Background: Therapeutic advances have extended survival for most children and adolescents with cancer beyond 5 years from diagnosis. However, excess mortality continues beyond 5 years, and a significant portion results from causes other than the primary cancer. Risk factors for these deaths are not currently known. Thus, the authors studied mortality in a cohort of adult survivors of childhood and adolescent cancer to determine whether survivor characteristics were associated with increased relative risk of death from other causes.
Methods: Using 3255 siblings as control subjects, the authors studied survival in a retrospective cohort study of 2319 adults who were at least 5-year survivors of cancer diagnosed before reaching 20 years of age and between 1945 and 1974 (the NCI Five Center Study). Follow-up occurred between 1980 and 1983 at a mean survivor age of 32 years (range, 21-55 years).
Results: Between cohort entry and follow-up, 292 (13%) survivors and 50 (2%) controls died. One-third of the deaths in survivors were from causes other than the primary malignancy. Compared with control subjects, between ages 21 and 40 years, survivors had a more than threefold risk of death from other causes. The relative risk (RR) for death from other causes was greatest for survivors treated with radiation and alkylating agents (RR = 6.1; 95% confidence interval [CI], 3.0-12.4) and for those treated with radiation alone (RR = 3.8; 95% CI, 2.3-6.2; Cox regression analysis containing terms for treatment and cancer diagnosis).
Conclusions: Adult survivors of childhood and adolescent cancer had a higher death rate than their siblings, even after removing the effect of primary cancer as the direct cause. Moreover, death from other causes was most strongly associated with increasing intensity of therapy, and this excess risk did not diminish with increasing age and duration of cancer survival. Because contemporary anticancer therapy is, in general, even more intensive than that received by the survivors described in this study, medical surveillance of cancer survivors is increasingly important to diagnose and treat potentially life-threatening complications that may occur decades after therapy has ceased.