Abstract T P182: Post-tPA ICH: INSTINCT Community Hospitals vs Pooled Analysis Data

Abstract

Background: The threat of ICH limits community physician use of tPA in stroke. Only the PH2 subtype, however, independently causes clinical deterioration. We explored ICH characteristics in the community-based INSTINCT trial and compared PH2 incidence rates in INSTINCT to the pooled analysis (PA) of the ECASS, ATLANTIS, NINDS and EPITHET trials. Methods: Secondary analysis of tPA-treated patients in the INSTINCT trial, a prospective, cluster-randomized, controlled trial conducted at 24 randomly selected Michigan community hospitals. Medical records of all tPA-treated strokes from 2007 to 2010 were initially reviewed by an independent physician panel to identify any evidence of possible ICH. In a second stage, 2 vascular neurologists, blinded to clinical data, independently categorized each possible ICH using ECASS criteria. Disagreements resolved by consensus. All images within 72h post-tPA were reviewed, with classification based on the most severe category. Data from pooled subjects with OTT of ≤180 minutes served as a historical control in the comparison. IRB approved. Results: Neuroimaging was obtained from 23 of 24 INSTINCT sites, with 462 tPA-treated subjects. 85 (18.4%) were identified at initial review as a possible ICH. 82/85 (96%) had post-treatment neuroimaging available. At second-stage review, 19 had no identified hemorrhage, leaving 63/462 (13.6%) with an ICH. ICH subtype is presented in Table 1. In the PA cohort, 464 tPA-treated subjects had OTT ≤ 180 minutes. 161 (34.7%; 95% CI 30.5, 39.1) had an ICH, with 22 (4.7%; 95% CI 3.1%, 7.1%) identified as PH2. In the comparison, the difference did not reach significance (1.1%; 95% CI -1.6%, 3.8%; p = 0.42). Conclusions: The overall INSTINCT ICH rate was low. Incidence of PH2 was not different from rates reported by investigators in large clinical research trials. This supports the safety of community treatment of stroke.