Diabetes Risk After Treatment for Childhood and Young Adult Cancer

Abstract

OBJECTIVE Diabetes is a potential late consequence of childhood and young adult cancer (CYAC) treatment. Causative treatments associated with diabetes have been identified in retrospective cohort studies but have not been validated in population-based cohorts. Our aim was to define the extent of diabetes risk and explore contrib-utory factors for its development in survivors of CYAC in the U.K. RESEARCH DESIGN AND METHODS Cancer registration data (n = 4,238) were linked to electronic health care data-bases to identify cases of diabetes through clinical coding or HbA1c values. Total effect of prespecified treatment exposures on diabetes risk was estimated using flexible parametric modeling and standardized cause-specific cumulative incidence functions (CIFs). RESULTS After median follow-up of 14.4 years, 163 individuals (3.8%) were identified with diabetes. Total body irradiation (TBI) increases diabetes risk over time, with a 40-year CIF reaching 21.0% (95% CI 13.8–31.9) compared with 8.4% (95% CI 6.1–11.5) without TBI. Survivors treated with corticosteroids had a 7.7% increased risk at 40 years after cancer diagnosis. Hematopoietic stem cell trans-plant (HSCT) survivors had markedly higher risk, with a 40-year CIF of 19.6% (95% CI 13.4–28.6) versus 8.2% (95% CI 6.0–11.3) for patients who had not under-gone HSCT. Among patients who received allogeneic HSCT, the 40-year CIF of diabetes was 25.7% (95% CI 17.4–38.0), compared with 7.9% (95% CI 3.3–19.1) in patients who received autologous transplants. CONCLUSIONS This evaluation of a hospital-based cohort of patients with CYAC identifies these patients’ increased long-term risk of developing diabetes and how this varies temporally according to treatment modalities. Notable contrasts in risk by treatment were detected as early as 10 years after cancer diagnosis. Findings should inform the development of risk-stratified evidence-based screening.