A trial can show a favorable secondary result while failing to establish a benefit on its primary endpoint. Ezard and colleagues’ 2025 methamphetamine-dependence trial illustrates why the observation window must remain attached to the conclusion.
What was tested?
Six Australian outpatient clinics randomized 164 adults with methamphetamine dependence and at least 14 use days in the preceding 28 days. Eighty were assigned lisdexamfetamine and 84 placebo.
The oral regimen involved induction to 250 mg/day, 12 weeks of maintenance and a two-week reduction, followed by follow-up. This was a specialist research regimen, not a dose recommendation or evidence that results from lower-dose ADHD treatment automatically transfer to dependence treatment.
Nine randomized participants never started medication and were excluded from analyses: five assigned lisdexamfetamine and four placebo. The treated analytic population was therefore 155, not all 164 randomized participants.
The primary endpoint
The primary efficacy outcome was methamphetamine-use days in the preceding 28 days at week 13. The adjusted difference was reported as 2.2 days, with a 95% confidence interval from −0.5 to 5.0. The trial did not establish a statistically significant primary-endpoint benefit.
This finding should not be replaced by the statement that the trial proved reduced use at treatment completion. The authors characterized evidence at that endpoint as weak.
The positive result covered another period
Across the whole 12-week maintenance phase, the lisdexamfetamine group reported fewer methamphetamine-use days in total: a difference of 8.8 days, 95% confidence interval 2.7–15.0, p = 0.005.
An accumulated difference across 12 weeks is not the same estimand as use in the final 28 days. Early reductions can contribute to a favorable total even when the final-period comparison is uncertain. Neither result establishes abstinence in every participant.
Retention and safety remain part of the result
Only 57% remained on study medication to the primary endpoint. Low retention reduces precision and raises concern about assumptions used for missing outcomes. The COVID-19 pandemic disrupted recruitment and study procedures, and medication adherence relied substantially on self-report.
Serious adverse events occurred in four lisdexamfetamine recipients, reported as 5%. A treatment-group adverse event is not automatically caused by the drug, but it should not disappear behind a general statement of tolerability.
What does the trial support?
The randomized evidence supports a possible reduction in use during treatment, with uncertainty about sustaining that difference at the primary endpoint. It does not establish a cure, an individual impairment threshold or a prescribing instruction. Primary and secondary outcomes must both remain visible when this study is discussed.
Scientific source
Ezard N, Clifford B, Siefried KJ, et al. Lisdexamfetamine in the treatment of methamphetamine dependence: A randomised, placebo-controlled trial. Addiction. 2025;120:1345–1359. doi:10.1111/add.16730. Original commentary on the supplied paper.
Leave a Reply