Zanzalintinib Plus Atezolizumab Misses OS End Point in Non–Liver Metastases Subgroup in Refractory mCRC

Treatment with zanzalintinib (XL092) in combination with atezolizumab (Tecentriq) failed to produce a significant overall survival (OS) benefit vs regorafenib (Stivarga) in the subset of patients with previously treated non-microsatellite instability (MSI)–high metastatic colorectal cancer (mCRC) who did not have active liver metastases (non–liver metastases; NLM), failing to meet a dual primary end point of the phase 3 STELLAR-303 trial (NCT05425940).¹

Data from the final OS analysis in this subgroup showed a non-statistically significant trend favoring the combination in the NLM subpopulation (stratified HR, 0.83; 95% CI, 0.66–1.05; P = .1185), with a median OS of 15.9 months with zanzalintinib plus atezolizumab vs 12.7 months with regorafenib.

In the pivotal analysis of the ITT population previously reported and published in The Lancet, zanzalintinib plus atezolizumab demonstrated a statistically significant reduction in the risk of death vs regorafenib (stratified HR, 0.80; 95% CI, 0.69-0.93; P = .0045) with a median OS of 10.9 months (95% CI, 9.9-12.1) vs 9.4 months (95% CI, 8.5-10.2) at a median follow-up of 18.0 months (interquartile range, 14.6–21.5).²

These findings supported a new drug application (NDA) submission for zanzalintinib plus atezolizumab for the treatment of adult patients with mCRC who have been previously treated with fluoropyrimidine-, oxaliplatin-, and irinotecan-based chemotherapy and, if RAS wild-type, an anti-EGFR therapy, which was accepted by the FDA in February 2026.³ The FDA's review of the NDA for zanzalintinib in mCRC is ongoing, with a Prescription Drug User Fee Act (PDUFA) target action date of December 3, 2026.

How was the STELLAR-303 trial designed?

STELLAR-303 was a global, multicenter, randomized, open-label phase 3 trial conducted at 121 centers in 16 countries.² Between September 7, 2022, and July 15, 2024, 901 patients were enrolled onto the study and randomly assigned 1:1 to receive either zanzalintinib at 100 mg orally once daily plus atezolizumab at 1200 mg intravenously every 3 weeks (n = 451) or regorafenib at 160 mg orally once daily on days 1 through 21 of each 28-day cycle (n = 450). Patients were stratified by geographic region, RAS mutation status, and the presence or absence of liver metastases at baseline.

Eligible patients were 18 years of age or older with confirmed metastatic adenocarcinoma of the colon or rectum whose tumors were documented to not have MSI-high or mismatch repair–deficient status, and who had previously received standard-of-care therapy including fluoropyrimidine-, oxaliplatin-, and irinotecan-based chemotherapy, as well as an anti-EGFR therapy if RAS wild-type.

The trial's dual primary end points were OS in the ITT population, which comprised all randomly assigned patients regardless of liver metastasis status; and OS in the NLM subgroup, defined as patients who did not have active liver metastases at baseline per investigator assessment.^1,2 Key secondary end points included progression-free survival (PFS), objective response rate (ORR), and duration of response (DOR) in both the ITT population and the NLM subgroup.¹

What efficacy data from the ITT population of STELLAR-303 were previously reported?

Results from the planned final OS analysis in the ITT population were presented at the 2025 ESMO Congress and published simultaneously in The Lancet.^1,2 At a data cutoff of April 30, 2025, the 12- and 24-month landmark OS estimates in the ITT population were 46% (95% CI, 41%-51%) and 20% (95% CI, 15%-26%), respectively, with zanzalintinib plus atezolizumab. These respective rates were 38% (95% CI, 34%-43%) and 10% (95% CI, 6%-16%) with regorafenib.² OS benefit with the combination was consistently observed across prespecified subgroups, including geographic region, RAS mutation status, and prior anti-VEGF therapy.

In the ITT population, zanzalintinib plus atezolizumab demonstrated a median PFS of 3.7 months vs 2.0 months with regorafenib (HR, 0.68; 95% CI, 0.59-0.79); formal statistical significance for PFS could not be claimed under the prespecified hierarchical testing strategy. The ORR was 4% (95% CI, 2%-6%) with the combination vs 1% (95% CI, 0%-3%) with regorafenib; all responses were partial responses. The disease control rate was 54% (95% CI, 49%-58%) vs 41% (95% CI, 36%-45%), respectively.²

In the prespecified interim analysis of the NLM subgroup, the stratified HR was 0.79 (95% CI, 0.61–1.03; P = .087) at a median follow-up of 16.8 months, with a median OS of 15.9 months vs 12.7 months with regorafenib. These outcomes failed to meet the prespecified significance threshold for that interim analysis. The final NLM OS analysis confirmed the NLM subgroup did not achieve statistical significance at either time point.¹

What was the safety profile of zanzalintinib STELLAR-303?

The safety profile of zanzalintinib plus atezolizumab in the NLM subgroup was consistent with that previously reported in the ITT population, with no new safety signals identified. In the ITT population, any-grade treatment-related adverse effects (TRAEs) occurred in 95% of patients in the combination arm vs 92% in the regorafenib arm.² The most common any-grade TRAEs with the combination included diarrhea (50%), hypertension (34%), fatigue (33%), nausea (31%), and decreased appetite (30%).

Grade 3 or higher TRAEs occurred in 60% of patients receiving zanzalintinib plus atezolizumab (n = 446) vs 37% of patients receiving regorafenib (n = 434). Serious TRAEs occurred in 26% vs 10% of patients, respectively. Treatment-related deaths occurred in 5 patients in the combination arm, including 2 events of intestinal perforation attributable to zanzalintinib, 1 pneumonitis event and 1 renal failure event attributable to atezolizumab, and 1 case of altered consciousness attributable to the combination. Only 1 patient in the regorafenib arm experienced a treatment-related death, which was due to jejunal perforation.

Detailed findings from the STELLAR-303 NLM subgroup analysis are expected to be submitted for presentation at a forthcoming medical conference.

References

1. Exelixis provides update on the phase 3 STELLAR-303 trial evaluating zanzalintinib in combination with an immune checkpoint inhibitor in patients with metastatic colorectal cancer. News release. Exelixis, Inc. June 21, 2026. Accessed June 22, 2026. https://ir.exelixis.com/news-releases/news-release-details/exelixis-provides-update-phase-3-stellar-303-trial-evaluating
2. Hecht JR, Park YS, Tabernero J, et al. Zanzalintinib plus atezolizumab vs regorafenib in refractory colorectal cancer (STELLAR-303): a randomised, open-label, phase 3 trial. Lancet. 2025;406(10517):2360-2370. doi:10.1016/S0140-6736(25)02025-2
3. Exelixis announces U.S. FDA accepted the new drug application for zanzalintinib in combination with an immune checkpoint inhibitor for patients with metastatic colorectal cancer. News release. Exelixis, Inc. February 2, 2026. Accessed June 22, 2026. https://ir.exelixis.com/news-releases/news-release-details/exelixis-announces-us-fda-accepted-new-drug-application