent sample as laboratory-confirmed abstinence. Its rule combined observed positive results with missing visits for outcome purposes. A report should therefore distinguish a convention used to calculate treatment success from an actual analytical detection.
Substantial dropout also matters. The assigned-group denominators keep the randomized comparison visible, whereas reporting only completers could answer a different, selected-sample question. Neither approach should be substituted without explanation.
What happened to cigarette smoking?
The investigators observed a greater decline in reported cigarette consumption with varenicline than placebo. That finding concerned a different substance and endpoint. It did not demonstrate that methamphetamine urine outcomes improved, nor did it validate the proposed biological mechanisms for treating methamphetamine dependence.
This is a useful example of why “the medication worked” is incomplete. A meaningful claim must identify the outcome, comparison and time window. A secondary benefit can coexist with failure to demonstrate the primary benefit being tested.
Did adherence explain away the null findings?
The study measured urinary varenicline for adherence analysis. Some associations suggested better outcomes with higher medication exposure, but they were not statistically conclusive. Adherence was not randomly assigned, so those analyses could not replace the medication-versus-placebo comparison.
The authors also calculated Bayes factors for the primary hypotheses. They leaned toward no treatment difference but did not cross the paper’s stated threshold for persuasive evidence. This is more nuanced than declaring proof that the treatments were identical or that every possible benefit was excluded.
How should this historical trial be used?
It did not demonstrate improved methamphetamine abstinence at the tested dose in this selected population. Its sample size, dropout, accompanying therapy and uncertainty constrain broader conclusions. It supplies evidence about a particular trial regimen, not a current prescribing recommendation or an estimate of benefit for an individual patient.
Source
Briones M and colleagues. Varenicline treatment for methamphetamine dependence: A randomized, double-blind phase II clinical trial. Drug and Alcohol Dependence. 2018;189:30–36. DOI: 10.1016/j.drugalcdep.2018.04.023. Full text.