Background: The size of adrenal tumour plays an important role in the indic
ations for surgical excision of nonfunctioning adrenal tumours and in selec
ting the best surgical approach. Computed tomography (CT) has been reported
to underestimate the real size of adrenal lesions. The accuracy of magneti
c resonance imaging (MRI) in predicting the true tumour size has not been p
reviously investigated. The present retrospective study investigates the ac
curacy of MRI and CT in the pre-operative determination of true adrenal tum
our size.
Methods: The medical records of 65 patients who underwent adrenalectomy for
an adrenal mass were reviewed. The size of adrenal tumours as determined b
y pre-operative MRI and/or CT was compared with the 'true' histopathologica
l size. The impact of histological diagnosis on size estimation was also in
vestigated.
Results: The median age at diagnosis was 42 years (range 1-82 years) and mo
re patients were female (60%). Five patients had bilateral adrenalectomy, t
hus giving rise to 70 adrenal specimens. The histopathological size of adre
nal tumours ranged from 0.9 to 26 cm with a mean of 5.96 cm and a median of
4.70 cm. For tumours larger than 3 cm, MRI significantly underestimated th
e real tumour size by 20% (P <0.001). CT also underestimated the size of su
ch tumours by 18.1% (P <0.003). Adrenal phaeochromocytomas were consistentl
y underestimated by both modalities.
Conclusions: MRI and CT significantly underestimated the true size of adren
al tumours larger than 3 cm by 20% and 18%, respectively. Surgeons and endo
crinologists should interpret the pre-operative size of adrenal lesions wit
h caution. (C) 2001 Harcourt Publishers Ltd.