The REQUITE project: Validating predictive models and biomarkers of radiotherapy toxicity to reduce side effects.

Abstract

276 Background: Recently the first replicated genetic associations for adverse reactions to radiotherapy (RT) have been reported. These will help to build predictive statistical models for optimising RT delivery or interventions to alleviate the side effects. It is now timely to start a project that aims to validate known predictors of adverse reactions and develop the statistical models to become clinically useful. REQUITE is a European Union funded FP7 project that aims to do this. Methods: REQUITE’s objectives: (1.) Perform a multi-centre, cohort study collecting: blood samples, epidemiology and treatment data, longitudinal side-effect and QOL data (before and after treatment, yrs 1 and 2). (2.) Produce a centralised biobank of DNA from 5,300 patients and a centralised data management system. (3.) Validate published biomarkers of radiosensitivity. (4.) Validate clinical predictors of RT toxicity in breast, prostate, and lung cancer and incorporate biomarker data. (5.) Design interventional trials to reduce long-term side effects. (6.) Provide a resource for dissemination and exploitation to the RT community. Results: REQUITE is funded for 60 mos and organised into 7 work packages. Overall management and scientific oversight is run by Manchester. The central activity of the project is a multi-centre, observational study. Enrolment will proceed for 2 yrs in 9 clinical centres, with 2 yrs follow-up. The primary endpoints are change in breast appearance; rectal bleeding (prostate); pneumonitis (lung). Blood samples will be collected before radiotherapy. Tracking, biobanking, DNA preparation and validation of biomarkers (genetic markers and apoptosis assays) as predictive factors will be carried out in WP3-4. Some clinical factors have suggested predictive value for RT side-effects, but there is no consensus. Validation of published models in existing cohorts will be performed, leading to replicated models that can be validated using the REQUITE cohorts. Conclusions: The predictive models will be used to design clinical interventional trials and produce protocols that seek to lower RT side-effects in those individuals at high risk of developing them without affecting tumour control. Dissemination will follow.