Risk of leukemia in children treated with human growth hormone: review and reanalysis

Data have suggested that any increased incidence of leukemia in growth-hormone (GH)-treated patients was limited to those with known risk factors for leukemia. However, previous studies may have overestimated the numbers of patient-years of risk by not excluding data from "positive-risk-factor" patients. This risk was reanalyzed by using data on children in the National Cooperative Growth Study (NCGS), with correction for this possible confounding factor.

The risk of leukemia in GH-treated patients without known risk factors was determined by using patient-years of GH therapy and patient-years since first exposure to GH therapy and the values obtained were compared with values from the Surveillance, Epidemiology, and End Results program of the National Cancer Institute.

Three cases of leukemia in patients without known risk factors were found in the NCGS database; 3.42 cases would be expected in the 119,846 patient-years in the analysis using time since GH exposure. Two of these cases of leukemia occurred during GH therapy (67,773 patient-years); 2.13 cases would be expected.

Excluding data on patients with known risk factors for leukemia provides a more accurate estimate of the risks in GH-treated patients. The incidence of leukemia in these patients is comparable to that in the general population of age-matched children.