肾上腺腺瘤与转移瘤的CT鉴别诊断

The role of CT in differentiation of adrenal adenomas from metastases

目的探讨ＣＴ对肾上腺腺瘤与转移瘤的鉴别诊断价值。方法１０９例共１３０个肾上腺腺瘤和转移瘤，分别观察肿瘤的大小、边界、内部结构及其与同侧肾上腺的关系，测量肿瘤的平扫和增强后ＣＴ值及增强程度，２０例做了像素矩阵图分析，应用接受器工作特性（ＲＯＣ）曲线对结果进行分析。结果肿瘤平扫ＣＴ值的ＲＯＣ曲线下面积是０．９０±０．０５，明显大于增强后ＣＴ值（０．８１±０．０４）、增强程度（０．７１±０．０８）和肿瘤大小（０．８０±０．０４）的ＲＯＣ曲线下面积；以肿瘤直径≤３．０ｃｍ，密度均匀，平扫ＣＴ值≤２０ＨＵ，增强后ＣＴ值≤３５ＨＵ，增强程度≤２０ＨＵ为标准，诊断腺瘤的敏感性分别是８６％、８５％、８８％、７６％、和７４％，特异性分别是６８％、６９％、９０％、７４％和５９％；肿瘤与肾上腺相连（６８％）、平扫ＣＴ值＜０ＨＵ、ＣＴ像素矩阵图上有轻度负ＣＴ值区域只见于腺瘤，边界模糊或（和）侵犯周围结构、呈不规则厚环状强化是转移瘤的特征。全组总体符合率为９３％。结论综合分析肿瘤的各种ＣＴ征象，ＣＴ能鉴别绝大多数肾上腺腺瘤和转移瘤。

Purpose To evaluate the diagnostic efficacy of CT in differentiating adrenal adenoma from metastases. Methods 109 patients with 130 adrenal adenomas and metastases were analyzed. Tumor size, margin, internal structure and the relation of the tumor to the adrenal gland were investigated, both precontrast and postcontrast CT attenuation values of the tumor were measured and the enhanced degree was calculated. CT pixel mapping of the tumor was done in 20 patients. Results were analyzed by using ROC curve. Results The area under the ROC curve for precontrast CT attenuation values (0.90 ± 0.05) was significantly larger than that for postcontrast CT attenuation values (0.81 ± 0.04), enhanced degree (0.71±0.08) and tumor sizes (0.80 ± 0.04). With a threshold of tumor size ≤3.0 cm, hemogeneity, precontrast CT attenuation value ≤ 20 HU, postcontrast CT attenuation value ≤ 35 HU or enhanced degree ≤ 20 HU, the sensitiveity for characterizing adenomas were 86%,85%,88%,76% and 74% and the specificity 68%,69%,90%,74% and 59% respectively. Connectivity with normal adrenal gland (68%), precontrast CT attenuation value <0 HU (17%), and area of slightly negative CT attenuation value on CT pixel mapping were only seen in adrenal adenomas, whereas poorly defined margin or(and) invasion of surrounding structure (28%), irreguler thick rim enhancement(24%) were characteristic of metastases. The overall correspondence rate was 93% in our group. Conclusion Most of the adrenal adenomas and metastases can be distinguished by comprehensive analysis of the CT features