RESUMEN
PURPOSE: Progression after first-line immunochemotherapy (ICT) for recurrent or metastatic nasopharyngeal carcinoma (R/M NPC) is a clinical concern due to subsequent limited treatment options. This study firstly predicted the progress outcome. METHODS: A cohort of 186 R/M NPC cases that received first-line ICT was included for developing a Cox regression model for progression-free survival (PFS) and risk stratification, which was verified by cross-validation. Discrimination and calibration were evaluated. Progression sites in risk groups was shown with a Sankey diagram. RESULTS: Baseline predictors including liver metastasis, trend of plasma Epstein-Barr virus DNA copies, lymphocyte-to-monocyte ratio, and level of platelet and lactate dehydrogenase were identified for model construction, which stratify the cohort into low, middle, and high-risk groups. The overall concordance index (C-index) was 0.67 (95% CI 0.62-0.73). The area under the curve (AUC) was 0.68 (95% CI 0.60-0.76), 0.74 (95% CI 0.66-0.82), 0.75 (95% CI 0.65-0.84) at predicting 12, 18, and 24 months PFS, indicating a moderate accuracy. Cross-validation showed the model performance was robust. Compared with the low-risk group (median PFS: 24.4 months, 95% CI 18.4 months to not reached), the high-risk group (median PFS: 7.1 months, 95% CI 6.4-10.1 months; hazard risk: 7.4, 95% CI 4.4-12.4, p < 0.001) progressed with more liver metastasis after ICT resistance. CONCLUSION: It was the first study that described the risk factors and progression characteristics in R/M NPC patients who received first-line ICT, investigating the progression patterns, which was helpful to identify patients with different risks and help guide personalized interventions.