Effect of Deferoxamine on Outcome According to Baseline Hematoma Volume: A Post Hoc Analysis of the i-DEF Trial
Chenchen Wei, Jeffrey Wang, Lydia D. Foster, Sharon D. Yeatts, Claudia Moy, J Mocco, Magdy Selim, Yuko Palesch, Jessica Griffin, Aaron Perlmutter, Casey Norton, Catherine Dillon, Andre Thornhill, Steven Greenberg, Lewis Morgenstern, Daniel Hanley, Guohua Xi, Thomas P Bleck, Robert Balk, Yuko Palesch, Jessica Griffin, Andre Thornhill, Aaron Perlmutter, Catherine Dillon, Wenle Zhao, Jaemyung Kim, Sandeep Kumar, Casey Norton, Megan Farinella, Erica Siwila-Sackman, Caroline Feigert, Gottfried Schlaug, Jasmine Wang, Clinton Wright, Scott Janis, Peter Gilbert, Carlos S Kase, Shannon Carson, Matthew Flaherty, Craig M Kessler, Leslie A McClure, Donna Bannon, Joseph Broderick, Amy Dill
Abstract
Background: Hematoma volume (HV) is a powerful determinant of outcome after intracerebral hemorrhage. We examined whether the effect of the iron chelator, deferoxamine, on functional outcome varied depending on HV in the i-DEF trial (Intracerebral Hemorrhage Deferoxamine). Methods: A post hoc analysis of the i-DEF trial; participants were classified according to baseline HV (small <10 mL, moderate 10–30 mL, and large >30 mL). Favorable outcome was defined as a modified Rankin Scale score of 0–2 at day-180; secondarily at day-90. Logistic regression was used to evaluate the differential treatment effect according to HV. Results: Two hundred ninety-one subjects were included in the as-treated analysis; 121 with small, 114 moderate, and 56 large HV. Day-180 modified Rankin Scale scores were available for 270/291 subjects (111 with small, 105 moderate, and 54 large HV). There was a differential effect of treatment according to HV on day-180 outcomes ( P -for-interaction =0.0077); 50% (27/54) of deferoxamine-treated patients with moderate HV had favorable outcome compared with 25.5% (13/51) of placebo-treated subjects (adjusted odds ratio, 2.7 [95% CI, 1.13–6.27]; P =0.0258). Treatment effect was not significant for small (adjusted odds ratio, 1.37 [95% CI, 0.62–3.02]) or large (adjusted odds ratio, 0.12 [95% CI, 0.01–1.05]) HV. Results for day-90 outcomes were comparable ( P -for-interaction =0.0617). Sensitivity analyses yielded similar results. Conclusions: Among patients with moderate HV, a greater proportion of deferoxamine- than placebo-treated patients achieved modified Rankin Scale score 0–2. The treatment effect was not significant for small or large HVs. These findings have important trial design and therapeutic implications. Registration: URL: https://www.clinicaltrials.gov ; Unique identifier: NCT02175225.