We explored the value of a currently developed multi-treatment patient-stratification technique that allows individual-based oncologic treatment selection through the determination of the most potent treatment available (including non-approved treatments) based on tumour sample cultures. In the absence of clinical data, modelling allowed us to generate evidence of the potential value of the technique in different situations (cancer localisation, treatment line, effect, treatment costs).

In each indication, we began by modelling the current best of care (probabilistic treatment) based on pivotal trials, clinical guidelines, medical resource consumptions, costs and patients’ quality of life. Scenarios including tool efficacy, cost, response rate in patients receiving individual treatment and impact of response on survival were then tested for every cancer localisation.