Superior Therapy Response Predictions For Patients With Pancreatic Cancer (Pdac) Using Cellworks Singula: Mycare-009-05.

e16766 Background: Current therapy for PDAC is modestly effective. Therapy selection guidelines are typically limited to single aberrations and ignore relevant patient-specific omics. These alterations create patient-specific resistance pathways that cancer cells use to resist one-mutation one-drug Physician Choice therapies. Singula™ utilizes the novel Cellworks Omics Biology Model (CBM), which predicts patient-specific biomarker and phenotype response of a personalized diseased cell to drug agents, radiation and cell signaling. We test the hypothesis that Singula is a more accurate predictor of patient-specific therapy response to targeted agents compared to Physician Choice. Methods: The performance of the Singula Classifier of patient therapy response was evaluated in an independent retrospective cohort of N = 52 Stage I - IV PDAC patients aged 35-85, whose omics data were included in TCGA and ICGC. The accuracy of Singula was compared to the accuracy of prescribed therapies and clinical outcomes. Comparisons in the accuracy were enabled using McNemar’s test to account for the correlation between Singula and physician recommendations. Logistic regression was used to model complete response (CR) as a function of age, prescribed therapies, and Singula against non-response (NR). Likelihood ratio tests were performed to assess if Singula provides predictive information beyond prescribed therapy and age alone. Similar analyses were performed for progression-free survival (PFS) and overall survival (OS) using proportional hazards regression. Results: Singula is a superior predictor of CR compared to Physician Choice (McNemar’s χ2 = 26.0, p-value < 0.0001), with an overall accuracy of 94.2% (84.1%, 98.8%) compared to 40.4% (27.0%, 54.9%) for Physician Choice. In regressions, Singula (p = 0.0023) remained a significant predictor of CR, but prescribed therapy (p = 0.3885) and age (p = 0.1148) were nonsignificant. Singula was also the only significant predictor of PFS (p = 0.0002) and OS (p = 0.0096). For all 28 true negatives, alternative therapy selections with predicted response were generated. Conclusions: Singula is a superior predictor of PDAC response to therapy compared to Physician Choice. The Singula report can validate therapy selection decision, potentially increasing median PFS and OS and provides alternative therapy selections for non-responders. Prospective validation is needed. [Table: see text]