From a population-based cohort of cases of first cancers diagnosed between 1987 and 2004, before the patient's age of 15 years, the authors conducted a nested case-control study, matching 64 patients who experienced a second malignant neoplasm (SMN) with 190 controls. SMNs comprised 10 leukemia or myelodysplastic syndromes, 5 lymphomas induced by Epstein-Barr virus after allograft, and 49 solid tumors, including mainly 25 carcinomas (17 of the thyroid), 9 bone sarcomas, and 7 central nervous system (CNS) tumors. The median latency occurrence was 6.5 years, and that of thyroid carcinomas induced by 12 Gy fractioned total body irradiation (TBI) was 7.6 years. The relative risk (RR) of an SMN was increased by genetic and family factors and increased 17 to 69 times according to the dose of radiotherapy administered in the region for the first cancer. Age younger than 4 years at the time of radiotherapy increased the risk of SMN. Chemotherapy adjusted according to the dose of radiotherapy administered in the field yielded a greater RR of an SMN only for cumulative doses exceeding 2 g/m2 of epipodophyllotoxin but not for alkylating agents or platinum compounds. The RR of secondary leukemia increased 10-fold following high doses of epipodophyllotoxin >2 g/m2 but was not affected by alkylating agents or anthracyclines. The crude RR of a solid SMN developing after radiotherapy was very high at 18 and reached 90.7 for thyroid carcinoma after TBI, whereas the authors observed no increased risk associated with chemotherapy. These results confirm the risk of secondary leukemia after epipodophyllotoxin and of solid tumor after radiotherapy.
Keywords: Case-control study; chemotherapy; childhood cancer; population-based registry; second cancer.