Agenus Inc. has announced that three-year survival findings from the Phase 1b C-800-01 study of botensilimab plus balstilimab in microsatellite-stable metastatic colorectal cancer will be presented at the ESMO Gastrointestinal Cancers Congress 2026. The update will provide another year of follow-up from 123 heavily pretreated patients without active liver metastases, extending a dataset that previously showed a 42% two-year overall survival rate and median overall survival of 20.9 months.
Why three-year follow-up could matter more than another response-rate update in refractory MSS colorectal cancer
The most important feature of the forthcoming presentation is not simply that the dataset has matured for another year. It is that longer follow-up could test whether the survival curve is developing the durable tail typically associated with successful cancer immunotherapy. That distinction matters in microsatellite-stable metastatic colorectal cancer because conventional checkpoint inhibitors have generally failed to produce meaningful or sustained activity in this immunologically resistant setting.
The earlier analysis of the 123-patient cohort showed a confirmed objective response rate of 20%, a disease control rate of 69% and a median duration of response of 16.6 months. Those figures were accompanied by a 42% probability of survival at two years. A three-year update could therefore reveal whether the patients alive at two years continued to derive prolonged benefit or whether the survival curve declined more sharply with additional follow-up.
That question is clinically more consequential than a modest movement in response rate. Patients receiving late-line colorectal cancer treatment often achieve temporary disease stabilisation without long-term survival, while durable immune-mediated control remains uncommon. Evidence of a sustained survival plateau would strengthen the hypothesis that botensilimab and balstilimab are altering the natural history of disease for a defined group of patients rather than merely delaying progression for several months.
However, survival maturity alone does not resolve the limitations of the evidence. C-800-01 is an early-phase, open-label study without a randomized control group, and its primary purpose was to establish safety, tolerability and appropriate dosing rather than demonstrate comparative survival. Longer observation can make the signal more credible, but it cannot independently prove that the treatment caused the apparent survival advantage.
Why the no-active-liver-metastases cohort remains both the clearest signal and the largest evidence gap
The restriction to patients without active liver metastases is central to interpreting the BOT+BAL programme. The strongest responses to botensilimab plus balstilimab have been concentrated in patients whose disease is outside the liver or whose previous liver metastases are no longer active. This is biologically plausible because the liver can create a highly immunosuppressive environment that weakens systemic T-cell activity and reduces the effectiveness of immune checkpoint blockade.
That biological distinction could allow Agenus to identify a clinically meaningful population that has historically been treated as part of a much broader MSS colorectal cancer category. Selecting patients by metastatic site would represent a different form of precision oncology, based on the immune characteristics of the organs involved rather than solely on tumour mutations or protein expression.
The same selection strategy also creates the programme’s most important generalisability problem. Patients without active liver metastases tend to have a more favourable prognosis than those with extensive liver involvement, and liver metastases are common in advanced colorectal cancer. The apparent survival advantage cannot therefore be compared directly with results from unselected late-line populations without adjusting for metastatic pattern, baseline disease burden, performance status and previous treatments.
The Phase 1b results may ultimately support a treatment option for a narrower subgroup rather than the entire MSS metastatic colorectal cancer population. That would still be clinically valuable, particularly because the non-liver-metastatic subgroup has few effective immunotherapy options. It would, however, reduce the immediately addressable population and require clinicians to define active liver involvement consistently in routine practice.
How BOT+BAL compares with established late-line treatments without overstating a single-arm trial
The historical context explains why the BOT+BAL survival findings have attracted attention. In the randomized SUNLIGHT study, trifluridine and tipiracil combined with bevacizumab produced median overall survival of 10.8 months in refractory metastatic colorectal cancer. Fruquintinib achieved median overall survival of 7.4 months in the FRESCO-2 trial, while regorafenib produced median overall survival of 6.4 months in the CORRECT trial.
The previously reported BOT+BAL median overall survival of 20.9 months appears substantially longer than those results. Its 42% two-year survival rate also suggests that a proportion of treated patients may experience unusually prolonged benefit. The 20% objective response rate is notable because established late-line treatments frequently generate low single-digit response rates and rely primarily on disease stabilisation.
