Targeted Therapy Using Polymyxin B Hemadsorption in Patients With Sepsis

A Post-hoc Analysis of the JSEPTIC-DIC Study and the EUPHRATES Trial

Itsuki Osawa; Tadahiro Goto; Daisuke Kudo; Mineji Hayakawa; Kazuma Yamakawa; Shigeki Kushimoto; Debra M. Foster; John A. Kellum; Kent Doi

Disclosures

Crit Care. 2023;27(245)

Abstract and Introduction

Abstract

Background: Polymyxin B hemadsorption (PMX-HA) reduces blood endotoxin levels, but characteristics of patients with sepsis likely to benefit from PMX-HA are not well known. We sought to identify patient subgroups likely to benefit from PMX-HA.

Methods: We retrospectively identified 1911 patients with sepsis from a retrospective observational study in Japan (the JSEPTIC-DIC study) and 286 patients with endotoxemic septic shock from a randomized controlled trial in North America that restricted patients to those with high endotoxin activity (the EUPHRATES trial). We applied the machine learning-based causal forest model to the JSEPTIC-DIC cohort to investigate heterogeneity in treatment effects of PMX-HA on 28-day survival after adjusting for potential confounders and ascertain the best criteria for PMX-HA use. The derived criteria for targeted therapy by PMX-HA were validated using the EUPHRATES trial cohort.

Results: The causal forest model revealed heterogeneity in treatment effects of PMX-HA. Since patients having higher treatment effects were more likely to have severe coagulopathy and hyperlactatemia, we identified the potential treatment targets of PMX-HA as patients with PT-INR > 1.4 or lactate > 3 mmol/L. In the EUPHRATES trial cohort, PMX-HA use on the targeted subpopulation (75% of all patients) was significantly associated with higher 28-day survival (PMX-HA vs. control, 68% vs. 52%; treatment effect of PMX-HA, + 16% [95% CI + 2.2% to + 30%], p = 0.02).

Conclusions: Abnormal coagulation and hyperlactatemia in septic patients with high endotoxin activity appear to be helpful to identify patients who may benefit most from PMX-HA. Our findings will inform enrollment criteria for future interventional trials targeting patients with coagulopathy and hyperlactatemia.

1 2 3 4 5 6 7

Next Section

Crit Care. 2023;27(245) © 2023 BioMed Central, Ltd.
Copyright to this article is held by the author(s), licensee BioMed Central Ltd. This is an Open Access article: verbatim copying and redistribution of this article are permitted in all media for any purpose, provided this notice is preserved along with the article's original citation.

Table 1. Patient characteristics in each quintile in order of estimated individual treatment effects, derivation cohort
Characteristic	Quintile 1 (n = 390)	Quintile 2 (n = 380)	Quintile 3 (n = 380)	Quintile 4 (n = 380)	Quintile 5 (n = 381)
Characteristic	Lower treatment effects	Quintile 2 (n = 380)	Quintile 3 (n = 380)	Quintile 4 (n = 380)	Higher treatment effects
Patient demographics
Age*	64 (53–69)	68 (58–76)	76 (70–82)	78 (71–84)	79 (71–84)
Female sex, n (%)*	156 (40%)	164 (43%)	168 (44%)	164 (43%)	165 (43%)
Infection site, n (%)
Lung	99 (25%)	101 (27%)	91 (24%)	99 (26%)	99 (26%)
Abdominal	104 (27%)	110 (29%)	126 (33%)	159 (42%)	143 (38%)
Blood culture positive, n (%)	288 (74%)	272 (72%)	304 (80%)	282 (74%)	284 (75%)
Patient severity
APACHE II score*	19 (14–25)	18 (14–23)	19 (15–23)	21 (17–25)	30 (23–37)
SOFA score*	8 (5–10)	7 (5–9)	7 (5–9)	8 (6–11)	12 (9–14)
Respiratory*	2 (1–3)	2 (1–3)	2 (1–3)	2 (1–3)	3 (2–3)
Coagulation*	0 (0–2)	0 (0–2)	1 (0–2)	1 (0–2)	1 (0–2)
Liver*	0 (0–1)	0 (0–1)	0 (0–1)	0 (0–1)	0 (0–1)
Cardiovascular*	1 (0–3)	2 (0–3)	2 (0–3)	2 (0–3)	3 (2–4)
Central nervous system*	1 (0–3)	1 (0–2)	1 (0–2)	1 (0–2)	3 (1–4)
Renal*	1 (0–2)	0 (0–1)	0 (0–1)	1 (0–2)	2 (1–3)
Laboratory tests
White blood cells [× 10³/μL]*	12 (6–22)	12 (6–19)	12 (6–17)	11 (5–16)	8 (3–16)
Platelet [× 10⁴/μL]*	15.0 (8.1–23.5)	14.9 (9.1–22.3)	14.0 (9.0–20.8)	13.8 (8.0–20.6)	10.3 (5.3–17.2)
PT-INR*	1.28 (1.13–1.44)	1.26 (1.12–1.42)	1.27 (1.15–1.46)	1.33 (1.18–1.51)	1.50 (1.29–1.85)
Fibrinogen [mg/dL]*	487 (366–650)	441 (346–553)	431 (341–524)	406 (277–499)	310 (185–426)
FDP [μg/mL]*	22 (12–33)	20 (11–29)	20 (11–31)	23 (14–40)	36 (18–66)
Lactate [mmol/L]*	2.3 (1.5–3.3)	2.3 (1.5–3.3)	2.3 (1.6–3.5)	4.1 (2.2–5.9)	5.9 (4.4–8.2)
Treatment effects of PMX-HA on 28-day survival [95% CI]	− 4.9% [− 24% to + 14%]	− 0.0% [− 9.3% to + 9.2%]	+ 1.8% [− 6.5% to + 10%]	+ 8.8% [+ 0.1% to + 18%]	+ 21% [+ 11% to + 32%]

