Diagnostic performance of CT versus MR in detecting aldosterone-producing adenoma in primary hyperaldosteronism (Conn's syndrome)

• Published in: European radiology 2004-10, Vol.14 (10), p.1787-1792
• Main Authors: Lingam, R K, Sohaib, S A, Rockall, A G, Isidori, A M, Chew, S, Monson, J P, Grossman, A, Besser, G M, Reznek, R H
• Format: Article
• Language: English
• Subjects: Adenoma - diagnosis; Adenoma - metabolism; Adrenal Gland Neoplasms - diagnosis; Adrenal Gland Neoplasms - metabolism; Adrenal Gland Neoplasms - secondary; Adrenal Glands - pathology; Adult; Aldosterone - metabolism; Female; Humans; Hyperaldosteronism - diagnosis; Hyperaldosteronism - etiology; Hyperplasia; Magnetic Resonance Imaging; Male; Medical research; Middle Aged; Observer Variation; Predictive Value of Tests; Retrospective Studies; ROC Curve; Sensitivity and Specificity; Tomography, X-Ray Computed
• ISSN: 0938-7994; 1432-1084
• DOI: 10.1007/s00330-004-2308-2

The aim of the present study is to compare the diagnostic performance of CT and MR imaging in detecting aldosterone-producing adenoma and to compare the interobserver variability in the detection of an aldosterone-producing adenoma on CT and MR. A retrospective study of 34 patients with primary hyperaldosteronism was performed. A total of 17 cases of aldosterone-producing adenoma and 17 cases of bilateral adrenal hyperplasia were included. The final diagnosis of an adenoma was made by surgery with histological confirmation, whereas that of bilateral adrenal hyperplasia was made on adrenal venous sampling or a good biochemical and clinical response following medical treatment alone and in the absence of a unilateral radiological abnormality. The CT (n=30) and MR (n=24) scans were reviewed independently by two radiologists experienced in adrenal imaging, who were unaware of the cause of the primary hyperaldosteronism. The diagnostic performances of both observers in detecting an aldosterone-producing adenoma on CT and MR imaging were compared. The 16 adenomatous nodules that were detected on imaging ranged from 1 to 4.75 cm in diameter. The calculated sensitivity and specificity for detecting aldosterone-producing adenoma were 87 and 93% for one observer and 85 and 82% for the other observer on CT, and 83 and 83% for one observer and 92 and 92% for the other observer on MR, respectively. Receptor operating characteristics curve analysis showed similar performances of both observers in detecting an aldosterone-producing adenoma on CT and MR imaging. There was good interobserver agreement on CT (k=0.71) and on MR (k=0.67). We have demonstrated comparable diagnostic performance and good interobserver agreement on CT and MR imaging for the detection of aldosterone-producing adenoma.