18F-FDG PET或PET/CT评估乳腺癌新辅助化疗疗效价值的meta分析

摘要

Objective To evaluate the accuracy and predictive value of 18F-FDG PET or PET/CT in the assessment of neoadjuvant chemotherapy in locally advanced breast cancer by meta-analysis.Methods Relevant studies were retrieved from PubMed,Embase,Cochrane and Wanfang,in English or Chinese.To ensure the homogeneity of all included studies,selection criteria were established and all the studies were scored according to a quality assessment of diagnostic accuracy studies (QUADAS) table.The combined Se,Sp,LR,diagnostic odds ratio (DOR),and AUC of PET and other conventional imaging techniques were compared.The assessment standard of neoadjuvant chemotherapy by conventional imaging techniques is tumor size change.Funnel plot and analysis with meta-regression were performed to explore the source of heterogeneity.Results Sixteen eligible studies were included with a total of 662 subjects.The combined results were:(1) PET:SePET 86.1％(247/287),SpET 69.6％(261/375),LR + PET 3.18,LR - PET 0.23,DORPET 17.26; (2) Other imaging techniques:Sed 57.2％(91/159),Spd 48.5％(50/103),LR+d 1.07,LR-d 0.87,DORd 1.30.ROC analysis showed that Q* values of PET ( QPET* ) and other imaging techniques ( Qd* ) were 0.8058 and 0.5328,respectively.Funnel plot showed that publication bias was not the main source of heterogeneity.Meta-regression results suggested that the year of publication might be one of the sources of heterogeneity (P=0.05).Conclusion This meta-analysis suggests that,in the assessment of treatment rosponse of neoadjuvant chemotherapy for breast cancer,FDG PET or PET/CT in monitoring the changes in glucose metabolism is more accurate than conventional imaging techniques in monitoring tumor size.%目的 用meta分析方法评价18F-FDG PET或PET/CT用于乳腺癌新辅助化疗疗效评估的准确性和预测价值.方法 检索PubMed、Embase、Cochrane以及万方数据库中相关的英文或中文文献.制定纳入文献的选择标准,并按照诊断性研究质量评价(QUADAS)评分表对文献质量进行评分,以保证纳入研究的同质性.对符合标准的研究结果进行meta分析,估计和比较PET和其他影像检查(以测量肿瘤体积改变情况为评估指标)的加权Se、Sp、LR、诊断优势比(DOR)与综合ROC(SROC)曲线,判断PET和其他影像检查评估新辅助化疗的准确性.用绘制漏斗图和meta回归的方法探讨纳入研究的异质性来源.结果 16篇文献符合全部纳入标准,含662例研究对象.Meta分析加权值:(1)PET:SePET 86.1％(247/287),SpPET 69.6％(261/375),LR + PET 3.18,LR - PET0.23,DORPET 17.26；(2)其他影像检查:Sed 57.2％ (91/159),Spd 48.5％(50/103),LR+d1.07,LR-d0.87,DORd1.30.绘制SROC曲线,得出QPET*(0.8058)＞Qd*(0.5328).漏斗图示发表偏倚对结果影响小.Meta回归分析发现,文献发表年份可能是异质性的来源之一(P=0.05).结论 监测葡萄糖代谢的改变比测量肿瘤大小的改变更能准确评估乳腺癌新辅助化疗的疗效.