To the Editor: Olanzapine treatment has been associated with impaired glucose metabolism, clinical diabetes mellitus, and ketoacidosis, as noted in a number of case reports (1–5). In many instances, glucose homeostasis improved or returned to normal after discontinuation of olanzapine. We found one report of ketoacidosis developing in a 12-year-old who stopped taking then restarted olanzapine; the child’s glucose level then returned to normal (6). We report on a patient who never stopped taking olanzapine after developing diabetic ketoacidosis and required gradually decreasing amounts of insulin until the illness was controlled by diet.
Mr. A was a 49-year-old Caucasian man with a 22-year history of undifferentiated schizophrenia and no family history of diabetes mellitus. He was also obese. He participated in a phase III olanzapine trial at a Department of Veterans Affairs medical center. Previous treatments had included haloperidol, lithium, and trihexyphenidyl. An initial work-up revealed a random glucose level of 89 mg/dl, normal urinalysis results, and a body mass index of 34.2 kg/m2. His history and the results of a physical examination did not suggest abnormal glucose metabolism. Mr. A’s condition was maintained with olanzapine treatment, 20 mg/day, which considerably improved his symptoms.
Eleven months later, tests conducted during a scheduled follow-up visit revealed an increase in body mass index to 37.8 kg/m2 and a 34-lb weight gain since Mr. A started taking olanzapine. He then experienced nausea, vomiting, polyuria, and polydipsia and went to the emergency room, where a serum glucose level of 766 mg/dl was detected. Mr. A left the hospital against medical advice and without treatment; he returned 8 days later. The results of a work-up at the time showed a serum glucose level of 368 mg/dl, a urinalysis with 4+ ketones and 4+ glucose, a pH of 7.10, and an anion gap of 29.0. He spent a day in the intensive care unit with ketoacidosis, received intravenous fluids and insulin, was transferred to a medical unit, and was ultimately discharged while receiving 94 U of regular and neutral protamine Hagedorn insulin per day.
After we obtained several low-normal serum glucose levels, we decreased his daily insulin dose to 78 U. His insulin requirement continued to decrease until he required no further medication for control of glucose levels. Since he started taking olanzapine, his body mass index has fluctuated between 34.2 and 37.8 kg/m2; his hemoglobin HbAlc levels have been measured at between 5.4% and 6.5%. Mr. A’s serum glucose levels in the last year have ranged from 112 to 164 mg/dl. His latest serum insulin level was 56.3 μU/ml (normal range=0–30), and his C-peptide level was 10.7 ng/ml (normal range=0.9–4.0). Mr. A continues to take his initial olanzapine dose of 20 mg/day and has experienced no new medical or psychiatric sequelae. He has never been treated with oral hypoglycemics and has subsequently been hospitalized only once, for psychosocial stabilization.
It was unclear to the investigators in the phase III study whether olanzapine could precipitate ketoacidosis. The dramatic improvement in the patient’s schizophrenia symptoms was noted by the patient and his family. This, along with the as-yet-unreported association of the drug with diabetes mellitus, made the risk-benefit factors weighted toward continuing olanzapine treatment. Weight gain, serotonin 5-HT1A and 5-HT2A/C receptor antagonism, insulin resistance, and elevated insulin, leptin, and serum lipid levels have all been implicated in this phenomenon (7). Although Mr. A’s prediabetic insulin and C-peptide levels were not available, this report suggests that olanzapine may have triggered a biphasic reaction, initially causing suppression of insulin secretion then subsequent rebound overcompensation. This also suggests possible future depletion of beta cells and a return to an insulin-dependent state. Further study of this phenomenon, including the role of peripheral insulin receptors, is warranted.