Cyclosporine treatment of ARDS due to severe COVID-19: The I-SPY COVID Adaptive Platform Trial
Angela Haczku, Ari Moskowitz, Martin Eklund, Andrew G Chapple, Fady A Youssef, Philip A Robinson, Sara C Auld, D Clark Files, Eliot Friedman, Richard A Lee, Derek W Russell, Kenneth K Wei, Timothy Obermiller, Jonathan L Koff, Carolyn S Calfee, Neil K Aggarwal, Ellen L Burnham, Nuala J Meyer, Rajiv Sonti, Richard G Wunderink, Daniel H Kett, Sandhya Samavedam, Paul A Berger, Malik M H S Khan, Jeremiah W A Hayanga, Michael A Bernstein, Santhi Kumar, Jeremy R Beitler, Timothy E Albertson, Robert C Hyzy, Caroline M Quill, Benjamin LaCombe, Andrea Discacciati, Lamorna Brown-Swigart, Michael A Matthay, Kathleen D Liu, Laura J Esserman, , Neil R Aggarwal, Ellen L Burnham, Carrie Higgins, Jeff McKeehan, Timothy Albertson, Angela Haczku, Erin Hardy, Richart Harper, Brian Morrissey, Christian Sandrock, Sara Auld, Philip Yang, Joshua Detelich, Gavin Harris, Katherine Nugent, Max Adelman, Jeremy R Beitler, Anita Darmanian, Amy L Dzierba, Ivan Garcia, Katarzyna Gosek, Purnema Madahar, Aaron M Mittel, Justin Muir, Amanda Rosen, John Schicchi, Alexis L Serra, Romina Wahab, Paul Berger, Carolyn S Calfee, Melissa Coleman, Alejandra Jauregui, Nathan Cobb, Rajiv Sonti, Amen M Hamed, Alessio Crippa, Andrea Discacciati, Martin Eklund, Laura Esserman, D Clark Files, Karl Thomas, Kevin W Gibbs, Leigha Landreth, Mary LaRose, Lisa Parks, Adina Wynn, Eliot Friedman, Derek W Russell, Donna Harris, Abhishek Methukupally, Siddharth Patel, Sheetal Gandotra, Kashif Khan, Se Fum Wong, Albert Yen, Jonathan Koff, Lindsie Boerger, John Kazianis, Santhi Kumar, Kathleen D Liu, Thomas R Martin, Mark M Wurfel, Michael A Matthay, Brian Daniel, Nuala J Meyer, Caroline A G Ittner, Nilam S Mangalmurti, John P Reilly, Timothy Obermiller, Philip A Robinson, Farjad Sarafian, Usman Shah, Richard G Wunderink, G R Scott Budinger, Helen K Donnelly, Benjamin D Singer, Fady A Youssef, Daniel Belvins, Catherine Nguyen, Alexis Suarez, Maged A Tanios, Scott Fields, James Hurst-Hopf, Lamorna Brown Swigart, Christina Creel-Bulos, Christina Spainhour, Ari Moskowitz, Praveen Vijhani, Adrienne M Casciato, Vaney Capetillo, Kenenth Wei, Tracie Huynh, Anna Rodriguez-Vasquez, Joseph L Nates, Jessica Suarez, Siddharth Nair, Sandya Samavedam, Michael Bernstein, Christopher Jordan, Daniel H Kett, Karla Leon Escalona, Richard A Lee, Kenneth Remy, Ali Zarrinpar, Robert Hyzy, Kristine Nelson, Caroline Quill, Emilio Mazza, Kristin Broderick, Jermiah Hayanga, Ana Costa Monteiro, Joseph Levitt, Ruixiao Lu, Paul Henderson, Adam Asare, Imogene Dunn, Alejandro Botello BarraganAbstract
Rationale
Aberrant host response to SARS-CoV-2 infection causes severe lung injury and acute respiratory distress syndrome (ARDS) with resultant multiorgan failure and death. Cyclosporine, a calcineurin inhibitor with potent immunosuppressant and direct antiviral properties, was hypothesized to reduce cytokine-driven injury in hospitalized patients with severe COVID-19 ARDS.
Objectives
To evaluate whether cyclosporine improves time to recovery and reduces mortality in critically ill patients with COVID-19, using the I-SPY COVID adaptive platform trial design.
Methods
In this phase 2, multicenter, open-label, randomized controlled trial, non-intubated and intubated adults with confirmed COVID-19 requiring ≥6L/min oxygen received oral cyclosporine 5mg/kg/day in two divided doses for 5 days, added to standard backbone therapy (dexamethasone and remdesivir), or standard therapy alone. Co-primary endpoints were time to recovery and time to death, assessed using Bayesian proportional-hazard models with pre-specified graduation thresholds.
Results
Of 238 randomized patients (110 cyclosporine, 128 control), 28-day recovery rates were 65.9% vs. 71.1% and 28-day mortality rates were 10.7% vs. 11.3%, respectively. The posterior probability that cyclosporine improved recovery was 6% (graduation threshold >97.5%) and that it reduced mortality was 59.6% (graduation threshold >90%), meeting neither pre-specified efficacy criterion. Serious adverse events were more frequent with cyclosporine (17.3% vs. 7.0%, p = 0.016). No significant differences in inflammatory plasma biomarkers were observed between groups.
Conclusions
Cyclosporine did not meet pre-specified efficacy thresholds for recovery or mortality in severe COVID-19 ARDS and was associated with a higher rate of serious adverse events. These results do not support its addition to standard care in this population.