Pilot Study to Explore the Accuracy of Current Prediction Equations in Assessing Energy Needs of Patients with Newly Diagnosed Glioblastoma Multiforme

Abstract

Glioblastoma multiforme (GBM) is rare, yet is the most common brain malignancy and has a poor prognosis. In GBM, there is a dearth of research on resting energy expenditure (REE) and accuracy of extant prediction equations. The aim of this cross-sectional study was to compare measured REE (mREE) to commonly-used prediction equations in newly-diagnosed GBM patients. REE was collected by indirect calorimetry in 20 GBM patients. Calculated REE was derived from Harris-Benedict (again with weight adjusted for obesity), Mifflin-St Jeor, and 20 kcal/kg. Paired t-tests and Bland-Altman analyses were used to compare group means, evaluate the bias, and find the limits of agreement. Clinical accuracy was assessed by determining the percentage of patients with predicted REE within ±10% of mREE. Subjects were evenly distributed with regard to gender, primarily Caucasian, largely overweight or obese, and had a mean age of 57 years. All equations overestimated mREE. Mifflin-St Jeor and adjusted Harris-Benedict had the narrowest limits of agreement and accurately predicted 60% and 65% of participants, respectively. Clinicians should be aware of the discrepancy between commonly used prediction equations and REE. More research is needed to verify these findings and decipher the cause and significance in the GBM population.