分析磁共振弥散加权成像（DWI）目测信号强度、量化信号强度（SI）值与透明细胞肾癌（CCRCC）的组织分化程度的关系，探讨DWI信号强度评价CCRCC的组织分化程度的价值。

回顾性收集经病理证实的CCRCC患者91例，并根据Fuhrman病理分级Ⅰ～Ⅳ级标准，分为高分化组（Ⅰ级和Ⅱ级，37例）、中分化组（Ⅲ级，32例）、低分化组（Ⅳ级，22例），所有患者均行中腹部MR平扫、增强和DWI检查（1.5 T，b=800 sec/mm²)，分别目测CCRCC的DWI信号强度、测量SI值，采用Kruskal-Wallis秩和检验比较CCRCC的DWI目测信号强度与组织分化程度差异；采用单因素方差比较CCRCC的SI值与组织分化程度的差异；采用Spearman等级相关检验分析CCRCC的组织分化程度与目测信号强度、SI值的相关性；并采用受试者工作特征ROC曲线评价CCRCC的SI值诊断高分化CCRCC、低分化CCRCC的效能。

91例CCRCC的DWI目测信号强度中，43.9%呈明显高信号，30.8%呈中等高信号，25.3%呈等/略高信号。明显高信号组与等/略高信号组的CCRCC的组织分化程度差异有统计学意义（P<0.05）。中等高信号组与等/略高信号组、中等高信号组与明显高信号组的CCRCC组织分化程度差异均无统计学意义（P>0.05）。CCRCC的DWI目测信号强度与组织分化程度呈中等的负相关（r_s=-0.552，P<0.05）。高分化CCRCC的SI值明显低于中、低分化CCRCC，中分化CCRCC的SI亦低于低分化CCRCC（P<0.05）。CCRCC的SI值与组织分化程度呈显著的负相关（r=-0.711，P<0.05）。受试者工作特征ROC曲线分析显示DWI的SI值诊断高分化CCRCC的最佳临界点值为273.7，相应的敏感度与特异度分别67.6%、98.2%；诊断低分化CCRCC的最佳临界点值为378.9，相应的敏感度与特异度分别91.3%、59.1%。

随DWI目测信号强度、SI值升高，CCRCC的组织分化程度降低。DWI目测信号强度及SI值预测CCRCC组织分化程度有一定的临床价值。

To evaluate the relationship between diffusion weighted imaging visual signal intensity and quantitative signal intensity of cell renal cell carcinoma (CCRCC) and histopathological differentiation of CCRCC.

This retrospective analysis included 91 patients with CCRCC confirmed by pathology. All patients were grouped according to the Fuhrman pathological grading system, from Ⅰto Ⅵ. Four grades were merged into three classifications consisting of 37 well-differentiated CCRCCs (Ⅰ and Ⅱ), 32 moderately-differentiated CCRCCs (Ⅲ) and 22 poorly-differentiated CCRCCs (Ⅳ). Magnetic resonance examinations of MR plain scan, LALA dynamic enhanced scan and DWI (1.5T, b value: 800 sec/mm²) were performed. The each visually signal intensity of CCRCC was evaluated and quantitative signal intensity of CCRCC was measured. The Kruskal-Wallis test was used to compare DWI visual signal intensity between the three different histopathological groups. ANOVA was used to compare SI values between the three different histopathological groups. Spearman correlation analysis was used to analyze the correlation between histopathological differentiation of CCRCC and DWI visual signal intensity and SI values. ROC analysis was performed to evaluate the diagnostic efficiency of SI values.

43.9% of CCRCC appeared as obviously hyperintense, 30.8% of CCRCC appeared as moderate hyperintense, and 25.3% of CCRCC appeared as isointense/slight hyperintense to the surrounding renal parenchyma. There was a significant difference between obviously hyperintense and isointense/slight hyperintense in histopathological differentiation (P<0.05). There was no significant difference between isointense/slight hyperintense and moderate hyperintense, and between moderate hyperintense and obviously hyperintense in histopathological differentiation (P>0.05).There was a moderate negative correlation between visually signal intensity and histopathological differentiation of CCRCC (r_s=-0.552; P<0.01). There was a significant difference in DWI signal intensity value among well- differentiated、moderately-differentiated and poorly-differentiated CCRCC (P<0.05). The SI value of moderately differentiated CCRCC was lower than that of poorly-differentiated CCRCC and there was significant difference between the SI value of moderately-differentiated and poorly-differentiated CCRCC. There was a significant negative correlation between SI value and histopathological differentiation of CCRCC (r=-0.711; P<0.01). The ROC curve showed the optimal cutoff point of SI value was 273.7 in diagnosing well-differentiated CCRCC. Taking 273.7 as the threshold value, sensitivity and specificity of differential diagnosis was 67.6% and 98.2%, respectively. The ROC curve showed the optimal cutoff point of SI value was 378.9 in diagnosing poorly-differentiated CCRCC. Taking 378.9 as the threshold value, sensitivity and specificity of differential diagnosis was 91.3% and 59.1%, respectively.

CCRCC tended to show a higher visual signal intensity and quantitative signal intensity on DWI with decreasing histopathologicaldifferentiation (P<0.05). DWI had some practical value in predicting histopathogical differentiation of CCRCC using signal intensity and quantitative signal intensity.