Diabetic ketoacidosis during therapy for pediatric acute lymphoblastic leukemia

Background. Hyperglycemia is common during therapy for acute lymphoblastic leukemia (ALL), but diabetic ketoacidosis (DKA) occurs rarely. Morbidity due to DKA in children with ALL has not been systematically studied. Procedures. We reviewed risk factors and clinical consequences of DKA in patients undergoing therapy for ALL at SJCRH between 1991 and 2006. Results. DKA occurred in 6 of 797 evaluable patients. Only older age at diagnosis of ALL was a risk factor for DKA. Four of six patients with DKA as compared to 232 of the other 791 patients were older than 10 years (P=0.03). Race, sex, body mass index, leukemia immunophenotype, ALL risk category, white blood cell count at diagnosis, and treatment protocol were not associated with DKA. All patients were managed with intravenous fluids, dietary modification, and short-term use of insulin. Patients were hospitalized for 4-12 days, with a median ICU stay of I day. in two patients, correction of hyperglycemia was too rapid, and two others experienced hypoglycemia due to insulin therapy. There were no permanent cornplications of DKA or its treatment. No patient required long-term insulin use. No patient had recurrent DKA; only one of the six patients had a subsequent hyperglycemia episode. All six patients are alive in remission 613 years after diagnosis. Conclusions. Patients with hyperglycemia during treatment for ALL should be screened for clinical evidence of DKA, which may require more intensive supportive care than those without ketoacidosis. The occurrence of DKA should not lead to alteration of ALL treatment.