LEXINGTON, Mass. — Agenus Inc. and Noetik reported retrospective data showing that Noetik’s artificial intelligence-based TARIO-2 model identified tumor microenvironment patterns associated with response and survival in patients treated with Agenus’ investigational BOT+BAL immunotherapy combination.
The data, scheduled for presentation May 30 at the 2026 American Society of Clinical Oncology Annual Meeting, evaluated whether TARIO-2 could analyze standard pretreatment hematoxylin and eosin pathology images to identify spatial tumor microenvironment patterns linked to clinical outcomes after treatment with botensilimab plus balstilimab.
Botensilimab, known as BOT, is Agenus’ investigational Fc-enhanced anti-CTLA-4 antibody. Balstilimab, known as BAL, is the company’s anti-PD-1 antibody. The combination is being studied as a potential treatment for solid tumors that have historically been difficult to treat with conventional immunotherapy.
The retrospective analysis included 113 efficacy-evaluable patients treated with BOT+BAL in the C-800-01 Phase 1b trial who had available pretreatment H&E images. The tumor cohorts included microsatellite stable metastatic colorectal cancer without active liver metastases, ovarian cancer and sarcomas.
In the MSS metastatic colorectal cancer subgroup without active liver metastases, TARIO-2 identified a patient group that had a 64% response rate to BOT+BAL, compared with 9% in the remaining cohort. Overall survival was also significantly improved in the AI-identified subgroup, with median overall survival not reached and a hazard ratio of 0.18 compared with the remaining cohort.
Agenus said TARIO-2 showed statistically significant predictive performance for both best overall response and overall survival in the MSS metastatic colorectal cancer cohort without active liver metastases. The company said supportive trends were also observed in ovarian cancer and sarcoma cohorts.
The model does not rely on a traditional single-marker biomarker approach. Instead, it uses AI-based spatial analysis of standard H&E pathology images, which are routinely generated during cancer diagnosis and clinical evaluation. Agenus said the approach could help support future patient stratification strategies if validated prospectively.
“Routine pathology images are already part of cancer care, but much of the biologic information they contain is difficult to interpret by eye alone,” said Ryan Dalton, Ph.D., Senior Computational Scientist at Noetik. “These data suggest that AI-based analysis of pretreatment H&E images may help identify spatial tumor microenvironment patterns associated with clinical benefit from BOT+BAL. The findings support prospective validation of TARIO-2 as a practical, image-based biomarker strategy.”
Agenus said BOT+BAL’s clinical activity has not been strongly associated with traditional biomarkers such as PD-L1 expression or tumor mutational burden, making broader tumor microenvironment-based approaches potentially important for identifying patients most likely to benefit.
“BOT+BAL is designed to engage the immune system in tumors that have historically been resistant to conventional immunotherapy, through differentiated mechanisms not fully captured by traditional biomarkers such as PD-L1 expression or tumor mutational burden,” said Dhan Chand, Ph.D., Vice President of Research at Agenus. “These data represent an important step toward aligning BOT+BAL’s differentiated biology with the patients most likely to benefit. Prospective validation will be an important next step as we continue to advance BOT+BAL clinical development.”
The companies said the findings support prospective validation of TARIO-2 as an H&E-based biomarker strategy for BOT+BAL, including further evaluation in MSS colorectal cancer and broader solid tumor datasets.
