Determinants of the effect of extracorporeal carbon dioxide removal in the SUPERNOVA trial: implications for trial design

Abstract

Purpose: To describe the variability and determinants of the effect of extracorporeal CO₂ removal (ECCO₂R) on tidal volume (V_t), driving pressure (ΔP), and mechanical power (Power_RS) and to determine whether highly responsive patients can be identified for the purpose of predictive enrichment in ECCO₂R trial design.

Methods: Using data from the SUPERNOVA trial (95 patients with early moderate acute respiratory distress syndrome), the independent effects of alveolar dead space fraction (ADF), respiratory system compliance (Crs), hypoxemia (PaO₂/FiO₂), and device performance (higher vs lower CO₂ extraction) on the magnitude of reduction in V_t, ΔP, and Power_RS permitted by ECCO₂R were assessed by linear regression. Predicted and observed changes in ΔP were compared by Bland-Altman analysis. Hypothetical trials of ECCO₂R, incorporating predictive enrichment and different target CO₂ removal rates, were simulated in the SUPERNOVA study population.

Results: Changes in V_t permitted by ECCO₂R were independently associated with ADF and device performance but not PaO₂/FiO₂. Changes in ΔP and Power_RS were independently associated with ADF, Crs, and device performance but not PaO₂/FiO₂. The change in ΔP predicted from ADF and Crs was moderately correlated with observed change in ΔP (R² 0.32, p < 0.001); limits of agreement between observed and predicted changes in ΔP were ± 3.9 cmH₂O. In simulated trials, restricting enrollment to patients with a larger predicted decrease in ΔP enhanced the average reduction in ΔP, increased predicted mortality benefit, and reduced sample size and screening size requirements. The increase in statistical power obtained by restricting enrollment based on predicted ΔP response varied according to device performance as specified by the target CO₂ removal rate.

Conclusions: The lung-protective benefits of ECCO₂R increase with higher alveolar dead space fraction, lower respiratory system compliance, and higher device performance. ADF and Crs, rather than severity of hypoxemia, should be the primary factors determining whether to enroll patients in clinical trials of ECCO₂R.

Keywords: Acute respiratory distress syndrome; Artificial ventilation; Extracorporeal carbon dioxide removal; Predictive enrichment; Ventilator-induced lung injury.