Gastrointestinal (GI) cancers present a significant challenge in oncology due to highly heterogeneous treatment responses. This heterogeneity makes therapy selection particularly difficult in relapsed or refractory (R/R) settings. Standard genomic profiling often fails to capture the full functional complexity of tumor biology. Consequently, clinical outcomes remain variable and unpredictable.
New research presented at ESMO Asia 2025 evaluates a functional approach to this problem. The study assesses the feasibility of the Optim.AI™ platform for drug sensitivity testing in gastrointestinal cancers. This method aims to generate individualized drug-response profiles to guide treatment when standard options are limited.
The Role of Functional Precision Medicine
Functional precision medicine offers a complementary strategy to genomic profiling. Instead of relying solely on genetic markers, this approach measures the ex vivo response of patient-derived tumor cells to therapeutic agents directly. This phenotypic readout creates a specific drug-response profile for each patient.
The study utilized the Optim.AI™ platform, which integrates functional drug sensitivity testing (DST) with data-driven analytics. This system identifies effective single agents or combinatorial treatments. The platform quantifies drug activity using normalized cell viability (NCV) values. A lower NCV indicates higher sensitivity to the drug.
Methodology: From Biopsy to Prediction
The research team analyzed 14 gastrointestinal cancer cases. The cohort included seven colorectal, five gastric, and two pancreatic cancer samples. Researchers collected various specimen types, including resections, biopsies, and ascitic fluid.
The workflow followed a specific protocol:
-
Tumor Isolation: The team isolated live tumor specimens to generate Patient-Derived Organoids (PDOs).
-
Ex Vivo Screening: These organoids were directly co-cultured with a custom panel of 12 cancer drugs.
-
AI Analysis: A proprietary small-data AI platform processed the cell-viability data.
This process allows for the rapid identification of cytotoxic effects on cancer cells. The study highlights that the average turnaround time from specimen receipt to report generation was only 6.6 days.
Key Findings in Drug Sensitivity
The study demonstrated that drug sensitivity testing in gastrointestinal cancers can successfully identify effective salvage therapies. The analysis revealed diverse sensitivity spectra across the evaluated cancers.
Notably, Gemcitabine-based combinations emerged as the most common top-ranked treatments among the samples. The platform predicted potentially effective doublets (two-drug combinations) that showed a mean normalized cell viability (NCV) of less than 0.5.
Case Studies in Personalized Medicine
The results provided concrete examples of how personalized medicine applies to refractory cases.
In one instance involving a colorectal cancer patient who relapsed after FOLFOX and FOLFIRI treatments, the platform identified alternative options. Gemcitabine-based combinations showed higher ex vivo sensitivity compared to next-line standard-of-care agents like Fruquintinib and Regorafenib.
Similarly, for a multiply refractory gastric cancer patient, the testing highlighted Mitomycin C-based treatment. This alternative demonstrated greater sensitivity than the available standard-of-care options.
Clinical Implications
This study confirms the feasibility of using functional DST in GI cancers. The platform provides interpretable drug sensitivity profiles within a practical timeframe of roughly one week.
Identifying tumor-specific drug activity is crucial when disease progression limits standard options. While further clinical validation is ongoing to establish concordance, these findings suggest that functional DST holds promise for guiding treatment decisions in complex GI cancer cases.