Ultra-Early, Short-Course Tranexamic Acid in Aneurysmal Subarachnoid Haemorrhage: An RCT-Only Meta-Analysis of Rebleeding Prevention Versus Ischaemic Harm, with Duration-Response and Trial Sequential Analysis

BackgroundAneurysmal subarachnoid haemorrhage (aSAH) carries a case fatality rate exceeding 35-40%, with early rebleeding occurring predominantly within the first 2-6 hours of ictus, representing the single most preventable determinant of catastrophic outcome. Tranexamic acid (TXA), a competitive plasminogen inhibitor that stabilises the perianeurysmal haemostatic clot, has been evaluated as a pre-occlusion bridge therapy for over four decades. Prior meta-analyses have been methodologically constrained by heterogeneous treatment-duration pooling, inclusion of non-randomised evidence, and the absence of prospective inferential boundary testing. The present analysis restricts inclusion exclusively to randomised controlled trials (RCTs), applies a pre-specified duration-response stratification (ultra-early/short-course [UE/SC] [≤]72 hours versus prolonged course), and deploys Trial Sequential Analysis (TSA) to determine whether cumulative randomised evidence has crossed definitive efficacy or harm thresholds. MethodsPRISMA 2020-compliant systematic review and meta-analysis restricted to RCTs, prospectively registered in PROSPERO (CRD42026133240). Electronic databases (MEDLINE, EMBASE, Cochrane CENTRAL, Web of Science) were searched from inception to December 2024. Eligible trials randomised adults with CT-confirmed aSAH to TXA versus placebo or standard care. Primary efficacy outcome: rebleeding incidence. Primary safety outcomes: delayed cerebral ischaemia (DCI), all-cause mortality, and poor functional outcome at 3-6 months. Effect sizes were pooled as risk ratios (RR) or odds ratios (OR) with 95% confidence intervals (CI) using random-effects meta-analysis (DerSimonian-Laird). Heterogeneity was assessed by I2 and Cochrans Q. Risk of bias was appraised via Cochrane RoB 2.0. GRADE certainty ratings were assigned. TSA was performed with two-sided alpha=0.05, power=80%, and an MCID of 30% relative risk reduction for rebleeding. ResultsSeven RCTs enrolling 2,194 patients in total were included. TXA significantly reduced rebleeding across all pooled RCTs (RR 0.58, 95% CI 0.44-0.76; I2=41%; GRADE: moderate certainty), with TSA confirming that the cumulative Z-curve crossed the adjusted monitoring boundary, establishing definitive evidence of efficacy. The duration-response contrast for DCI was unambiguous: ultra-early/short-course TXA ([≤]72 hours) conferred no excess DCI risk (RR 0.97, 95% CI 0.78-1.19), whereas prolonged-course TXA significantly elevated DCI risk (RR 1.42, 95% CI 1.12-1.81; TSA harm boundary crossed). Despite rebleeding reduction, neither UE/SC nor prolonged TXA improved all-cause mortality (RR 0.98, 95% CI 0.84-1.14) or good functional outcome (OR 0.86, 95% CI 0.66-1.12 in the UE/SC stratum). TSA confirmed insufficient cumulative evidence for functional benefit across all subgroups. ConclusionsTreatment duration is the critical and sole modulator of TXAs risk-benefit ratio in aSAH. Ultra-early, short-course TXA eliminates the DCI penalty while preserving a robust rebleeding-reduction signal. However, no adequately powered RCT demonstrates improvement in functional outcome or survival with any TXA regimen. These findings definitively argue against prolonged TXA and support its selective, time-limited use as a haemostatic bridge during unavoidable pre-occlusion delays, pending subgroup-stratified trials powered for functional endpoints.