When a stroke trial changes its primary outcome midway through enrollment, the integrity of its central conclusion deserves scrutiny — especially when that conclusion is being used to argue for an adjunctive treatment in time-critical emergency medicine. That is exactly the concern raised here about a randomized clinical trial evaluating intra-arterial alteplase delivered immediately after successful mechanical thrombectomy in large vessel occlusion stroke.
The trial by Renú and colleagues tested whether infusing alteplase — a thrombolytic agent — directly into the affected artery following thrombectomy could further improve functional recovery beyond the mechanical reperfusion alone. The study reported a positive signal, suggesting adjunctive intra-arterial alteplase may enhance outcomes. However, a critical methodological problem undercuts this finding: the primary outcome measure was modified during the course of the trial, raising the possibility that the reported result reflects post-hoc optimization rather than a pre-specified hypothesis test.
Outcome switching in clinical trials — even when transparently disclosed — substantially inflates the risk of false-positive findings. In stroke research particularly, where functional outcome scales like the modified Rankin Score can be analyzed across multiple thresholds, the number of possible analytical choices creates fertile ground for inadvertent or deliberate p-value optimization. Regulatory bodies and methodology experts have long flagged mid-trial primary endpoint changes as a serious threat to internal validity, regardless of the scientific plausibility of the intervention itself.
Alteplase post-thrombectomy is a biologically coherent idea — residual microvascular thrombus burden may limit tissue reperfusion even after large vessel recanalization — and earlier exploratory data have been modestly encouraging. But coherent mechanisms do not rescue compromised trial design. Until a confirmatory trial with a stable, pre-registered primary outcome replicates this signal, clinicians should treat the functional recovery claim as hypothesis-generating rather than practice-changing. This letter is an important methodological corrective, not a dismissal of the underlying biology.