快速B1不敏感MR指纹技术联合扩散MRI表征外周带临床显著前列腺癌
结果显示csPCa的T2和ADC均显著低于CIL,T2联合ADC模型AUC达0.90,进一步加入PSAD后AUC提升至0.94。
要点
- MR指纹T2联合ADC鉴别csPCa与CIL的AUC为0.90,加入PSAD后升至0.94。
- MR指纹T2单独阈值58 msec时AUC 0.87、敏感度73%、特异度93%。
- ADC单独阈值761×10⁻⁶ mm²/sec时AUC 0.83、敏感度82%、特异度79%。
关键影像表现/征象
- csPCa(n=22)的平均T2低于CIL(n=29)(56 msec±12 vs 80 msec±20,P<.001)。
- csPCa的平均ADC低于CIL(668×10⁻⁶ mm²/sec±111 vs 876×10⁻⁶ mm²/sec±181,P<.001)。
- 共分析51个PI-RADS v2.1评分≥3分的外周带病灶。
诊断标准与关键数值
- MR指纹T2单变量最优阈值58 msec,鉴别csPCa与CIL的AUC为0.87(95%CI:0.76,0.97),敏感度73%(16/22),特异度93%(27/29)。
- ADC单变量最优阈值761×10⁻⁶ mm²/sec,鉴别csPCa与CIL的AUC为0.83(95%CI:0.71,0.95),敏感度82%(18/22),特异度79%(23/29)。
- T2联合ADC多变量模型AUC为0.90(95%CI:0.81,1.00;P<.001)。
- 进一步纳入血清PSAD后AUC提升至0.94(95%CI:0.86,1.0;P=.01)。
临床应用与阅片要点
- 对PI-RADS≥3分外周带病灶,可参考MR指纹T2和ADC定量值辅助鉴别csPCa与CIL。
- MR指纹T2阈值58 msec特异度较高(93%),有助于减少假阳性。
- ADC阈值761×10⁻⁶ mm²/sec敏感度较高(82%),有助于减少漏诊。
- 联合T2、ADC及PSAD可进一步提升诊断性能。
局限性
- csPCa组仅22例,CIL组29例,统计效能有限。
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
Background MR fingerprinting (MRF) has shown utility in focal prostate lesion characterization as a complement to standard prostate MRI. Purpose To determine the utility of a rapid, B1-insensitive MRF technique, combined with standard diffusion MRI, in the characterization of focal prostate lesions in the peripheral zone (PZ), with MRI-guided transrectal US fusion biopsy-derived histopathologic analysis as the reference standard. Materials and Methods Between February 2021 and April 2024, consecutive adults with clinical suspicion of prostate cancer (PCa) who had not undergone biopsy underwent MRI, including rapid, B1-insensitive MRF (15 seconds per section) on a 3.0-T MRI unit. T1 and T2 from MRF and apparent diffusion coefficient (ADC) from diffusion MRI were assessed as independent variables to differentiate clinically significant PCa (csPCa) from clinically insignificant lesions (CILs) (benign lesions and Gleason score of ≤3 + 3 PCa). Both univariable and multivariable logistic regression analyses with receiver operating characteristic curve analyses were performed to calculate the area under the receiver operating curve (AUC), sensitivity, and specificity. Results Fifty-two males (mean age, 65 years ± 11 [SD]) with 51 PZ lesions (Prostate Imaging Reporting and Data System version 2.1, score ≥3) were included. The mean T2 and ADC were lower for csPCa (n = 22) than for CILs (n = 29) (T2, 56 msec ± 12 vs 80 msec ± 20, respectively [P < .001]; ADC, 668 × 10-6 mm2/sec ± 111 vs 876 × 10-6 mm2/sec ± 181, respectively [P < .001]). At univariable analyses, an optimal T2 cutoff of 58 msec yielded an AUC of 0.87 (95% CI: 0.76, 0.97) (sensitivity, 73% [16 of 22]; specificity, 93% [27 of 29]), whereas an ADC cutoff of 761 × 10-6 mm2/sec yielded an AUC of 0.83 (95% CI: 0.71, 0.95) (sensitivity, 82% [18 of 22]; specificity, 79% [23 of 29]). A multivariable model incorporating both T2 and ADC achieved a higher AUC (AUC, 0.90; 95% CI: 0.81, 1.00; P < .001). Further improvement in diagnostic performance was observed when serum prostate-specific antigen density (PSAD) was included (AUC, 0.94; 95% CI: 0.86, 1.0; P = .01). Conclusion Rapid, B1-insensitive prostate MRF-derived T1 and T2 combined with ADC and serum PSAD improved PCa characterization in the PZ. © RSNA, 2026 Supplemental material is available for this article. See also the editorial by Margolis and Gulani in this issue.
引用
- Futela D, Yin Z, Mirak SA, et al. Characterization of Clinically Significant Prostate Cancer in the Peripheral Zone Using Rapid B1-Insensitive MR Fingerprinting. Radiology. 2026. PMID: 41944725. doi:10.1148/radiol.252547
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。