EFFECTS OF BODY MASS INDEX ON RELAPSE RISK IN PEDIATRIC HODGKIN LYMPHOMA PATIENTS

Abstract

Background. Rates of childhood obesity in the United States have risen dramatically in recent decades, with more than 31% of children currently classified as overweight or obese. This raises concerns about the effects of weight on outcomes for pediatric illness, including cancer. There is some evidence of poorer outcomes for pediatric leukemia patients who are overweight or obese, and studies in adults have suggested negative impacts of obesity in numerous cancer types. To date, there are no studies investigating outcomes in overweight and obese children with Hodgkin lymphoma (HL). Our hypothesis was that higher body mass index (BMI) at diagnosis is associated with increased risk for HL relapse. Methods. We conducted a retrospective cohort study of 101 pediatric HL patients treated between 1980 and 2010 at Phoenix Children’s Hospital, a large pediatric oncology referral center in the Southwestern United States. Data was abstracted from electronic and paper medical charts as well as survival clinic follow‐up records. We performed logistic regression and conducted a survival analysis to test whether body mass index (BMI) at diagnosis was associated with time to disease relapse. For this pilot study, we conducted a primary analysis as well as several exploratory secondary analyses with the goal of generating hypotheses to be tested in future large studies of this population. Results. In the primary analysis comparing underweight and normal children to overweight and obese children, none of the patient characteristics – sex, race, age, clinical risk level, or radiation status – were significantly associated with BMI group. In the univariate analysis of HL relapse, children in the overweight/obese group had an increased unadjusted odds ratio of 1.58 (95% CI: 0.50‐5.28), but this was not statistically significant. Exploratory analyses categorizing BMI groups in various ways also suggested an association between increased BMI and risk for HL relapse, though this failed to reach statistical significance. No potential confounders were associated with HL relapse except radiation status (p=0.004), although we were unable to calculate an odds ratio due to a lack of patients in some subgroups. In the survival analysis, radiation was the only variable significantly associated with time to HL relapse. Kaplan‐Meier curves of relapse‐free survival time did not show a significant difference between BMI groups in the primary analysis, but secondary analyses suggested a nonsignificant trend toward decreased long‐term disease‐free survival in patients with higher BMI. Discussion. The relatively small sample size for this pilot study precluded demonstration of statistically significant differences in HL relapse risk or time to relapse between BMI groups. However, exploratory analyses suggested a trend toward increased risk for relapse and shorter disease‐free survival in patients with higher BMI, and these results merit further investigation in larger studies. Multi‐center collaborative studies will be required to attain sufficient sample sizes to accurately assess clinical prognosis in this patient population. Improving our understanding of how BMI affects pediatric cancer outcomes is an important step toward identifying patients at increased risk and determining how best to individualize treatment and monitoring plans for overweight and obese children.