Atezolizumab Combo Falls Short in Phase III CRC Trial

The combination of atezolizumab and cobimetinib failed to demonstrate a significant benefit compared with regorafenib in a phase III study of heavily pretreated patients with advanced colorectal cancer.

Cathy Eng, MD

The combination of atezolizumab (Tecentriq) and cobimetinib (Cotellic) failed to demonstrate a significant benefit compared with regorafenib (Stivarga) in a phase III study of heavily pretreated patients with advanced colorectal cancer (CRC).1

Neither the combination regimen nor single-agent atezolizumab showed a statistically significant improvement in overall survival (OS) versus regorafenib, failing to meet the primary endpoint of the ongoing phase III IMblaze370 trial (NCT02788279). The median OS was 8.87 months (95% CI, 7.00-10.61) with atezolizumab and cobimetinib versus 7.10 months (95% CI, 6.05-10.05) for atezolizumab alone and versus 8.51 months (95% CI, 6.41-10.71) for regorafenib.

The PD-L1/MEK inhibitor combination regimen was previously studied in a phase Ib trial that demonstrated the potential antitumor activity of the treatment in patients with metastatic CRC. Atezolizumab and cobimetinib induced a median progression-free survival (PFS) of 1.9 months and a median OS of 10 months, with responses seen in 3 patients with microsatellite-stable (MSS) disease.2The positive findings led to the initiation of the phase III study.

“Despite the rationale supported by preclinical data, our results suggest that dual inhibition of the PD-L1 immune checkpoint and MAPK-mediated immune suppression is insufficient to generate antitumor immune responses in immune-excluded tumors, such as MSS metastatic CRC,” the study authors, led by Cathy Eng, MD, of The University of Texas MD Anderson Cancer Center, wrote in a report of the results of the phase III trial published inThe Lancet Oncology.

IMblaze370 is a multicenter, open-label, randomized, controlled trial that assessed atezolizumab with or without cobimetinib versus regorafenib in patients with unresectable, locally advanced or metastatic colorectal cancer. The study was conducted at 73 academic medical centers and community oncology practices across 11 countries.

Eligible patients had an ECOG performance status of 0 or 1 and disease progression on or intolerance of ≥2 prior systemic chemotherapy regimens containing fluorouracil, oxaliplatin, and irinotecan in the metastatic setting. Prior EGFR-targeted therapy was allowed, but prior immunotherapy, MEK/ERK inhibitors, and regorafenib were not.

A total of 363 patients were randomized 2:1:1 to either atezolizumab 840 mg by intravenous administration (IV) every 2 weeks plus cobimetinib 60 mg orally once a day for days 1 through 21 of a 28-day cycle (n = 183), atezolizumab monotherapy at 1200 mg IV every 3 weeks (n = 90), or regorafenib at 160 mg orally once a day for days 1 through 21 of a 28-day cycle (n = 90). Patients were further stratified byRASstatus and time since diagnosis of first metastasis (±18 months).

The number of patients with microsatellite instability (MSI) was capped at 5% of patients and the number ofRASwild-type patients was capped at 50%. Additionally, the investigators did not enrich enrollment for PD-L1 positivity.

Cobimetinib dose modifications or reductions were allowed for adverse events (AEs). Treatment was continued until loss of clinical benefit or unacceptable toxicity, and patients were even allowed to continue treatment despite radiographic progression if they wished; however, no crossover was allowed between the treatment arms.

The primary endpoint of the trial was OS for the combination regimen versus regorafenib and for atezolizumab monotherapy versus regorafenib. Secondary endpoints included investigator-assessed objective response rate (ORR), duration of response (DOR), and PFS by RECIST v1.1. Additional endpoints of quality of life and pharmacokinetic biomarker studies have not yet been reported.

Across the 3 treatment groups, the ages of patients ranged from 51 to 67 years and the majority of patients were male, white, and European. Most patients (approximately 69%) had gone 18 months or more since the diagnosis of their first metastasis and 65% of patients had liver metastases. Additionally, most patients (approximately 74%) had received no more than 3 lines of previous therapy in the metastatic setting, which typically included a prior targeted therapy.

The primary tumor site was the left side of the colon in more than half of patients across the 3 groups. Fifty-four percent of patients wereRASmutant and only 3% had aBRAFmutation. The majority of patients had MSS or MSI-low disease. PD-L1 expression, as assessed by the Ventana SP142 assay, was ≥1% in 34% to 43% of patients across the 3 groups, and PD-L1 expression was <1% in 44% to 47%.

Patients were followed for a median of 7.3 months (range, 3.7-13.6). The stratified hazard ratio (HR) for OS for the combination versus regorafenib was 1.00 (95% CI, 0.73-1.38;P= .99). In the comparison of atezolizumab monotherapy versus regorafenib, the HR for OS was 1.19 (95% CI, 0.83-1.71;P= .34).

In the intent-to-treat (ITT) population, the 6-month OS rate for atezolizumab and cobimetinib was 64.7% (95% CI, 57.6%-71.8%), 61.3% (95% CI, 50.9%-71.6%) for atezolizumab, and 63.5% (95% CI, 53.0%-73.9%) for regorafenib. At 1 year, the OS rates were 38.5% (95% CI, 31.2%-45.8%), 27.2% (95% CI, 17.6%-37.8%), and 36.6% (95% CI, 26.2%-47.0%) for the combination, atezolizumab alone, and regorafenib, respectively.

