Effect of Decompressive Craniectomy According to Location of Deep Intracerebral Hemorrhage: A SWITCH Trial Analysis

BACKGROUND: Decompressive craniectomy (DC) seemed to reduce the risk of death or profound disability (modified Rankin Scale score, 5–6) after deep intracerebral hemorrhage (ICH) by an absolute 13% (95% CI, 0%–26%) in the SWITCH trial (Swiss Trial of Decompressive Craniectomy versus Best Medical Treatment of Spontaneous Supratentorial Intracerebral Hemorrhage). Whether the effect of DC differs by ICH location is unknown. METHODS: Post hoc analysis of participants with supratentorial severe deep ICH from the intention-to-treat population of the SWITCH randomized controlled trial. We categorized ICH as involving (1) the basal ganglia (BG) alone, (2) BG and the posterior limb of the internal capsule (PLIC), or (3) BG, PLIC, and thalamus. We examined the interaction between ICH location and DC’s effect on primary (modified Rankin Scale score, 5–6) and secondary outcomes (death; full modified Rankin Scale score range) at 180 days using unadjusted and adjusted logistic or survival models. RESULTS: Of 197 participants comprising the trial’s intention-to-treat population, 184 were available for analysis (median age, 61 years; 59 women; 91 randomized to DC plus best medical treatment; and 93 to best medical treatment). ICH involved BG alone in 26 (14%), BG+PLIC in 94 (51%), and BG+PLIC+thalamus in 64 participants (35%). The marginal risk of the primary outcome after adjustment for age, ICH severity, and volume was lower with DC by 15.6% (95% CI, −49.2% to 18.1%) in participants with ICH of BG alone, by 11.4% (−29.3% to 6.6%) in those with ICH of BG+PLIC, and by 9% (−31% to 12.9%) in those with ICH of BG+PLIC+thalamus, without evidence for treatment-by-location interaction ( P =0.95). Secondary outcome analyses yielded consistent results. CONCLUSIONS: The potential benefits of DC seemed preserved regardless of the location of severe deep ICH. REGISTRATION: URL: https://www.clinicaltrials.gov ; Unique identifier: NCT02258919.