PurposeThe EUPHRATES trial examined the impact of polymyxin B hemoperfusion (PMX) on mortality in patients with septic shock and endotoxemia, defined as EAA ≥ 0.60. No difference was found in 28-day all-cause mortality. However, the trial showed that in some patients with septic shock the burden of endotoxin activity was extreme (EAA ≥ 0.9). In a post hoc analysis, we evaluated the impact of PMX use in patients with septic shock and endotoxin activity measured between 0.6–0.89.MethodsPost-hoc analysis of the EUPHRATES trial for the 194 patients with EAA ≥ 0.6–0.89 who completed two treatments (PMX or sham). The primary end point was mortality at 28 days adjusted for APACHE II score and baseline mean arterial pressure (MAP). Additional end points included changes in MAP, cumulative vasopressor index (CVI), median EAA reduction, ventilator-free days (VFD), dialysis-free days (DFD) and hospital length of stay. Subpopulations analyzed were site and type of infection and those with norepinephrine dose > 0.1 mcg/kg/min at baseline.ResultsAt 28 days, 23 patients of 88 (26.1%) in the PMX group died versus 39 of 106 (36.8%) in the sham group [risk difference 10.7%, OR 0.52, 95% CI (0.27, 0.99), P = 0.047]. When unadjusted for baseline variables, P = 0.11. The 28-day survival time in the PMX group was longer than for the sham group [HR 0.56 (95% CI 0.33, 0.95) P = 0.03]. PMX treatment compared with sham showed greater change in MAP [median (IQR) 8 mmHg (− 0.5, 19.5) vs. 4 mmHg (− 4.0, 11) P = 0.04] and VFD [median (IQR) 20 days (0.5, 23.5) vs. 6 days (0, 20), P = 0.004]. There were no significant differences in other end points. There was a significant difference in mortality in PMX-treated patients with no bacterial growth on culture [PMX, 6/30 (20%) vs. sham, 13/31 (41.9%), P = 0.005]. The median EAA change in the population was − 12.9% (range: increase 49.2%–reduction 86.3%). The mortality in the above median EAA change group was PMX: 6/38 (15.7%) vs. sham 15/49 (30.6%), P = 0.08.ConclusionsThese hypothesis-generating results, based on an exploratory post hoc analysis of the EUPHRATES trial, suggest measurable responses in patients with septic shock and an EAA ≥ 0.6 to 0.89 on changes in mean arterial pressure, ventilator-free days and mortality.Trial registrationClinicaltrials.gov Identifier: NCT01046669. Funding Spectral Medical Incorporated.Electronic supplementary materialThe online version of this article (10.1007/s00134-018-5463-7) contains supplementary material, which is available to authorized users.