Triamcinolone-augmented glucose tolerance in non-diabetic patients with chronic pancreatitis.

Seven non-diabetic patients with chronic pancreatitis were shown to have a diminished acute insulin secretory response after intensive beta cell and intravenous tolbutamide stimulation. In an attempt to unmask their 'latent' diabetic state, triamcinolone-augmented glucose tolerance tests were performed some days after documenting normal standard 50 g oral glucose tolerance tests. A matched group of non-diabetic controls was similarly investigated. Although the steroid-augmented glucose tolerance tests showed marked impairment in the patients, becoming frankly diabetic in three cases, the normal control subjects reacted in a similar though less striking fashion. There was no significant difference between the mean glucose values in the two groups. The ability of patients with chronic pancreatitis to maintain normal glucose tolerance in the face of diminished insulin output is commented on. We conclude that, as an ancillary investigation for diagnosing chronic pancreatitis, the triamcinolone glucose tolerance test is unreliable.