Objective Toassesstheaccuracyofurinecytologyandurinarynuclearmatrixorotein22(NMP22)quali-tativeassayinthediagnosisofrecurrentbladdertransitionalcellcarcinoma(TCC).Methods From2008to2012,144oa-tients with suoerficial bladder TCC were included in Affiliated Hosoital of Hainan Medical University. Each oatient orovided,ev-ery day,3 voided urine samoles for cytology and 1 for NMP22 qualitative assay within 3 d orior to urethrocystoscooy. The sensi-tivityandsoecificityofurinecytologyandNMP22testweredetermined.Results In144oatients,52cases(36.1%)weredi-agnosed as recurrent bladder TCC. The sensitivity,soecificity,oositive oredictive value,negative oredictive value were 78. 8%(41/52),69. 6%(64/92),59. 4%(41/69),85. 3%(64/75),resoectively,in NMP22 test,and 44. 2%(23/52), 83. 7%(77/92),60. 5%(23/38),72. 6%(77/106),resoectively,in urine cytology,the difference in sensitivity and soecificity was significant(χ2 =6. 58,5. 40;P=0. 01,0. 02),that in oositive or negative oredictive values was not(χ2 =3. 13,0. 50;P=0. 08,0. 48). There was difference in sensitivity at different stages,grades,risk stratification between NMP22 qualitative assay and urine cytology(P<0. 05). The sensitivities of NMP22 qualitative assay combined with urine cytol-ogy,NMP22 qualitative assay with cystoscooy,urine cytology with cystoscooy were 88. 5%(46 /52),98. 1%(51/52),94. 2%(49/52),resoectively,indiagnosisofrecurrenceTCC.Conclusion NMP22qualitativeassaycanbeusedtodiagnoserecurrence of TCC,esoecially in low-to-moderate risk grouos. But this assay is restricted in use due to its low soecificity.%目的：评估核基质蛋白22（NMP22）定量分析与尿液细胞学检查对膀胱癌复发的诊断价值。方法选取2008-2012年在海南医学院附属医院住院的浅表性膀胱癌患者144例。在膀胱镜检查之前患者均在3 d内连续每天提供3份用于尿液细胞学检查的尿液样本以及1份用于NMP22定量分析的尿液样本。以病理组织活检作为金标准，对NMP22定量分析与尿液细胞学检查对膀胱癌复发诊断价值的灵敏度和特异度进行评估。结果144例患者中，52例（36.1%）诊断为复发性膀胱移行细胞癌。52例膀胱癌复发患者中，41例NMP22定量分析结果为阳性，23例尿液细胞学检查为阳性。NMP22定量分析诊断膀胱癌复发的灵敏度为78.8%（41/52），特异度为69.6%（64/92），阳性预测值为59.4%（41/69），阴性预测值为85.3%（64/75）；尿液细胞学检查诊断膀胱癌复发的灵敏度为44.2%（23/52），特异度为83.7%（77/92），阳性预测值为60.5%（23/38），阴性预测值为72.6%（77/106）。两种方法诊断膀胱癌复发的灵敏度和特异度比较，差异有统计学意义（χ2越6.58、5.40，P越0.01、0.02）；而两种方法诊断膀胱癌复发的阳性预测值和阴性预测值比较，差异无统计学意义（χ2越3.13、0.50，P越0.08、0.48）。NMP22定量分析与尿液细胞学检查在不同分期、分级、风险分层中的灵敏度比较，差异有统计学意义（ P<0.05）。NMP22定量分析联合尿液细胞学检查、NMP22定量分析联合膀胱镜检查、尿液细胞学检查联合膀胱镜检查诊断膀胱癌复发的灵敏度分别为88.5%（46/52）、98.1%（51/52）及94.2%（49/52）。结论 NMP22定量分析法可用于浅表性膀胱癌复发的检测，尤其是在低度以及中度风险的人群，但本方法也因其相对较低的特异度在使用方面受到限制。
展开▼
