SEARCH

SEARCH BY CITATION

Abstract

Background

Size is considered the single best predictor of malignancy in evaluating adrenal neoplasms identified incidentally during abdominal imaging, yet small adrenal cortical cancers have been reported previously from several centres.

Methods

This was a retrospective evaluation of the value of tumour size and other preoperative clinical parameters in predicting the presence or absence of malignancy in adult patients who underwent adrenalectomy for an adrenal tumour at this institution.

Results

One hundred and sixteen patients underwent adrenalectomy. Diagnoses were adrenal cortical carcinoma (n = 38), phaeochromocytoma (n = 27), metastasis to the adrenal gland (n = 11) and benign tumours (n = 40) (cortical adenoma, 23; cyst, two; cortical hyperplasia, five; ganglioneuroma, four; myelolipoma, six). The median size of the adrenal cortical carcinomas was 9·2 (range 1·7–30) cm; five tumours were smaller than 5·0 cm. The median overall size of the benign tumours (excluding phaeochromocytomas) was 4·0 cm; the median size of the adrenal cortical adenomas was 3·0 cm. Eleven benign tumours were larger than 5·0 cm, including two cortical adenomas. The radiographic features of three of five small adrenal cancers correctly predicted malignancy; the remaining two patients with small cancers had symptoms related to functioning tumours. Conversely, the radiographic features of seven of 11 large benign adrenal tumours correctly predicted benign histology, including five of five myelolipomas.

Conclusion

While size remains a good predictor of histology and clinical behaviour of functioning and non-functioning adrenal neoplasms, both small adrenal cortical cancers and large benign tumours occur with measurable frequency. High-quality imaging studies (computed tomography and magnetic resonance imaging) may be helpful in identifying relatively small yet radiographically atypical adrenal tumours more likely to be cancers, as well as radiographically characteristic benign myelolipomas which may be followed selectively. © 2000 British Journal of Surgery Society Ltd