Second Primary Cancers After Intensity-Modulated vs 3-Dimensional Conformal Radiation Therapy for Prostate Cancer

Intensity-modulated radiation therapy (IMRT) is commonly used for patients with prostate cancer because it allows dose escalation to the tumor while reducing radiation exposure to surrounding healthy tissues such as the bladder and rectum.^1,2 This reduction may be at the expense of increased radiation exposure to more distant tissues from scatter radiation, particularly the red bone marrow, compared with the exposure from 3-dimensional conformal radiotherapy (3D-CRT), the previous standard radiotherapy technique.¹ Simulation studies have suggested this reduced radiation exposure could double the risk of second primary cancers.³ To date, however, no observational studies have directly compared second cancer rates after IMRT to 3D-CRT for prostate cancer.⁴ We compared the risks of leukemia and myelodysplasia (of particular concern given the potentially higher bone marrow dose and because they can occur as early as 2 years after exposure⁵) and second solid cancers after IMRT vs 3D-CRT in a large cohort of prostate cancer patients.

We conducted a retrospective cohort study using SEER (Surveillance, Epidemiology, and End Results) Medicare data. The cohort included men diagnosed between 2002 and 2009 with nonmetastatic prostate cancer who were aged 66 through 84 years and who received IMRT or 3D-CRT, but no chemotherapy, within the first year after diagnosis and survived at least 2 years after treatment initiation. As previously reported,⁶ SEER data were used to collect demographic, cancer, and death information. Medicare billing records were used to obtain information on cancer treatments and comorbidities. Eligible individuals were followed up from radiotherapy initiation until the earliest of second cancer diagnosis, death, 90th birthday, or December 31, 2011.

Relative risks (RRs) of second primary cancers after IMRT vs 3D-CRT were estimated by Poisson regression to simultaneously account for attained age, time since exposure, and calendar time. Person-years at risk were accumulated from RT initiation + 2 years (for hematopoietic tumors) or 5 years (for solid cancers), to account for minimal time intervals to develop radiation-related cancers,⁵ up to study end. The analyses were stratified by attained age, time since diagnosis, and calendar year, and adjusted for tumor grade, race, Charlson comorbidity score, smoking history, receipt of chemotherapy (≥1 year after diagnosis), hormonal therapy, and brachytherapy. Sensitivity analyses excluding men diagnosed in 2002-2003 were conducted to account for possible treatment misclassification in the early period of IMRT use.

The cohort included 39 028 patients with a median follow-up of 5.2 years (range, 2.0-10.0 years) (Table). A total of 2901 men developed second cancers: 1691 (6.1%) in the IMRT group and 1210 (10.9%) in the 3D-CRT group. There was no difference in the risk of leukemia or myelodysplasia after IMRT vs 3D-CRT (Figure). Risks of colon cancer (RR, 0.59; 95% CI, 0.43-0.81) and rectal cancer (RR, 0.58; 95% CI, 0.36-0.93) were significantly lower after IMRT. The risks of other solid cancers and lymphomas did not differ significantly between IMRT and 3D-CRT. Receipt of chemotherapy, brachytherapy, hormonal therapy, or surgery did not confound or significantly modify the results. In sensitivity analyses, the results did not differ meaningfully from the main analyses.

In this large cohort of prostate cancer patients, IMRT was not associated with an early elevated risk of leukemia or myelodysplasia. There was some preliminary evidence of reduced risks of colon and rectal cancers compared with 3D-CRT, which is potentially consistent with lower radiation doses from IMRT to these organs.^1,3 No association of RT modality with lung cancer risk was observed, suggesting that residual confounding by smoking is unlikely to account for the inverse associations observed for colon and rectal cancers. The study had sufficient follow-up to evaluate early incidence of leukemia and myelodysplasia, which might occur as soon as 2 years after radiation exposure, but was currently limited to evaluate the risks of solid cancers, which usually occur 5 to 10 years after radiation exposure.⁵ Further follow-up is needed to continue to monitor the potential impact of IMRT on second cancer risks.

Corresponding Author: Neige M. Y. Journy, PhD, Radiation Epidemiology Branch, Division of Cancer Epidemiology and Genetics, National Cancer Institute, National Institutes of Health, Room 7E556, MSC 9778, 9609 Medical Center Dr, Bethesda, MD 20892-9778 (neige.journy@nih.gov).

Published Online: July 14, 2016. doi:10.1001/jamaoncol.2016.1368.

Author Contributions: Dr Journy had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study concept and design: Morton, Kleinerman, Bekelman, Berrington de Gonzalez.

Acquisition, analysis, or interpretation of data: All authors.

Drafting of the manuscript: Journy, Berrington de Gonzalez.

Critical revision of the manuscript for important intellectual content: All authors.

Statistical analysis: Journy, Kleinerman, Bekelman.

Obtained funding: Bekelman.

Conflict of Interest Disclosures: None reported.

Funding/Support: This study was supported by the National Cancer Institute (NCI) intramural research program, National Institutes of Health. Dr Bekelman was supported by grant NCI K07-CA16316. This study used the linked SEER-Medicare database.

Role of the Funder/Sponsor: The funding sources and sponsor had no role in design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.

Disclaimer: The interpretation and reporting of these data are the sole responsibility of the authors.

Additional Contributions: We acknowledge the efforts of the National Cancer Institute; the Office of Research, Development and Information, CMS; Information Management Services (IMS) Inc; and the SEER Program tumor registries in the creation of the SEER-Medicare database. We also thank Dale Preston, PhD (HiroSoft, Inc), and Jeannette Wong-Siegel, MPH (Washington University in St Louis), who provided assistance in computer programing. These individuals received no compensation.