Positive Phase III Data For Refractory Colorectal Cancer — But Where Does Zanzalintinib Plus Atezolizumab Fit?

For nearly a decade, the conversation around third-line and beyond treatment in metastatic colorectal cancer has been, frankly, a difficult one to have with patients. Regorafenib and trifluridine-tipiracil (the two workhorses of the salvage setting) offered median overall survival figures that rarely exceeded 7 or 8 months, and came with toxicity profiles that further eroded quality of life in patients already depleted by multiple lines of chemotherapy. Another option was desperately needed, and the STELLAR-303 trial was designed to determine if zanzalintinib plus atezolizumab could be the answer.

While the trial data shows promise, with the combination leading to longer survival in the majority of patients, questions about sequencing, patient selection, and where exactly in the treatment course this regimen could be of best use still remain.

“The sequence after chemotherapy is not clear, but more treatment options beyond the second line is better for our patients,” Dr. Maen Abdelrahim, Section Chief of Gastrointestinal Medical Oncology at Houston Methodist’s Dr. Mary and Ron Neal Cancer Center, tells SurvivorNet Connect.

STELLAR-303 Shows Promise

STELLAR-303 was a global, randomized, open-label phase III trial enrolling 901 patients with previously treated metastatic colorectal adenocarcinoma across 121 centers in 16 countries.

The trial had a clear eligibility focus:

• Patients must have progressed on or been intolerant to fluoropyrimidine, irinotecan, and oxaliplatin (with or without anti-VEGF and anti-EGFR agents where applicable).
• Patients must have been documented MSI-H or dMMR negative, a criterion that excluded roughly 5% of the metastatic CRC population and defined the study’s target as the other 95%, the microsatellite stable majority for whom immunotherapy has historically offered nothing.

Patients were randomized 1:1 to zanzalintinib (100 mg oral daily) plus atezolizumab (1200 mg IV every 3 weeks) versus regorafenib (160 mg oral daily, days 1-21 of a 28-day cycle). The dual primary endpoints were overall survival in the intention-to-treat population and in the subset without liver metastases.

In the ITT population, zanzalintinib-atezolizumab demonstrated a statistically significant overall survival benefit over regorafenib:

• Median OS of 10.9 months versus 9.4 months, with a stratified hazard ratio of 0.80 (95% CI 0.69–0.93; p=0.0045). That translates to a 20% reduction in the risk of death.
• At 24 months, the OS estimates were 20% versus 10%, suggesting that the tail of the curve was meaningfully different.
• Progression-free survival also favored the combination: median 3.7 months versus 2.0 months, HR 0.68.

Formal statistical superiority for PFS cannot yet be claimed under the prespecified hierarchical testing plan, as the nlmITT OS endpoint (patients without liver metastases) had not yet met significance at the time of this analysis.

“The overall survival for the patient improved in zanzalintinib-atezo compared to the regorafenib arm,” Dr. Abdelrahim explains. “The hazard ratio for this trial was 0.8, meaning that there’s a 20% reduction in death in patients treated with zanzalintinib with atezo compared to rego.

“Also, the trial was positive for the secondary objective which is progression-free survival. The median PFS for rego was two months and for zanzalintinib plus atezo 3.8 months, with 30% reduction of death and disease progression.”

Why This Mechanism Is Different

Understanding why this combination appears to work where others have failed requires a closer look at zanzalintinib’s kinase inhibition profile. Zanzalintinib is a second-generation, multitargeted tyrosine kinase inhibitor that, like regorafenib, inhibits VEGF receptors (thereby suppressing angiogenesis). But it also targets MET, AXL, and the full TAM kinase family (TYRO3, AXL, MER). Those additional targets are not incidental.

“Zanzalintinib can modulate the microenvironment or the immuno-microenvironment of the tumor,” Dr. Abdelrahim notes. “That’s why this has a little bit of excitement over the other or the old TKI, because not only is it targeting VEGF and decreasing angiogenesis, but it also changes the microenvironment and makes it more susceptible … priming the immuno-microenvironment to immunotherapy.”

This mechanistic differentiation likely explains why STELLAR-303 succeeded where LEAP-017 did not. In LEAP-017, lenvatinib plus pembrolizumab failed to improve overall survival over regorafenib or trifluridine-tipiracil. Lenvatinib targets VEGF receptors and FGFR but does not inhibit the TAM kinases or MET in the same profile as zanzalintinib. Whether that difference is pharmacologically decisive is a hypothesis the data support but cannot definitively confirm, because the trial was not designed to deconvolute the contributions of each component.

The Sequencing Problem

Here is where the clinical debate gets substantive. The post-chemotherapy landscape in metastatic CRC no longer has one or two options. It has several, and no prospective head-to-head data to organize them. The currently available or guideline-referenced options in third-line and beyond include regorafenib, trifluridine-tipiracil (alone or with bevacizumab, as in SUNLIGHT), fruquintinib, and now potentially zanzalintinib-atezolizumab pending FDA review.

For specific molecular subgroups, targeted agents like HER2-directed therapies may also be relevant.

Dr. Abdelrahim highlights that while there’s no clear sequencing answer, having more options in later treatment lines gives patients more chances to manage their disease.

