Prostacyclin for Patients with Septic Shock: Promise, Progress, and Perspective

Septic shock is characterized by hypoperfusion and organ dysfunction contributing to high mortality in these patients. Endothelial and microvascular dysfunction has long been proposed as a key mechanism in this process, based on decades of basic and translational research (1–3). Logically, endogenous vasodilators produced by the healthy endothelium have been suggested to improve endothelial function and microvascular flow and, hence, improve tissue perfusion and restore organ function, which may improve outcomes in patients with septic shock. Supporting this rationale, favorable effects were found with infusion of the prostacyclin analog, iloprost (4), on clinical hypoperfusion in a small case series of patients with septic shock (5) and on biomarkers of endothelial dysfunction and on organ dysfunction in a small, randomized trial of patients with septic shock (6).

Based on these early findings, the I-MICRO (Ilomedin for Treatment of Septic Shock with Persistent Microperfusion Defects) trial by Legrand and colleagues (7), published in this issue of the Journal (pp. 1211–1219), is an important step forward. I-MICRO was an investigator-initiated, publicly funded, placebo-controlled, blinded, randomized trial on the infusion of iloprost (dose, 0.5–1.5 ng/kg/min) for septic shock in 26 French ICUs. It enrolled 240 patients who, after initial resuscitation, had persistent hypoperfusion (defined as mottling or prolonged capillary refill time) within 24 hours of septic shock. These clinical markers may have enrolled a population with increased likelihood of benefit from an intervention that may improve endothelial and microvascular function.

The primary outcome of the I-MICRO trial, changes at Day 7 in Sequential Organ Failure Assessment score, was −4 in the iloprost group, versus −5 in the placebo group (median difference, 1; 95% confidence interval [CI], 0 to 3; P = 0.12). At 90 days, 48 (42%) patients had died in the iloprost group, versus 47 (39%) patients in the placebo group (relative risk, 1.08; 95% CI, 0.80 to 1.5). Serious adverse events occurred in 15% of patients in the iloprost group versus 7% of patients in the placebo group (P = 0.06) (7).

The I-MICRO trial was well conducted and reported (7). The investigators appropriately concluded that iloprost did not reduce the severity of organ failure in patients with septic shock. By design, the estimates for the more patient-important outcomes remain uncertain, but the signal was not in favor of iloprost.

Another recent trial reinforces these findings (8). The COMBAT SHINE (Combat Septic Shock Induced Endotheliopathy) trial randomized patients with septic shock and severe endotheliopathy (based on high plasma concentrations of soluble thrombomodulin, a marker of endothelial dysfunction) to iloprost (dose, 1.0 ng/kg/min) or placebo in six Danish ICUs; infusion was commenced within 12 hours of shock. COMBAT SHINE was stopped for futility when 279 patients had been enrolled; the primary outcome, mean Sequential Organ Failure Assessment score in the ICU, was 11 in both the iloprost and placebo groups. Mortality at 90 days in the iloprost group was 57%, versus 51% in the placebo group (adjusted relative risk, 1.12 [95% CI, 0.91–1.40]; P = 0.33), and serious adverse events occurred in 18% versus 15%, respectively (P = 0.52) (8).

When interpreted together, these two trials increase our confidence in the conclusion. Iloprost is unlikely to benefit patients with septic shock and possible endothelial and microvascular dysfunction, and it may harm them. Of course, questions remain around patient selection, timing, dosing, and statistical uncertainty. But this line of development appears to follow that of many others in sepsis—compelling in theory, but not in clinical practice.

This is the nature of drug development. No matter how compelling the mechanistic rationale, interventions should be assessed in randomized clinical trials with patient-important outcomes before being used in clinical practice (9). Sepsis is complex, and harm is a constant companion for the interventions given to these patients. Along this line, it remains to be proven that more granular detailing of the pathophysiology and, hence, precision medicine will provide better and faster drug development for these patients.

Another major challenge is the time it takes to advance a hypothesis to clinically relevant trial results. Both the I-MICRO and COMBAT SHINE programs took several years to conclude (7, 8); the trials were first registered in the European Union Drug Regulating Authorities Clinical Trials database in 2018 and 2019, respectively. These long timelines delay progress, increase research waste, and reduce opportunities for iterations. This harms patients, healthcare systems, and societies. The reasons are multifactorial, but they include that clinical trials are extremely regulated, detached from clinical practice, and funded on a trial-by-trial basis by partners external to the core task—delivery of health care (10). Consequently, most time and resources are spent on establishing the legal, operational, and controlling frameworks around single trials and less on lasting structures that facilitate adaptation and ongoing learning.

This must change. We echo the many colleagues who advocate for broad adoption of adaptive platform trials as practical framework for better, faster, and more cost-effective improvements in health care (11, 12). It is encouraging that academic groups within critical care are evolving adaptive platform trials across a broad range of patient conditions and scopes (12). But the potential for platform trials will only be fully realized with broader commitment from regulators, industry, and healthcare systems and, essentially, from clinicians and patients themselves.