LAUSR

We read the article by Marina et al on the long-term follow-up of Ewing sarcoma (ES) survivors with great interest because, with improving treatments, more and more patients are surviving this disease. It is interesting to compare their findings regarding the risk of second malignant neoplasms (SMNs) with those identified by Sultan et al, who used a population-based data set from the Surveillance, Epidemiology, and End Results (SEER) program to estimate the risk of SMNs in a cohort of 1166 ES patients identified between 1973 and 2005. Thirty-five of these patients developed SMNs (excluding nonmelanoma skin cancers), and this meant an observed-to-expected (O/E) ratio of 4.01 (95% confidence interval [CI], 2.80-5.58; excess risk [ER], 33.18 per 10,000 persons) for all cancers in ES patients. We updated the SEER analysis with the most recent data set (1973-2014). Sixty cases developed SMNs in a total cohort of 1431 ES patients with a similar risk (O/E, 3.64; 95% CI, 2.78-4.68; P < .05; ER, 35.64). Among these cases, 41 were diagnosed with solid SMNs, and 19 had hematologic SMNs. Bone tumors were the most common SMNs identified (n 5 8; O/E, 66.26; 95% CI, 28.6130.55; P < .05; ER, 6.46), and they were followed by soft-tissue sarcomas (n 5 7; O/E, 34.16; 95% CI, 13.7370.38; P < .05; ER, 5.57), breast cancer (n 5 8; O/E, 3.27; 95% CI, 1.41-6.45; P < .05; ER, 4.55), and leukemia (n 5 17; O/E, 30.03; 95% CI, 17.49-48.08; P< .05; ER, 13.46). There are obviously several key differences in the 2 analytical approaches. Marina et al chose to focus on the delayed incidence of SMNs (more than 5 years after the diagnosis), whereas in our analysis (and in Sultan et al’s analysis), a 5-year lag time before the development of an SMN was not allowed. We observed an increase in the O/E ratio for hematologic SMNs in the SEER cohort (O/E for acute myeloid leukemia, 71 vs 29). The difference may reflect the 5-year lag time implemented by Marina et al. The data used by Marina et al come from a retrospective cohort of patients treated at selected institutions. However, they were able to examine the effects of different treatments, such as chemotherapy and radiation therapy, and associated chronic medical conditions, none of which are available in the SEER database. Overall, the results are fairly consistent and demonstrate that survivors of ES have a significantly increased risk of both solid and hematologic SMNs.