Endocrine Abstracts

A 71-year-old man was diagnosed with poorly differentiated T4N0M0 gastric adenocarcinoma. He received neoadjuvant chemotherapy, followed by elective subtotal gastrectomy. Mean arterial pressure was maintained above 70 mmHg throughout the 5-hour operation. On post-operative day 3, the patient became confused, pyrexial, hypotensive with new-onset atrial fibrillation. CT head was unremarkable and CT chest/abdomen/pelvis (CAP) showed bilateral pleural effusions with normal adrenal glands. He was treated for chest sepsis with IV antibiotics with improvement. On post-operative day 5, a repeat CT CAP, which was done to rule out gastric leak from the surgery site, due to worsening sepsis, showed new, bilateral adrenal haemorrhages. The patient had no history of tuberculosis or recent travel outside the UK.

Investigations: Adrenal function was interrogated with a short synacthen test (SST).

Results and treatment: Baseline serum cortisol was 474 nmol/l with an ACTH of 28.6 ng/l. The SST showed serum cortisol increment to 524 nmol/l and 599 nmol/l at 30 and 60 minutes respectively. Albumin at the time was 24 g/l (NR 35–50) and total protein 49 g/l (NR 60–80). The patient subsequently became severely septic with persistent hypotension, despite on-going antibiotic and fluid therapy, and was started on 100 mg intravenous hydrocortisone qds. His repeat SST whilst still on ITU 6 weeks later shows baseline cortisol 1416 nmol/l, cortisol 1333 nmol/l at 30 minutes and 1259 nmol/l at 60 minutes, in the context of albumin of 16 g/l. A repeat CT shows reabsorption of the previous haemorrhages and has revealed bilateral adrenal adenomas.

Conclusions and discussion: At first glance, the baseline serum cortisol was within normal reported range. However, that is for tests performed under ‘normal’ circumstances, ie not in critical illness. On the other hand, serum total cortisol levels in the presence of hypoproteinaemia, which may be present in critical illness, can be highly variable and the reported incidence of ‘adrenal insufficiency’ in sepsis and septic shock is between 30 and 70%. Furthermore, there is evidence for treating patients with septic shock and adrenal insufficiency with stress doses of steroids. The actual cortisol increment in this patient after administration of synachthen was suboptimal, both by ‘normal’ reported values (requires increment >=150 nmol/l) but also in the context of critical illness (requires increment >=250 nmol/l). The finding of adrenal haemorrhages further contributed to the decision for steroid therapy. Free cortisol index calculations can further assist in decisions towards steroid therapy.