The Comprehensive Geriatric Assessment (CGA) is, still today, barely established in the standard practice. Although the different scientific societies recommend a CGA in elderly population affected with cancer at first visit, the evidence regarding its transcendence is limited. On the other hand, the components of this evaluation and the translation of their findings in changes in the clinical practice are not stablished. This project intends to implement a CGA in a group of patients to compare this group with an historical cohort that reflects the standard practice, to evaluate the differences between both approaches. The study also evaluate the feasibility and utility of this CGA. All these questions aims to support the hypothesis that CGA impacts on the cancer treatment for elderly population affected with cancer.

This study compares two groups of elderly patients affected with stage IV non-small lung cancer (NSCLC), stage III colon cancer and stage IV colorectal cancer (CRC): a prospective experimental cohort and an historic control cohort. The sample size is 60 patients, 30 prospective and 30 retrospective. Of each cohort, 10 patients have stage IV NSCLC, 10 have stage III colon cancer and 10 have stage IV CRC. Prospective patients will be recruited longitudinally from January 2013 and retrospective patients are recorded longitudinally backward in time from December 2012. Both groups are patients of Medical Oncology Service of Hospital General Universitario de Elche, and both groups sign an informed consent form.

A CGA is performed in the experimental group after the first visit with oncologist and prior to therapeutic decision. This CGA includes:

• Nutritional Assessment: Mini Nutritional Assessment (MNA)
• Functional Assessment: Get up and Go, Activities of Daily Living (ADL), Instrumental Activities of Daily Living (IADL), Karnofsky Scale, Walking one Block, Number of Falls in last 6 months and Hearing Loss.
• Cognitive Assessment: Mini-Mental State Examination (MMSE-30)
• Psychological status: Geriatric Depression Scale (GDS)
• Social Support: Medical Outcomes Study Social Support Survey (MOS-SSS)
• Comorbidity and Severity of Comorbidities: Charlson Comorbidity Index and Adult Comorbidity Evaluation (ACE-27)

The evaluation includes too age, haemoglobin, Creatinine Clearance (CrCl) and presence of geriatric syndromes.

The results of this evaluation allow to calculate:

• Early mortality risk according to Soubeyran et al (JCO 2012, 30: 1829-1834) (MNA and Get up and Go):
• risk of early mortality
• no risk of early mortality.
• Toxicity to chemotherapy according to Hurria et al. (J Clin Oncol 2011, 29: 3457-3465) (age, type of cancer, dose and number of chemotherapy drugs, haemoglobin, CrCl, hearing loss, number of falls in last 6 months, ability to take medication without help (IADL), walking one block and social support (MOS-SSS):
• low risk
• medium risk
• high risk.
• Frailty group according Köhne et al (The Oncologist 2008, 13: 390-402) (ADL, IADL, comorbidity and severity of comorbidities and presence of geriatric syndromes). This information dictates the therapeutic strategy for each experimental patient:
• Group 1: standard treatment.
• Group 2: modified treatment
• Group 3: no treatment

Experimental patients will be followed for six months from CGA and their toxicity and its severity, occurrence of treatment modifications, number of hospitalizations and reasons for hospitalizations and mortality and reason of mortality will be compared with the same items of historic controls, followed too for six months from the moment the physician made the therapeutic decision.

Other data to record for both groups: sex, age, origin, leucocytes, platelets, LDH, albumin, CEA, Cr Cl, haemoglobin, subjective weight loss, corporal mass index, comorbidities, primary tumor: stage, metastases location, treatment intent: adjuvant, palliative, number of drugs, number of concomitant medications, PS according to ECOG and Karnofsky (some data are expected to be missing for historic group).

Only for historic controls, physicians should answer one question : In wich category would include the treatment recommendation? according to Köhne, The Oncologist 2008: Balducci, The Oncologist 2000; J Clin Oncol 2012:

• standard treatment
• modified treatment
• no treatment despite treatment is recommended
• no treatment

Only for historic controls two more questions should be answered:

• Has a Geriatric Evaluation been performed? Yes or No
• What sort of evaluation?

The trial records too how long does it take each test and how long does it take to complete CGA.

All information is recorded in a database with anonymous code for each subject.

Statistical analysis will consist of a descriptive analysis of sociodemographic and clinical characteristics of both cohorts. It will be performed also non-parametric analysis comparing this characteristics between the experimental groups. A descriptive analysis of daily life characteristics in experimental group will be performed, and also a non-parametric analysis comparing these variables between the experimental groups.

Changes due to CGA incorporation will be evaluated and the variables measuring the toxicity and early mortality should be correlated to the coincidence or not of the treatment performed with the one concluded by the geriatric evaluation.

Contribution of the different elements of CGA will be evaluated. A multiple regression will be performed in order to carry out an evaluation of the degree of contribution of each of the elements of the CGA.

A significance "p" less than 0.05 will be considered significant.

There are some study limitations: size, missing data for historic patients, monocentric study.

Inclusion Criteria:

• Age more than 70 years
• Outpatients
• Patients diagnosed with stage III colon cancer, stage IV colorectal cancer or stage IV NSCLC
• No previous diagnosis of cancer.
• No previous chemotherapy for current diagnosis of cancer
• Signed informed consent

Exclusion Criteria:

• Not fulfilling inclusion criteria