Diabetic ketoacidosis (DKA) is a cardinal feature of type 1 diabetes. However, there is a strong, almost dogmatic, errant perception by physicians that DKA is a complication that only occurs in patients with type 1 diabetes. This is not true. DKA does occur in type 2 diabetes; however, it rarely occurs in the absence of a precipitating event.
R.T., a 25-year-old African-American man with type 2 diabetes presented with a 5-day history of nausea and vomiting. He also reported a 2-week history of polyuria and polydipsia and a 10-lb weight loss. A review of symptoms was pertinent for a 5-day history of persistent lower back pain.
The patient was diagnosed with type 2 diabetes 5 years ago when he presented to a different hospital with symptoms of polyuria, polydipsia, and weight loss. He was given a prescription for a sulfonylurea, which he says he took until his initial prescription ran out 1 month later. He had not taken any other medication since that time.
Physical examination revealed an afebrile, obese man (BMI 40 kg/m2) with prominent acanthosis nigricans, no retinopathy by direct funduscopic exam, and a normal neurological exam, including motor function and sensation. The patient had no tenderness to palpation over the lumbrosacral spine or paraspinous muscles despite his complaint of lower back pain.
The laboratory data showed an anion gap, metabolic acidosis, and hyperglycemia (pH of 7.14, anion gap of 24, bicarbonate 6 mmol/l, urinary ketones 150 mg/dl, glucose 314 mg/dl) consistent with the diagnosis of DKA. His white blood count was 20,400/μl. Urinalysis demonstrated no evidence of infection. The patient's hemoglobin A1c (A1C) was 13.5%.
The patient was admitted and treated aggressively with intravenous fluid and an insulin-glucose infusion. A non-contrast magnetic resonance imaging(MRI) of the lumbosacral spine (L-spine) was obtained because of the patient's persistent complaint of lower back pain. The L-spine MRI results were negative for pathology. However, R.T. reported increasing discomfort and now noted weakness and numbness in his bilateral lower extremities.
Neurology was consulted, and during their assessment, the patient became incontinent and was found to have 0/5 strength in the lower extremities,severely compromised sensation, and decreased rectal tone. A contrast MRI of both the thoracic and lumbar spine was ordered, and the patient was found to have a T10-T12 epidural abscess (Figure 1).
The patient's antibiotic coverage was broadly expanded, high-dose intravenous steroids were initiated, and neurosurgery was urgently consulted. Emergent evacuation of the epidural abscess with laminectomies of T10-T12 was performed without complication.
R.T.'s neurogenic bladder resolved without further intervention. After intensive inpatient rehabilitation, he had 3/5 strength in bilateral lower extremities and was still unable to ambulate.
S.D., a 39-year-old white man with type 2 diabetes and mild mental retardation, presented with a 3-week history of polyuria and polydipsia, as well as dysuria, left hip pain, and a feeling of incomplete bladder emptying. Because of the severity of his left hip discomfort, the patient required a cane to ambulate.
The patient was diagnosed with type 2 diabetes 4 years ago on the basis of an elevated fasting blood glucose level during a routine medical examination. He was started on oral hypoglycemic agents, but he discontinued them after 1 month because he was unable to pay for them.
On physical exam, S.D. was afebrile but tachycardic (heart rate 131 bpm)and hypertensive (blood pressure 192/118 mmHg). General examination revealed a wasted, severely volume-depleted man. Thrush was observed on oropharyngeal exam. Cardiopulmonary and abdominal examinations were unremarkable. The patient had point tenderness on the anterior aspect of his left hip. Rectal examination revealed a non-tender prostate.
The laboratory data showed an anion gap, metabolic acidosis, and hyperglycemia (pH 7.24, bicarbonate 9 mmol/l, anion gap 24, urinary ketones 150 mg/dl, and glucose 322 mg/dl) consistent with the diagnosis of DKA. Urinalysis was remarkable for large blood, 4+ bacteria, and > 400 white blood cells. S.D.'s serum white blood count was 22,200, and his erythrocyte sedimentation rate was 109 mm/hour. His A1C result was 12.6%.
