(BSPOT26) Association between childhood exposure to nuclear medicine and the risk of malignancy: sibling control and population-based cohort study

Background: Prior research has indicated a relationship between exposure to nuclear medicine and the subsequent development of malignancy. However, there has been scant evidence specifically addressing nuclear medicine exposure during childhood. In addition, environmental and genetic factors associated with malignancy were rarely been considered in prior studies.

Objectives: To assess the risk of developing malignancy following exposure to nuclear medicine during childhood.

Methods: We conducted a population-based retrospective cohort study using data from the National Health Insurance Research Database from 2004-2020.
We selected patients under 18 years old who had been exposed to nuclear medicine as our treatment group and randomly selected 2 million patients who had not been exposed to nuclear medicine serve as our control group. For the treatment group, we defined the date of first nuclear examination as index date while randomly assigned index date for the control group. After excluding patients receiving positron emission tomography, radiotherapy, iodine131, and computed tomography before index date, the remaining patients were matched in a 1:4 manner based on age, sex, and index date. We followed patients from the two years after the index date till the occurrence of solid/hematological malignancies, death, or end of the study period, to avoid possible protopathic bias. Additionally, we implemented the sibling control analyses where treated patients were compared with controls selected from their unexposed siblings. The sibling control analyses were used to address the invariant unmeasured confounders within-family levels.

Results: We enrolled a total of 30,292 patients received nuclear medicine and 99,579 patients as our control group after matching. The incidence rates of solid malignancy were 2.70 and 2.35 per 100,000 person-years in the treatment group and control group, respectively. For the hematological malignancies, the incidence rates were 0.11 and 0.03 per 100,000 person-years in the treatment group and control group, respectively. The adjusted hazard ratios (HR) were 1.10 (95% CI 1.02-1.19) and 1.76 (95% CI 0.91-3.39) for solid malignancies and hematological malignancies, respectively. In the sibling control analyses, the association between nuclear examination and solid malignancies was no longer present (HR: 1.04; 95% CI 0.91-1.20) whereas the association with hematological malignancies remained consistent (HR: 2.40; 95% CI 1.07-5.40).

Conclusions: Our study indicated an increase in hematological malignancies risk following childhood exposure to nuclear medicine but found no increased risk for solid malignancies. Physicians should carefully assess the benefit-risk balance when using nuclear medicine in pediatrics.