Preventive Tirofiban for Early Neurological Deterioration in Acute Ischemic Stroke Without Thrombolysis or Thrombectomy

BACKGROUND AND PURPOSE: We assessed whether preventive intravenous tirofiban reduces early neurological deterioration (END) in acute ischemic stroke without thrombolysis or thrombectomy, and how population, timing and comparator differences affect interpretation. METHODS: A focused evidence update included PubMed searching through October 13, 2026, screening of 520 records and citation checks. Randomized and observational evidence were separated. Risk ratios (RRs) were analyzed using restricted maximum-likelihood models, with fixed-effect, Hartung-Knapp, prediction-interval and influence analyses. Abstract-level evidence was evaluated separately where eligibility was uncertain. RESULTS: Two randomized trials (1,395 participants) yielded discordant END estimates: RR 0.50 (95% CI 0.32-0.77) and 0.91 (0.67-1.22). The exploratory pooled RR was 0.69 (0.38-1.24; I²=79.6%; τ²=0.143), with marked model dependence. Four trials contributed 2,017 observed 90-day modified Rankin Scale 0-1 outcomes: RR 1.12 (1.05-1.19); modified Hartung-Knapp CI 1.01-1.23 and prediction interval 0.98-1.28. Excluding ESCAPIST because of discrepant published counts widened the modified Hartung-Knapp interval to 0.96-1.27. Separately, the BRANT abstract reported RR 1.01 (0.90-1.13); target-population eligibility was unconfirmed. Treatment-by-time interactions and mediation of functional benefit through END prevention were not established. CONCLUSIONS: Preventive END effects are trial-specific. The functional signal is sensitive to source and model assumptions and does not establish END-mediated benefit. Imaging-guided selection and incremental benefit over optimized oral therapy require direct randomized evaluation, with standardized neurological, functional and bleeding outcomes.