Values represent median (IQR), unless otherwise indicated. Higher treatment effects on 28-day survival are equivalent to higher absolute risk reduction of 28-day mortality
APACHE II score, Acute Physiology and Chronic Health Enquiry II score; SOFA score, Sequential Organ Failure Assessment score; PT-INR, Prothrombin Time and International Normalized Ratio; FDP, Fibrinogen/Fibrin Degradation Products; PMX-HA, Polymyxin B Hemadsorption
*Variables were used for adjusting the causal forest model

Table 2. Estimated treatment effects of Polymyxin B hemadsorption in the derivation and validation cohorts
	Treatment effects on 28-day survival [95% CI]	Hazard ratio for 28-day mortality [95% CI]	Treatment effects on 28-day survival without RRT initiation [95% CI]
Derivation cohort
Full sample (n = 1911)	+ 5.3% [− 1.9% to + 13%]	–	–
Targeted subpopulation (n = 1203)	+ 9.2% [+ 1.0% to + 17%]	–	–
Full sample (n = 286)	+ 7.4% [− 4.3% to + 19%], (PMX-HA vs. control, 70% vs. 62%)	0.71 [0.48 to 1.06]	+ 2.9% [− 9.0% to + 15%] (PMX-HA vs. control, 39% vs. 36%)
Validation cohort
Targeted subpopulation (n = 214)	+ 16% [+ 2.2% to + 30%] (PMX-HA vs. control, 68% vs. 52%)	0.62 [0.40 to 0.95]	+ 12% [− 1.4% to + 25%] (PMX-HA vs. control, 37% vs. 25%)
Patients on RRT at baseline (n = 54)	+ 7.4% [− 24% to + 39%] (PMX-HA vs. control, 47% vs. 40%)	0.56 [0.25 to 1.27]	–
Patients not on RRT at baseline (n = 160)	+ 15% [− 0.3% to + 31%] (PMX-HA vs. control, 73% vs. 58%)	0.68 [0.40 to 1.16]	+ 11% [− 5.8% to + 27%] (PMX-HA vs. control, 45% vs. 35%)

Higher treatment effects on 28-day survival are equivalent to higher absolute risk reduction of 28-day mortality. "Targeted subpopulation" indicates patients with PT-INR > 1.4 or lactate > 3 mmol/L on ICU admission. The estimate of each treatment effect of PMX-HA in the derivation cohort are computed using the augmented inverse probability weighting formula with the causal forest model. The estimated treatment effects in the validation cohort were calculated using the per-protocol analysis utilizing the nature of a RCT
RRT, Renal Replacement Therapy; PT-INR, Prothrombin Time and International Normalized Ratio; PMX-HA, Polymyxin B Hemadsorption; RCT, Randomized Controlled Trial