MES2008 Poster Presentations (1) (41 abstracts)
A 60-year-old gentleman who had previously undergone a right nephrectomy for renal cell carcinoma was admitted electively for a left adrenalectomy due to metastatic disease. Prior to this he had been treated with immunotherapy (Sunitinib) and radiotherapy for pulmonary and bony metastases respectively. He was given perioperative cover with hydrocortisone. A short synacthen test (SST) performed the morning after discontinuing hydrocortisone showed a baseline cortisol of 406 nmol/l rising to 469 nmol/l at 30 min and 555 nmol/l at 60 min. He was clinically euadrenal, so it was decided that he did not require further hydrocortisone. He was discharged home 3 days later. Three weeks later, he presented with profound nausea and tiredness. On examination he had postural hypotension. Plasma sodium was 129 mmol/l, potassium 4.4 mmol/l and random cortisol 689 nmol/l. He was started on hydrocortisone 20 mg, 10 mg, 10 mg and fludrocortisone 100 mcg once daily. His symptoms resolved over the next 48 h, and the hydrocortisone was reduced to 10 mg, 5 mg, 5 mg prior to review in endocrinology out-patients. At out-patient review, it was noted that the baseline ACTH from his earlier SST was elevated at 52.0 ng/l suggesting subclinical hypocortisolaemia. The dose of his hydrocortisone was increased to 15 mg, 10 mg, 5 mg, as he is about to start further immunotherapy (Bevacizmab). We present a case of what we believe to be adrenal toxicity due to Sunitinib, with clinical findings and investigations suggesting predominantly mineralocorticoid deficiency. Adrenal toxicity with Sunitinib has been reported previously in animals but not humans. Such subclinical toxicity may be difficult to detect unless unmasked by physiological stressors. We highlight the importance of knowing both a baseline ACTH and peak cortisol when interpreting a SST, and also the difficulty in determining what is an appropriate random cortisol in an acutely unwell patient. Another potential explanation for the discordance between cortisol results and his clinical state is abnormal cortisol binding globulin levels, which must be considered.