Exposure to magnetic resonance imaging (MRI) procedures in relation to increased risk of early-onset colorectal cancer.

Abstract

e15676 Background: The incidence of early-onset (< 50 years of age) colorectal cancer (EOCRC) has been rising at an alarming rate in developed countries in recent decades, while late-onset (≥ 50 years of age) CRC (LOCRC) has been steadily declining. This stark contrast suggests the presence of a new environmental risk factor that has a greater impact on younger than older adults. One such emerging risk factor is increased exposure to magnetic non-ionizing radiation from widespread use of the devices that emit non-ionizing radiation. One of such emitting sources is magnetic resonance imaging (MRI) whose usage has increased dramatically, including among pediatric patients, since the 1990s parallel to the rise in EOCRC. Given that intestinal epithelial cells are among the tissue cells most sensitive to radiation exposure, the current study was designed to determine whether exposure to MRI procedures, especially to the abdominal region, is associated with increased risk of EOCRC. Methods: We conducted a population-based matched case-control study among Kaiser Permanente Northern California (KPNC) members. EOCRC cases (N=837) were identified through KPNC’s Cancer Registry. We randomly selected 120,492 age- and sex-matched controls without CRC among all eligible controls. Exposure to MRI procedures was ascertained from KPNC electric health record (EHR) databases. Conditional logistic regression was used to estimate the association after controlling for confounders, including duration of follow-up time for ascertaining MRI. To avoid possible reverse causality, we excluded all MRI procedures performed within 24 months prior to the initial EOCRC diagnosis or corresponding reference dates for matched controls. In addition to the overall association, we assessed whether (1) MRI procedures applied to the abdominal region, where the colon and rectum were directly exposed, had a stronger association compared to MRI applied to non-abdominal regions, and (2) there was a dose-response relationship. Results: After controlling for follow-up duration and confounders including other imaging procedures, exposure to MRIs was associated with 90% higher risk of EOCRC compared to those without exposure to MRIs: adjusted odd ratio (aOR)=1.9, 95% confidence interval (CI): 1.6-2.3. Importantly, the association was stronger for MRI procedures applied to abdominal regions (aOR=3.2) than MRI procedures applied to non-abdominal regions of the body (aOR=1.9). The observed association also showed a dose-response relationship: aOR=3.0 (95%CI: 2.1-4.2) for ≤ 2 MRI procedures and aOR=9.1 (3.9-21.6) for 3+ MRI procedures. Conclusions: Exposure to magnetic non-ionizing radiation from MRI procedures was associated with increased risk of EOCRC characterized by a stronger association for abdominal exposure and a dose-response relationship.