Medication induced diabetes during induction in pediatric acute lymphoblastic leukemia: prevalence, risk factors and characteristics.

INTRODUCTION

Medication induced diabetes (MID) during induction therapy (MIDi) in patients with acute lymphoblastic leukemia (ALL) is not well characterized in children, with recent studies yielding conflicting results.

PURPOSE

The purpose of the study was to describe the prevalence of MIDi and risk factors for its development.

METHODS

We retrospectively gathered demographic, disease course and treatment data on 363 patients aged 1 to 17.9 years diagnosed with ALL at a pediatric tertiary care hospital between 1998 and 2005. MIDi was defined as blood glucose ≥200 mg/dL (11.1 mmol/L) on at least 2 separate days during induction.

RESULTS

Fifty-seven subjects (15.7%) developed MIDi during the study period. Patients ≥10 years were more likely to develop MIDi than those <10 years (odds ratio [OR] 9.6, 95% confidence interval [CI] 5.1-17.8). BMI percentile among those with MIDi (mean ± SD 58.2 ± 31.0) did not differ from those without MIDi (52.2 ± 32.0, P = 0.429). The presence of Trisomy 21 (OR 3.6, 95% CI 1.1-11.4, P = 0.030) and CNS involvement at diagnosis (OR 3.8, 95% CI 1.4-10.1, P = 0.009) were associated with an increased risk of MIDi. After adjustment for potential confounding variables, age ≥10 years and the presence of CNS disease at diagnosis remained significantly associated with MIDi.

CONCLUSIONS

Older age and CNS involvement at diagnosis increase the risk of MIDi. In contrast to previous studies, higher BMI was not associated with MIDi in our population.