Secondary Malignant Neoplasms after Osteosarcoma: Early Onset and Cumulative Alkylating Agent Dose Dependency

Purpose To analyze the impacts of chemotherapeutic agent exposures on the development of secondary malignant neoplasms (SMP) after osteosarcoma. Methods Of 132 patients who had been treated for high-grade extremity osteosarcoma from September 1992 to September 2008, 90 survivors were retrospectively reviewed. Fifty-eight of the survivors received a doublet of doxorubicin (ADR) and cisplatin (DDP), and 32 received a triplet of ADR, DDP, and ifosfamide (Ifos). On the basis of the dose distributions in the study cohort, the association between SMN and the cumulative dose of each agent was evaluated. Results After a mean of 13.1 years of follow-up, six SMNs were noted, three in each regimen. There were no SMNs among 42 patients who died of osteosarcoma. In Kaplan–Meier estimates, the triplet regimen group showed a higher cumulative incidence and shorter latency for SMNs than the doublet group (log rank P = 0.032). Fifteen years’ cumulative incidence of SMNs in the triplet and doublet regimen group was 9.4and 3.8 %, respectively. In the independent t test, the mean latency to SMN in the triplet group (3.7 ± 1.3 years) was shorter than that in the double group (13.1 ± 2.8 years) ( P = 0.017). In Cox regression, the alkylating agent score (AAS) [hazard ratio (HR) = 2.459, P = 0.015] and DDP (HR = 1.779, P = 0.046) showed a significant relationship with SMNs, whereas ADR (HR = 0.896, P = 0.664) and Ifos (HR = 3.694, P = 0.119) did not. AAS was also significant after adjusting for ADR and DDP (HR = 3.319, P = 0.020). Conclusions High cumulative AAS is an independent risk factor for SMN and its early-onset development after osteosarcoma.