From the Departments of Radiology (S.B., J.T., S.D., H.P.S.) and Medical Physics in Radiology (F.B.L.), German Cancer Research Center, Heidelberg Im Neuenheimer Feld 280, 69120 Heidelberg, Germany; Radiological Clinic at the ATOS Clinic Heidelberg, Heidelberg, Germany (W.L.); Radiology Center Mannheim, Mannheim, Germany (H.D.); and Department of Clinical and Interventional Radiology, University Hospital Heidelberg, Heidelberg, Germany (A.S.).
Radiology. 2016 Mar;278(3):689-97. doi: 10.1148/radiol.2015150425. Epub 2015 Sep 29.
To evaluate the ability of a diagnostic abbreviated magnetic resonance (MR) imaging protocol consisting of maximum intensity projections (MIPs) from diffusion-weighted imaging with background suppression (DWIBS) and unenhanced morphologic sequences to help predict the likelihood of malignancy on suspicious screening x-ray mammograms, as compared with an abbreviated contrast material-enhanced MR imaging protocol and a full diagnostic breast MR imaging protocol.
This prospective institutional review board-approved study included 50 women (mean age, 57.1 years; range, 50-69 years), who gave informed consent and who had suspicious screening mammograms and an indication for biopsy, from September 2014 to January 2015. Before biopsy, full diagnostic contrast-enhanced MR imaging was performed that included DWIBS (b = 1500 sec/mm(2)). Two abbreviated protocols (APs) based on MIPs were evaluated regarding the potential to exclude malignancy: DWIBS (AP1) and subtraction images from the first postcontrast and the unenhanced series (AP2). Diagnostic indexes of both methods were examined by using the McNemar test and were compared with those of the full diagnostic protocol and histopathologic findings.
Twenty-four of 50 participants had a breast carcinoma. With AP1 (DWIBS), the sensitivity was 0.92 (95% confidence interval [CI]: 0.73, 0.98), the specificity was 0.94 (95% CI: 0.77, 0.99), the negative predictive value (NPV) was 0.92 (95% CI: 0.75, 0.99), and the positive predictive value (PPV) was 0.93 (95% CI: 0.75, 0.99). The mean reading time was 29.7 seconds (range, 4.9-110.0 seconds) and was less than 3 seconds (range, 1.2-7.6 seconds) in the absence of suspicious findings on the DWIBS MIPs. With the AP2 protocol, the sensitivity was 0.85 (95% CI: 0.78, 0.95), the specificity was 0.90 (95% CI: 0.72, 0.97), the NPV was 0.87 (95% CI: 0.69, 0.95), the PPV was 0.89 (95% CI: 0.69, 0.97), and the mean reading time was 29.6 seconds (range, 6.0-100.0 seconds).
Unenhanced diagnostic MR imaging (DWIBS mammography), with an NPV of 0.92 and an acquisition time of less than 7 minutes, could help exclude malignancy in women with suspicious x-ray screening mammograms. The method has the potential to reduce unnecessary invasive procedures and emotional distress for breast cancer screening participants if it is used as a complement after the regular screening clarification procedure.
评估一种包括扩散加权成像（DWIBS）最大强度投影（MIP）和未增强形态序列的诊断性缩短磁共振成像（MR）方案的能力，以帮助预测可疑筛查 X 射线乳房 X 光片中恶性肿瘤的可能性，与缩短的对比增强 MR 成像方案和完整的诊断性乳房 MR 成像方案相比。
本前瞻性机构审查委员会批准的研究包括 50 名女性（平均年龄 57.1 岁；范围，50-69 岁），她们在 2014 年 9 月至 2015 年 1 月期间知情同意并进行了可疑筛查乳房 X 光检查和活检指征。在活检之前，进行了完整的诊断性对比增强磁共振成像，包括 DWIBS（b = 1500 sec/mm(2)）。评估了两种基于 MIP 的缩短方案（AP）排除恶性肿瘤的能力：DWIBS（AP1）和来自初次对比增强后和未增强系列的减影图像（AP2）。使用 McNemar 检验检查两种方法的诊断指标，并与完整的诊断方案和组织病理学发现进行比较。
50 名参与者中有 24 名患有乳腺癌。使用 AP1（DWIBS），灵敏度为 0.92（95%置信区间[CI]：0.73，0.98），特异性为 0.94（95%CI：0.77，0.99），阴性预测值（NPV）为 0.92（95%CI：0.75，0.99），阳性预测值（PPV）为 0.93（95%CI：0.75，0.99）。平均阅读时间为 29.7 秒（范围，4.9-110.0 秒），在 DWIBS MIP 上没有可疑发现时，阅读时间少于 3 秒（范围，1.2-7.6 秒）。使用 AP2 方案，灵敏度为 0.85（95%CI：0.78，0.95），特异性为 0.90（95%CI：0.72，0.97），NPV 为 0.87（95%CI：0.69，0.95），PPV 为 0.89（95%CI：0.69，0.97），平均阅读时间为 29.6 秒（范围，6.0-100.0 秒）。
未增强的诊断性 MR 成像（DWIBS 乳房 X 线摄影），NPV 为 0.92，采集时间少于 7 分钟，可帮助可疑 X 射线筛查乳房 X 光片中的女性排除恶性肿瘤。如果该方法在常规筛查澄清程序之后用作补充，则有可能减少乳腺癌筛查参与者的不必要的侵入性程序和情绪困扰。
