In this randomized controlled trial involving patients with ARDS, we addressed the question of whether inhaled prostacyclin would improve the lung function, as measured by oxygenation in the blood. We were able to show improved oxygenation on Day 5 of treatment in a population with ARDS however, the effect was not significant. The observed effect of prostacyclin was not associated with improved secondary outcomes in the intervention group, and neither the overall outcome nor the incidence of secondary complications was significantly different between groups.

In addition to extensive inflammation within the alveolar space, the central hallmark of ARDS is hypoxia [13, 14]. Prone positioning and the use of extracorporeal membrane oxygenation (ECMO) have been shown to reduce hypoxia and to increase oxygenation [15, 16]. ECMO therapy, however, is limited to expert centers and cannot be used widespread in all hospitals caring for these patients, since it involves a significant logistical effort and expert knowledge. Therefore, pharmaceutical approaches to improve pulmonary function are still very important. We described in this trial that a prostacyclin intervention only showed a non-significant tendency toward exerting a positive effect on oxygenation in critically ill patients with ARDS. In a small case study of twenty patients, Sawheny et al. showed that oxygenation in patients with ARDS was improved by administering inhaled prostacyclin [9]. Johanssen et al. showed that the intravenous administration of prostacyclin in COVID-19 ARDS with endotheliopathy was not associated with a significant reduction of mortality, but a point estimate analysis done after the end of the trial favored the prostacyclin group [10]. However, these two studies performed were either done without a control group, did not employ a randomized prospective design or used a different administration strategy for prostacyclin. Therefore, data about the role of prostacyclin acquired with a RCT design in patients with ARDS are still sparse to date.

As mentioned above, this randomized study documents the effect of prostacyclin on patients with ARDS including COVID-19-induced ARDS. COVID-19-induced ARDS is an entity characterized by additional features compared to classical ARDS. Patients with COVID-19 present widespread pulmonary microthrombi and inflammatory infiltrates with diffuse pulmonary fibrosis [8, 17]. In addition, endothelial dysfunction and a severe inflammatory response are indicators of COVID-19-induced pulmonary failure. Furthermore, hypoxemia that is unrelated to lung mechanics is present in patients with COVID-19-induced ARDS [18]. These pathological features are patterns that could be influenced by prostacyclin. Prostacyclin controls platelet aggregation and aggregability, preventing thrombus formation in an environment with a damaged endothelium [19, 20]. In addition, prostacyclin interacts with and enhances the effect of nitric oxide on the vascular surface [21]. As a result, endothelial function is improved, microthrombi are prevented, and the inflammatory response is reduced by administering prostacyclin to these patients. All of the described effects could have beneficial functions in patients with ARDS, especially in patients with COVID-19-induced ARDS.

Of course, our trial also has several limitations. First, the trial was started before the COVID-19 pandemic to evaluate the effects of prostacyclin on oxygenation and outcomes of critically ill patients with ARDS. Then, shortly after the start of the trial, the first wave of patients with COVID-19-induced ARDS were treated in Germany and German ICUs, including ours. Given the potential differences in the pathologies of ARDS and COVID-19-induced ARDS, this factor might have significant implications for therapy with prostacyclin. However, we decided to include all patient groups with ARDS and not exclude patients with COVID-19, since our trial should also take advantage of the opportunity to compare patients with different ARDS etiologies and their responses to prostacyclin treatment. Second, our sample size was moderate, and our study was probably underpowered. This interpretation seems justified, as we obtained the expected effect, i.e., a superiority of 21 mmHg in PaO2/FiO2, but the standard deviations were much larger, as expected (80 mmHg in the controls, 91 mmHg in the prostacyclin group vs. 40 mmHg assumed). Third, the intervention group and the control group differed significantly in age, which could have a potential effect on the overall outcome in this patient group. The average age was older in the intervention group, and therefore, one would expect this factor to have a potential negative effect if any effect at all, based on the literature [22, 23]. However, in our sample, no significant association of age with the primary outcome was observed. We also included patients receiving ECMO in this trial, which is particularly important because we measured oxygenation as the primary outcome. We recorded a nonsignificant difference between 21 patients treated with ECMO in the control group and 14 patients treated with ECMO in the treatment group, but of course, ECMO is important for the oxygenation levels measured. This is remarkable since the larger number in the control group would potentially skew the oxygenation toward the control group on Day 6, but we did not observe this result. The treatment groups still performed better when analyzing the primary outcome oxygenation and supported the positive effect of prostacyclin on oxygenation. Fourth, although the study medication assignment was randomized, we did not blind the investigators to the study medication, which was not possible due to the complex nature of the preparation of the prostacyclin in a blinded manner in our setting; therefore, we did not pursue this approach. Fifth, we included patients who had ARDS due to multiple reasons, and patients with and without COVID-19. However, impaired oxygenation is the common cardinal symptom of patients with all forms of ARDS, and most clinical approaches to improve oxygenation in all patients were tested in heterogeneous clinical ARDS groups, since we wanted to identify a commonly used intervention that would improve the poor oxygenation status. Therefore, we included all patients who met the inclusion criteria.

In conclusion, among patients with severe ARDS, inhaled prostacyclin showed a tendency to improve oxygenation, especially in COVID-19-induced ARDS. This change was not associated with a survival benefit but was associated with an improvement of secondary outcomes in the treated patient population. Larger clinical trials will evaluate the effect of prostacyclin on the overall outcomes of patients with ARDS.