Clinical Procedure Exposures and Rare or Poorly Understood Cancers: A Case-control Study Using Electronic Health Records

Rare and poorly understood cancers (RPUC) represent a heterogeneous group of diseases for which risk factors remain unexplored. We evaluated associations between four selected clinical exposures and a composite outcome of 12 RPUC. We conducted four matched case-control analyses using electronic health record data from the University of Missouri Health Care system. Composite outcome cases were matched 1 : 3 to controls on sex, race, Hispanic ethnicity, marital status, age group, and diagnosis year group. Conditional logistic regression estimated associations for radiation procedures, hormone therapy, orthopedic implants, and breast implants (matched OR [mOR]). Models adjusted for body mass index, smoking, healthcare utilization, Charlson Comorbidity Index, and additional breast implant analysis covariates. The radiation analysis included 679 cases and 2,037 controls; hormone therapy, 1,019 cases and 3,057 controls; orthopedic implants, 726 cases and 2,178 controls; and breast implants, 1,344 cases and 4,032 controls. Radiation exposure was not significantly associated with case status after adjustment (mOR, 0.94; 95% CI, 0.75-1.18). Hormone therapy was positively associated with case status (mOR, 1.32; 95% CI, 1.05-1.66), as were orthopedic implants (mOR, 2.68; 95% CI, 1.38-5.21). Breast implant exposure was not significantly associated with case status (mOR, 0.67; 95% CI, 0.18-2.50). In conclusion, the composite outcome was positively associated with hormone therapy and orthopedic implant. Given the heterogeneity of the cancers included, simplified exposure definitions, potential residual confounding, and incomplete electronic health record ascertainment, these findings should be interpreted as exploratory (hypothesis-generating) and warrant cancer-specific evaluation.