A randomized, double-blind, active-controlled phase 3 multicenter trial of rHSA for cirrhotic ascites

In this multicenter, randomized, double-blind, active-controlled phase 3 trial (NCT06553456), we assessed the equivalence of recombinant human serum albumin (rHSA) expressed in Pichia pastoris to plasma-derived human serum albumin (pHSA) for elevating serum albumin (ALB) levels and non-inferiority regarding ascites improvement in patients with cirrhotic ascites. Using a rigorous dual-endpoint design, the primary endpoint was the change from baseline in serum ALB immediately after the final intravenous infusion. The key secondary endpoint was the ascites improvement rate after the final administration. Ultimately, 364 out of 390 patients completed the study. Statistical equivalence was established for the primary endpoint, with least-squares mean ALB changes of 11.16 ± 0.364 g/L (rHSA) and 10.95 ± 0.395 g/L (pHSA) (LSM difference: 0.21 g/L [95% CI: -0.75, 1.16]). For the secondary endpoint, ascites improvement was non-inferior, demonstrating a 58.0% response rate for rHSA versus 48.7% for pHSA (difference: 9.4% [95% CI: -0.56%, 19.10%]). Both endpoints of the trial were successfully achieved. The overall incidence of adverse events was similar in the two groups; crucially, no anti-drug antibodies were detected in the rHSA group. Exploratory analysis revealed that rHSA significantly reduced glycated ALB (-12.24%), whereas pHSA increased it ( + 12.52%) (P < 0.001). Corroborated by LC‒MS characterization, silver-staining purity assessment, and four biochemical quality attributes (free thiol, Hcy, AGEs, and carbonyl levels), this divergence highlights the preserved molecular integrity of rHSA. This trial demonstrated that in patients with cirrhotic ascites, rHSA was statistically equivalent to pHSA in elevating serum ALB and non-inferior in ascites improvement, with a favorable safety profile.