PLoS ONE (Jan 2017)

Efficacy of polymyxin B-immobilized fiber hemoperfusion for patients with septic shock caused by Gram-negative bacillus infection.

  • Nobuyuki Saito,
  • Kazuhiro Sugiyama,
  • Testu Ohnuma,
  • Takashi Kanemura,
  • Michitaka Nasu,
  • Yuya Yoshidomi,
  • Yuta Tsujimoto,
  • Hiroshi Adachi,
  • Hiroyuki Koami,
  • Aito Tochiki,
  • Kota Hori,
  • Yukiko Wagatsuma,
  • Hisashi Matsumoto

DOI
https://doi.org/10.1371/journal.pone.0173633
Journal volume & issue
Vol. 12, no. 3
p. e0173633

Abstract

Read online

Septic shock-associated mortality in intensive care units (ICUs) remains high, with reported rates ranging 30-50%. In particular, Gram-negative bacilli (GNB), which induce significant inflammation and consequent multiple organ failure, are the etiological bacterial agent in 40% of severe sepsis cases. Hemoperfusion using polymyxin B-immobilized fiber (PMX), which adsorbs endotoxin, is expected to reduce the inflammatory sepsis cascade due to GNB. However, the clinical efficacy of this treatment has not yet been demonstrated. Here, we aimed to verify the efficacy of endotoxin adsorption therapy using PMX through a retrospective analysis of 413 patients who received broad spectrum antimicrobial treatment for GNB-related septic shock between January 2009 and December 2012 in 11 ICUs of Japanese tertiary hospitals. After aligning the patients' treatment time phases, we classified patients in two groups depending on whether PMX hemoperfusion (PMXHP) therapy was administered or not within 24 hours after ICU admission (PMXHP group: n = 134, conventional group: n = 279). The primary study endpoint was the mortality rate at 28 days after ICU admission. The mean age was 72.4 (standard deviation: 12.6) years, and the mean Sequential Organ Failure Assessment score at ICU admission was 9.9 (3.4). The infection sites included intra-abdominal (38.0%), pulmonary (18.9%), and urinary tract (32.2%), and two thirds of all patients had GNB-related bacteremia. Notably, the mortality at 28 days after ICU admission did not differ between the groups (PMXHP: 29.1% vs. conventional: 29.0%, P = 0.98), and PMXHP therapy was not found to improve this outcome in a Cox regression analysis (hazard ratio = 1.16; 95% confidence interval, 0.81-1.64, P = 0.407). We conclude that PMX-based endotoxin adsorption within 24 hours from ICU admission was not associated with mortality among patients with septic shock due to GNB. TRIAL REGISTRATION:University Hospital Medical Information Network Clinical Trial Registry (UMIN-CTR ID: UMIN000012748).