PALO ALTO, Calif.--(BUSINESS WIRE)--Endpoint Health, Inc., a precision-first™ therapeutics company dedicated to addressing urgent needs in immune-driven critical and chronic illnesses, today announced new retrospective analyses that provide additional insights into the company’s artificial intelligence-driven approach to identifying biologically-distinct groups within diseases predictive of outcomes of different treatment strategies in adults with acute respiratory distress syndrome (ARDS). These insights build on previous work that identified two clinically distinct subgroups (subphenotypes) of adults with ARDS. A total of four retrospective analyses were presented at the 2022 Virtual Critical Care Congress conducted by the Society of Critical Care Medicine (SCCM), April 18-21. ARDS is a life-threatening condition that impacts an estimated 150,000 Americans each year and is a common cause of death in patients with COVID-19.5,6
The retrospective analyses were based on recently published research in BMJ Open. That research identified two new subphenotypes in adults with ARDS, based on just nine, widely available clinical variables that are collected in the routine care of all ARDS patients: Subphenotype A, characterized by low inflammation and low mortality, and subphenotype B, associated with high inflammation and high mortality.7 The new retrospective analyses presented at SCCM showed that ARDS patients in these two subphenotypes had meaningful differences in clinical outcomes.
Heterogeneity of Effect of PEEP Strategy by Subphenotypes Derived from Clinical Data in ARDS (Star Research Presentation No. 12)
A retrospective evaluation of two randomized clinical studies in 1559 adult patients with ARDS (ALVEOLI and ART) by Kast R, et al. found that high positive end-expiratory pressure (PEEP) resulted in higher risk for 28-day mortality compared to low PEEP in patients who were in the subphenotype A group (OR, 1.66 [95% CrI, 1.13 to 2.47]; probability of harm of 99.4%) but not in subphenotype B (OR, 0.94 [95% CrI, 0.65 to 1.34]; probability of benefit of 63.9%).1 These effects were not modified when patients were stratified by either PaO2/FiO2 ratio or driving pressure, when the trials were assessed separately or using different priors.1
“In this retrospective analysis, we found that based on two previously identified ARDS subphenotypes derived from commonly available clinical variables, we could predict probabilities of response to different PEEP treatment strategies in patients with ARDS,” said Rodrigo Deliberato, MD, PhD, Endpoint Health’s Vice President of Research and Translational Medicine. “Our promising findings, which need to be further validated in prospective randomized controlled trials, may improve clinical classification of patients and lead to successful application of individualized treatment strategies in ARDS patients.”
Findings From Additional Retrospective Analyses
A retrospective analysis by Bulgarelli, et al. (Abstract No. 1124), which analyzed full enteral vs. trophic feeding in patients with ARDS from the EDEN study, showed a different probability of benefit of full enteral feeding in subphenotype A (OR, 0.78 [95% CrI, 0.49 to 1.22], probability of benefit of 86.3%) compared to subphenotype B (OR, 1.05 [95% CrI, 0.66 to 1.67], probability of benefit of 42.1%).2 Bulgarelli, et al. (Abstract No. 1089) also conducted a separate retrospective evaluation of the ROSE trial, finding that subphenotype A patients had a 94.6% chance to benefit from cisatracurium, a neuromuscular blocking agent commonly used in patients with ARDS, compared to 42.2% of subphenotype B patients.3 A fourth retrospective analysis by Kast R, et al. (Abstract No. 281) of CoDEX, a randomized study evaluating dexamethasone against the usual care of patients with COVID-19-associated ARDS, found the same differential clinical, laboratory and outcomes characteristics in the two subphenotypes as were found in ARDS patients without SARS-CoV-2 infection.4
“We are encouraged by these findings that add to the body of work demonstrating the potential utility of our proprietary clustering approach to classifying ARDS patients, with a view toward eventually applying personalized therapeutic strategies to patients in need,” said Diego Rey, PhD, Chief Scientific Officer at Endpoint Health. “The identification of subphenotypes based on nine commonly available clinical variables could streamline clinical trials and ultimately inform the future development of targeted therapeutic approaches for this heterogeneous condition with high unmet need.”
The four retrospective analyses had some limitations. All evaluations were based on patients enrolled in randomized controlled clinical studies, and the limitations of these original trials influence the retrospective analyses. Considering possible selection bias in the subphenotyping process, HTE in the two previously identified ARDS subgroups could be present due to other characteristics that were not assessed. The results of the retrospective analyses only apply to patients with a complete set of variables needed to run the subphenotyping algorithm. The findings need to be validated in prospective randomized controlled clinical trials. As such, no final conclusions can be made regarding the ability to predict probabilities of response to different PEEP treatment strategies in patients with ARDS.
