From the Russell H. Morgan Department of Radiology and Radiological Science (F.D.G., T.S., M.A., C.M., L.M.F.) and Department of Orthopaedic Surgery (K.W.), Johns Hopkins Hospital, 601 N Caroline St, Baltimore, MD 21287.
Radiology. 2014 May;271(2):499-511. doi: 10.1148/radiol.13130844. Epub 2014 Feb 1.
To determine the added value of functional magnetic resonance (MR) sequences (dynamic contrast material-enhanced [DCE] and quantitative diffusion-weighted [DW] imaging with apparent diffusion coefficient [ADC] mapping) for the detection of recurrent soft-tissue sarcomas following surgical resection.
This retrospective study was approved by the institutional review board. The requirement to obtain informed consent was waived. Thirty-seven patients referred for postoperative surveillance after resection of soft-tissue sarcoma (35 with high-grade sarcoma) were studied. Imaging at 3.0 T included conventional (T1-weighted, fluid-sensitive, and contrast-enhanced T1-weighted imaging) and functional (DCE MR imaging, DW imaging with ADC mapping) sequences. Recurrences were confirmed with biopsy or resection. A disease-free state was determined with at least 6 months of follow-up. Two readers independently recorded the signal and morphologic characteristics with conventional sequences, the presence or absence of arterial enhancement at DCE MR imaging, and ADCs of the surgical bed. The accuracy of conventional MR imaging in the detection of recurrence was compared with that with the addition of functional sequences. The Fisher exact and Wilcoxon rank sum tests were used to define the accuracy of imaging features, the Cohen κ and Lin interclass correlation were used to define interobserver variability, and receiver operating characteristic analysis was used to define a threshold to detect recurrence and assess reader confidence after the addition of functional imaging to conventional sequences.
There were six histologically proved recurrences in 37 patients. Sensitivity and specificity of MR imaging in the detection of tumor recurrence were 100% (six of six patients) and 52% (16 of 31 patients), respectively, with conventional sequences, 100% (six of six patients) and 97% (30 of 31 patients) with the addition of DCE MR imaging, and 60% (three of five patients) and 97% (30 of 31 patients) with the addition of DW imaging and ADC mapping. The average ADC of recurrence (1.08 mm(2)/sec ± 0.19) was significantly different from those of postoperative scarring (0.9 mm(2)/sec ± 0.00) and hematomas (2.34 mm(2)/sec ± 0.72) (P = .03 for both).
The addition of functional MR sequences to a routine MR protocol, in particular DCE MR imaging, offers a specificity of more than 95% for distinguishing recurrent sarcoma from postsurgical scarring.
确定功能磁共振（MR）序列（动态对比增强[DCE]和定量扩散加权[DW]成像与表观扩散系数[ADC]图）在检测手术后软组织肉瘤复发中的附加价值。
本回顾性研究经机构审查委员会批准，豁免了获得知情同意的要求。研究了 37 例因软组织肉瘤切除后接受术后监测的患者（35 例为高级别肉瘤）。在 3.0T 进行常规（T1 加权、液体敏感和对比增强 T1 加权成像）和功能（DCE MR 成像、ADC 图 DW 成像）序列成像。通过活检或切除确认复发。通过至少 6 个月的随访确定无病状态。两位读者分别记录常规序列的信号和形态特征、DCE MR 成像的动脉增强、手术床的 ADC 值。比较常规 MR 成像检测复发的准确性与添加功能序列的准确性。采用 Fisher 精确检验和 Wilcoxon 秩和检验确定成像特征的准确性，采用 Cohen κ 和 Lin 组内相关系数确定观察者间的可变性，采用受试者工作特征分析确定检测复发的阈值，并评估在常规序列中添加功能成像后读者的信心。
37 例患者中有 6 例经组织学证实为复发。常规序列检测肿瘤复发的敏感性和特异性分别为 100%（6 例均为阳性）和 52%（31 例中 16 例为阳性），添加 DCE MR 成像后分别为 100%（6 例均为阳性）和 97%（31 例中 30 例为阳性），添加 DW 成像和 ADC 映射后分别为 60%（5 例中 3 例为阳性）和 97%（31 例中 30 例为阳性）。复发的平均 ADC 值（1.08mm²/sec±0.19）与术后瘢痕（0.9mm²/sec±0.00）和血肿（2.34mm²/sec±0.72）明显不同（均 P=.03）。
在常规 MR 方案中添加功能 MR 序列，特别是 DCE MR 成像，对区分复发性肉瘤与术后瘢痕具有超过 95%的特异性。