“Whatever you choose in the third line, you can choose fourth line and fifth line … each line actually gives more life,” he says, adding that “in some patients you can choose between chemo-based in the third line and beyond, or maybe a TKI and chemotherapy-free regimen. So it’s going to be case by case, based on molecular profiling. Also based on the patient performance status.”

Essentially, the debate is not really “regorafenib vs. zanzalintinib-atezolizumab vs. trifluridine-tipiracil-bevacizumab” as discrete competitors for a single slot. It is about building a treatment sequencing and the availability of a chemotherapy-free and immunotherapy-containing options, particularly for patients who are approaching the limits of their tolerance for cytotoxic therapy. The choice at each decision point should be driven by molecular profile, performance status, prior exposure, and the strategic goal of preserving options downstream.

“Based on molecular profiling and the patient performance status, if there are certain mutations and targeted therapies specifically for certain mutations or alterations, those will be implemented in the third line or maybe the second line depending on the case,” Dr. Abdelrahim adds.

The Benefit Of A Chemo-Free Option

The SUNLIGHT comparison is worth examining more carefully, even acknowledging the limitations of cross-trial comparisons. In SUNLIGHT, trifluridine-tipiracil plus bevacizumab produced a median OS of 10.8 months in the overall population, strikingly similar to the 10.9 months seen in STELLAR-303 with zanzalintinib-atezolizumab.

However, among the subset of SUNLIGHT patients who were bevacizumab-pretreated (the majority of real-world patients), median OS dropped to 9.0 months. In STELLAR-303, median OS among patients with prior anti-VEGF exposure was 10.5 months with zanzalintinib-atezolizumab (a potentially meaningful difference), though one that comes with all the usual caveats.

For clinicians sequencing these agents, a practical consideration is that zanzalintinib-atezolizumab is chemotherapy-free. In patients who have already been exposed to two lines of cytotoxic therapy, often FOLFOX or FOLFIRI-based regimens, followed by a second-line alternative, the option to shift entirely away from chemotherapy carries clinical weight.

“This is a chemo-free regimen,” Dr. Abdelrahim notes. “You have a targeted therapy with TKI inhibitors and immunotherapy, in a patient population that is microsatellite stable, which is around 95% of colorectal cancer patients. This is a way where we’re giving immunotherapy after priming the tumor microenvironment with a TKI that makes microsatellite stable tumors hot, and maybe have some response to immunotherapy. Otherwise those patients could be on single-agent or double-agent regimens with minimal outcome.”

The flip side of this should be considered as well. The combination carried a higher rate of serious adverse events than regorafenib. Grade 3 or worse treatment-related adverse events occurred in 60% of patients on zanzalintinib-atezolizumab versus 37% on regorafenib. There were five treatment-related deaths in the combination arm (versus one with regorafenib), including two from intestinal perforation attributed to zanzalintinib and one from pneumonitis attributed to atezolizumab.

These are not numbers to gloss over. The safety profile (hypertension, proteinuria, fatigue, diarrhea, rash, hepatotoxicity) is broadly consistent with other TKI-checkpoint inhibitor combinations and manageable with dose modifications, but it demands careful patient selection and monitoring.

“In general the regimen was well tolerated by most patients,” Dr. Abdelrahim says, “but as a class effect (this is a TKI) you expect to see a little bit of hypertension, some fatigue, proteinuria because we’re targeting the VEGF receptor pathway, sometimes diarrhea and GI symptoms. But mostly it’s low grade and well tolerated for most patients.”

Notably, palmar-plantar erythrodysesthesia (one of the most burdensome adverse effects of regorafenib) occurred far less frequently with the combination (16% any grade, 3% grade 3) than with regorafenib (50% any grade, 9% grade 3), which may be a meaningful quality-of-life advantage for certain patients.

Where Does This Land in Clinical Practice?

STELLAR-303 has not yet resulted in FDA approval. The data have been presented and published, and the regulatory pathway is in process. Until approval, the combination is not a standard-of-care option outside of clinical trial settings.

When it is approved, the question of where it slots into the sequence will remain genuinely open.

“The key is how those patient populations who are microsatellite stable (for whom immunotherapy is not part of their regimen in the earlier line of treatment) when we give them the opportunity to have immunotherapy in combination with targeted therapy and they have a response, this is an exciting time, this is an exciting treatment option,” Dr. Abdelrahim says.

He also raises the possibility, still speculative, that the regimen could eventually move earlier in the sequence for specific patient subsets, particularly those who cannot tolerate further chemotherapy.

“If a treatment did a good job in that setting and there’s an opportunity to bring it to an earlier line of treatment for somebody who might not be eligible for chemotherapy, this might be an option for them. Is it the guideline? Not yet. But in the future, clinical trials could be conducted to prove its efficacy in the earlier line of treatment,” he says.

For now, the practical framework is this: zanzalintinib-atezolizumab is a third-line or later option for patients with microsatellite stable metastatic CRC who have received prior fluoropyrimidine, irinotecan, and oxaliplatin. It offers a chemotherapy-free, immunotherapy-containing regimen with a statistically significant overall survival advantage over regorafenib and a toxicity profile that, while not trivial, is consistent with its drug class and manageable in an experienced center.

In a disease where the five-year survival rate for stage 4 cancer remains approximately 15%, the sequencing debate is a good problem to have.