Abstract
The aim of this study was to investigate the prognostic value of baseline PET/CT parameters alone and combined with clinical features in Extranodal Natural killer/T-cell lymphoma (ENKTL) patients treated with P-GEMOX regimen (pegaspargase, gemcitabine and oxaliplatin). A total of 97 patients were retrospectively evaluated. The relationships between baseline PETCT metabolic parameters and survival were tested using Cox regression analysis and receiver operating characteristic(ROC) curve analysis was employed to evaluate the optimal cut-off value of these parameters. Kaplan-Meier curves with log-rank tests were used for survival analysis. At a median follow-up of 49 months, the 3-year PFS and OS were 62.9% and 70.1%. SUVmean, SUVmax, and SUVpeak were related to both PFS and OS in univariate analysis(P < 0.05 for all). Further multivariate analysis including PET/CT parameters and clinical parameters revealed that SUVmean was an independent prognostic factor and seemed to be slightly superior to SUVmax and SUVpeak. The low SUVmean was significantly associated with a better prognosis (3-year OS 85.1% vs.65.0%, P = 0.014; 3-year PFS 76.8% vs.62.1%, P = 0.032). SUVmean was able to further separate patients with a low-risk PINK/PINKE of < 2(n = 85, 79, separately) into two subgroups with significantly different outcomes. Moreover, the metabolic-parameter-contained m-PINK/PINKE model was constructed and showed superior predictive performance in the whole cohort. Conclusions. SUVmean was an independent prognostic factor in patients with ENKTL treated with P-GEMOX chemotherapy. Adding SUVmean to the PINK or PINKE model could improve the predictive value and further distinguish patients with poor outcomes.