From the Departments of Neuroradiology (P.K., S.H., M.B., A.R.), Neuro-oncology (B.W., W.W.), and Neuropathology (F.S.), University of Heidelberg Medical Center, Heidelberg, Germany; German Cancer Consortium (DKTK), Clinical Cooperation Unit Neuro-oncology, German Cancer Research Center (DKFZ), Heidelberg, Germany (B.W., W.W.); DKTK, Clinical Cooperation Unit Neuropathology, DKFZ, Heidelberg, Germany (F.S.); Department of Radiology, DKFZ, Heidelberg, Germany (M.R., H.P.S.); and Department of Neuro-oncologic Imaging (E012), Department of Radiology, DKFZ, Heidelberg, Germany (A.R.).
Radiology. 2014 Sep;272(3):843-50. doi: 10.1148/radiol.14132740. Epub 2014 May 3.
To compare multiparametric diagnostic performance with diffusion-weighted, dynamic susceptibility-weighted contrast material-enhanced perfusion-weighted, and susceptibility-weighted magnetic resonance (MR) imaging for differentiating primary central nervous system lymphoma (PCNSL) and atypical glioblastoma.
This retrospective study was institutional review board-approved and informed consent was waived. Pretreatment MR imaging was performed in 314 patients with glioblastoma, and a subset of 28 patients with glioblastoma of atypical appearance (solid enhancement with no visible necrosis) was selected. Parameters of diffusion-weighted (apparent diffusion coefficient [ADC]), susceptibility-weighted (intratumoral susceptibility signals [ITSS]), and dynamic susceptibility-weighted contrast-enhanced perfusion-weighted (relative cerebral blood volume [rCBV]) imaging were evaluated in these 28 patients with glioblastoma and 19 immunocompetent patients with PCNSL. A two-sample t test and χ(2) test were used to compare parameters.The diagnostic performance for differentiating PCNSL from glioblastoma was evaluated by using logistic regression analyses with leave-one-out cross validation.
Minimum, maximum, and mean ADCs and maximum and mean rCBVs were significantly lower in patients with PCNSL than in those with glioblastoma (P < .01, respectively), whereas mean ADCs and mean rCBVs allowed the best diagnostic performance. Presence of ITSS was significantly lower in patients with PCNSL (32% [six of 19]) than in those with glioblastoma (82% [23 of 28]) (P < .01). Multiparametric assessment of mean ADC, mean rCBV, and presence of ITSS significantly increased the probability for differentiating PCNSL and atypical glioblastoma compared with the evaluation of one or two imaging parameters (P < .01), thereby correctly predicting histologic results in 95% (18 of 19) of patients with PCNSL and 96% (27 of 28) of patients with atypical glioblastoma.
Combined evaluation of mean ADC, mean rCBV, and presence of ITSS allowed reliable differentiation of PCNSL and atypical glioblastoma in most patients, and these results support an integration of advanced MR imaging techniques for the routine diagnostic workup of patients with these tumors.
比较扩散加权、动态对比增强灌注加权和磁敏感加权磁共振成像(MR)在鉴别原发性中枢神经系统淋巴瘤(PCNSL)和非典型胶质母细胞瘤中的多参数诊断性能。
本回顾性研究经机构审查委员会批准,并豁免了知情同意。对 314 例胶质母细胞瘤患者进行了预处理 MR 成像,选择了一组 28 例具有非典型表现(无可见坏死的实性增强)的胶质母细胞瘤患者。在这 28 例胶质母细胞瘤患者和 19 例免疫功能正常的 PCNSL 患者中,评估了扩散加权(表观扩散系数 [ADC])、磁敏感加权(肿瘤内磁敏感信号 [ITSS])和动态对比增强灌注加权(相对脑血容量 [rCBV])成像的参数。使用两样本 t 检验和 χ²检验比较参数。使用具有留一交叉验证的逻辑回归分析评估区分 PCNSL 和胶质母细胞瘤的诊断性能。
PCNSL 患者的最小、最大和平均 ADC 以及最大和平均 rCBV 明显低于胶质母细胞瘤患者(P<.01),而平均 ADC 和平均 rCBV 的诊断性能最佳。PCNSL 患者的 ITSS 存在率(32% [19 例中的 6 例])明显低于胶质母细胞瘤患者(82% [28 例中的 23 例])(P<.01)。与评估一个或两个成像参数相比,平均 ADC、平均 rCBV 和 ITSS 存在的多参数评估显著增加了区分 PCNSL 和非典型胶质母细胞瘤的概率(P<.01),从而正确预测了 95%(19 例中的 18 例)的 PCNSL 患者和 96%(28 例中的 27 例)的非典型胶质母细胞瘤患者的组织学结果。
平均 ADC、平均 rCBV 和 ITSS 存在的综合评估可在大多数患者中可靠地区分 PCNSL 和非典型胶质母细胞瘤,这些结果支持整合先进的 MR 成像技术用于这些肿瘤患者的常规诊断。