中医因素参与的弥漫大B细胞淋巴瘤治疗后进展预测模型  

A Prognostic Model for the Progression of Diffuse Large B-cell Lymphoma after Treatment Involving Traditional Chinese Medicine Factors

作　　者：胡明 司华玟 蔡晶晶 魏聪敏 张红玉[1] 鲍计章[1,3] 胡琦[1] HU Ming;SI Huawen;CAI Jingjing;WEI Congmin;ZHANG Hongyu;BAO Jizhang;HU Qi(Department of Hematology,Shanghai Municipal Hospital of Traditional Chinese Medicine,Shanghai University of Traditional Chinese Medicine,Shanghai 200071,China;Shanghai University of Traditional Chinese Medicine,Shanghai 200071,China;Institute of Hematology,Shanghai Municipal Hospital of Traditional Chinese Medicine,Shanghai University of Traditional Chinese Medicine,Shanghai 200071,China)

机构地区：[1]上海中医药大学附属市中医医院血液病科,上海200071 [2]上海中医药大学,上海200071 [3]上海中医药大学附属市中医医院临床血液学中心,上海200071

出　　处：《世界中医药》2024年第19期2978-2984,共7页World Chinese Medicine

基　　金：国家自然科学基金面上项目(81774258);上海市科委中医引导类科技支撑项目(18401903900);上海市血液病学临床专科能力提升项目(B01B1);上海申康医院发展中心临床科技创新项目(SHDC22021313)。

摘　　要：目的:构建一个中医因素参与的弥漫大B细胞淋巴瘤(DLBCL)化疗后进展风险模型,为DLBCL治疗提供依据。方法:回顾性分析2014年1月至2020年12月上海市中医医院的初发DLBCL患者110例,依据复发情况分为观察组与对照组,其中观察组32例,对照组78例,随访2年。将年龄、Karnofsky功能状态评分(KPS)、美国东部肿瘤协作组功能状态评分(ECOG评分)、基础疾病数、中药服用时间、性别、分期、B症状(a.不明原因发热,非感染性发热,体温>38℃;b.夜间盗汗;c.6个月内体质量下降10%,以上3条满足1条即符合B症状。)、受累淋巴结≥5个、大包块、结外病变、生发中心、初次治疗结局和服用中药13个变量纳入分析,基于LASSO回归与多变量Cox回归的结果,开发列线图模型,分析是否服用中药及服用中药时间是否可改善DLBCL患者预后。通过受试者工作特征曲线(ROC)、ROC曲线下面积(AUC)、临床决策曲线(DCA)、校准曲线等评估模型的效能,并通过Bootstrap法自抽样验证模型的稳定性。结果:LASSO回归与Cox回归结果表明中医因素:中药服用时间为DLBCL无进展生存时间(PFS)保护因素(P<0.05),基础疾病数、大包块、初次治疗结局是DLBCL PFS的危险因素(P<0.05)。整合到模型中,6、12、24个月模型的ROC曲线下面积分别为0.85、0.93、0.92,均大于单个危险因素;校准曲线接近理想曲线。结论:本研究根据LASSO回归与Cox回归筛选的DLBCL患者化疗后进展的影响因素:中药服用时间、基础疾病数、大包块、初次治疗结局构建了列线图预测模型,经Bootstrap内部验证后,该预测模型具有较好的效能与稳定性,能够较为准确地预测DLBCL患者化疗后6、12、24个月复发的发生风险。Objective:To establish a risk model for progression after chemotherapy in diffuse large B-cell lymphoma(DLBCL)involving traditional Chinese medicine(TCM)factors,providing a basis for DLBCL treatment.Methods:A retrospective analysis was conducted on 110 newly diagnosed DLBCL patients from Shanghai Municipal Hospital of Traditional Chinese Medicine between January 2014 and December 2020.Patients were divided into an observation group(32 cases)and a control group(78 cases)based on relapse status,with a 2-year follow-up.The analysis incorporated 13 variables,including age,Karnofsky Performance Status(KPS),Eastern Cooperative Oncology Group(ECOG)performance status,number of comorbidities,duration of Chinese medicine usage,sex,disease stage,B symptoms(a.unexplained fever,non-infectious fever,body temperature>38℃;b.night sweats;c.weight loss of≥10%in the past 6 months.Any one of these indicates B symptoms),number of involved lymph nodes≥5,large masses,extranodal involvement,germinal center,initial treatment outcome,and Chinese medicine usage.LASSO regression and multivariate Cox regression were used to develop a nomogram model,analyzing whether Chinese medicine usage and duration could improve the prognosis of DLBCL patients.Model performance was assessed using receiver operating characteristic(ROC)curves,area under the curve(AUC),decision curve analysis(DCA),and calibration curves,with Bootstrap resampling to verify model stability.Results:LASSO and Cox regression results indicated that TCM factors,specifically Chinese medicine usage duration,were protective factors for progression-free survival(PFS)in DLBCL(P<0.05),while the number of comorbidities,large masses,and initial treatment outcome were risk factors for DLBCL PFS(P<0.05).The model integrated these variables,with ROC curve AUCs at 6,12,and 24 months of 0.85,0.93,and 0.92,respectively,all higher than individual risk factors.The calibration curve was close to the ideal line.Conclusion:This study established a nomogram prognostic model for post-chemotherapy

关 键 词：弥漫大B细胞淋巴瘤 无进展生存期 临床预测模型 中医因素 诺模图 

分 类 号：R273[医药卫生—中西医结合]