Heather Wakelee, MD, on NSCLC: IMpower010 Survival Results After Long-Term Follow-up of Atezolizumab vs Best Supportive Care

IMpower010, which looked at adjuvant atezolizumab for patients with completely resected early stage non-small cell lung cancer, we presented the long-term final disease-free survival as well as the second interim overall survival data, now with over 60 months of follow-up. The study had a complex hierarchical statistical design such that first group were those with PD-L1 expression stage 2 to 3A, next group, all stage 2 to 3A, then all comers, including about 12% of patients with 1B, and only if all three of those were met could we formally test overall survival. As was seen in the first interim analysis, the second interim now with the final disease-free survival in that first group with PD-L1 expression stage 2 to 3A remains robust with a hazard ratio of 0.7, translating into nearly a 30-month difference in disease-free survival. In the second group, when we look at all stage 2 to 3A, continue to see a disease-free survival benefit, which does reach statistical significance, hazard ratio of 0.85. But when we bring in that stage 1B, that drops to... Sorry, that was 0.83. ... 0.85, when we bring in that other group, no longer statistically significant. So overall survival could not be formally tested, but we do continue to see this robust disease-free survival benefit, particularly in those with PD-L1 expression. And when we look at those with the highest PD-L1 expression, greater than 50%, that disease-free survival benefit hazard ratio was 0.48. We haven't reached the median for that group of patients. And, in that group, the overall survival hazard ratio was 0.47. However, we can't formally statistically test that because of the way the trial is designed. So in conclusion, when we look at this final disease-free survival result and second interim overall survival result from IMpower010, we do continue to see robust clinical benefit for patients who have gone through surgical resection with early stage lung cancer, stage 2 to 3A, where their tumor expresses PD-L1 expression. And we do believe this continues to be a very reasonable treatment option for that subgroup of patients where there is PD-L1 expression, particularly in those with the highest PD-L1 expression of greater than 50%.