Comparison of the impact of vaginal and oral administration of combined hormonal contraceptives on hepatic proteins sensitive to estrogen

Document Type

Article (peer-reviewed)

Publication Date



Objective: We evaluated the effects of a new combined hormonal contraceptive vaginal ring (CVR) delivering the nonandrogenic progestin Nestorone® (NES) and ethinyl estradiol (EE) on several key estrogen-sensitive hepatic proteins that may be markers for the risk of arterial or venous disease events and on blood pressure (BP). Because the pharmacologic androgenicity of the progestin in these formulations influences the hepatic impact of EE, we selected an oral contraceptive (OC) delivering the androgenic progestin levonorgestrel (LNG) and EE as the comparator. We also investigated the effect of delivery route, which is known to modify the hepatic effects of estradiol, but has not been widely studied with EE. Study Methods: Women, aged 18-34 years, with no contraindications to the use of combined OCs, were randomized to three cycles of treatment with a CVR delivering NES/EE (150/15 μg/day) or a combined OC providing LNG and EE (150/30 μg per tablet). Each cycle consisted of 21 days of active treatment, followed by 7 days without treatment. During the last weeks of the pretreatment and third treatment cycles, blood samples were obtained for determinations of plasma concentrations of angiotensinogen, an estrogen-sensitive hepatic protein, and serum concentrations of sex hormone-binding globulin (SHBG), total cholesterol (TC), high-density lipoprotein cholesterol (HDL-C), low-density lipoprotein cholesterol (LDL-C), triglycerides (TG) and estrogen- and androgen-sensitive proteins. BP was also measured. Results: Of 47 women randomized, 45 completed the study (CVR: 23; OC: 22). Within-group comparisons over time by repeated-measure analysis of variance demonstrated statistically significant changes over time with both treatments for all hepatic proteins (p < .02) but not for TC. The within-group effects, presented as relative percent difference [95% confidence interval (CI)], were greatest for angiotensinogen [CVR: 227% (195-262%); OC: 251.3% (218-288%)] and SHBG [CVR: 306% (237-389%); OC: 55% (30-86)]. Both treatments were associated with small changes in systolic BP and diastolic BP (DBP), but only the within-group change in DBP for the OC group was statistically significant (p=.04). Between-treatment comparisons of third treatment cycle mean values were performed by analysis of covariance (baseline values as covariate). No statistically significant between-treatment differences were found for angiotensinogen, sensitive only to estrogen, or BP. Statistically significant treatment differences were found for all estrogen- and androgen-sensitive proteins (p ≤ .002) but not for TC. When presented as relative percent difference between the effects of treatment (CVR-OC/OC; 95% CI of percent difference), the difference was largest for SHBG (159% [117-210%]); smaller relative percent differences were found for HDL-C [31.9% (18.5-46.8%)], LDL-C [23.6% (33.4% to -2.4%)] and TG [39.0% (14.0-69.4%)], but not TC. Conclusion: Vaginal delivery of a combined hormonal contraceptive did not reduce the EE-associated changes in estrogen-sensitive hepatic proteins observed after use of a combined OC. Significant treatment differences between the NES/EE CVR and the LNG/EE OC were found for SHBG, HDL-C, LDL-C, and TG, proteins sensitive to androgen as well as estrogen. No treatment difference was observed for angiotensinogen, which is sensitive only to estrogen. The observed treatment differences were therefore most likely due to the difference in androgenicity between NES and LNG.