Functioning adrenocortical neoplasms in children

To describe the clinical features, treatment and outcome of functioning adrenocortical tumours in children, and to assess the value of tumour size as a marker for malignant potential.

Twenty-one children (mean age at presentation 4.9 years, range 0.6-11; 12 girls and nine boys) who presented with a functioning adrenal mass over a 29-year period (1972-2000) were assessed. Clinicopathological features and outcome information were obtained from a structured review of the case-notes. Before 1980 the diagnosis and staging was mainly obtained by intravenous pyelography and plain X-ray. Since 1980, ultrasonography, computed tomography and more recently magnetic resonance imaging were the methods of choice.

The mean (range) duration of symptoms before diagnosis was 31 (2-108) months. The symptoms and signs at presentation comprised virilization (18 children), Cushingoid features (three), precocious puberty (three) and hypertension (four). Three children had malignant neoplasms, one presented with metastasis and of the other two, one died from the disease within 6 months, despite adjuvant chemotherapy and radiotherapy. The third girl is disease-free 11 years after complete excision of the neoplasm. Eighteen children had a benign neoplasm and all are alive and free of recurrence. In most of the children surgical extirpation was relatively simple, blood loss minimal, recovery uneventful and the hospital stay short (mean 5 days). The virilizing effects of all the neoplasms gradually resolved after surgery within a median (range) of 27 (6-108) months.

Functioning adrenocortical neoplasms in children are rare, with a peak incidence in the first decade, are predominantly unilateral and more common in girls. Most are benign, but both benign and malignant lesions can present with virilizing and Cushingoid features. The malignant neoplasms have an extremely poor prognosis, especially if excision is incomplete, despite adjuvant chemotherapy and radiotherapy. Although there is an association between tumour size and malignancy, this cannot be used as a reliable individual discriminator.