Ex vivo 3D micro-tumor testing platform for predicting clinical response to platinum-based therapy in patients with high-grade serous ovarian cancer.

Abstract

e17555 Background: While over 200,000 women are diagnosed with high-grade serous ovarian cancer (HGSOC) annually, primary treatment has not changed for decades and consists of cytoreductive surgery and platinum-based therapy. However, around 20% of patients do not respond to this treatment. Unfortunately, predictive biomarkers for response to treatment are lacking. We present an innovative high throughput ex vivo 3D micro-tumor testing platform that predicts patient-specific response to different systemic treatments. Methods: Patients with HGSOC eligible for platinum-based neoadjuvant chemotherapy (NACT) were included in the study between 2019 and 2023 in the Netherlands (IRB P18.032, N=104). Micro-tumors enriched from ascites were embedded in hydrogel and exposed to carboplatin, paclitaxel and alternative therapies doxorubicin, gemcitabine, topotecan, olaparib and niraparib. Screening plates were imaged using a high-content 3D screening platform. Morphological features were extracted and fitted as dose-response curves. A linear regression model was trained on the AUCs of carboplatin and paclitaxel to predict the patient’s CA125 decay rates (N=38). Predicted CA125 decay rates were correlated to clinical outcomes (patient CA125 decay rate, change in target lesion size and progression-free survival). Throughput time was measured. Results: Ovarian cancer markers found in patient tumors were represented by isolated micro-tumors. CA125 decay rates predicted by ex vivo micro-tumor testing correlated to clinical patient responses with a correlation coefficient of 0.77 (R 2 =0.59). Moreover, predicted CA125 decay rates were significantly related to responses categorized according to gynaecological cancer intergroup (GCIG) guidelines (p<0.01 when comparing responders (CR/PR) to stable disease (SD)). Patients with high predicted sensitivity to NACT demonstrated significantly increased progression-free survival (p=0.02) and decreased target lesion size (R =-0.49). Complementary, patient-specific responses for second-line therapies were determined and presented in integrated reports. Sensitivity results could be generated within 2 weeks after sample collection. Conclusions: An ex vivo 3D micro-tumor testing platform has been established. Results are generated within 2 weeks after sample collection, aligning with the clinical time frame for treatment decision-making. The platform enabled prediction of clinical response to NACT in ovarian cancer patients with high accuracy (CR/PR vs SD/PD p < 0.01) and measured patient-specific responses to second-line therapies. In addition, patients with predicted high sensitivity to NACT demonstrated a significant survival benefit. Ongoing prospective evaluation of the clinical performance will further validate the application of the platform to support informed treatment decisions. Clinical trial information: P18.032 .