AUTHORS
Mei Wang, MSc, Byron Seiya Sigel, MSc2, Kristen Marie Sanfilippo, MD, MPHS3, Theodore Thomas, MD4, J. Heimlich, MD, PhD5, Michael Levin6, Daphne R. Friedman, MD7, Martin Schoen, MD, PMH8, Su-Hsin Chang, PhD9
AFFILIATIONS
1 St. Louis Veterans Affairs Medical Center, St. Louis, Missouri, United States, Washington University School of Medicine, St. Louis, Missouri, United States, Washington University School of Medicine, St. Louis, Missouri, United States
2 Washington University School of Medicine, St. Louis, Missouri, United States
3 St. Louis Veterans Affairs Medical Center, St. Louis, Missouri, United States, Washington University School of Medicine, St. Louis, Missouri, United States
4 St. Louis Veterans Affairs Medical Center, St. Louis, Missouri, United States, Washington University School of Medicine, St. Louis, Missouri, United States
5 Nashville Veterans Affairs Hospital, Nashville, Tennessee, United States, Vanderbilt University Medical Center, Nashville, Tennessee, United States
6 Philadelphia Veterans Affairs Medical Center, Philadelphia, Pennsylvania, United States, University of Pennsylvania Perelman School of Medicine, Philadelphia, Pennsylvania, United States
7 Duke University School of Medicine, Durham, North Carolina, United States, Durham Veterans Affairs Health Care System, Durham, North Carolina, United States, National Oncology Program, Veterans Health Administration, Washington DC, District of Columbia, United States
8 Saint Louis University School of Medicine, St. Louis, Missouri, United States, St. Louis Veterans Affairs Medical Center, St. Louis, Missouri, United States
9 St. Louis Veterans Affairs Medical Center, St. Louis, Missouri, United States, Washington University School of Medicine, St. Louis, Missouri, United States
ABSTRACT
Introduction
Monoclonal gammopathy of undetermined significance (MGUS), a common premalignant plasma cell disorder, has been shown to increase cardiovascular disease (CVD) risk. Studies have also shown that glucagon-like peptide-1 receptor agonists (GLP-1RAs) may reduce cardiovascular risk in patients with or without type 2 diabetes mellitus (T2DM). However, whether this benefit differs by MGUS remains unknown. Clarifying this potential effect modification could inform risk-adapted prevention. Using a target trial emulation, we quantified the association between GLP-1RA and CVD risk among T2DM adults with and without a prior MGUS diagnosis.
Methods
Using nationwide Veterans Health Administration (VHA) database, we emulated a target trial among individuals with T2DM who newly initiated either a GLP-1RA or a non-GLP-1RA glucose-lowering medication (including insulin, dipeptidyl peptidase 4-inhibitors, sulfonylureas, thiazolidinediones, alpha-glucosidase inhibitors, and amylin analog) from 2016-2025. Index date was defined as the date of treatment initiation. Eligibility was restricted to patients with no history of CVD and were receiving metformin monotherapy before the index date. Individuals who received ≥1 sodium-glucose cotransporter-2 inhibitor during the follow-up were excluded due to the FDA-approved cardiovascular benefits of this drug class. GLP-1RA users and nonusers were 1:2 matched on duration from T2DM diagnosis to the index date, calendar year, and age at the index date (all <±1 year). The exposure was time-varying GLP-1RA use because patients may use it intermittently. The outcome was incident CVD, defined as a composite of acute myocardial infarction, ischemic stroke, heart failure, atrial fibrillation, and peripheral artery disease. Participants were followed for up to 5 years after the index date.
To adjust for confounding, we estimated propensity scores for GLP-1RA initiation using multivariable logistic regression. Covariates included prior MGUS diagnosis, sex, race/ethnicity, body mass index category, marital status, rurality of residence, 10-year atherosclerotic CVD risk score calculated using the pooled cohort equation (Goff et al, 2014), serum creatinine, hemoglobin A1c, neuropathy diagnosis, smoking status, number of VA visits, age, calendar year, Charlson Comorbidity Index, and use of any lipid-lowering drug, all ≤1 year before the index date. Stabilized inverse probability of treatment weighting (IPTW) was applied in multivariable Fine-Gray subdistribution hazard models stratified by matched pairs with death as the competing event to estimate the association between time-varying GLP-1RA use and incident CVD. Robust variance estimators were used to compute 95% confidence intervals (CIs). Effect modification by MGUS status was evaluated to be statistically significant in the matched cohort via an interaction term. Therefore, same analyses were conducted stratified by MGUS status to estimate the multivariable-adjusted hazard ratios (aHRs).
Results
We identified 4,401 GLP-1RA users and 16,992 nonusers before matching. The matched cohort comprised 3,457 users and 6,914 nonusers (median age=49.4 years, IQR: 41.1-57.6; median follow-up time=19.4 months, IQR: 9.0-33.7). Prior MGUS was present in 2.3% of GLP-1RA users and 1.6% of nonusers. CVD incidence was the lowest among GLP-1RA users without prior MGUS (7.2%) compared to users with prior MGUS (7.7%), nonusers without prior MGUS (10.8%), and nonusers with prior MGUS (21.3%). After IPTW, baseline covariates were balanced between treatment groups (all standardized mean differences <0.1). In the overall matched cohort, the association between time-varying GLP-1RA use and incident CVD differed significantly by MGUS status (p for interaction=0.030). In the subgroup analyses, GLP-1RA use was associated with lower CVD risk in both groups, with the magnitude of association being stronger among patients with MGUS (aHR=0.417, 95% CI: 0.184-0.948), compared to those without MGUS (aHR=0.821, 95% CI: 0.753-0.895).
Conclusion
In this target trial emulation, GLP-1RA use was associated with reduced incident CVD risk versus non-GLP-1RA glucose-lowering agents, with a stronger association among patients with MGUS (58% relative risk reduction), compared to those without MGUS (18%). Our results suggest that in T2DM patients with MGUS, GLP-1RAs may offer larger cardiovascular benefits compared to those without MGUS, warranting validation in prospective studies.
DISCLOSURES
Byron Seiya Sigel, MSc: No relevant conflicts of interest to declare.