AUTHORS
Priyanshi Maurya, MBBS, Mohamed Doma, MD2, Varshini Srinivasan, MBBS3, Vishnu Sudhakaran, MD4, Karthik Rayapureddi, MBBS5, Sukul Khanna6, Adham Ramadan, MD7, Simran Chandra8, Ayesha Laraib, MBBS9, Sawsan El sweilmein, MD10, Osama Mosalem, MD11
AFFILIATIONS
1 University of Missouri - Kansas City, Kansas City, Missouri, United States
2 Massachusetts General Hospital, Boston, United States
3 University of Missouri - Kansas City, Kansas City, Missouri, United States
4 University of Missouri - Kansas City, Kansas City, Missouri, United States
5 Armed Forces Medical College, Pune, India
6 University of Missouri - Kansas City, Kansas City, Missouri, United States
7 Boston Medical Center, Boston, United States
8 University of Missouri - Kansas City, Kansas City, Missouri, United States
9 University of Missouri - Kansas City, Kansas City, Missouri, United States
10 University of Missouri - Kansas City, Kansas City, Missouri, United States
11 Saint Luke's Cancer Institute, Kansas City, United States
ABSTRACT
Introduction
Thrombopoietin receptor agonists (TPO-RAs) are the preferred second-line therapy for immune thrombocytopenia (ITP), but their use is associated with a paradoxical increase in thromboembolic risk despite reducing bleeding. Although class-level thrombotic risk has been established, comparative safety among individual agents (romiplostim, eltrombopag, and avatrombopag) remains uncertain. We aimed to compare thromboembolic, bleeding, and mortality outcomes across TPO-RAs in adults with ITP.
Methods
We conducted a retrospective cohort study using the TriNetX Global Collaborative Network, including adults (≥18 years) with ITP who initiated romiplostim, eltrombopag, or avatrombopag between January 2018 and December 2024. Patients exposed to more than one TPO-RA were excluded. Pairwise 1:1 propensity score matching was performed using demographic characteristics, comorbidities, cardiovascular risk factors, nicotine and alcohol use, BMI, HbA1c, creatinine, prior venous and arterial thrombotic diagnoses, and baseline anticoagulant, antiplatelet, and corticosteroid use. The primary outcomes at 1 and 2 years were venous thromboembolism (VTE), arterial thromboembolism (ATE), composite thromboembolism, major bleeding, and all-cause mortality. Hazard ratios (HRs) were estimated using Cox proportional hazards models.
Results
Propensity score matching achieved baseline balance across all comparisons (standardized differences <0.11). Matched cohorts included 776-783 pairs for eltrombopag versus avatrombopag, 2,654 pairs for eltrombopag versus romiplostim, and 794 pairs for romiplostim versus avatrombopag.
At 1 year, compared with avatrombopag, eltrombopag was associated with similar thromboembolic risk but higher rates of major bleeding (HR 1.45, 95% CI 1.17-1.80) and all-cause mortality (HR 1.46, 95% CI 1.06-2.00). Compared with romiplostim, eltrombopag was associated with lower risks of VTE (HR 0.79, 95% CI 0.68-0.92), ATE (HR 0.80, 95% CI 0.67-0.96), composite thromboembolism (HR 0.80, 95% CI 0.70-0.90), major bleeding (HR 0.65, 95% CI 0.59-0.71), and all-cause mortality (HR 0.68, 95% CI 0.59-0.78). Romiplostim and avatrombopag had comparable thromboembolic risk, although romiplostim was associated with higher major bleeding (HR 1.91, 95% CI 1.56-2.33) and all-cause mortality (HR 1.96, 95% CI 1.47-2.63).
At 2 years, compared with avatrombopag, eltrombopag was associated with a lower risk of VTE (HR 0.68, 95% CI 0.50-0.93), while major bleeding (HR 1.08, 95% CI 0.87-1.32) and all-cause mortality (HR 1.28, 95% CI 0.97-1.70) were no longer significantly different. Compared with romiplostim, eltrombopag continued to demonstrate lower risks of VTE (HR 0.81, 95% CI 0.70-0.94), composite thromboembolism (HR 0.83, 95% CI 0.74-0.94), major bleeding (HR 0.67, 95% CI 0.61-0.74), and all-cause mortality (HR 0.74, 95% CI 0.65-0.83), whereas differences in ATE were no longer significant (HR 0.86, 95% CI 0.73-1.01). Romiplostim and avatrombopag continued to have similar thromboembolic risk, while romiplostim remained associated with significantly higher major bleeding (HR 1.74, 95% CI 1.44-2.09) and all-cause mortality (HR 1.95, 95% CI 1.51-2.52).
Conclusion
In this propensity score-matched analysis, safety outcomes differed across individual TPO-RAs. Etrombopag was associated with lower long-term venous thromboembolic risk compared to romiplostim and avatrombopag. Romiplostim was associated with higher bleeding and mortality. Early differences in bleeding risk between eltrombopag and avatrombopag attenuated over time. These findings suggest that TPO-RAs may differ in their long-term safety profiles and support prospective head-to-head studies to guide individualized treatment selection.
DISCLOSURES
Priyanshi Maurya, MBBS: No relevant conflicts of interest to declare.