Predicting therapy efficacy and revealing tumor heterogeneity using patient-derived 3D bioprinted ovarian cancer models

Ovarian cancer is the most lethal gynecologic malignancy, characterized by tumor heterogeneity and a high recurrence rate. Patient-derived in vitro tumor models offer a promising strategy for individualized drug screening to overcome limitations of systemic therapy. Among existing modeling methodologies, 3D bioprinting exhibits advantages including high-throughput, high fidelity, and a drug screening timeline of 8 days. Here, we present experimental data of a cohort of novel 3D bioprinted patient-derived ovarian cancer (3DP-OC) models. We established 3DP-OC models by mixing primary ovarian cancer cells with Gelatin Methacryloyl (GelMA) and photoinitiator, and bioprinting in a layer-by-layer manner. In total, 3DP-OC from 79 patients were successfully established, including 61 high-grade serous ovarian cancer patients, 9 ovarian clear cell carcinoma patients, 4 ovarian sarcoma patients, 4 ovarian endometrioid carcinoma patients, and 1 ovarian neuroendocrine cancer patient. 113 3DP-OC models were constructed from different tissue origins (primary lesion and metastatic sites) with high cell viability maintained throughout bioprinting and prolonged in vitro culture. Bulk RNA sequencing and immunohistochemistry confirmed that key molecular markers and Ki-67 levels in 3DP-OC models closely resembled those of their paired tumor tissues, demonstrating that 3DP-OC can serve as a patient avatar for drug sensitivity testing. On days in vitro 5, 3DP-OC models were exposed to gradient concentrations of 15 frequently used chemotherapeutic and targeted drugs in ovarian cancer including paclitaxel, carboplatin, olaparib, etc . Cell viability was quantified to calculate IC 50 values of different anti-tumor drugs. Drug sensitivity testing revealed substantial interpatient heterogeneity in therapeutic responses. To further investigate whether this response heterogeneity has clinical relevance, we conducted a prospective observational cohort study that enrolled 41 stage III/IV newly diagnosed ovarian cancer patients. Patients were divided into “3DP-OC identified sensitive group” or “3DP-OC identified resistant group” according to 3DP-OC IC 50 values of anti-tumor drugs they received. The median follow-up time for all patients was 580.5 days. The 3DP-OC identified sensitive group exhibited significantly prolonged progression-free survival compared to the 3DP-OC identified resistant group ( P < 0.05), with disease progression observed in 16.7% (4/24) and 52.9% (9/17) of patients, respectively. In this study, we established a 3D bioprinted patient-derived cancer model with high success rate of establishment, low intra-batch heterogeneity, and high biological fidelity. 3DP-OC model holds promise for predictive utility in precision oncology as well as the potential to serve as an innovative platform that bridges fundamental cancer research and clinical practice.