Abstract 7214: Real-world concordance analysis of a combinatorial functional precision medicine platform, Optim.AI™, in both solid and hematological cancers

Abstract PURPOSE: Functional precision medicine aims to expand the landscape of effective drug treatment options beyond standard of care through additional insights from ex vivo phenotypic drug response data. In our previous work, we have demonstrated the clinical feasibility of Optim.AI™, a combinatorial functional precision medicine platform, in identifying effective and resistant treatment options for non-Hodgkin lymphoma and acute myeloid leukemia (AML). This study further investigates the clinical utility and concordance of Optim.AI™ in predicting treatment sensitivity or resistance in a variety of hematological and solid cancers processed at KYAN’s clinical laboratory, with notable case studies highlighted. METHODS: Optim.AI™ utilizes efficiently designed experiments of less than 200 ex vivo combinations to interrogate over 530, 000 possible treatment permutations from a 12-drug panel. Optim.AI™ then ranks all clinically actionable treatments within the tested drug panel based on predicted sensitivity or resistance. In this study, we investigated 66 cases from a total of 20 cancer types (5 hematological and 15 solid cancer) and a range of tissue sample sources, including bone marrow aspirate, peripheral blood, tissue biopsy or resections, and pleural or peritoneal fluid. Retrospective concordance analysis was performed by evaluating Optim.AI™-generated tumor cell viability results against the prior lines of treatment for each sample. Prospective data of patients administered with Optim.AI™-predicted treatment were also collected, where applicable. RESULTS: With an established viability cut-off of 0.5, Optim.AI™ demonstrated a correlation of 88% with prior therapies, which included a range of 36 targeted treatments and chemotherapy agents. Among the 37 solid cancer cases evaluated, 86.5% (p-value < 0.0001) demonstrated concordance to prior line(s) of treatment, while for hematological cancers, 89.7% of the 29 cases (p-value = 0.0005) were concordant with previously observed resistance. Additionally, platform-predicted treatments were associated with positive clinical responses in several cases of hematological and solid cancers, where prospective data were available. CONCLUSIONS: Our retrospective concordance results demonstrate the clinical applicability and robustness of Optim.AI™ in predicting sensitivity or resistance to cancer treatments. In line with previous prospective clinical studies, these findings illustrate the platform’s potential to positively impact treatment recommendations across a broad range of hematological and solid cancers. Future prospective studies to evaluate Optim.AI™-guided treatments are warranted to further confirm the predictive value of the platform and provide evidence for incorporating such functional assays in clinical oncology. Citation Format: Masturah Mohamed Abdul Rashid, Jhin Jieh Lim, Weng Tong Ho, Su Pin Choo, Donald Yew Hee Poon, Daryl Chen Lung Tan, William Ying Khee Hwang, Yvonne Loh, Jeffry Beta Tenggara, Chin Hin Ng, Hugo Saavedra, Edward Kai-Hua Chow. Real-world concordance analysis of a combinatorial functional precision medicine platform, Optim.AI™, in both solid and hematological cancers [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2025; Part 1 (Regular Abstracts); 2025 Apr 25-30; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2025;85(8_Suppl_1):Abstract nr 7214.