[Skip to Content]
[Skip to Content Landing]
Views 192
Citations 0
Comment & Response
June 20, 2017

Temporal Changes in Subsequent Malignancies Among Childhood Cancer Survivors—Reply

Author Affiliations
  • 1University of Minnesota Medical School, Minneapolis
  • 2Department of Epidemiology and Cancer Control, St Jude Children’s Research Hospital, Memphis, Tennessee
JAMA. 2017;317(23):2451-2452. doi:10.1001/jama.2017.6028

In Reply Dr Tanimoto and colleagues query whether treatment with HCT was considered in our analysis of temporal changes in subsequent neoplasms in the Childhood Cancer Survivor Study (CCSS). They raise the point that treatment modalities used in the context of HCT, such as the dose and intensity of chemotherapy, total body irradiation, or both, are unique and also point out the potential importance of chronic GVHD as a risk factor for subsequent neoplasms.

Our ability to perform a detailed assessment of subsequent neoplasms among HCT recipients in the CCSS is limited only to descriptive analysis due to the relatively small number of survivors who received transplantation. Furthermore, we do not have the ability to distinguish between allogeneic vs autologous HCT or to identify survivors with GVHD. Among the 23 603 survivors included in the cohort, 893 had HCT, and of those, only 75 experienced some type of subsequent neoplasm. By decade of diagnosis, for the 1970s, of 44 survivors who received HCT, 14 experienced 71 subsequent neoplasms (9 malignancies, 1 meningioma, and 61 nonmelanoma skin cancers [1 squamous cell carcinoma]); for the 1980s, of 270 survivors who received HCT, 40 experienced 87 subsequent neoplasms (25 malignancies, 1 meningioma, and 61 nonmelanoma skin cancers [0 squamous cell carcinomas]); for the 1990s, of 579 survivors who received HCT, 21 experienced 27 subsequent neoplasms (19 malignancies, 2 meningiomas, and 6 nonmelanoma skin cancers [1 squamous cell carcinoma]). The cumulative incidence of subsequent neoplasms and subsequent malignant neoplasms at 15 years since childhood cancer diagnosis, with 95% confidence intervals, are provided in the Table. However, within the limitations of these small numbers of transplanted patients, we have insufficient power for formal subgroup analysis.