Back

FDA-approved drug repurposing in zebrafish identifies thyroid hormone and other compounds as potential antithrombotics

Yaman, M.; Su, H.; Lee, J. K.; Ferguson, A. C.; Sowell, K. M.; Xun, J.; Wu, D.; Habiger, C. J.; Hanauer, D. A.; Clasby, M. C.; Rech, J. C.; Shavit, J. A.

2025-10-20 pharmacology and toxicology
10.1101/2025.09.23.677920 bioRxiv
Show abstract

Venous thromboembolism (VTE) is a highly prevalent medical condition with limited therapeutic options and an incomplete understanding of its acquired and inherited subtypes. The zebrafish is a model with the benefits of external development, fecundity, optical transparency, and hemostasis that demonstrates conservation with mammals. We utilized zebrafish as a phenotypic screening tool to identify novel therapeutic options for preventing VTE. A library of FDA-approved compounds was screened for suppression of acquired (elevated estrogen) and spontaneous (protein C deficiency) thrombosis. We found that thyroid hormone, receptor tyrosine kinase (RTK) inhibitors, and proton-pump inhibitors (PPIs) effectively modulated levels of thrombosis, particularly in the estrogen-induced model. These also showed a more favorable hemostatic profile than standard therapies, suggesting alternative mechanisms. Genome editing of thyroid hormone receptor proved that thyroid hormone action is on target. A retrospective electronic health record (EHR) analysis found that thyroid-hormone prescriptions in hormonal contraceptive users correlated with a higher VTE risk, potentially limiting direct repurposing but highlighting thyroid signaling as a pathway involved in estrogen-induced thrombosis. Together, these data identify several drug classes that can be tailored to specific subtypes of VTE and help elucidate distinct pathways driving thrombosis.

Matching journals

The top 4 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.