Risk of Subsequent Central Nervous System Neoplasms in Childhood Cancer Survivors after Radiation Therapy: Results from the Pediatric Normal Tissue Effects in the Clinic (PENTEC) Collaboration

Abstract

<h3>Purpose/Objective(s)</h3> A PENTEC analysis of published investigations of central nervous system (CNS) subsequent neoplasms (CNS-SN) in childhood cancer survivors who received radiation therapy (RT) to the brain was performed to estimate the effect of RT dose and gender on the risk of CNS-SN following RT. <h3>Materials/Methods</h3> Through the PENTEC initiative, a systematic literature review was performed to identify published data on CNS-SN after prior cranial RT in childhood cancer survivors. Using the Covidence platform 2,156 studies were screened for potential inclusion. The incidences of CNS-SNs, RT dose, age, gender, primary cancer diagnosis, and latent time from primary diagnosis to CNS-SN were extracted, to assess the factors influencing risk for subsequent meningiomas or subsequent malignant CNS tumors (e.g., gliomas). The odds ratio for CNS-SNs in different dose intervals were calculated and excess odds ratio (EOR) per Gy of developing subsequent meningiomas or malignant tumors was estimated using inverse-variance weighted linear regression to model the risk for CNS-SN versus dose. <h3>Results</h3> Forty studies of independent patient cohorts provided information on 736 subsequent malignant tumors with average median latency 10.3 years, and 32 studies provided information on 1,035 subsequent meningiomas with average median latency 20.5 years. Dose-response was derived from 6 studies of 248 subsequent malignant tumors and 7 studies of 557 subsequent meningiomas. The pooled EOR/Gy was 0.45 (95% CI: 0.25, 0.66) for meningiomas and 0.16 (95% CI: 0.11, 0.20) for malignant CNS tumors. The average cumulative incidence of developing a meningioma or malignant CNS tumor at 15 years of follow-up was 2.4% (range, 1.2-6.3%) or 0.9% (range, 0.4-1.8%), respectively. Females had a higher risk of meningioma than males (OR=1.5, 95% CI: 1.2, 1.8; 6 studies; 50,346 patients) whereas no gender difference was seen in risk of malignant tumors (OR=0.9, 95% CI: 0.7, 1.2; 4 studies; 32,446 patients). <h3>Conclusion</h3> This PENTEC systematic review shows a significant radiation dose-response relationship and higher risk among females for meningioma, akin to the general population, and a highly significant but somewhat less steep relationship for subsequent malignant tumors with no effect of gender. Further evaluation of the effect of age and chemotherapy in relation to dose and gender is necessary to elucidate the risk of CNS-SN after RT.