1197P - Ex vivo basket study reports patient-specific sensitivity to carboplatin versus cisplatin in lung, ovarian and bladder cancer

Background

Platinum therapies cisplatin (CP) and carboplatin (CB) are the backbone of chemotherapy regimens. While the mechanism of action and survival rates are comparable, the toxicity profiles, response rates and administration differ. This causes clinicians to question when to favor CP or CB as best treatment for their cancer patients. Here, we aimed to identify patient-specific sensitivities to CP and CB in multiple indications using fresh ex vivo microtumors.

Methods

Fresh tumor tissue was collected from 212 patients (132 ovarian (OC), 49 bladder (BC) and 31 lung cancer (LC)). Microtumors with native tumor microenvironment were exposed to up to 12 chemotherapies and AUCs were calculated. 5 hierarchical factors were modelled: tissue sensitivity (f1), treatment strength (f2), tissue-specific treatment effects (f3), patient-specific overall chemosensitivity score (f4) and patient-specific drug sensitivity (f5). 46 samples (17 OC, 20 BC and 9 LC) were tested ex vivo for both CP and CB. Plasma CA125 levels (30 OC patients) and radiological responses (RECIST 1.1) (7 BC patients) were used for clinical correlation.

Results

The patient chemosensitivity score (f4) significantly correlated with CA125 decay in OC patients (r = 0.58, 95% CI 0.31-0.79, p < 0.001) and radiological response in BC patients. At tissue-level (f3), BC patients were more sensitive to CP (but not CB) compared to OC and LC patients (p = 0.05, Cohen’s d 1.9). Using f5 to extract personal treatment recommendations, we identified that 16/46 patients demonstrated significant (p <0.05) differences in CP vs CB treatment sensitivity, of which 56% (9/16) deviates from the general guidelines (67% in BC (6/9), 60% in OC (3/5), 0% in LC (0/2)).

Conclusions

This work presents an ex vivo basket study employing clinically validated functional tumor testing to identify multi-level responses to CP vs CB. Ex vivo tumor testing and integrated data analysis demonstrates that even if CP and CB perform on average equally well in large cohorts, differential tissue- and patient-specific responses are present. In this way, ex vivo tumor testing can provide personalized sensitivity profiles to support clinical decision making.

Clinical trial identification

IRB P18.032.

Editorial acknowledgement

Legal entity responsible for the study

The authors.

Funding

VitroScan.

Disclosure

E. Koedoot, D. van der Meer, L.J. Ceton, T. Sijsenaar, W. Vader: Financial Interests, Personal, Full or part-time Employment: VitroScan. All other authors have declared no conflicts of interest.