动态对比增强MRI升级前列腺影像报告和数据系统第2版3类外周带观察灶按15毫米大小阈值分层
本研究评估动态对比增强MRI(DCE-MRI)升级前列腺影像报告和数据系统第2版(PI-RADSv2)外周带3类观察灶,按15毫米大小阈值分层。两名盲法放射科医生独立评估301例患者326个临床显著肿瘤(Gleason评分≥7),计算升级观察灶的临床显著癌检出率。结论:基于DCE-MRI异常升级的外周带观察灶临床显著癌可能性高,但读者间一致性低;≥15毫米的升级肿瘤无假阳性诊断且前列腺外侵犯率更高,提示可归为PI-RADSv2第5类。
要点
- DCE-MRI升级的PI-RADSv2外周带3类观察灶中,临床显著癌真阳性率为91.2%(31/34),假阳性率为8.8%(3/34)。
- ≥15毫米的DCE-MRI升级观察灶无假阳性诊断,且前列腺外侵犯率高于<15毫米者(84.6% [11/13] vs 38.1% [8/21];p=0.016)。
- PI-RADSv2评分总体读者间一致性为中等(κ=0.42),仅考虑DCE-MRI升级观察灶时为一般(κ=0.23)。
核心结论
- 基于DCE-MRI异常升级的外周带观察灶临床显著癌可能性高,但读者间一致性低;≥15毫米的升级肿瘤无假阳性诊断且前列腺外侵犯率更高,提示可归为PI-RADSv2第5类。
研究设计与样本
- 回顾性评估2012至2017年根治性前列腺切除术前多参数MRI的301例患者326个临床显著肿瘤(Gleason评分≥7),两名盲法放射科医生独立评分,分歧经共识解决,以病理为参考标准。
诊断标准与关键数值
- DCE-MRI升级的PI-RADSv2外周带3类观察灶中,临床显著癌真阳性率91.2%(31/34)、假阳性率8.8%(3/34);≥15毫米观察灶假阳性诊断0例,前列腺外侵犯率84.6%(11/13),<15毫米为38.1%(8/21),差异有统计学意义(p=0.016);两组Gleason评分无差异(p=0.354)。
临床应用与阅片要点
- 作者认为,基于DCE-MRI异常升级且≥15毫米的外周带观察灶可归为PI-RADSv2第5类。
局限性
- PI-RADSv2评分总体读者间一致性为中等(κ=0.42),仅考虑DCE-MRI升级观察灶时一致性为一般(κ=0.23)。
以上解读依据论文摘要(全文未开放获取或许可不允许翻译转载)。
英文原文(摘要)· 点击展开
OBJECTIVE. The purpose of this study is to evaluate dynamic contrast-enhanced (DCE) MRI (DCE-MRI)-upgraded Prostate Imaging Reporting and Data System version 2 (PI-RADSv2) peripheral zone (PZ) observations stratified by a size threshold of 15 mm. MATERIALS AND METHODS. Two blinded radiologists independently assessed 301 patients with 326 clinically significant tumors (Gleason score [GS] ≥ 7) using multiparametric MRI performed before radical prostatectomy (RP) between 2012 and 2017 and then assigned PI-RADSv2 scores for the tumors. PI-RADSv2 category 3 PZ observations upgraded on the basis of abnormal DCE-MRI findings were tabulated, agreement was calculated, and discrepancies were resolved by consensus. The rate of detection of clinically significant cancer among upgraded observations was calculated. Size was measured at consensus review and was compared with pathologic outcomes on the basis of the PI-RADSv2 size threshold of 15 mm or more, with the use of chi-square tests. RESULTS. Reader 1 identified 5.2% (17/326) of DCE-MRI-upgraded PZ observations, and reader 2 identified 8.3% (27/326) of such observations. Interobserver agreement for PI-RADSv2 scoring was moderate (κ = 0.42) overall, but it was fair (κ = 0.23) when only DCE-MRI-upgraded observations were considered. Of the upgraded observations, which had a mean (± SD) size of 14 ± 6 mm (range, 6-29 mm), 10.4% (34/326) were agreed on after consensus review. Size smaller than 15 mm was noted for 61.8% (21/34) of observations. Among DCE-MRI-upgraded PZ observations, true- and false-positive detection rates for significant cancer were 91.2% (31/34) and 8.8% (3/34), respectively. Observations 15 mm or larger had no false-positive diagnoses and higher rates of extraprostatic extension (84.6% [11/13] vs 38.1% [8/21]; p = 0.016); however, there was no difference in GS (p = 0.354) compared with observations less than 15 mm in size. CONCLUSION. PZ observations upgraded on the basis of abnormal DCE-MRI findings have a high likelihood of being clinically significant cancer; however, agreement between readers was low. DCE-MRI-upgraded tumors of 15 mm or larger had no false-positive diagnoses and higher rates of extraprostatic extension, suggesting that they could be assigned to PI-RADSv2 assessment category 5.
引用
- Abreu-Gomez J, Krishna S, Narayanasamy S, et al. Dynamic Contrast-Enhanced MRI-Upgraded Prostate Imaging Reporting and Data System Version 2 Category 3 Peripheral Zone Observations Stratified by a Size Threshold of 15 mm. AJR Am J Roentgenol. 2019. PMID: 31120786. doi:10.2214/AJR.18.21005
供执业医师工作参考,不替代临床判断;数值与分级以指南原文为准。