Patient-derived organoid pharmacotyping as a predictive tool for therapeutic selection in pancreatic ductal adenocarcinoma

Objective: We integrate a new approach to chemosensitivity data for clinically-relevant regimen matching, and demonstrate the relationship with clinical outcomes in a large PDO biobank. Summary Background Data: Pancreatic ductal adenocarcinoma (PDAC) usually recurs following potentially curative resection. Prior studies related patient-derived organoid (PDO) chemosensitivity with clinical responses. Methods: PDOs were established from pre-treatment biopsies in a multi-institution clinical trial (n=21) and clinical specimens at a high-volume pancreatectomy center (n=74, of which 48 were pre-treated). PDO in vitro chemosensitivities to standard-of-care chemotherapeutics (pharmacotypes) were matched to potential clinically-relevant regimens by a weighted nearest-neighbors analysis. Clinical outcomes were then compared for patients who had well-matched versus poorly-matched treatment according to this metric. Results: Our function matched 91% of PDOs to a standard-of-care regimen (9% pan-resistant). PDOs poorly-matched to the neoadjuvant regimen received would have matched to an alternative in 34% of cases. Patients receiving neoadjuvant chemotherapy well-matched to their pharmacotype experienced improved CA 19-9 response (60% decreased to normal when well-matched, 29% when poorly-matched, p<0.05) and lymph node down-staging (33% N0 after poorly-matched, 69% after well-matched, p<0.05). Patients receiving both well-matched neoadjuvant and adjuvant chemotherapy experienced improved recurrence-free- and overall survival (median RFS 8.5 months poorly-matched, 15.9 months well-matched, p<0.05; median OS 19.5 versus 30.3 months, p<0.05). Conclusion: In vitro PDO pharmacotyping can inform PDAC therapy selection. We demonstrate improved outcomes including survival for patients treated with regimens well-matched to their PDO chemosensitivities. A subsequent prospective study using PDO pharmacotype matching could improve oncologic outcomes and improve quality of life by avoiding therapies not expected to be effective.