AUTHORS
Ashish Sharma, MD, Yara Hamadah, MD2, Kevin Zhu, M.D.3, Rohit Singh, MD4, Sukhmani Sidhu, MBBS5, Saloni Haldule, MBBS6, Sajjad Ahmed Ahmed Khan7, Victoria Forbes, MD8
AFFILIATIONS
1 Yale New Haven Hospital, New Haven, Connecticut, United States
2 Univeristy of Connecticut, Farmington, Connecticut, United States
3 The University of Vermont Medical Center, South Burlington, Vermont, United States
4 The University of Vermont Medical Center, South Burlington, Vermont, United States
5 Univeristy of Connecticut, Farmington, Connecticut, United States
6 Univeristy of Connecticut, Farmington, Connecticut, United States
7 Birat Medical College Teaching Hospital, Morang, Nepal
8 UConn Health Center, Farmington, Connecticut, United States
ABSTRACT
Background:
Myelodysplastic syndromes (MDS) are clonal hematopoietic disorders with variable risk of progression to acute myeloid leukemia (AML). Glucagon-like peptide-1 receptor agonists (GLP-1RAs) are widely used for diabetes and obesity and may influence inflammatory and metabolic pathways relevant to clonal myeloid disease. Whether GLP-1RA exposure after MDS diagnosis is associated with AML transformation or survival is unknown.
Methods:
We performed a retrospective cohort study using the TriNetX Research Network. Adults with MDS (ICD-10-CM D46.x) were identified across 114 healthcare organizations. The exposed cohort included patients with first recorded GLP-1RA exposure (semaglutide, tirzepatide, liraglutide, dulaglutide, or exenatide) on or after the first recorded MDS diagnosis. Controls had MDS without recorded GLP-1RA exposure. One-to-one propensity score matching was performed using demographics, broad diagnosis categories, and medication classes. Outcomes were assessed from 1 day after the index event through 1,095 days. The primary outcome was AML transformation (ICD-10-CM C92.0), excluding patients with prior AML. Secondary outcomes included all-cause mortality, hemoglobin distribution, and platelet outcomes. Kaplan-Meier analysis and Cox proportional hazards models were used for time-to-event outcomes.
Results:
Before matching, 1,904 GLP-1RA-exposed patients and 121,240 controls met cohort criteria. After matching, 1,873 patients were included in each group. Mean age at index was 64.9 years in GLP-1RA users and 64.4 years in controls; sex, race, ethnicity, and broad comorbidity categories were closely balanced. Median follow-up was longer in the GLP-1RA cohort than in controls (650 vs 426 days). After exclusion of patients with prior AML, AML transformation occurred in 51 of 1,704 GLP-1RA users and 134 of 1,648 controls (3.0% vs 8.1%; odds ratio [OR], 0.349; 95% CI, 0.251-0.485). Three-year AML-free survival was 95.7% versus 88.0%, respectively (hazard ratio [HR], 0.320; 95% CI, 0.232-0.442; log-rank P<0.001). All-cause mortality occurred in 225 of 1,873 GLP-1RA users and 563 of 1,873 controls (12.0% vs 30.1%; OR, 0.318; 95% CI, 0.268-0.377). Three-year survival was 81.8% versus 57.6% (HR, 0.328; 95% CI, 0.281-0.383; log-rank P<0.001), although the proportional hazards assumption was not met for mortality. Mean hemoglobin during follow-up was higher in GLP-1RA users (11.49 vs 10.07 g/dL; P<0.001), while incident thrombocytopenia was more frequently recorded (26.5% vs 13.3%; OR, 2.346; 95% CI, 1.438-3.826).
Conclusions:
In this propensity score-matched real-world analysis, GLP-1RA exposure after MDS diagnosis was associated with lower observed AML transformation and all-cause mortality over 3 years. The AML signal was consistent across risk and time-to-event analyses and is hypothesis-generating. Residual confounding, differential follow-up, non-user comparator design, lack of IPSS-R/cytogenetic/molecular data, and violation of proportional hazards for mortality limit causal interpretation. These findings support further study using active comparators and prospectively defined MDS cohorts with metabolic comorbidities.
DISCLOSURES
Ashish Sharma, MD: No relevant conflicts of interest to declare.