The patient was admitted and treated with intravenous fluids and an insulin-glucose infusion. Cultures were obtained. S.D. was started empirically on ticarcillin/clavulanic acid because of concern for left hip osteomyelitis and complicated urinary tract infection. An MRI of the left hip was ordered to evaluate for suspected osteomyelitis. Unexpectedly, it revealed left hip myonecrosis and a large loculated prostatic abscess(Figure 2).
Urology was consulted, and the patient underwent transurethral drainage of the prostatic abscess. Methicillin-sensitive Staphylococcus aureusgrew from both blood and urine cultures. S.D. was treated with intravenous antibiotics per culture sensitivities. The myonecrosis was treated conservatively.
The patient recovered well. He was started on subcutaneous insulin and discharged home to complete a 2-week course of intravenous antibiotics.
What is the mechanism of DKA?
Why does DKA occur in type 2 diabetes?
DKA is a cardinal feature of type 1 diabetes, which has led to the widespread errant perception that it is a complication unique to type 1 diabetes. However, it has been repeatedly reported that DKA does occur in patients with type 2 diabetes.1-5 Moreover, as the cases presented here illustrate, it can occur even in patients who were previously insulinindependent.
A recent study evaluating 138 consecutive admissions for DKA at a large academic center observed that 21.7% had type 2 diabetes.6 Nearly 70% of the admissions involved discontinuation of medications, and almost half had an identifiable infection when an intensive search was undertaken.
A review of the mechanism of DKA is important. Ketoacidosis occurs as a function not only of severe insulin deficiency, but also of elevated glucagon levels. Insulin is an anabolic hormone. Severe insulin deficiency results in decreased glucose utilization by muscle and an unregulated increase in lipolysis. This leads to an enhanced delivery of gluconeogenetic precursors(glycerol and alanine) to the liver. Furthermore, removal of the normal suppressive effect of insulin causes glucagon elevation.7,8 Glucagon is a catabolic hormone. Glucagon promotes gluconeogenesis, decreases oxidation of free fatty acids to triglycerides, and promotes hepatic ketogenesis.9
Importantly, the concentration of insulin required to suppress lipolysis is only one-tenth of that required to promote glucose utilization.10 Typically, moderate insulin deficiency (as observed in patients with type 2 diabetes) is associated with sufficient insulin to block lipolysis (and therefore ketoacid formation), but not enough to promote glucose utilization. This leads to hyperglycemia without formation of the ketoacids.
When DKA occurs in patients with type 2 diabetes, the presumed mechanism of ketoacidosis is the combination of relative insulin deficiency and increased secretion of glucagon (as well as other counteregulatory hormones such as cortisol, catecholamines, and growth hormone) in response to stress from 1) overwhelming infection, 2) infarction of tissue, or 3) other severe illness. The elevated catecholamines further suppress insulin secretion to perpetuate a downward spiral. The increased glucagons-to-insulin ratio causes a mismatch that promotes unregulated lipolysis and proteolysis with subsequent uninterrupted formation of ketoacids.
To summarize, DKA is not a unique feature of type 1 diabetes. Though much more common in type 1 diabetes, it does occur in patients with type 2 diabetes, as illustrated by these case reports. However, it is rare for DKA to occur in type 2 diabetes in the absence of some precipitating event. When DKA occurs in an individual with type 2 diabetes, the clinician should “look under the sheets” and initiate an intensive search for the precipitating factor. Once identified, the trigger should be treated promptly and appropriately.
DKA does occur in type 2 diabetes.
DKA in type 2 diabetes rarely occurs without a trigger.
When it does, an intensive search for the precipitating factor should be undertaken.
Brian J. Welch, MD, and Ivana Zib, MD, are fellows in the Division of Endocrinology and Metabolism at the University of Texas Southwestern Medical Center in Dallas.
The authors thank Philip Raskin, MD, for his support and guidance.