These comparisons are hypothesis-generating rather than definitive. The approved-treatment studies were randomized Phase 3 trials involving differently selected populations, while the BOT+BAL result came from a non-randomized cohort that excluded active liver metastases. Differences in eligibility, prognostic characteristics, subsequent treatment and assessment schedules could explain part of the apparent gap.
BOT+BAL would also enter a treatment environment with multiple established options rather than an empty market. Trifluridine and tipiracil plus bevacizumab has become an important late-line standard, while fruquintinib and regorafenib offer orally administered alternatives. A two-antibody intravenous immunotherapy regimen would need to deliver sufficiently durable survival and quality-of-life benefits to justify additional infusion visits, immune-related toxicity monitoring and potentially higher treatment costs.
The meaningful commercial comparison will therefore not be based on headline median survival alone. It will depend on randomized overall survival, the proportion of patients alive after two and three years, treatment-free intervals, symptom control, hospital utilisation and the ability to identify likely beneficiaries before therapy begins.
Why the BATTMAN Phase 3 trial must turn a promising survival tail into randomized evidence
The global BATTMAN study is now the decisive test of the BOT+BAL hypothesis. The Phase 3 trial is expected to enroll approximately 830 patients with refractory, unresectable MSS or mismatch repair-proficient colorectal cancer across more than 100 sites in Canada, France, Australia and New Zealand. Participants are being randomized to botensilimab plus balstilimab and best supportive care or best supportive care alone, with overall survival serving as the primary endpoint.
This design addresses the largest weakness of the earlier development programme by directly testing whether BOT+BAL helps patients live longer. It also allows the treatment effect to be assessed in the overall population and in clinically important subgroups defined by the presence or absence of liver metastases. That subgroup analysis will be especially important because the most mature survival signal currently comes from patients without active liver disease.
The randomized design also reduces the risk that patient selection, performance status or access to subsequent therapies is driving the apparent benefit. Patients entering BATTMAN must have exhausted or been unable to tolerate available treatments, but they are still required to have relatively preserved performance status and organ function. The resulting population may therefore remain healthier than many patients seen in everyday late-line practice.
A best-supportive-care control can produce a clear estimate of survival benefit after available agents have been exhausted. It may nevertheless create operational challenges because treatment availability and sequencing differ across participating countries. Investigators will need to document prior exposure to trifluridine and tipiracil, bevacizumab, fruquintinib, regorafenib and molecularly targeted therapies to ensure that the control group represents a genuinely refractory population.
Regulatory discussions have already shifted the programme away from relying on response rate as a surrogate for approval and toward randomized survival confirmation. The removal of a botensilimab monotherapy arm has simplified the Phase 3 study, but the central regulatory requirement remains unchanged. BOT+BAL must demonstrate that the combination produces a clinically meaningful overall survival improvement that outweighs its toxicity and treatment burden.
What the BOT+BAL safety profile means for patients who have already exhausted multiple therapies
The broader Phase 1 colorectal cancer experience showed treatment-related adverse events in 89% of patients. Fatigue, diarrhoea and fever were among the most frequently observed events, while 12% of treated patients discontinued therapy because of a treatment-related adverse event. No treatment-related grade 5 events were reported in the peer-reviewed analysis, and subsequent updates have not identified a new safety signal.
The absence of unexpected toxicity is reassuring, but immune-mediated gastrointestinal toxicity remains particularly relevant in colorectal cancer. Patients may already have altered bowel function, previous abdominal surgery, peritoneal disease or complications from extensive chemotherapy. Diarrhoea and immune-mediated colitis can therefore affect hydration, nutrition, hospitalisation risk and willingness to remain on treatment.
Botensilimab is designed as an Fc-enhanced anti-CTLA-4 antibody, while balstilimab blocks PD-1. Combining these mechanisms may be necessary to generate immune activity in a cold tumour, but it also increases the need for experienced immune-toxicity management. Early recognition, corticosteroid treatment and selective immunosuppression must be available across participating centres if the regimen is eventually adopted beyond specialist oncology institutions.
