Diabetic ketoacidosis in acromegaly: indication for long‐term insulin therapy?

Abstract

Conventionally, diabetic ketoacidosis (DKA) is regarded as the hallmark of type 1 diabetes and indicates pancreatic beta-cell failure and the need for life-long insulin therapy. Though uncommon, DKA may also occur in type 2 diabetes and insulin resistant states such as acromegaly1, 2 and acanthosis nigricans.3 The underlying pathophysiology for DKA in such cases is not well understood. Whether long-term insulin therapy is indicated following DKA in insulin resistant states will, to a large extent, depend on the pancreatic reserve in these individuals. In 1997, we reported in this journal a case of acromegaly presenting with DKA.' We would like to report briefly on the subsequent progress of this patient, which may add to our existing knowledge regarding the pathophysiology of DKA in insulin resistant states. Our patient is a Chinese woman who newly presented (age 22) with severe DKA (pH 6.93, plasma glucose 27 mmol/l and ketonuria) in 1993.1 She was found to have acromegalic features, and an extended oral glucose tolerance test confirmed acromegaly. She was initially treated with octreotide and subsequently had transphenoidal surgery followed by radiotherapy one year later. She was subsequently maintained on bromocriptine (growth hormone level ranged between 4 and 6 mU/l). Her other medication included thyroxine and Nordette. Her daily insulin requirement approached 90 units. Since starting insulin in 1993 (BMI 26), she gained 8 kg of weight (BMI 29 in 2001). In 1998, she developed episodes of hypoglycaemia, and insulin dosage was gradually reduced and later withdrawn completely. A normal short synacthen test ruled out hypocortisolism (cortisol at baseline 91.7 nmol/l, 30 min 727.1 nmol/l). Random growth hormone levels were 4.2–6.0 mU/l, and IgFl was 259 μg/l (NR 84–327 μg/l). On diet alone, her fasting glucose was 8.8 mmol/l, fasting c-peptide 1421 pmol/l (NR 200–960 pmol/l) and HbA1c 7.2%. The unusual feature in this case is that, despite presenting with DKA initially, she became non-insulin dependent after her growth hormone excess was controlled. Her diabetes was adequately controlled on diet alone. Previous case reports of DKA occurring in acromegalic patients documented significant reduction of insulin and, in some cases, complete discontinuation of insulin therapy after normalisation of growth hormone levels.4, 5 Our case provides further evidence that subjects who developed DKA in insulin resistant states may not be insulin-dependent in the long term. The precise pathophysiology underlying DKA in these cases is not fully understood. The development of DKA in acromegaly may indicate a severe insulin resistant state with relative insulin deficiency. This is supported by the elevated fasting c-peptide after insulin withdrawal in our case. Additionally, excessive glucagon, which cannot be suppressed by glucose load in acromegaly, may also play a role in the development of DKA, because it reduces hepatic fructose 2,6-biphosphate, a metabolite that inhibits hepatic gluconeogenesis.6 Furthermore, the increase in glucagon to insulin ratio in portal blood enhances hepatic ketogenesis.7 These mechanisms are likely to be pivotal in the pathogenesis of DKA in acromegaly, because normalisation of growth hormone significantly reduces insulin requirement and, in our case, total withdrawal of insulin was possible. The understanding of the pathophysiology is essential in the management of diabetes. Inappropriate use of insulin in insulin resistant states may lead to weight gain, which may result in worsening glucose intolerance and insulin resistance.