Erythropoietin to Reduce Mortality in Traumatic Brain Injury: A Post-hoc Dose-effect Analysis.

Abstract

We aimed to assess whether the dosing regimen of erythropoietin shows a relationship to mortality in critically ill patients with traumatic brain injury (TBI). Erythropoietin may decrease mortality in patients with TBI; however, the optimal dosing regimen remains uncertain. We conducted a post-hoc analysis of a multicenter, randomized trial of weekly erythropoietin versus placebo in patients with moderate and severe TBI admitted to intensive care. We assessed whether the cumulative dosage of erythropoietin was differentially associated with all-cause patient mortality evaluated at 6 months after injury. There was a nonlinear relationship between dose and mortality (P = 0.008) that remained after adjustment for site and severity of illness (P = 0.01). Six-month mortality was lower in randomized patients who received 1 [adjusted hazard ratio (HR) 0.58, 95% confidence interval (CI) 0.33-1.01; P = 0.06] or 2 doses of erythropoietin (HR 0.31, 95% CI 0.12-0.80; P = 0.02) compared with those who received no doses. No benefit was seen with 3 doses (HR 1.55, 95% CI 0.66-3.62; P = 0.33). There was no differential effect of dose on functional neurological outcomes. Results across subgroups and secondary intention to treat analyses were consistent with primary findings. This post-hoc, hypothesis-generating analysis found potential reductions in mortality following 1 or 2 weekly doses of 40,000 IU of erythropoietin in intensive care unit patients with moderate or severe TBI, but not with 3 doses. These findings will inform the design of future trials of erythropoietin in critically ill patients with TBI and trauma.