Clinical value of intratumoral metabolic heterogeneity in [¹⁸F]FDG PET/CT for prediction of recurrence in patients with locally advanced colorectal cancer

BACKGROUND: The aim of this study was to find a reliable predictor of recurrence in patients with locally advanced colorectal cancer. METHODS: We enrolled 96 patients for this retrospective study. We investigated metabolic (SUV_max, metabolic tumor volume, total lesion glycolysis, and heterogeneity), clinical (age, sex, stage, and CEA level) and pathologic (Ki-67, p53, CD31, COX-2, E-cadherin and EGFR) parameters. The coefficient of variation (COV) was chosen to assess heterogeneity of [¹⁸F]FDG uptake by dividing the standard deviation of the SUV by SUV_mean. Recurrence-free survival was compared with each metabolic, clinical and pathologic parameters by using univariate and multivariate survival analysis. RESULTS: Among 96 patients, 19 patients (19.8%) showed disease recurrence. In the ROC analysis, the optimal cutoff values of SUV_max, metabolic tumor volume (cm³), total lesion glycolysis (cm³), and metabolic heterogeneity were determined as 17.6, 10.05, 232.46, and 0.48, respectively. In univariate analysis, probability of recurrence was statistically increased in those with metabolic tumor volume >10.05 (P=0.045), and those with metabolic heterogeneity >0.48 (P=0.031). In multivariate analysis, metabolic heterogeneity was the only independent prognostic factor (HR 4.56, 95% CI 1.57-13.23, P=0.006). CONCLUSIONS Intratumoral metabolic heterogeneity assessed by COV is a reliable predictive factor for disease recurrence in patients with locally advanced colorectal cancer. Therefore, its application could be an important step for personalized management of colorectal cancer.