Treatment-Related Weight Change and Overall Survival in United States Veterans with Follicular Lymphoma

Introduction

Weight loss during chemotherapy has been associated with decreased overall survival (OS) in various solid tumors. While weight loss >10% in the 6 months leading up to diagnosis is a known adverse prognostic factor for non-Hodgkin's lymphoma (one of the B symptoms), the association between weight loss during chemotherapy and survival in follicular lymphoma (FL) patients is not well understood. Few studies have looked at long-term weight change patterns following chemotherapy treatment in this patient population. We investigated short and long-term weight change trends, predictors, and association with OS and disease-specific survival (DSS) in a cohort of FL patients.

Methods

FL patients diagnosed and treated with CHOP (cyclophosphamide, doxorubicin, vincristine, and prednisone) +/- rituximab, CVP (cyclophosphamide, vincristine, and prednisone) +/- rituximab, or rituximab monotherapy regimens between 1998 and 2010 were identified in the Veterans Health Administration database. Data on weight 1 year prior to treatment, at time of first treatment (baseline), and up to 5 years after treatment initiation was obtained. Additional data on height, age, stage, race, comorbidities, date of diagnosis, LDH, and B-symptoms was obtained. Body Mass Index (BMI) at diagnosis was categorized according to World Health Organization criteria. Weight change during treatment is calculated as difference between baseline weight and 3 months after start of treatment. Long-term weight change is calculated as difference between baseline weight and 24 months after start of treatment. Logistic regression identified factors associated with long-term weight gain. Landmark Cox analysis evaluated prognostic significance of weight loss during treatment among patients who survived at least 6 months after treatment initiation.

Results

1022 patients met inclusion criteria out of 2235. Mean and median age at diagnosis was 63.6 and 63.0 years respectively, 95.9% were men, and 72.7% had Stage III/IV disease. The mean Charlson co-morbidity score was 2.3. B symptoms were present in 37.9% and LDH was elevated in 26.8%.

Mean and median weight change during treatment was -1.4kg (-1.5%) and -0.4kg (-0.6%), with a majority of patients losing weight (56%) and 23% of patients losing >5% of their baseline weight. In contrast, mean and median weight change at 24 months after treatment initiation was +1.2kg (+1.6%) and +1.3kg (+1.6%), with a majority of patients (57%) gaining weight and 14% of patients gaining >10% of their baseline weight after treatment completion.

Logistic regression analysis identified factors associated with increased risk of weight gain >10% at 24 months after treatment initiation. These included: weight loss >5% in the year prior to treatment (Odds Ratio (OR) 6.82, 95% Confidence Interval (CI) 3.09-15.05), weight gain between 0-5% during treatment (OR 2.53, 95% CI 1.21-5.27), and weight gain >5% during treatment (OR 9.43, 95% CI 3.85-23.14).

Kaplan-Meier survival analysis showed that weight loss >5% during treatment was associated with decreased OS (p<0.0001) and disease specific survival (DSS) (p=0.0027) compared to weight loss <5% or weight gain. A landmark Cox analysis controlling for age, disease stage, comorbidities, elevated LDH, B symptoms, BMI at diagnosis, and treatment type showed that weight loss >5% during treatment was independently associated with worse OS (Hazard Ratio (HR) 1.71, 95% CI 1.32-2.22) and DSS (HR 1.61, 95% CI 1.11-2.34).

Conclusions

In patients with FL, weight loss >5% during treatment is independently predictive of worse overall survival and disease-specific survival. Weight loss could be considered in conjunction with other dynamic variables (such as PET positivity and nodal size) to assess prognosis at the end of therapy. 14% of patients experience long-term weight gain >10% of baseline. Patients who gained 5% or more during treatment are at highest risk (OR=9.4) for long-term weight gain-this subset of patients could be targeted for weight loss interventions to prevent future obesity-related comorbidities.