荧光原位杂交和尿细胞学检查对膀胱癌诊断意义的Meta分析

Meta-analysis of fluorescence in situ hybridization and cytology for diagnosis of bladder cancer

背景与目的:目前膀胱癌疗效和监测的主要方法是膀胱镜和尿细胞学检查,前者为侵入性检查,令患者感到不适;后者虽无创且特异性高,但敏感性太低,且受主观因素影响大。本研究拟对中、英文有关比较荧光原位杂交(fluorescence in situ hybridization,FISH)和尿细胞学检查诊断膀胱癌研究的结果进行系统分析,以明确FISH对膀胱癌的诊断意义。方法:采用Cochrane系统评价方法,MEDLINE(1966年1月～2008年6月)、EMBASE(1988年1月～2008年6月)、Cochrane图书馆、中国生物医学期刊文献数据库(CMCC,1979年～2008年6月)、CNKI数字图书馆(1979年1月～2008年6月)进行有关FISH和尿细胞学检查诊断膀胱癌文献的检索、质量评价和资料提取,采用MetaDiSc1.4软件进行Meta分析。结果:共检索到相关研究242篇,排除230篇,符合纳入标准12篇进入Meta分析,涉及研究对象3430例。异质性检验提示无阈值效应,但存在其它原因导致的异质性。按随机效应模型进行Meta分析,FISH和尿细胞学诊断膀胱癌的准确度指标敏感度、特异度、阳性似然比、阴性似然比以及诊断优势比等汇总及95%CI分别为74%(71%～77%)vs.57%(54%～61%)、88%(86%～90%)vs.85%(83%～87%)、6.18(3.56～10.73)vs.4.15(2.78～6.20)、0.29(0.19～0.45)vs.0.51(0.41～0.63)及24.17(9.33～62.64)vs.9.59(5.91～15.57)。FISH和尿脱落细胞学检查的敏感度随肿瘤分级、分期的升高而增高。综合受试者工作特征曲线下面积分别为0.8938、0.8247,Q*值分别为0.7847、0.7226。结论:FISH诊断膀胱癌的准确度较高,但对高分期的敏感度较细胞学低,目前尚不能取代传统的尿细胞学检查,但可作为膀胱癌术前诊断、术后监测和随访的指标。

Background and Objective. Currently, cystoscopy and urine cytology are standard modalities in therapy monitoring and follow-up of bladder cancer. Cystoscopy is an invasive and uncomfortable procedure while cytology has a limited value because it is operator-dependent and has a low sensitivity. This study was to assess the accuracy of fluorescence in situ hybridization （FISH） in detecting bladder cancer by comparing it with cytology using systemic analyses of studies published in both English and Chinese. Methods： Cochrane systematic evaluation was used to search through MEDLINE, EMBASE, Cochrane Library, CMCC and CNKI for studies regarding FISH and cytology in the detection of bladder cancer. Data were extracted and analyzed using software MetaDiScl.4. Results. In total 242 relevant studies were searched, of which 12 studies were enrolled and 3430 patients were included. Heterogeneity, except for threshold effects, was found within these studies. A meta-analysis was performed using the random effect model. Pooled accuracy indicators like sensitivity, specificity,positive likelihood ratio （LR）, negative LR and diagnostic odds ratio of FISH and cytology were 74% （71%-77%） vs. 57% （54%-61%）, 88 % （86%- 90%） vs. 85 % （83%-87%）, 6.18 （3.56-10.73） vs. 4.15 （2.78-6.20）,0.29 （0.19-0.45） vs. 0.51 （0.41-0.63） and 24.17 （9.33-62.64） vs. 9.59 （5.91-15.57）, respectively. The sensitivity of both FISH and cytology were increased with the increase of tumor grade and stage. The areas under summary receiver operating characteristics （SROC） curve were 0.8938 and 0.7847, and the Q indices were 0.8247 and 0.7226 for FISH and cytology, respectively. Conclusions. FISH has a high accuracy in detecting bladder cancer. But its sensitivity in detecting high-stage bladder cancer is lower than that of cytology. Although FISH can not replace traditional urine cytology, it can be used as an important adjunct in the preoperative detection and postoperative monitoring of bladder cancer.