Abstract T MP8: Faster Angiographic Reperfusion is Associated with Improved Outcomes Across the Full Rankin Distribution: An Ordinal Analysis of the IMS III Trial
INTRODUCTION: In the NIH-funded IMS III trial, faster angiographic reperfusion led to a greater likelihood of good clinical outcome based on the dichotomized modified Rankin Score (mRS) score of 0-2 (minimal or no disability) vs. 3-6. From 240 to 390 minutes, among 1000 patients, every 15 minute acceleration in reperfusion was associated with 32 more patients with mRS 0-2. We hypothesized that faster reperfusion would lead to better clinical outcomes across the full mRS distribution (0, 1, 2, 3, 4, 5-6).
METHODS: IMS III randomized subjects to IV/endovascular vs IV tPA alone. We analyzed the endovascular cohort with M1, M2 and ICAT occlusions reperfused (TICI 2-3) during the procedure (up to 7 hours). Time to reperfusion was defined as time from stroke onset to procedure termination. The ordinal 90-day mRS as a function of time to reperfusion was assessed using a proportional odds model with adjustment for pre-morbid mRS>0, ASPECTS 0-4, and NIHSS strata (10-19 vs ≥20). “Typical patient” covariate values were used to obtain mRS distributions at prespecified time points. An algorithmic joint table analysis was applied to generate number-needed-to-treat values.
RESULTS: Among 177 M1, M2, and ICAT occlusion cases with angiographic reperfusion (after excluding 2 cases with missing 90-day mRS and 3 with missing baseline ASPECTS), the common OR for improvement across all levels of the mRS was 0.78 (95% 0.66-0.91) adjusted and 0.77 (95% 0.66-0.89) unadjusted for each 30 minute increase in time. See Figure for adjusted probabilities based on the fitted model. From 240 to 390 minutes, among 1000 patients, every 15 minute acceleration in reperfusion was associated with 57 more patients with improvement of ≥ 1 mRS disability level.
CONCLUSION: Faster time to reperfusion improves clinical outcomes for substantially more patients when considering the potential effects across all mRS levels, as opposed to considering a single transition between binary endpoints 0-2 vs 3-6.
Author Disclosures: P. Khatri: Research Grant; Modest; Dr. Khatri’s Dept of Neurology receives support for her role DSMB member from Biogen, Inc.. Research Grant; Significant; Dr. Khatri’s Dept of Neurology receives support for her roles as: (1) Lead PI of the PRISMS trial from Genentech, Inc and (2) Neurology PI of the THERAPY trial from Penumbra, Inc. S.D. Yeatts: Research Grant; Significant; NIH - funding for role as unblinded statistician in IMS III. Consultant/Advisory Board; Significant; Genentech for role as member of PRISMS Trial Steering Committee. L. Fan: None. J.P. Broderick: Research Grant; Modest; PRISMS Trial - Genentech. Honoraria; Modest; Boehringer Ingelheim. Consultant/Advisory Board; Modest; Pfizer. Other Research Support; Significant; Study medication for IMS III Trial Genentech. J.L. Saver: Other; Significant; Dr. Saver is an employee of the University of California.
- © 2015 by American Heart Association, Inc.