For older cancer patients, Comprehensive Geriatric Assessment (CGA) is recommended in order to help the oncologist in his decision making. However, the implementation of the CGA in oncologic setting presents major limitations; The CGA is time consuming, costly in terms of resources and is not standardized. Moreover, recent studies show that the CGA, used as the gold standard, may have a ceiling effect in detecting vulnerability in older cancer patients. Several authors suggest that a more sensitive approach, using frailty markers may be a better way to detect potential vulnerability in older cancer patients. In this study, for each patient, a brief screening assessment, a full CGA and the assessment of frailty markers will be completed at inclusion. The brief screening assessment will be based on self report questionnaire (4 items of instrumental Activities of Daily Living + 2 items of nutritional assessment) and one physical measure (one-leg standing balance test). The CGA will be based on seven domains (and their assessment tools): functional status, comorbidities, objective physical performance, nutrition, cognition, depression, and social support. Five frailty markers (as described by Fried and al) will be evaluated: nutrition, mobility, energy, physical activity and grip strength. CGA and frailty markers will be completed at 3, 6, 12 and 18 months after the beginning of chemotherapy as well as oncologic criteria (treatment toxicities, treatment modification such as decrease or change or end of chemotherapy, percentage of chemotherapy dose received, cancer related death) and geriatric criteria for adverse outcomes (functional, nutritional or cognition decline, hospitalization or consultation with their general practitioner, death for other causes).