Prognostic value of primary gross tumor volume and standardized uptake value of (18)F-FDG in PET/CT for distant metastasis in locoregionally advanced nasopharyngeal carcinoma.

Distant metastasis has become the predominant model of treatment failures in patients with locoregionally advanced nasopharyngeal carcinoma. Effort should therefore be made to stratify locoregionally advanced nasopharyngeal carcinoma patients into different groups based on the risk of metastasis to improve prognosis and tailor individualized treatments. This study aims to assess the value of primary gross tumor volume and the maximum standardized uptake value for predicting distant metastasis-free survival of patients with locoregionally advanced nasopharyngeal carcinoma. A total of 294 locoregionally advanced nasopharyngeal carcinoma patients who were identified from prospectively maintained database and underwent fluor-18-fluorodeoxyglucose positron emission tomography/computed tomography imaging before treatment were included. The maximum standardized uptake value was recorded for the primary tumor (SUVmax-P) and neck lymph nodes (SUVmax-N). Computed tomography-derived primary gross tumor volume was measured using the summation-of-area technique. At 5 years, the distant metastasis-free survival rate was 83.7%. The cut-off of the SUVmax-P, SUVmax-N, and primary gross tumor volume for distant metastasis-free survival was 8.95, 5.75, and 31.3 mL, respectively, by receiver operating characteristic curve. In univariate analysis, only SUVmax-N (hazard ratio: 7.01; 95% confidence interval: 1.70-28.87; p < 0.01) and clinical stage (hazard ratio: 3.03; 95% confidence interval: 1.67-5.47; p = 0.007) were confirmed as independent predictors of distant metastasis-free survival. A prognostic model was derived by SUVmax-N and clinical stage: low risk (SUVmax-N < 5.75 regardless of clinical stage), medium risk (stage III and SUVmax-N ≥ 5.75), and high risk (stage IV and SUVmax-N ≥ 5.75). Multivariate analysis revealed that SUVmax-N and the prognostic model remained independent prognostic factors for distant metastasis-free survival (p = 0.023 and p < 0.001, respectively), but the clinical stage became insignificant (p = 0.133). Furthermore, the adjusted hazard ratios for the prognostic model were higher than SUVmax-N (hazard ratio = 6.27 vs 5.21, respectively). In summary, compared with SUVmax-P, SUVmax-N may be a better predictor of distant metastasis-free survival for patients with locoregionally advanced nasopharyngeal carcinoma. Combining SUVmax-N with clinical stage gives a more precise picture in predicting distant metastasis.