Supplementary value of primary tumor volume for prognostic stratification in patients with T3‑4 nasopharyngeal carcinoma: A retrospective cohort study

The present retrospective study evaluated the supplementary value of gross tumor volume of the primary nasopharyngeal tumor (GTVnx) for overall survival (OS) risk stratification in patients with T3‑4 nasopharyngeal carcinoma (NPC). A total of 118 patients with T3‑4 NPC who received definitive intensity‑modulated radiotherapy at Guilin People's Hospital (Guilin, China) between March 2020 and June 2025 were included. GTVnx was obtained from the radiotherapy planning system. A time‑dependent receiver operating characteristic (ROC) curve was used to determine the exploratory cut‑off for predicting 3‑year OS. Survival differences were assessed using the Kaplan‑Meier method and log‑rank test, and Cox proportional hazards models were used to evaluate the association between GTVnx and OS. Clinical and extended models were compared using the concordance index (C‑index), Akaike information criterion (AIC) and time‑dependent area under the curve (AUC), and sensitivity analyses were performed to assess robustness. The median follow‑up time was 34.1 months and 18 deaths occurred (15.3%). The 36‑month time‑dependent ROC analysis identified 78 cm3 as the exploratory cut‑off for GTVnx, with an AUC of 0.709. Patients with high GTVnx had poorer OS than those with low GTVnx (log‑rank P<0.001). In a multivariable Cox analysis, high GTVnx remained associated with poorer OS after adjustment for T stage and pretreatment Epstein‑Barr virus DNA (EBV DNA; hazard ratio, 4.375; 95% confidence interval, 1.474‑12.986; P=0.008). Adding GTVnx to the clinical model improved the C‑index from 0.759 to 0.817, reduced the AIC from 135.151 to 133.085 and increased the 36‑month AUC from 0.748 to 0.787. Sensitivity analyses supported the consistency of the main findings. GTVnx may provide supplementary prognostic information beyond T stage and pretreatment EBV DNA in patients with T3‑4 NPC. However, the identified cut‑off value and incremental prognostic value require validation in independent cohorts.