Case Study: Tea-Colored Urine in a Patient With Diabetic Ketoacidosis
C.S., a previously healthy 26-year-old Cantonese woman, presented with 6 days of nausea, vomiting, occasional fevers, and nonspecific abdominal pain that progressed to myalgias, polyuria, polydipsia, fatigue, and delirium. She had been asleep in bed for many hours preceding admission. Her only medication was an oral contraceptive, and she did not use herbals or alcohol. Her father had type 2 diabetes.
In the emergency room, she was mildly delirious, hyperventilating, afebrile, and dehydrated. She weighed 133 lb, her pulse was 110 bpm, and her blood pressure was 102/62 mmHg. The physical exam was otherwise normal.
Urinalysis revealed 4+ glucose and ketones, large blood without red blood cells, and no infection. The urine was tea-colored. Plasma glucose was 809 mg/dl, betahydroxybutyrate 6.9 mmol/l, lactate 3.2 mmol/l, sodium 126 mmol/l, potassium 4.9 mmol/l, creatinine 2.2 mg/dl, carbon dioxide 9 mmol/l, leukocytes 24,800/ml, arterial pH 7.06, and partial pressure of arterial carbon dioxide (PaCO2) was 15 mmHg. Toxicology screen was negative, phosphate was 3.0 mmol/l, alanine aminotransferase was 299 IU/l, and aspartate aminotransferase was 782 IU/l, with normal bilirubin and alkaline phosphatase levels. Creatine kinase (CK) was 46,305 IU/l. Troponin I peaked at 5.8 ng/ml, but C.S. reported no chest pain, and electrocardiogram showed only sinus tachycardia. Islet cell, insulin, and glutamic acid decarboxylase antibodies were negative. Her hemoglobin A1c was 6.7%.
Treating physicians diagnosed diabetic ketoacidosis (DKA) and rhabdomyolysis and initiated intravenous insulin and fluids with bicarbonate. Within 18 hours, C.S. was alert and oriented, and the metabolic abnormalities had been corrected. She was then transitioned to glargine insulin with pre-meal lispro insulin. CK peaked at 51,330 IU/l, but serum creatinine normalized. Blood and urine cultures and serology for acute cytomegalovirus and Epstein-Barr virus infections were negative.
How often is rhabdomyolysis associated with DKA?
What is the mechanism of DKA-mediated muscle injury?
Should the management of DKA be modified when there is co-existing rhabdomyolysis?
Do patients who present with DKA and rhabdomyolysis have a worse prognosis than patients presenting with DKA only?
Rhabdomyolysis occurs in as many as 50% of patients presenting with DKA or the hyperglycemic hyperosmolar nonketotic syndrome (HHNK) and varies in severity from mildly elevated CK levels with no symptoms to markedly elevated CK with acute renal failure, possibly requiring hemodialysis.1-3 DKA and HHNK patients with rhabdomyolysis have higher blood glucose concentrations, serum osmolalities, and serum creatinine measurements than do those without rhabdomyolysis.1-3 However, the mechanism of DKA-mediated muscle injury is uncertain. Theories include insufficient energy delivery to muscle, hyperosmolar effects, and underlying metabolic defects, such as McArdle's (myophosphorylase deficiency causing glycogen accumulation and reduced muscle ATP generation, usually characterized by fatigue, exercise intolerance, and myalgias).1
Although CK measurements are often obtained in older patients to rule out myocardial infarction as a precipitant for DKA, CK levels are not routinely tested in younger patients because their likelihood of myocardial ischemia is low. Furthermore, troponin measurements, which are more specific markers for myocardial injury, are replacing CK testing in many centers. Therefore, rhabdomyolysis associated with DKA may be overlooked, resulting in further complications, such as renal failure.
Nonetheless, rhabdomyolysis is important to diagnose because significant potential complications, such as acute renal failure, may be averted with appropriate therapy. Although controlled clinical trials are lacking, case series and animal data support the use of bicarbonate therapy, aggressive fluid resuscitation, and possibly mannitol infusion in patients with severe rhabdomyolysis to prevent acute renal failure, which may occur in as many as one-fourth of patients with rhabdomyolysis and HHNK.3-5
Rhabdomyolysis may also be associated with a high 1-week mortality rate in patients with DKA and HHNK. One report2 described a fourfold higher 1-week mortality in patients with DKA and rhabdomyolysis compared to patients with DKA only (38.5 vs. 9.7%). Patients with HHNK and rhabdomyolysis also had a higher mortality rate compared to patients with HHNK only (35.5 vs. 25.4%).
C.S.'s relative immobilization possibly contributed to the extreme CK elevation. Perhaps she had a Coxsackie virus infection, characterized by fever, nausea, vomiting, myocarditis, rhabdomyolysis, and new-onset diabetes. Unfortunately, we considered this diagnosis when laboratory testing to confirm it was no longer possible.
Rhabdomyolysis occurs commonly in patients presenting with DKA but is usually subclinical
Patients who present with DKA and severe rhabdomyolysis may have higher short-term mortality rates than patients presenting with DKA only.
Serum CK is recommended when DKA patients have other risk factors for rhabdomyolysis (prolonged bed rest, significant alcohol consumption, drug use, and toxin exposure) or signs and symptoms of rhabdomyolysis (myalgias, urine dipstick with heme but no red blood cells, or tea-colored urine).
Consider adding bicarbonate and mannitol to insulin and intravenous fluids when DKA patients present with severe rhabdomyolysis.
Hylton V. Joffe, MD, is a clinical and research fellow in the Division of Endocrinology, Diabetes, and Hypertension at the Brigham and Women's Hospital and Harvard Medical School, and Martin J. Abrahamson, MD, is acting chief medical officer at the Joslin Diabetes Center and an associate professor of medicine at Harvard Medical School, in Boston, Mass.
Note of disclosure: Dr. Abrahamson has received honoraria for speaking engagements from Aventis, which makes glargine, and Eli Lilly, which makes lispro insulin.
- American Diabetes Association