Second Neoplasms in Children Following Radiotherapy in the Modern Era: A Bicentric Report from the University of Florida and Memorial Sloan Kettering Cancer Center

Abstract

<h3>Purpose/Objective(s)</h3> Over the past two decades, eliminating secondary radiation-induced neoplasms has been a primary aim of our field. This study quantifies the results of that effort with a focus on children who received intensity-modulated radiotherapy (IMRT) and proton therapy (PT). <h3>Materials/Methods</h3> A total of 1880 pediatric patients with > 2 years follow-up received radiotherapy between 1999 and 2019, including 348 treated with IMRT and 1532 treated with PT (97% passive-scatter). The median age was 9.2 years old, including 623 patients ≤5 years old. Overall, 31 patients had a known genetic tumor predisposition syndrome, most commonly neurofibromatosis (n=15); 1044 had central nervous system tumors; 191 patients were treated with large volume-radiation, including craniospinal irradiation (CSI) in 135; and 988 patients received alkylating chemotherapy. Patients treated with IMRT were more likely to receive alkylating chemotherapy (76.4% vs 47.1%, p <0.001), but received a lower median radiation dose (50.4 vs 54 Gy, p <0.001) and were marginally less likely to have a tumor predisposition syndrome (0.6% vs 1.9%, p = 0.10). Cumulative incidence method provided estimates of secondary neoplasms including benign and malignant solid tumors as well as leukemia. Multiple variables were assessed using proportional hazard regression for competing risks. The median follow-up duration was 6.5 years (range, 2-21.9) and included 296 and 773 patients with >5 years follow-up in the IMRT and proton cohorts, respectively. <h3>Results</h3> The 5-year local control rate was similar between the 2 cohorts at 90.7% following PT vs 91.4% following IMRT. For all patients, the 5- and 10-year rates of developing a second neoplasm were 1.2% and 3.3%, respectively. On multivariate analysis, the presence of a tumor predisposition syndrome (10.1% vs 1.1%), the use of alkylating chemotherapy (2.1% vs 0.3%), and the use of IMRT (2% vs 1%) were significantly associated with the development of a second neoplasm 5 years after treatment (all p < 0.05). When patients with tumor predisposition syndromes were excluded from the analysis, the use of alkylating chemotherapy (1.9% vs 0.2%) and the use of IMRT (2.0% vs 0.8%) independently retained a significant association with second neoplasms (both p < 0.05). The difference in secondary malignant solid tumors was non-significant; however, 4/14 (29%) malignant solid tumors occurred outside the target region in the patients treated with PT compared to 5/10 (50%) in patients treated with IMRT. <h3>Conclusion</h3> Compared to IMRT, early data reveal a significant reduction in second neoplasms among children treated with PT, most pronounced in those who do not have underlying tumor predisposition syndromes. This could be due to a reduction in leukemia and malignant solid tumors developing in tissue exposed to low-dose radiation outside the target region.