Cai Sunqin, Huang Zongwei, Chen Zihan, Li Ying, Su Jingjing, Chen Ronghui, Xu Siqi, Wang Jing, Qiu Sufang
Department of Radiation Oncology, Clinical Oncology School of Fujian Medical University, Fujian Cancer Hospital, Fuzhou, China.
Department of Radiation Oncology, The Second Affiliated Hospital of Fujian Medical University, Quanzhou, China.
Radiat Oncol. 2025 Jun 4;20(1):93. doi: 10.1186/s13014-025-02672-1.
This research focused on determining the optimal cycles of induction chemotherapy (IC) in high-risk locoregionally advanced nasopharyngeal carcinoma (LA-NPC).
The retrospective analysis was conducted on 885 patients. Potential bias was minimized by propensity score matching (PSM). Overall survival (OS) served as the primary endpoint. Survival outcomes were analyzed using Kaplan-Meier curves, with statistical comparisons performed via the log-rank test. Prognostic determinants were identified through multivariate cox regression analysis. A nomogram model was constructed to quantify individualized prognosis.
Patients were divided into 2/3-cycle (IC = 2/3) and 4-cycle IC (IC = 4) groups. After PSM, 446 patients remained and were categorized into distinct risk groups according to independent predictors, including clinical stage and pre-treatment Epstein-Barr virus DNA (pre-EBV DNA). For the high-risk cohort (stage IVa with pre-EBV DNA ≥ 4000 copies/mL), the IC = 4 regimen showed higher 5-year OS (70.4% vs. 54.7%, P = 0.036) than the IC = 2/3 regimen. In the low- and middle-risk cohorts, the IC = 2/3 regimen exhibited OS comparable to the IC = 4 regimen. The established nomogram model demonstrated superior prognostic power compared to individual factors. Given the adverse effects, the IC = 4 regimen was associated with significantly higher rates of grade 3-4 neutropenia (24.6% vs. 15.5%, P = 0.017) and thrombocytopenia (8.0% vs. 3.7%, P = 0.049) compared to the IC = 2/3 regimen.
The developed nomogram offers personalized guidance on selecting individual IC cycles for LA-NPC patients.
本研究聚焦于确定高危局部晚期鼻咽癌(LA-NPC)诱导化疗(IC)的最佳周期数。
对885例患者进行回顾性分析。通过倾向评分匹配(PSM)将潜在偏倚降至最低。总生存期(OS)作为主要终点。采用Kaplan-Meier曲线分析生存结局,并通过对数秩检验进行统计学比较。通过多变量cox回归分析确定预后决定因素。构建列线图模型以量化个体预后。
患者分为2/3周期(IC = 2/3)组和4周期IC(IC = 4)组。PSM后,剩余446例患者,并根据包括临床分期和治疗前爱泼斯坦-巴尔病毒DNA(pre-EBV DNA)在内的独立预测因素分为不同风险组。对于高危队列(IVa期且pre-EBV DNA≥4000拷贝/mL),IC = 4方案的5年总生存率(70.4%对54.7%,P = 0.036)高于IC = 2/3方案。在低风险和中等风险队列中,IC = 2/3方案的总生存率与IC = 4方案相当。所建立的列线图模型显示出比个体因素更强的预后预测能力。考虑到不良反应,与IC = 2/3方案相比,IC = 4方案的3-4级中性粒细胞减少症(24.6%对15.5%,P = 0.017)和血小板减少症(8.0%对3.7%,P = 0.049)发生率显著更高。
所开发的列线图为LA-NPC患者选择个体IC周期提供了个性化指导。