Priola Adriano Massimiliano, Priola Sandro Massimo, Giraudo Maria Teresa, Gned Dario, Fornari Alessandro, Ferrero Bruno, Ducco Lorena, Veltri Andrea
Department of Diagnostic Imaging, San Luigi Gonzaga University Hospital, Regione Gonzole 10, 10043, Orbassano (Torino), Italy.
Department of Mathematics "Giuseppe Peano", University of Torino, Via Carlo Alberto 10, 10123, Torino, Italy.
Eur Radiol. 2016 Jul;26(7):2126-38. doi: 10.1007/s00330-015-4031-6. Epub 2015 Oct 1.
To evaluate the usefulness of diffusion-weighted magnetic resonance for distinguishing thymomas according to WHO and Masaoka-Koga classifications and in predicting disease-free survival (DFS) by using the apparent diffusion coefficient (ADC).
Forty-one patients were grouped based on WHO (low-risk vs. high-risk) and Masaoka-Koga (early vs. advanced) classifications. For prognosis, seven patients with recurrence at follow-up were grouped separately from healthy subjects. Differences on ADC levels between groups were tested using Student-t testing. Logistic regression models and areas under the ROC curve (AUROC) were estimated.
Mean ADC values were different between groups of WHO (low-risk = 1.58 ± 0.20 × 10(-3)mm(2)/sec; high-risk = 1.21 ± 0.23 × 10(-3)mm(2)/sec; p < 0.0001) and Masaoka-Koga (early = 1.43 ± 0.26 × 10(-3)mm(2)/sec; advanced = 1.31 ± 0.31 × 10(-3)mm(2)/sec; p = 0.016) classifications. Mean ADC of type-B3 (1.05 ± 0.17 × 10(-3)mm(2)/sec) was lower than type-B2 (1.32 ± 0.20 × 10(-3)mm(2)/sec; p = 0.023). AUROC in discriminating groups was 0.864 for WHO classification (cut-point = 1.309 × 10(-3)mm(2)/sec; accuracy = 78.1 %) and 0.730 for Masaoka-Koga classification (cut-point = 1.243 × 10(-3)mm(2)/sec; accuracy = 73.2 %). Logistic regression models and two-way ANOVA were significant for WHO classification (odds ratio[OR] = 0.93, p = 0.007; p < 0.001), but not for Masaoka-Koga classification (OR = 0.98, p = 0.31; p = 0.38). ADC levels were significantly associated with DFS recurrence rate being higher for patients with ADC ≤ 1.299 × 10(-3)mm(2)/sec (p = 0.001; AUROC, 0.834; accuracy = 78.0 %).
ADC helps to differentiate high-risk from low-risk thymomas and discriminates the more aggressive type-B3. Primary tumour ADC is a prognostic indicator of recurrence.
• DW-MRI is useful in characterizing thymomas and in predicting disease-free survival. • ADC can differentiate low-risk from high-risk thymomas based on different histological composition • The cutoff-ADC-value of 1.309 × 10 (-3) mm (2) /sec is proposed as optimal cut-point for this differentiation • The ADC ability in predicting Masaoka-Koga stage is uncertain and needs further validations • ADC has prognostic value on disease-free survival and helps in stratification of risk.
评估扩散加权磁共振成像在根据世界卫生组织（WHO）和正冈-古贺（Masaoka-Koga）分类区分胸腺瘤以及通过表观扩散系数（ADC）预测无病生存期（DFS）方面的作用。
41例患者根据WHO（低风险与高风险）和Masaoka-Koga（早期与晚期）分类进行分组。为评估预后，将7例随访时复发的患者与健康受试者单独分组。采用Student-t检验比较各组间ADC水平的差异。估计逻辑回归模型和ROC曲线下面积（AUROC）。
WHO分类组间平均ADC值不同（低风险组 = 1.58 ± 0.20×10⁻³mm²/秒；高风险组 = 1.21 ± 0.23×10⁻³mm²/秒；p < 0.0001），Masaoka-Koga分类组间也不同（早期组 = 1.43 ± 0.26×10⁻³mm²/秒；晚期组 = 1.31 ± 0.31×10⁻³mm²/秒；p = 0.016）。B3型胸腺瘤的平均ADC值（1.05 ± 0.17×10⁻³mm²/秒）低于B2型（1.32 ± 0.20×10⁻³mm²/秒；p = 0.023）。WHO分类中区分各组的AUROC为0.864（截断点 = 1.309×10⁻³mm²/秒；准确率 = 78.1%），Masaoka-Koga分类中为0.730（截断点 = 1.243×10⁻³mm²/秒；准确率 = 73.2%）。逻辑回归模型和双向方差分析对WHO分类有显著意义（优势比[OR] = （此处原文似乎有误，推测应为0.93），p = 0.007；p < 0.001），但对Masaoka-Koga分类无显著意义（OR = 0.98，p = 0.31；p = 0.38）。ADC水平与DFS显著相关，ADC≤1.299×10⁻³mm²/秒的患者复发率更高（p = 0.001；AUROC，0.834；准确率 = 78.0%）。
ADC有助于区分高风险和低风险胸腺瘤，并鉴别侵袭性更强的B3型。原发肿瘤ADC是复发的预后指标。
• 扩散加权磁共振成像在胸腺瘤特征描述和预测无病生存期方面有用。• ADC可根据不同组织学构成区分低风险和高风险胸腺瘤。• 提出1.309×10⁻³mm²/秒的ADC截断值作为这种区分的最佳截断点。• ADC预测Masaoka-Koga分期的能力不确定，需要进一步验证。• ADC对无病生存期有预后价值，有助于风险分层。
