Zanzalintinib (XL092) alone or in combination with atezolizumab in patients (pts) with refractory metastatic colorectal cancer (mCRC): Results from an expansion cohort of the phase 1 STELLAR-001 study.

127 Background: VEGFR tyrosine kinase inhibitors (TKIs) in combination with immune checkpoint inhibitors (ICIs) have demonstrated clinical activity in pts with previously treated mCRC, particularly in pts without liver metastases (LM). Zanzalintinib (zanza) is a novel, oral, multitargeted TKI with activity against VEGFR, MET, and TAM kinases. Here, we present results from the randomized expansion cohort of pts with previously treated non-MSI-H/dMMR mCRC receiving zanza ± atezolizumab (atezo) in the phase 1 STELLAR-001 study (NCT03845166). Methods: Adults (aged ≥18 y) with locally advanced or metastatic CRC that was RAS -wildtype, who were refractory/intolerant to prior standard of care (5-FU–based treatment ± targeted therapies) and had an ECOG PS of 0/1, were enrolled. Pts with MSI-high or dMMR CRC, or who had received prior treatment with a PD-L1/PD-1 targeting ICI, regorafenib, and/or TAS-102, were excluded. Pts were randomized 1:1 to receive zanza+atezo (100 mg QD + 1200 mg Q3W) or single-agent zanza (100 mg QD). Objectives were to evaluate median (m)OS, investigator-assessed ORR, and mPFS per RECIST 1.1, and safety. Results: As of May 6, 2024, 107 pts were enrolled (zanza+atezo, n=54; zanza, n=53). In the zanza+atezo/zanza groups, median age was 61/60 years, 39%/51% had an ECOG PS of 1, pts had a median 2.5/3.0 (range, 1–5) prior lines of therapy, and 31%/32% did not have LM. With a median follow-up of 15 months across both arms, the mOS was 14.3 and 11.1 months for zanza+atezo and zanza, respectively (HR 0.75, 95% CI 0.45–1.26), confirmed ORR was 7.4% and 1.9% (all partial responses), mPFS was 4.0 and 3.0 months (HR 0.68, 0.44–1.04), and DCR was 59% and 55%. In pts without LM, mOS was not reached and 12.5 months (HR 0.46, 0.15–1.36) with 6-month survival rates of 88% and 65%, confirmed ORR was 18.0% and 5.9%, mPFS was 8.2 and 3.3 months (HR 0.40, 0.16–1.0), and DCR was 76% and 59%. The most common treatment-related adverse events (TRAEs) were diarrhea (zanza+atezo, 52%; zanza, 49%), nausea (54%, 36%), and decreased appetite (41%, 36%). Grade 3/4 TRAEs occurred in 48% of pts with zanza+atezo and 40% with zanza (Grade 4 in 3.7% and 0); a Grade 5 TRAE occurred in 1 pt (2%) in each group. 89% and 94% of pts had discontinued study treatment. Zanza was discontinued due to TRAEs in 11% and 8% of pts; atezo was discontinued due to TRAEs in 9% (zanza+atezo). Biomarker analysis is ongoing and results will be presented. Conclusions: Clinical activity was observed with zanza both as a single agent and in combination with atezo in this cohort of heavily pretreated pts with mCRC. Greater activity was seen with zanza+atezo versus zanza alone, particularly in pts without LM. Both treatments were generally well tolerated. Evaluation of zanza+atezo versus regorafenib is ongoing in the phase 3 STELLAR-303 study in non-MSI-H/dMMR mCRC (NCT05425940). Clinical trial information: NCT03845166 .