Atezolizumab Combination Not Superior to Regorafenib for Refractory mCRC

A phase III study comparing atezolizumab plus cobimetinib and atezolizumab monotherapy vs standard-of-care regorafenib in chemotherapy-refractory metastatic colorectal cancer (mCRC) did not reveal any significant differences in outcomes, reported a late-breaking session at ESMO World Congress on Gastrointestinal Cancer (June 20-23; Barcelona, Spain).

The primary analysis results from IMblaze370, a global, multi-center, open-label, randomized phase III trial, were presented by Johanna Bendell, Sarah Cannon Research Institute/Tennessee Oncology (Nashville, TN).

Patients with chemotherapy-refractory microsatellite stable (MSS) mCRC have limited treatment options and relatively short survival. Researchers hypothesized that combining atezolizumab, a PD-L1 inhibitor, with cobimetinib, a MEK1/MEK2 inhibitor, may allow better immune recognition and generate greater anti-tumor effects than either agent alone in MSS/microsatellite instability-low (MSI-L) mCRC.

Patients with previously treated, unresectable locally advanced or mCRC were randomized 2:1:1 to receive atezolizumab plus cobimetinib, atezolizumab monotherapy, or regorafenib. Atezolizumab was administered twice weekly (840 mg IV) in the combination arm and three times weekly (1200 mg IV) in the monotherapy arm. Cobimetinib (60 mg orally) was administered on a 21-days-on/7-days-off schedule, and regorafenib (160 mg orally) was administered on a 21-days-on/7-days-off schedule. The primary endpoint was overall survival (OS) in intention-to-treat (ITT) patients; secondary endpoints included investigator-assessed progression-free survival (PFS), overall response rate (ORR), and duration of response (DOR) per RECIST v1.1. 

A total of 363 patients were evaluated for efficacy and safety in the reported analysis. Patient had a median age of 58 years, and 26% had received more than 3 lines of prior treatment in the metastatic setting. Of the patients enrolled, 1.7% were identified as having MSI-high mCRC, and 54.3% had RAS-mutated mCRC.

IMblaze370 did not meet its primary endpoint; atezolizumab plus cobimetinib and atezolizumab monotherapy did not demonstrate statistically significant prolonged OS benefit vs regorafenib in the ITT population. Median OS was 8.9 months with atezolizumab + cobimetinib vs 8.5 months with regorafenib (HR = 1.00, 95% CI: 0.73, 1.38; P= 0.987) and was 7.1 months with atezolizumab monotherapy (vs regorafenib, HR = 1.19, 95% CI: 0.83, 1.71). An exploratory analysis demonstrated a numerically increased median OS for atezolizumab plus cobimetinib vs atezolizumab monotherapy (HR = 0.81; 95% CI: 0.60, 1.10), Dr Bendell reported.

In addition, PFS and ORR were similar across treatment arms (atezolizumab plus cobimetinib vs regorafenib, HR = 1.25, 95% CI: 0.94, 1.65; atezolizumab monotherapy vs regorafenib, HR = 1.39, 95% CI: 1.00, 1.94). ORRs were 2.7%, 2.2% and 2.2% with atezolizumab plus cobimetinib, atezolizumab monotherapy, and regorafenib, respectively.

Dr Bendell also reported that there were no statistically significant differences in OS by clinical or biomarker subgroups, including patients with MSS or extended RASmutation disease. Partial response was observed in two of three patients with MSI high status in the atezolizumab plus cobimetinib arm and in one of three patients with MSI high status in the atezolizumab monotherapy arm; however, patient numbers were too limited to draw conclusions.

No new safety signals were observed, and the safety profiles of atezolizumab plus cobimetinib combination and atezolizumab monotherapy were consistent with previous findings. Treatment-related Grade 3-4 AEs were reported in 45% of patients who received atezolizumab plus cobimetinib, 10% who received atezolizumab monotherapy, and 49% who received regorafenib. Treatment-related AEs of any grade with >30% occurrence were diarrhea (56%), rash (42%) and nausea (32%) with atezolizumab plus cobimetinib; none with atezolizumab monotherapy; and palmar-plantar erythrodyesthesia (51%), fatigue (43%), diarrhea (35%) and decreased appetite (34%) with regorafenib.

Dr Bendell suggested that the lack of clinical activity may be due to the immune-excluded phenotype of mCRC. “Dual inhibition of the PD-L1 immune checkpoint and MAPK-mediated immune suppression may not be sufficient to generate anti-tumor immune responses in immune-excluded tumors, such as mCRC,” she explained. The study group is continuing to perform biomarker evaluations, including gene expression analyses, she added.—Kara Rosania