Objective: To analyze the effects of the sequence of radiotherapy and chemotherapy on the efficacy of early-stage extranodal NK/T-cell lymphoma (nasal type, ENKTCL) patients, and to provide a quantitative evaluation method for individualized radiotherapy and chemotherapy. Methods: The Chinese Lymphoma Collaborative Group (CLCG) collected the clinical data of 2 008 patients with early-stage Ⅰ/Ⅱ ENKTCL who received radiotherapy and chemotherapy from January 2000 to early September 2019 from 21 hospitals across the country, including 1 417 males and 591 females, aged 2 to 83 (42±14) years. According to the sequence of radiotherapy and chemotherapy, patients were divided into radiotherapy-first group (388 cases) and chemotherapy-first group (1 620 cases). Survival rate was estimated using Kaplan-Meier method, and multivariate Cox proportional risk model was used to screen and identify independent prognostic factors. The prognostic prediction models of the two therapies were constructed separately, and the models were used to predict the individualized mortality risk of all patients to determine the appropriate radiotherapy and chemotherapy regimen for each patient. Results: The 5-year overall survival rate was 74.2% (95%CI: 69.6%-79.2%) in the radiotherapy-first group and 69.7% (95%CI: 67.1%-72.4%) in the chemotherapy-first group. Although the 5-year overall survival rate of patients in the radiotherapy-first group was numerically higher than that of the chemotherapy-first group, the difference was not statistically significant (χ2= 2.26, HR=0.84 (95%CI: 0.68-1.05), P=0.133). Six variables including age, gender, ECOG score, LDH, Ann Arbor staging, and PTI (primary tumor invasion) were screened out as independent prognostic factors (the chemotherapy-first group: HR were 1.01, 1.25, 2.07, 0.77, 1.34, 1.49, respectively, all P<0.05; radiotherapy-first group: HR were 1.02, 1.31, 1.66, 0.78, 1.37, 1.29, all P>0.05). The mean 5-year predicted mortality risk for all patients receiving radiotherapy-first regimen was lower than those receiving chemotherapy-first regimen (26.8% vs 30.2%, P<0.001). There were individualized differences in the predicted mortality risk of patients with different clinical characteristics who received radiotherapy-first regimen or chemotherapy-first regimen. Conclusion: Patients with stage Ⅰ/Ⅱ ENKTCL treated with radiotherapy-first regimen had a better expected prognosis than patients treated with chemotherapy-first regimen. The quantitative assessment of the differential effects of the sequence of radiotherapy and chemotherapy on the mortality risk of individual patients based on their clinical characteristics was helpful for the clinical development of the optimal radiotherapy and chemotherapy plan for each patient.
目的： 分析放疗和化疗顺序对早期结外鼻型NK/T细胞淋巴瘤（ENKTCL）患者疗效的影响，为个体化制定放化疗方案提供量化评估方法。 方法： 中国淋巴瘤协作组（CLCG）从全国21家医院收集2000年1月至2019年9月初治的2 008例接受放化疗的早期（Ⅰ/Ⅱ期）ENKTCL患者的临床资料，男1 417例，女591例，年龄2~83（42±14）岁。根据放化疗顺序将患者分为先放疗组（388例）和先化疗组（1 620例）。生存估计采用Kaplan-Meier法，使用多变量Cox比例风险模型筛选和识别影响预后的独立预测因素。分别构建两种疗法的预后预测模型，并使用该模型预测全部患者的个体化死亡风险，以确定适宜每例患者的放化疗方案。 结果： 先放疗组患者的5年总生存率为74.2%（95%CI：69.6%~79.2%），先化疗组患者的5年总生存率为69.7%（95%CI：67.1%~72.4%）；先放疗组患者的5年总生存率虽然数值上高于先化疗组患者，但差异无统计学意义［χ2=2.26，HR=0.84（95%CI： 0.68~1.05），P=0.133］。筛选出年龄、性别、美国东部合作肿瘤组（ECOG）评分、乳酸脱氢酶（LDH）、Ann Arbor分期、原发肿瘤侵犯（PTI）等6个变量作为独立预后因素（先化疗组：HR分别为1.01、1.25、2.07、0.77、1.34、1.49，均P<0.05；先放疗组：HR 分别为1.02、1.31、1.66、0.78、1.37、1.29，均P>0.05）。全部患者接受先放疗方案的5年期望死亡风险的均值低于接受先化疗方案（26.8%比30.2%，P<0.001）。不同临床特征的患者采用先放疗方案和先化疗方案的预测死亡风险存在个体化差异。 结论： Ⅰ/Ⅱ期ENKTCL患者采用先放疗方案的期望预后优于采用先化疗方案。根据个体患者的临床特征量化评估放化疗顺序对其死亡风险的差异性影响，有助于临床为该患者制定最佳的放化疗方案。.