Treatment convenience will also influence adoption. The Phase 3 regimen gives botensilimab every six weeks for a limited number of doses, while balstilimab continues every three weeks until progression. That schedule could be acceptable if long-term survival is accompanied by manageable toxicity and periods of durable control. It would be more difficult to justify if the benefit is limited to a small response subset that cannot be identified in advance.
Why compassionate access and manufacturing readiness cannot substitute for regulatory proof
BOT+BAL is already being provided through authorized access mechanisms in selected countries, including reimbursed hospital access in France and paid named-patient pathways elsewhere. These programmes indicate substantial physician interest and provide Agenus with experience in treatment delivery, pharmacovigilance, manufacturing and international supply before any broad commercial launch.
Early access may also generate supportive real-world evidence on treatment duration, safety and outcomes outside conventional clinical trials. Such evidence could help regulators and health technology assessment bodies understand how the combination performs across different healthcare systems. It may also expose operational issues involving infusion capacity, reimbursement, immune-toxicity management and the cost of prolonged balstilimab treatment.
These programmes do not provide a substitute for randomized evidence. Patients receiving medicines through compassionate or named-patient pathways are highly selected, follow-up can vary and outcomes may be influenced by differences in supportive care. Real-world experience can strengthen the overall evidence package, but it cannot resolve whether survival would have been longer without treatment or with another late-line therapy.
Manufacturing readiness presents a related challenge. A successful BOT+BAL launch would require reliable production of two biologic medicines, consistent global release testing and adequate inventory for both clinical trials and commercial supply. The programme’s expanding manufacturing partnerships reduce some execution risk, although scalability will only become commercially meaningful if BATTMAN produces a sufficiently strong treatment effect to support approval and reimbursement.
What the ESMO GI 2026 poster must disclose to make the three-year update clinically meaningful
The headline three-year survival percentage will attract the most attention, but several supporting details will determine whether the result changes the programme’s clinical credibility. The poster needs to show the number of patients still at risk at each time point, the confidence intervals around landmark survival estimates, median follow-up and the extent of censoring. A high survival estimate based on a small remaining denominator would be less persuasive than a stable curve supported by substantial follow-up.
Updated duration-of-response data will also matter. The earlier median duration of response was 16.6 months, so the new analysis should clarify how many responders remain in remission, whether treatment has stopped and whether disease control continues after the final botensilimab dose. Evidence of persistent benefit after limited CTLA-4 exposure would support the proposed immune-memory mechanism.
The analysis should also describe outcomes by treatment line, metastatic location, performance status, tumour burden and previous therapy. Patients treated in the fourth line or later appeared to retain meaningful benefit in the previous analysis, but small subgroups can produce unstable estimates. Consistency across prespecified clinical categories would make the survival signal more credible and help define the population most appropriate for Phase 3 treatment.
Safety follow-up is equally important because immune-related events can occur late or require prolonged management. The absence of new signals would support continued development, while cumulative gastrointestinal toxicity, treatment discontinuation or long-term immunosuppression could affect the regimen’s benefit-risk profile.
Agenus has built a credible durability hypothesis, but Phase 3 remains decisive
The BOT+BAL programme has moved beyond an isolated response-rate signal. A 20.9-month median overall survival result, 42% two-year survival and responses lasting well beyond one year create a credible hypothesis that intensified CTLA-4 and PD-1 blockade can generate durable immune control in a subgroup of MSS metastatic colorectal cancer.
The three-year update could make that hypothesis considerably stronger, especially if the survival curve remains stable and the number of patients at risk remains meaningful. It could also reinforce the use of metastatic organ pattern as a practical biomarker for immunotherapy sensitivity.
The evidence still stops short of establishing a new standard of care. The Phase 1b cohort is selected, non-randomized and restricted to patients without active liver metastases. BATTMAN must now demonstrate that the survival signal persists against a concurrent control, remains clinically meaningful across important subgroups and is accompanied by acceptable safety and quality of life.
ESMO GI 2026 may therefore strengthen the rationale for BOT+BAL, but the conference update will not settle the programme’s future. The long-term value of botensilimab plus balstilimab will be determined by whether durable survival in an early cohort can be reproduced at Phase 3 scale.
