Key points are not available for this paper at this time.
Introduction: IMbrave150 established first-line atezolizumab plus bevacizumab as a global standard of care for unresectable hepatocellular carcinoma (HCC). We report exploratory analyses of associations between overall survival (OS) and depth of response (DpR) or duration of response (DoR). Methods: IMbrave150 was a phase III randomized study of atezolizumab plus bevacizumab versus sorafenib in patients with unresectable HCC. DpR was defined as maximum tumor shrinkage from baseline based on the sum of longest diameters per independent review facility (IRF)-assessed RECIST 1.1. DoR was defined as time from first complete/partial response by IRF-assessed RECIST 1.1 until progression or death. Associations between OS and DpR or DoR were evaluated by scatterplot in both arms; OS and PFS were evaluated by DpR in atezolizumab plus bevacizumab-treated patients. To minimize immortal time bias, the DpR analysis included patients who survived ≥6 months. Results: Of 312 and 140 patients with baseline measurable disease in the atezolizumab plus bevacizumab and sorafenib arms, respectively, 264 and 99 surviving ≥6 months were included in the DpR analysis, and 97 and 18 in the DoR analysis. Tumor shrinkage occurred in 230/312 (74%) patients in the atezolizumab plus bevacizumab arm and 76/140 (54%) in the sorafenib arm; their mean (SD) DpR was -42.5% (32.4%) and -25.0% (21.9%), respectively. Atezolizumab plus bevacizumab-treated ≥6-month survivors with DpR <0% had improved OS versus those with DpR ≥0% (HR: 0.29; 95% CI: 0.19-0.44). Those with deeper responses (DpR -100% to -60%) had longer OS than those with DpR ≥20% (unstratified HR: 0.08; 95% CI: 0.03-0.21). In scatterplots, DpR and DoR were generally associated with OS in both arms; interpretation was limited by censored patients. Conclusions: DpR and DoR to atezolizumab plus bevacizumab and sorafenib were associated with OS in patients with unresectable HCC. More longer, deeper responses occurred with atezolizumab plus bevacizumab.
Kudo et al. (Mon,) studied this question.