About the Retrospective Analyses
All retrospective analyses were based on two recently identified clinically distinct subgroups in adults with ARDS. These subphenotypes were determined using a novel, proprietary, unsupervised clustering model from Endpoint Health. The final model included nine clinical variables that are routinely collected during ARDS care: heart rate, mean arterial pressure, respiratory rate, bilirubin, bicarbonate, creatinine, partial pressure of oxygen (PaO2), arterial pH, and fraction of inspired oxygen (FiO2).7 Patients in subphenotype B had consistently higher mortality, lower numbers of ventilator-free days at day 28, and longer duration of ventilation compared to subphenotype A patients.7 They also presented increased levels of pro-inflammatory markers.7
The retrospective analysis by Kast R, et al. used data from the ALVEOLI and ART trials, two randomized clinical studies in 1559 adult patients with ARDS.1 The purpose of the retrospective analysis was to predict the treatment effects of high vs. low PEEP strategies, according to EPH subphenotypes.1 The primary endpoint was 28-day mortality based on subphenotype and heterogeneity of treatment effect (HTE) of PEEP strategy.1 The secondary endpoints were total ventilator days and total ventilator-free days at day 28.1
The retrospective analysis of full vs. trophic enteral feeding by Bulgarelli, et al. was based on data from 777 patients from the EDEN trial, a randomized clinical study assessing the effects of trophic or full enteral feeding in 1000 adult patients with ARDS.2 The primary endpoint was 60-day mortality based on subphenotype and heterogeneity of treatment effect (HTE) of nutrition strategy.2
The second retrospective analysis by Bulgarelli, et al. evaluated HTE of neuromuscular blockade with cisatracurium vs. no neuromuscular blockade.3 This evaluation utilized data from 653 participants of the ROSE trial, a randomized clinical study assessing the benefits of early continuous neuromuscular blockade in 1006 adults with ARDS.3 The primary endpoint was 90-day mortality based on subphenotype and heterogeneity of treatment effect (HTE) of cisatracurium3
The retrospective analysis by Kast R, et al. was designed to determine whether patients with ARDS and COVID-19 share similar clinical characteristics to ARDS patients without SARS-CoV-2 infection.4 The research used data from 165 patients who participated in the COVID-19 Dexamethasone (CoDEX) trial, a pivotal multicenter randomized clinical study of 229 patients to evaluate the efficacy of intravenous dexamethasone vs. usual care in patients with moderate to severe ARDS due to COVID-19.4 The primary outcome was the identification of subphenotypes in adults with ARDS and COVID-19.4 The secondary outcome was the difference in 28-day mortality between subphenotypes.4
The scientific information discussed in this release related to the Endpoint Health pipeline of therapies and therapy-guiding tests is preliminary and investigative. Product and diagnostic candidates are not approved by the U.S. Food and Drug Administration, and no conclusions can or should be drawn regarding the safety or effectiveness of the product or diagnostic candidates.
About ARDS
Acute Respiratory Distress Syndrome (ARDS) is a life-threatening condition, which leads to inflammation and buildup of fluid in the lungs.5 These clinical manifestations prevent the lungs from adequately providing enough oxygen into the blood.5 ARDS can be caused by direct injuries, such as pneumonia, inhalation of harmful fumes and chest injuries, or indirect injuries like sepsis (severe and widespread bacterial infection in the body) and acute pancreatitis (inflammation of the pancreas).5 ARDS is a common cause of death in patients with COVID-19.6 Altogether, ARDS impacts an estimated 150,000 Americans each year. Despite intensive therapy, approximately 40% of people with ARDS die from lung failure.5
About Endpoint Health
Endpoint Health is a precision-first™ therapeutics company dedicated to addressing urgent needs in immune-driven critical and chronic illnesses by building a pipeline of therapies designed to personalize treatment to each patient’s biology. We combine therapeutics, therapy-guiding tests, and AI to develop targeted therapies for patients with inflammatory illnesses, such as acute respiratory distress syndrome (ARDS), sepsis, rheumatologic illnesses, and gastroenterologic illnesses. Our vision is a world in which all patients get the best treatment possible for their unique biology and disease. For more information, visit www.endpointhealth.com.
Follow us on Social Media: LinkedIn, Twitter
References
[1] Kast R, et al. Heterogeneity of Effect of PEEP Strategy by Subphenotypes Derived from Clinical Data in ARDS. Abstract no. 12. Presented at Society of Critical Care Medicine virtual 2022 Critical Care Congress, April 18-21, 2022.
[2] Bulgarelli L, et al. Heterogeneity of Effect of Nutrition Strategies by Subphenotypes Derived from Clinical Data in ARDS. Abstract no. 1124. Presented at Society of Critical Care Medicine virtual 2022 Critical Care Congress, April 18-21, 2022.
[3] Bulgarelli L, et al. Heterogeneity of Effect of Cisatracurium by Subphenotypes Derived from Clinical Data in ARDS. Abstract no. 1089. Presented at Society of Critical Care Medicine virtual 2022 Critical Care Congress, April 18-21, 2022.
[4] Kast R, et al. Subphenotypes in COVID-19 ARDS: Secondary Analysis of a Randomized Clinical Trial. Abstract no. 281. Presented at Society of Critical Care Medicine virtual 2022 Critical Care Congress, April 18-21, 2022.
[5] American Thoracic Society. What is Acute Respiratory Distress Syndrome? Updated April 2020. Available at https://www.thoracic.org/patients/patient-resources/resources/acute-respiratory-distress-syndrome.pdf. Last accessed April 2022.
[6] Hasan S, Capstick T, Ahmed R, et al. Mortality in COVID-19 patients with acute respiratory distress syndrome and corticosteroids use: a systematic review and meta-analysis. Expert Rev Respir Med. 2020:1-15. Published online 2020 Sep 29. doi: 10.1080/17476348.2020.1804365.
[7] Duggal A, Kast R,, Van Ark E, et al. Identification of Acute Respiratory Distress Syndrome subphenotypes de novo using routine clinical data: a retrospective analysis of ARDS clinical trials. BMJ Open 2022;12:e053297. doi:10.1136/bmjopen-2021-053297.