The median PFS was 1.91 months (95% CI, 1.87-1.97) with atezolizumab and cobimetinib, 1.94 months (95% CI, 1.91-2.10) with atezolizumab monotherapy, and 2.00 months (95% CI, 1.87-3.61) with regorafenib. The HR for PFS for the combination versus regorafenib was 1.25 (95% CI, 0.94-1.65) and the HR for atezolizumab versus regorafenib was 1.39 (95% CI, 1.00-1.94).

The ORR in the ITT population was 3% in the combination group, 2% in the atezolizumab monotherapy group, and 2% in the regorafenib group, which consisted of all partial responses. Stable disease was achieved by 23%, 19%, and 32% of patients in the 3 groups, respectively, for a total disease control rate of 26%, 21%, and 34%, respectively.

“These results underscore the challenge of expanding the benefit of immunotherapy to patients whose tumors have lower baseline levels of immune inflammation, such as those with microsatellite-stable metastatic colorectal cancer,” Eng et al wrote. “The immunomodulatory effects of the addition of a MEK inhibitor to an anti—PD-L1 [agent] might have not been adequate to overcome the immune resistance in microsatellite-stable metastatic colorectal cancer, a tumor type known for its non-inflamed, nonimmunogenic phenotype.”

The DOR was 11.4 months (95% CI, 2.7-not estimable) with atezolizumab and cobimetinib, 4.8 months (95% CI, 3.8-5.8) with atezolizumab alone, and 9.2 months (95% CI, not estimable) with regorafenib.

Median duration of treatment for cobimetinib was 59.0 days (range, 42.0-112.0) and 1.9 months (range, 1.4-3.7) for atezolizumab in the combination arm, 1.4 (range, 1.0-3.5) months for atezolizumab alone, and 51.5 (range, 42.5-161-5) days for regorafenib.

A post-hoc analysis looked at ORR according to patient subgroups. ThreeRAS-mutant patients achieved a response, one in each group. Additionally, half of all MSI-high patients (n = 6) achieved a response, and 4 patients with high PD-L1 expression (n = 145) achieved a response, 3 from treatment with the combination and 1 from atezolizumab alone. In the subgroup of patients with PD-L1—low tumors (n = 166), regorafenib outperformed the other treatment arms in terms of both OS and PFS.

The study authors noted that “this study was not designed to assess the combination in different biomarker subsets, which might have been more sensitive to this combination.”

According to further analyses looking at OS in the patient subgroups, younger patients tended to favor treatment with the combination arm (HR, 0.84) whereas older patients favored regorafenib alone (HR, 1.50). Improved survival with regorafenib was also negatively impacted by patients who had a larger number of prior treatments (HR, 1.58). In addition, PD-L1—high expression slightly favored the combination (HR, 0.80) whereas PD-L1–low expression slightly favored regorafenib (HR, 1.26).

Fourteen patients who were randomized did not receive any treatment and were not included in the safety analysis. All-cause AEs were observed in 99% of patients who received the combination regimen, 92% who received atezolizumab alone, and in 98% of patients who received regorafenib. Grade 3/4 events were seen in 61% of patients with combination treatment, 31% with atezolizumab, and 58% with regorafenib. The most common grade 3/4 AEs in the combination regimen arm were diarrhea in 11%, anemia in 6%, increased blood creatine phosphokinase in 7%, and fatigue in 4%.

“The 2 drugs have very distinct AE profiles, and overlapping AEs such as diarrhea or colitis do not seem to be increased in frequency or severity by the combination,” Eng et al wrote.

Serious AEs were reported in 40% of patients in the combination group, 17% in the atezolizumab arm, and in 23% in the regorafenib group. The most common serious AE in the combination arm was pyrexia in 7% of patients. In the combination arm, 24% of patients had an AE of special interest that required a systemic corticosteroid compared with 9% in the atezolizumab-alone group and 10% in the regorafenib group. The most common AE requiring steroids was rash.

Twenty-one percent of patients discontinued from the atezolizumab plus cobimetinib group due to AEs, 4% discontinued in the atezolizumab monotherapy group, and 9% in the regorafenib arm. AEs led to death for 5 patients in the combination-regimen group, 2 of which were caused by treatment-related sepsis; and for 2 patients in the regorafenib group, 1 of which was due to treatment-related intestinal perforation. There were no fatal AEs with atezolizumab monotherapy in this cohort.


  1. Eng C, Kim TW, Bendell J, et al. Atezolizumab with or without cobimetinib versus regorafenib in previously treated metastatic colorectal cancer (IMblaze370): a multicentre, open-label, phase 3, randomised, controlled trial [published online April 16, 2019]. Lancet Oncol. doi: 10.1016/S1470-2045(19)30027-0.
  2. Bendell J, Bang YJ, Chee CE, et al. A phase Ib study of safety and clinical activity of atezolizumab (A) and cobimetinib (C) in patients (pts) with metastatic colorectal cancer (mCRC). J Clin Oncol. 2018;35(suppl 4; abstr 560). doi: 10.1200/JCO.2018.36.4_suppl.560.