摘要
目的建立弥漫大B细胞淋巴瘤(DLBCL)R-CHOP化疗的生存预后评估模型,以指导临床进行风险分层。方法选取2018年1月至2023年1月在金华市中心医院确诊的DLBCL患者210例为研究对象,根据2:1抽样比例随机分为建模集140例和验证集70例。所有患者均接受R-CHOP化疗方案至少4个疗程(21 d为1个疗程),常规随访至2023年6月。根据生存预后分为良好组和不良组。比较两组患者的临床资料[包括性别、年龄、BMI、基础疾病、血清乳酸脱氢酶(LDH)、美国国家综合癌症网络国际预后指数(NCCN-IPI)]、病理特征[包括Hans分型、Ann Arbor分期、原发部位、Ki-67阳性表达率、MYC和B细胞淋巴瘤(Bcl)-2蛋白阳性表达以及骨髓侵犯]、治疗(化疗疗程和并发症)和随访时间。结果建模集中良好组105例和不良组35例,预后不良发生率25.0%;验证集良好组55例和不良组15例,不良发生率21.4%。建模集中不良组年龄大于良好组,LDH水平、NCCN-IPI、Ann Arbor分期Ⅲ~Ⅳ比例、Ki-67阳性表达率、MYC和Bcl-2蛋白双阳性表达率和骨髓侵犯比例均高于良好组,差异均有统计学意义(均P<0.05)。多因素Cox回归分析显示,NCCN-IPI(HR=2.526,95%CI:2.001~3.125,P<0.001)、Ann Arbor分期Ⅲ~Ⅳ(HR=5.021,95%CI:4.125~5.998,P<0.001)、MYC和Bcl-2蛋白双阳性(HR=3.859,95%CI:3.256~4.754,P<0.001)均是DLBCL患者R-CHOP化疗预后不良的危险因素。建立预测模型Y=0.056+1.032×(NCCN-IPI)+1.986×(Ann Arbor分期)+1.434×(MYC和Bcl-2蛋白双阳性)。ROC曲线显示,模型预测建模集与验证集预后不良的AUC分别为0.923和0.866(均P<0.01)。结论DLBCL患者化疗前NCCN-IPI、Ann Arbor分期升高以及MYC和Bcl-2蛋白双阳性表达与R-CHOP化疗预后不良密切相关,通过建立量化预测模型能够辅助临床早期、准确识别预后不良的高危群体,有较好的应用价值。
Objective To establish a survival prognosis model for diffuse large B-cell lymphoma(DLBCL)after RCHOP chemotherapy(rituximab,cyclophosphamide,doxorubicin,vincristine,prednisolone)to guide the risk stratification of patients.Methods A total of 210 patients diagnosed as DLBCL in Jinhua Municipal Central Hospital were retrospectively enrolled from January 2018 to January 2023.They were randomly divided into a training set of 140 patients and a validation set of 70 cases based on a ratio of 2:1.All patients received R-CHOP chemotherapy for at least 4 courses of treatment(21 days as 1 course),and were followed up routinely until June 2023.According to the survival prognosis,patients were divided into the favorable and adverse groups.Their clinical information[including gender,age,body mass index,chronic comorbidity,serum lactate dehydrogenase(LDH),National Comprehensive Cancer Network international prognostic index(NCCN-IPI)],pathological characteristics[including Hans classification,Ann Arbor staging,primary sites,Ki-67 positive expression rate,MYC and B-cell lymphoma-2(Bcl-2)protein positive expression,and bone marrow invasion],treatment landscape(chemotherapy course and complications),and follow-up time were compared.Results In the training set,there were 105 cases in the favorable group and 35 cases in the adverse group,with the adverse prognosis rate of 25.0%.In contrast,the validation set consisted of 55 cases in the favorable group and 15 cases in the adverse group,with the adverse prognosis rate of 21.4%.Univariate analyses found that in the training set,patients of the adverse group were older than the favorable group,and their LDH level,NCCN-IPI,ratio of Ann Arbor stages III-IV,positive expression rate of Ki-67,double positive expressions of MYC and Bcl-2 proteins,and ratio of bone marrow invasion were all significantly higher than the favorable group(all P<0.05).Multivariate Cox regression analysis showed that NCCN-IPI(HR=2.526,95%CI:2.001-3.125,P<0.001),Ann Arbor stages III-IV(HR=5.021,95%CI:4.125-5.998,P<0.001),and double positive expression of MYC and Bcl-2 proteins(HR=3.859,95%CI:3.256-4.754,P<0.001)remained as the significant adverse predictors for DLBCL patients received R-CHOP chemotherapy.Finally,we established a predictive model:Y=0.056+1.032×(NCCN-IPI)+1.986×(Ann Arbor staging)+1.434×(MYC and Bcl-2 double expression).The receiver operating curve(ROC)showed that the area under the curve(AUC)of the training and validation sets in predicting adverse prognosis were 0.923 and 0.866,respectively(both P<0.01).Conclusion Elevated NCCN-IPI,advanced Ann Arbor staging,and double expression of MYC and Bcl-2 proteins before chemotherapy are closely related to the poor clinical outcomes of DLBCL patients after R-CHOP chemotherapy.This prognostic model can assist in clinical practice for early and accurate identification of patients with high risk for poor outcomes,which has good application value.
作者
翁翔
赵明哲
胡慧仙
WENG Xiang;ZHAO Mingzhe;HU Huixian(Department of Hematology,Jinhua Municipal Central Hospital,Jinhua 321000,China)
出处
《浙江医学》
CAS
2024年第2期172-176,共5页
Zhejiang Medical Journal
基金
金华市公益性技术应用研究项目(2022-4-094)。
