Back to search

BEHANDLING-God og treffsikker diagnostikk, behandling og rehabilitering

AI for new signatures and models for tailored organ preservation approaches in laryngeal and hypopharyngeal cancer

Alternative title: Kunstig intelligens for nye signaturer og modeller for skreddersydd organbevaring ved kreft ii strupehode og svelg

Awarded: NOK 3.1 mill.

Locally advanced voice box lower throat cancer may be treated with induction chemotherapy, potentially followed by radiotherapy for voice box preservation, as an alternative to total excision of the voice box. However, not all patients benefit from this strategy, and up to 30-40% have a total excision; we miss biomarkers for patient selection and treatment optimization, as well as for the exploitation of new therapeutic options. In this project, we will personalize patient management to increase the rate of preservative treatment in these types of cancer, by maximizing the probability of response to chemotherapy. Main objectives are to assess features that may be predictive of response to chemotherapy, and to define alternative treatment options to be tackled in patients non-responding to chemotherapy. The Preserve project will collect and integrate a large series of clinically annotated data from these types of cancer patients receiving preserving therapy, to assess a molecular features of response to such treatment and to define alternative therapeutic routes. Transcriptomic analysis, molecular data on cell lines and medical imaging evaluation will be main among the main features assessed. Our predictive models integrated into an intuitive clinical decision support system will be validated in a phase II feasibility trial with 49 patients treated with tailored preservative treatments, according to the features identified, providing evidence for clinical translation. We will implement ethical and legal frameworks for FAIR (open) data management and transnational data exchange. A cost-utility analysis of our personalized treatment in these cancer types, integrating quality of life measures will assess sustainability of precision medicine in aclinical setting. We have involved patients’ associations for the quality of life measurements, evaluation of our approach and wide dissemination of study results.

Locally advanced laryngeal (LAR) and hypopharyngeal (HYPO) squamous cell carcinoma may be treated with induction chemotherapy (IC) followed (in case of response > 50%) by radiotherapy (RT) for larynx preservation (LP) as an alternative to total laryngectomy (TL). However, not all patients (pts) benefit from LP strategy and up to 30-40% have a TL; we miss biomarkers for pts selection and treatment optimization, as well as for the exploitation of new therapeutic options. The current proposal will personalize pts management to increase the rate of LP in LAR/HYPO cancer, by maximizing the probability of response to induction treatment. Main objectives are to assess a multimodal signature predictive of response to IC and to define alternative pathways to be tackled in pts non-responding to IC. Preserve will collect and integrate a large series of clinically annotated data from LAR/HYPO cancer pts treated with IC followed by RT, to assess a multi-omic signature of response to IC and to define alternative pathways. Transcriptomic analysis, molecular data on cell lines and radiomic evaluation will be main components of this signature. Our predictive models integrated into an intuitive clinical decision support system will be validated in a phase II feasibility trial with 49 pts treated with tailored systemic induction treatments, according to the discovered signature, providing evidence for clinical translation. Main endpoints of the trial are: overall response rate to tailored induction treatment in at least 80% of the pts, with overall grade > 3 toxicities in less than 20% of the pts. We will implement ethical and legal frameworks for FAIR data management and transnational data exchange. A cost-utility analysis of our personalized treatment in LAR/HYPO cancer integrating QoL measures will assess sustainability of PM in clinical setting. We have involved patients’ associations for QoL measurement, evaluation of our approach and wide dissemination of study results.

Funding scheme:

BEHANDLING-God og treffsikker diagnostikk, behandling og rehabilitering