Body Mass Index (BMI) and Quality of Life (QoL) in Cancer Patients

Health-related quality of life (HRQoL) might have a varied association with body weight in patients with solid cancer[1].

On one hand, low body weight may reflect cancer-related anorexia and weight loss and cachectic syndrome, which are associated with impaired performance status, deteriorated general conditions, and advanced cancer[2]. In such patients, the probability of recording a low Body Mass Index (BMI) and, concomitantly, an inferior health-related quality of life (HRQoL, as a patient reported outcome, PRO) is high.

On the other hand, some of the hormonal mediators found at increased concentration in obese patients, such as insulin-like growth factor (IGF), have been demonstrated to be involved in biological pathways that favor improved HRQoL[3].

An additional apparent contradiction is that obesity and high BMI may represent in some cancer settings an adverse feature. In particular, obesity is associated with an increased risk of developing certain tumor types and, in some cancer patients with radically resected primary tumors, of cancer relapse[4]. Moreover high BMI is often associated with an increased burden of comorbidity (e.g. cardiovascular and metabolic diseases)[5] and reduced physical activity. All these factors may reduce HRQoL.

Studies investigating specific associations between BMI and HRQoL in specific cancer settings are therefore warranted.

The present prospective observational cohort study aims to investigate the relationship between BMI and PRO-HRQoL as measured by the EORTC-QLQ-C30 or the EQ-5D-5L questionnaire [6] in different primary tumors (breast, lung, colorectal and others) and in different cancer stages (localized vs metastatic). Patients will also be stratified according to the presence of cardiovascular and metabolic comorbidities, the Karnofsky Performance status, and the oncological treatment received (systemic treatment vs radically resected patients on follow-up). If available, retrospective data will be used to train possible predictive models.

STUDY PROCEDURES Study participation will be offered to all consecutive patients with a histologically confirmed diagnosis of solid tumor referred to the Medical Oncology Units of the S.I.C.O.G. cooperative group (http://www.sicog.it/). Upon acceptance, patients will sign an informed consent forms and be asked to fill out the EORTC QLQ C30 or EQ-5D-5L questionnaire.

All common anthropometric, demographic, clinical, and biochemical variables will be recorded at the time of first referral (within three months).

Re-assessable variables, including re-administration of the EORTC QLQ-C30/EQ-5D-5L questionnaire, will be recorded every 4-6 months thereafter. All data will be stored in a prospectively maintained database.

Among recorded data will be: age, sex, weight, height, occupation, marital status, primary tumor site, tumor stage, possible metastatic sites, past and current type of oncological treatment, concomitant medications, Karnofsky Performance Status, vital signs, and routine and clinically relevant blood tests.

Patients will be oncologically managed according to standard practice. If indicated, patients will also be asked to provide consent for peripheral blood sampling for immune response evaluation. Association between BMI and PROs will be assessed using regression analyses across the different clinical settings identified.

STATISTICAL CONSIDERATIONS The design of the study hypothesizes that in metastatic patients an improved HR-QoL is associated with high BMI (non-cachectic patients) An 'exact single-stage design' will be followed [7]. According to historical data (endometrial cancer), 50% of patients with BMI < 30 have a high global health status score (GHS) of the EORTC QLQ C30 (i.e. a GHS score ≥ 80%, which would correspond to a EQ-5D-5L index ≥ 0.85) [8].

The hypothesis to be tested will be H0, P < P0 vs H1, P > P1, where P is the percentage of patients with GHS ≥ 80%. A one-tail alpha error of 0.05 and false-negative (beta) rate of 0.2 will be considered. P0 will be set at 50% and P1 at 65%, looking for a 15% increase in the percentage of high GHS scores among patients with BMI >30. The H0 hypothesis will be rejected and H1 accepted with a statistical power of 80% if at least 42 patients out of 69 with BMI > 30 report a GHS score ≥ 80.

Since BMI > 30 is observed in about 10% of all metastatic patients, a total of 690 metastatic patients will be required. Since metastatic patients represent about half of all cancer patients referred to Medical Oncology Units, a final sample size of 1380 cancer patients (all stages) will be set as the target number. Secondary analyses will evaluate the extent to which clinical and biochemical characteristics documented during the observation period moderate the association between BMI and HRQoL. Additionally, these analyses will quantify the relative contributions of both BMI and HRQoL to anticancer treatment outcomes and overall survival.