Final DFS analysis in the Impower010 trial

Professor Wakelee from Stanford, United States, presented the final DFS analysis of the IMpower010 trial at ASCO 2024. IMpower010 is a randomised phase 3 trial investigating adjuvant atezolizumab for patients with completely resected stage IB to IIIA NSCLC). Patients received up to four cycles of cisplatin-based chemotherapy and were then randomised to either one year of atezolizumab or best supportive care. The focus of this presentation was the five-year follow-up.

The results confirm previous findings, showing a robust DFS benefit for patients with stage II to IIIA disease who have PD-L1 expression of at least 1%. The hazard ratio for this group is 0.7, indicating a substantial 30-month improvement in DFS at 60 months of follow-up. The study employed a complex statistical design, with the first endpoint focusing on this specific PD-L1 positive group.  The second endpoint evaluated DFS regardless of PD-L1 expression in stages II to IIIA, yielding a statistically significant hazard ratio of 0.83. When including all patients, including those with stage IB disease, the hazard ratio for DFS is 0.85. However, the 95% confidence interval for this group is 1.01, narrowly missing statistical significance.

Despite not meeting the formal overall survival (OS) testing criteria, the OS curves are noticeably separated, particularly in patients with high PD-L1 expression. For patients with PD-L1 expression of at least 50% in stage II to IIIA, the hazard ratio for DFS is 0.48, and for OS, it is 0.47. While these results are not statistically testable due to not hitting the formal endpoint, they suggest clinically meaningful benefits.

Adjuvant atezolizumab thus remains a viable option for patients with completely resected NSCLC and PD-L1 expression, providing significant DFS benefits, especially in those with higher PD-L1 expression.