Signal of venous thromboembolism with ethinyl estradiol vs natural estrogens in oral contraceptives: analysis of the EudraVigilance and Food and Drug Administration Adverse Events Reporting System databases

Abstract

Background

Disproportionality analyses in the Food and Drug Administration Adverse Events Reporting System (FAERS) and EudraVigilance reported lower reporting of venous thromboembolism (VTE) with natural estrogen–based combined oral contraceptives (COCs) than with ethinylestradiol/levonorgestrel. However, using ethinylestradiol/levonorgestrel as the reference showed reduced reporting only for natural estrogen–based products. We therefore reanalyzed both databases separately, using natural estrogen–based COCs as the reference.

Methods

Previously extracted individual case safety report datasets and published disproportionality analyses were re-examined, using natural estrogen–based contraceptives as the comparator. Proportional reporting ratios (PRRs) were recalculated separately for each database. COCs were categorized as ethinylestradiol-based or natural estrogen–based formulations. Drospirenone-only pills were included for comparison. Analyses used pooled natural estrogen contraceptives, estetrol/drospirenone, or estradiol/dienogest as reference groups. Disproportionality was defined as PRR >2, at least 3 cases, and χ² ≥ 4.

Discussion

Consistent patterns were observed in EudraVigilance and FAERS. Compared with natural estrogen–based COCs, ethinylestradiol-based COCs showed a disproportionate reporting of VTE (EudraVigilance: PRR, 5.10 [95% CI, 4.68-5.56]; FAERS: PRR, 4.21 [95% CI, 3.49-5.08]). In EudraVigilance, when using estetrol/drospirenone as the reference, all ethinylestradiol-based formulations showed disproportionality according to the conventional PRR threshold. A similar trend was noted in FAERS, though not statistically significant, with the drospirenone-only formulation comparable to estetrol/drospirenone.

Conclusion

This analysis revealed a consistent and disproportionate reporting of VTE associated with ethinylestradiol-based COCs compared with natural estrogen–based alternatives across 2 pharmacovigilance databases. These results, together with prior biological and observational research, support further reevaluation of COC options given the estrogenic component’s potential influence on VTE risk.