A subgroup analysis of 124 EGFR-positive patients in the IMpower150 trial showed improved survival under a regimen of atezolizumab plus bevacizumab plus carboplatin plus paclitaxel versus the standard-of-care.
Why this matters
The IMpower150 trial included among its 1202 patients with stage IV or recurrent metastatic nonsquamous NSCLC approximately 10% with alterations of the epidermal growth factor receptor (EGFR) gene. For those patients, treatment options are limited once tyrosine kinase inhibitors (TKIs) have failed.
Analysis of the subpopulation of patients with EGFR in the three arms of IMpower150. Arm A: atezolizumab + chemotherapy (n=45), arm B: atezolizumab + bevacizumab + chemotherapy (n=34), arm C: bevacizumab + chemotherapy (n=45).
- Median age in the 3 arms was approx 62 years, with 38% male patients in arm A and approximately 50% in arms B and C. ECOG 0 was 44%, 53%, and 60%. Never smokers 64%, 59%, and 44%. Liver metastases 20%, 12%, and 16% respectively.
- Median progression free survival (PFS) was 10.2 months in arm B vs 6.9 months in arm C for a hazard ratio of 0.61 and a 95%-confidence interval (CI) of 0.36–1.03.
- Hazard ratios were better and significant among the 58 patients (73%) who harbored sensitizing EGFR mutations (HR 0.41; 95% CI 0.23–0.75), and for those 50 patients (63%) who had received prior TKI therapy (HR 0.42; 95% CI 0.22–0.80).
- Median overall survival (OS) in arm B was not estimable; it was 18.7 months in arm C. The HR was 0.61 for the total EGFR-mutated population (95% CI 0.29–1.28). It was significant for sensitizing EGFR mutations (HR 0.31; 95% CI 0.11–0.83) but barely missed significance for those receiving prior TKI (HR 0.39; 95% CI 0.14–1.07).
- Median duration of response: arm A 5.6 months; arm B 11.1 months; arm C 4.7 months.
- There was no difference in PFS and OS between arms A and C.
- Serious adverse events lead to withdrawal in 14%, 33%, and 16% in arms A, B, and C respectively.
Subgroup analysis was done on a relatively small number of patients with greater variation in baseline characteristics compared to the full population study arms.
F. Hoffmann-La Roche, Ltd.