2298P - Feasibility of ex vivo drug sensitivity testing in urothelial cancer: EVITA trial

Background

Patients with bladder cancer (BC) are treated with chemotherapy, either intravesically or systemically. Predictive biomarkers to guide clinical decision-making on chemotherapy are lacking, leading to potential ineffective treatment and toxicity. We hypothesized that an established ex vivo tumor testing assay could identify chemoresponsive BC prior to treatment and we performed a prospective feasibility trial (OZBS62.21051).

Methods

Fresh tumor tissue was collected and characterized by histological evaluation of H&E slides and hotspot mutation analysis of hTERT, FGFR3 and PIK3CA genes. For ex vivo drug testing, tumor clusters were isolated, seeded, and exposed to single-agent chemotherapy: carboplatin, cisplatin, gemcitabine and targeted therapy (erdafitinib). Sensitivity evaluation was performed by extraction of morphological features from high-throughput 3D imaging data, followed by AUC comparison of fitted dose-response curves.

Results

Ex vivo drug testing was successful for 28 out of 48 patients (58%), generating results within two weeks. Absence of muscle-invasiveness was key in assay success, with higher success in non-muscle invasive BC (NMIBC) (17/20, 85%, (OR: 3.8, 95% CI: 1.7 – 9.1) than in muscle-invasive BC (11/28, 39%). The assay identified a subgroup of NMIBC patients with exceptional ex vivo gemcitabine response. Ex vivo sensitivity to FGFR inhibitor erdafitinib was observed in two patients with somatic FGFR3-Y375C mutations but not for FGFR3-WT patients (P = 0.03). Table: 2298P

Conclusions

We report the feasibility of ex vivo 3D tumor testing for BC. A correlation with clinical response to chemotherapy could not be established due to a lack of successfully screened patients with matching clinical chemotherapy treatment. For NMIBC, however, differential ex vivo sensitivity to gemcitabine, and a preliminary correlation for ex vivo erdafitinib sensitivity and FGFR3 mutation status offers a potential solution to select NMIBC patients for effective (novel) treatment options.

Clinical trial identification

Editorial acknowledgement

Legal entity responsible for the study

The authors.

Funding

PPP allowance: Health∼Holland (in cash), VitroScan B.V (in kind).

Disclosure

L. Ceton, D. van der Meer, F. Grillet, M. Garcia Montero: Financial Interests, Personal, Full or part-time Employment: VitroScan B.V. W. Vader: Financial Interests, Personal, Full or part-time Employment: VitroScan B.V; Financial Interests, Personal, Stocks/Shares: VitroScan B.V. J.L. Boormans: Financial Interests, Personal and Institutional, Financially compensated role, Consultancy: Janssen, Bristol Myers Squibb, AstraZeneca, Merck, MSD; Financial Interests, Institutional, Other, Research collaboration: Janssen, Merck. All other authors have declared no conflicts of interest.