前列腺影像报告和数据系统中4、5分病变的假阳性率评价

Evaluation of False Positive Rate of 4,5 Points Lesions in Prostate Imaging Report and Data System

目的:评价前列腺影像报告和数据系统中4、5分病变的假阳性率。方法:回顾性分析2015年9月至2019年5月于深圳市第二人民医院接受前列腺MRI检查患者163例,对168个病灶进行了MRI超声融合靶向活检,放射科医师回顾性地采用PI–RADS v2.1进行鉴定和评估。在PI–RADS 4、5分病变中,确定临床显著性前列腺癌(定义为Gleason评分≥7)的发生率。临床数据首先通过t检验比较分类变量,并通过Fisher精确检验比较连续变量,多变量逻辑回归模型用于确定与良性病理结果相关的因素。结果:用mpMRI认知融合活检评估的36个PI–RADS 4分病变中,80.55%(29/36)为前列腺癌,19.45%(7/36)为良性病变;132个PI–RADS 5分病变中,93.94%(124/132)为前列腺癌,6.06%(8/132)为良性病变。PI–RADS 4分病灶表现出61.11%(22/36)为有临床意义的前列腺癌;PI–RADS 5分病灶表现出87.88%(116/132)为有临床意义的前列腺癌。此外,15个良性病灶的其他地方也没有临床意义上的前列腺癌。在单变量分析中,与恶性病理结果的PI–RADS 4、5分病变相比,良性病理结果的PI–RADS 4、5分病变与较低的前列腺特异性抗原(PSA)密度有明显的关系,OR=0.017,P=0.002。在多变量分析中,与良性特征相关的因素是较低的PSA密度,OR=0.016,P=0.005,95%CI为〔0.453,3.794〕。对15个病理结果为良性的病变进行二次复查,53.33%(8/15)病理为基质性良性前列腺增生,46.67%(7/15)病理为炎症。结论:有良性病理结果的PI–RADS 4、5分病变发生频率很低。笔者的结果提示临床参数(PSA密度)可能有助于帮助临床决策,考虑将其纳入未来PI–RADS评分中。

Objective The purpose of this study is to evaluate the false positive rate of 4,5 points lesions in the prostate imaging report and data system.Methods A retrospective analysis of 163 patients undergoing prostate MRI examination at Shenzhen second people’s hospital from September 2015 to May 2019,MRI ultrasound fusion targeted biopsy was performed on 168 lesions,and radiologists retrospectively adopted PI-RADS v2.1 for identification and evaluation.In PI–RADS 4 and 5 lesions,the incidence of clinically significant prostate cancer(defined as Gleason score≥7)was determined.The clinical data first compared categorical variables by t-test,and compared continuous variables by Fisher’s exact test.Multivariate logistic regression models are used to determine factors related to benign pathological results.Result Among 36 PI–RADS 4-point lesions assessed by mpMRI cognitive fusion biopsy,80.55%(29/36)were prostate cancer,and 19.45%(7/36)were benign lesions;among 132 PI–RADS 5-point lesions,93.94%(124/132)were prostate cancer,and 6.06%(8/132)were benign lesions.In general,PI–RADS 4-point lesions showed 61.11%(22/36)as clinically significant prostate cancer;PI–RADS 5-point lesions showed 87.88%(116/132)as clinically significant prostate cancer.In addition,there are no clinically meaningful prostate cancers elsewhere in these 15 benign lesions.In univariate analysis,compared with PI–RADS 4 and 5 with malignant pathological results,PI–RADS 4 and 5 with benign pathological results were significantly related to lower PSA density(P=0.002).In the multivariate analysis,the factor associated with benign characteristics is the lower PSA density(OR=0.016;P=0.005).Fifteen lesions with benign pathological results were re-examined again,53.33%(8/15)pathologies were stromal benign prostatic hyperplasia,and 46.67%(7/15)pathologies were inflammation.Conclusion PI–RADS types 4 and 5 with benign pathological results occur very low.Our results suggest that clinical parameters(PSA density)may help clinical decision-making,and consider including them in future PI–RADS scores.