Geriatric analysis from PRODIGE 20 randomized phase II trial evaluating bevacizumab + chemotherapy versus chemotherapy alone in older patients with untreated metastatic colorectal cancer

Older patients have frailty characteristics that impair the transposition of treatment results found in younger patients. Predictive factors are needed to help with treatment choices for older patients. The PRODIGE 20 study is a randomized phase II study that evaluated chemotherapy associated with bevacizumab (BEV) or not (CT) in patients aged 75 years or older. Patients underwent a geriatric assessment at randomization and at each evaluation. The predictive value of geriatric and oncologic factors was determined for the primary composite end-point assessing safety and efficacy of treatment (BEV or CT) simultaneously and also progression-free survival (PFS) and overall survival (OS). 102 patients were randomized (51 BEV and 51 CT; median age 80 years [range 75–91]). On multivariate analysis, baseline normal independent activity of daily living (IADL) score and no previous cardiovascular disease predicted the primary end-point. High (versus low) baseline Köhne score predicted short PFS and baseline Spitzer quality of life (QoL) score 2 LN levels above normal and high baseline Köhne score predicted short OS. Survival without deteriorated QoL and autonomy was similar with BEV and CT. On subgroup analyses, the benefit of bevacizumab seemed to be maintained in patients with baseline impaired IADL or nutritional status. Normal IADL score was associated with a good efficacy and safety of both BEV and CT. Köhne criteria may be relevant prognostic factors in older patients. Adding bevacizumab to chemotherapy does not impair patient autonomy or QoL. •A normal independent activity of daily living is predictive of efficacy of front line chemotherapy.•Köhne criteria for the first time is validated as a predictor for overall survival in older patients.•Adding bevacizumab to chemotherapy does not impair older patient's autonomy or quality of life.