Endotoxin Removal Tied to Lower Mortality Risk in Septic Shock Subtype

CHICAGO — Blood purification using polymyxin B showed a mortality reduction signal in patients with septic shock and high endotoxin activity, researchers reported here.

In the phase III Tigris trial, 28-day mortality was numerically lower in patients receiving standard of care plus two sessions of polymyxin B hemoadsorption (39%) versus standard of care alone (45%), with a baseline-severity-adjusted odds ratio (aOR) of 0.80 (95% credible interval [CrI] 0.38-1.49), according to Javier A. Neyra, MD, of the University of Alabama at Birmingham.

At 90 days, the absolute difference in mortality in favor of the hemodialysis intervention reached 18% (aOR 0.50, 95% CrI 0.23-0.94), findings presented at the Society of Critical Care Medicine annual meeting and published in Lancet Respiratory Medicine indicated.

Tigris had a prespecified Bayesian analysis framework to pool evidence from a cohort from a prior phase III trial (EUPHRATES) that used the same study protocol. This analysis showed a 95.3% posterior probability of benefit with polymyxin B hemoadsorption for mortality at 28 days (aOR 0.67, 95% CrI 0.39-1.08), meeting predetermined criteria for success.

“In patients with endotoxic septic shock defined by high endotoxic activity and multiorgan failure, polymyxin B hemoadsorption was associated with a high probability of lower mortality at 28 days and at 90 days,” Neyra said.

The posterior probability of benefit at 90 days with the endotoxin removal approach reached 99.4% (aOR 0.54, 95% CrI 0.32-0.87), and there was only a small increase in treatment-related serious adverse events (AEs) in the polymyxin B group.

Endotoxic septic shock “represents a specific subtype of septic shock affecting about 10% of patients with sepsis, characterized by overwhelming endotoxin-driven inflammation, vascular dysfunction, and subsequent multiorgan failure,” explained Tobias Schupp, MD, PhD, and colleagues at Heidelberg University in Mannheim, Germany, in an accompanying editorial. “Endotoxin activity, which can be measured by point-of-care assays, can contribute to a robust host-immune response and subsequent multiorgan failure with short-term mortality rates reaching 60%.”

Pharmaceutical approaches targeting endotoxin — a lipopolysaccharide of Gram-negative bacteria — in septic shock have not been shown to improve survival, they noted.

Polymyxin B binds to endotoxin through tightly packed polystyrene fibers, according to the researchers, and one cartridge of the investigational device used in the study (Toraymyxin PMX-20R) can remove approximately 10-20 μg of endotoxin over 2 hours. Endotoxin remains within the cartridge while purified blood returns to the patient. In a previous subgroup analysis of EUPHRATES, polymyxin B showed potential benefit among the patients with levels of endotoxin considered high but treatable (0.60 to 0.89 on an endotoxin activity assay).

Schupp and colleagues acknowledged the challenge of identifying eligible patients for the intervention in daily clinical practice. Tigris screened nearly 15,000 patients for septic shock, but ultimately only randomized 157 who met eligibility criteria, which included endotoxin levels within the treatable range, need for vasopressors, multiorgan dysfunction, and other requirements to match the EUPHRATES patient population.

Of note, they said, Gram-negative bacteria — pathogens most associated with endotoxic septic shock — were present in fewer than 30% of the Tigris patients.

Schupp and co-authors concluded that the trial “might serve as an important signpost towards modern precision medicine, shaping the way for a promising therapeutic approach for a select subgroup of patients with septic shock, which warrants special attention in view of an overall very poor prognosis.”

Spectral Medical, the company that makes the polymyxin B hemofiltration device, said it intends to submit the trial data to the FDA for premarket approval in the coming months.

From 2019 to 2025, the open-label, phase III Tigris trial enrolled adults with septic shock at 19 intensive care units (ICUs) in the U.S., randomizing 106 to standard of care plus polymyxin B via hemodialysis (two sessions, 22 hours apart, at a blood flow rate of 80-120 mL per minute for 90 to 120 minutes per session) and 51 to standard of care alone. ORs for mortality were adjusted for baseline Acute Physiology and Chronic Health Evaluation (APACHE-II) scores.

Participants had an average age of 61 years, 58% were male, over two-thirds were white, 17% were Black, and 8% were Hispanic. The intervention and control groups were similar in terms of APACHE-II scores, multiple organ dysfunction score, Sequential Organ Failure Assessment score, presence of bacteremia, norepinephrine dose, cumulative vasopressor index, endotoxin level, and types of microorganisms present.

Among the patients alive at day 28, no significant differences were observed in their location:

• ICU: 14% of the polymyxin B group vs 21% of controls
• Hospital, non-ICU: 31% vs 29%
• Discharged to a nursing facility: 23% vs 18%
• Discharged home: 32% vs 32%

Among the patients alive and in the ICU at 28 days, six of the nine (67%) in the polymyxin B group survived to 90 days compared with one of six (17%) control patients. Researchers noted a numerically larger difference in 28-day mortality for the patients who received a baseline dose of norepinephrine greater than 0.1 μg/kg per minute (38% among the intervention group vs 49% among controls).

The pooled results included 179 treatable patients from EUPHRATES (range of 0.60 to 0.89 on an endotoxin activity assay) and in Bayesian analyses showed a mortality difference of 10.3% (95% CrI -1.7 to 22.3) at 28 days favoring polymyxin B treatment and 15.5% (95% CrI 3.6-27.1) at 90 days, with numbers needed to treat to prevent one death of 9.7 and 6.5, respectively.

Serious AEs in Tigris occurred in 30% of the patients receiving polymyxin B and 22% of controls. Of the serious AEs in the polymyxin B group, two (2%) were considered related to the intervention — one due to polymyxin B and the other due to catheter placement.