Abstract
Introduction:
Radiotherapy is a cornerstone of cancer treatment, known to improve local tumor control and survival; however, its long-term risks, particularly the development of second primary malignancies, are increasingly relevant in cancer survivors. In this study, we used data from the SEER cancer registries (2001–2022) to estimate the proportion of second hematological malignancies attributable to radiotherapy across a broad range of first solid cancer sites
Methodology
We conducted a population-based cohort study using SEER data from 17 U.S. cancer registries, including adults aged ≥15 years diagnosed between January 1, 2001, and December 31, 2022. The cohort comprised patients with one of 15 solid tumors commonly treated with radiotherapy (oral cavity and pharynx, salivary gland, rectum, anus, larynx, lung, soft tissue, breast, cervix, endometrium, prostate, testis, eye and orbit, brain and CNS, and thyroid). To assess long-term outcomes and reduce surveillance bias, we included patients who survived at least five years after their first cancer diagnosis. Patients were followed for the development of second hematological malignancies. Relative risks (RR) comparing patients treated with radiotherapy to those not treated were estimated using Poisson regression, adjusted for age, stage, year of diagnosis, and other relevant confounders. In the database, radiotherapy exposure was categorized as beam radiation, radioactive implants (including brachytherapy), radioisotopes, a combination of beam with implants or isotopes, and radiation with a method or source not otherwise specified
Result:
We identified 2,221,476 adult cancer patients in the cohort who survived for five years or longer after their first cancer diagnosis, with a mean follow-up duration of 11 years (SD 4.5, range 5-22 years). During this period, 303,762 individuals (13.6%) developed a second primary cancer. Of these, 262,250 (11.8%) were second solid tumors, while 28,202 (1.3%) were second hematological malignancies.
Among patients with second hematological cancers, beam radiation was the most common associated modality (72.1%, n=9,350), followed by radioactive implants including brachytherapy (13.8%, n=1,782), and a combination of beam with implants or isotopes (7.4%, n=951). Across all radiation modalities, non-Hodgkin lymphoma was the most frequent second hematological malignancy, followed by Multiple Myeloma. Acute myeloid leukemia (AML) ranked third across all radiation types except in the group treated with radioactive implants, where chronic lymphocytic leukemia (CLL) was ranked third.
When stratified by the site of the first cancer, the relative risk (RR) of developing a second hematological malignancy following radiotherapy varied. Radiotherapy was associated with a statistically significant increased risk of second hematological cancers in patients whose first cancer involved the thyroid (RR 1.222, 95% CI: 1.100 - 1.358), endometrium (RR 1.163, 95% CI: 1.033 - 1.310), breast (RR 1.140, 95% CI: 1.085 - 1.198), and prostate (RR 1.048, 95% CI: 1.014 - 1.083). In contrast, a statistically significant decreased risk was observed following cancer of the anus, anal canal, and anorectum (RR 0.675, 95% CI: 0.489 - 0.932). For other first cancer sites, including oral cavity, lung, cervix, and CNS, the relative risks were not statistically significant
Conclusion:
Only a small proportion (1.3%) of second hematological malignancies were associated with prior radiotherapy to first cancer, suggesting that most second cancers are likely associated to other factors such as genetic predisposition or lifestyle-related exposures. In this large population-based cohort of over 2.2 million adult cancer survivors, radiotherapy was associated with an increased risk of second hematological malignancies following treatment for some first solid tumors, particularly of the thyroid, endometrium, breast, and prostate. Non-Hodgkin lymphoma and multiple myeloma were the most frequently observed second hematological cancers across all radiation modalities. Limitations of this study include potential residual confounding, absence of granular data on radiation dose, field, and fractionation.
