P894: infection risk before diagnosis of monoclonal gammopathy of undetermined significance: results from the screened population-based istopmm study

Topic: 13. Myeloma and other monoclonal gammopathies - Biology & Translational Research Background: Infections have often been hypothesized to play an etiological role in multiple myeloma (MM) and monoclonal gammopathy of undetermined significance (MGUS), possibly through a dysregulated immune response to infective pathogens or chronic antigen stimulation. Although the risk of infections seems to be increased in clinical cohorts of individuals with MGUS, it is still unclear if infections increase the risk of developing the precursor condition, if infections are a manifestation of an underlying immune disturbance, or if this association reflects a selection bias in clinical cohorts of individuals with MGUS. Aims: To determine if infections and/or use of antimicrobials increase the risk of MGUS diagnosed through a nationwide screening program. Methods: Data was obtained from the ongoing study Iceland Screens, Treats, or Prevents Multiple Myeloma (iStopMM), a population-based screening study for MM precursors in Iceland. A total of 75,422 persons ≥40 years old were screened using serum protein electrophoresis and a free light chain assay in 2016-2020. Data on infectious disease diagnoses (ICD-10 codes) was acquired from the Icelandic Patient Registry, which includes all outpatient visits and inpatient admissions in Iceland from January 1st, 2004. Data on antimicrobial use (ATC codes) was acquired from the Icelandic Medicines Registry which includes all prescriptions in Iceland from January 1st, 2002. Weighted Poisson regression was used to calculate incidence rate ratios (IRR) for infections with 95% confidence intervals (CI) and exposure time as offset. Adjustments were made for age, sex, and time of year of the screening. Results: Overall, 5,310 individuals were identified with MGUS in the cohort, thereof 1,957 with light-chain (LC)- MGUS. A total of 342,430 ICD-10 codes and 925,963 ATC codes were collected, thereof 23,296 and 69,688 for individuals with MGUS overall, respectively. There was no significant difference in overall infection rates between MGUS and controls (Figure 1). When heavy-chain (HC)- and LC-MGUS were analyzed separately no difference was observed for HC-MGUS but there was a higher rate of infections for LC-MGUS compared to controls, IRR 1.069 (CI: 1.016-1.125). Conversely, there were significantly more prescriptions of antimicrobials for the MGUS group, IRR 1.040 (CI: 1.012-1.068). When analyzed separately, this was only significant for LC-MGUS, IRR 1.057 (CI: 1.011-1.100), but not for HC-MGUS. There was a higher rate of antibacterial prescriptions, IRR 1.044 (CI: 1.017-1.072), for the MGUS group compared to controls, no significant difference was seen for antimycotic or antiviral medications. In a sensitivity analysis with a more restricted LC-MGUS diagnostic criteria the associations of increased prescriptions in the MGUS group were no longer significant. Conclusion In our large, population-based screened cohort, including >75,000 individuals, we did not find an increased risk of infections in individuals with MGUS compared to controls before a positive screening for MGUS. However, there was a 4% higher prescription rate of antimicrobials for MGUS compared to controls, explained by higher prescription rates for individuals with LC-MGUS, but not for HC-MGUS. These observed differences were not present in sensitivity analyses on LC-MGUS and prescriptions, suggesting a spurious association and not a true causality. We conclude based on a large screening cohort that there is no association between the number of previous infections and the presence of MGUS.Keywords: Screening, MGUS, Infection