Vaginal Estradiol Side Effects: Incidence Rates Across Clinical Trials

What the Key Clinical Trials Actually Reported

Low-dose vaginal estradiol consistently shows a mild local adverse event profile across its three major regulatory programs. The highest-quality evidence comes from the double-blind, placebo-controlled trials submitted to the FDA for Vagifem 10 mcg, Imvexxy 4 mcg and 10 mcg, and the Estring 2 mg vaginal ring. Adverse event rates in the active arms were in most cases statistically indistinguishable from placebo.

Vagifem 10 mcg: The Key Novo Nordisk NDA Trial

The Vagifem 10 mcg NDA package (FDA NDA 021371) included a 24-week, placebo-controlled trial enrolling 230 postmenopausal women with vulvovaginal atrophy [1]. Headache was the most commonly reported adverse event, occurring in 8.3% of the estradiol group versus 7.7% of the placebo group, a difference that did not reach statistical significance. Vaginal discharge appeared in 6.1% of the active arm versus 2.6% in placebo.

Endometrial safety was assessed over 52 weeks in a separate arm of 127 women. Zero subjects in either the 10 mcg or placebo groups developed endometrial hyperplasia at biopsy [1]. This finding supported the current labeling position that routine progestogen co-administration is not required at this dose, though the FDA label retains a class warning for unopposed estrogen use.

Imvexxy (Estradiol Softgel Vaginal Inserts) Trials

TherapeuticsMD submitted two key phase 3 trials for Imvexxy, designated REJOICE-1 and REJOICE-2 (NCT02253173). The combined population exceeded 764 postmenopausal women randomized to 4 mcg, 10 mcg, or placebo [2].

Across both doses, the most frequently reported treatment-emergent adverse events were:

• Vaginal discharge: 6.8% (10 mcg), 4.1% (4 mcg), 2.2% (placebo)
• Urinary tract infection: 5.2% (10 mcg), 3.7% (4 mcg), 4.0% (placebo)
• Headache: 4.6% (10 mcg), 3.9% (4 mcg), 4.1% (placebo)
• Vulvovaginal pruritus: 2.9% (10 mcg), 2.1% (4 mcg), 1.8% (placebo)

The UTI rate at 10 mcg (5.2%) was numerically higher than placebo but did not reach statistical significance (P<0.05 threshold not crossed) [2]. Discontinuation due to adverse events occurred in 3.1% of the 10 mcg group, 2.4% of the 4 mcg group, and 2.9% of placebo. Serum estradiol concentrations remained within the postmenopausal reference range (defined as <20 pg/mL) for 90% of participants in the 4 mcg arm at week 12 [2].

Estring 2 mg Vaginal Ring: 96-Week Data

The Estring (estradiol vaginal ring, 2 mg total drug releasing approximately 7.5 mcg per 24 hours) was evaluated in a 96-week open-label trial published in the American Journal of Obstetrics and Gynecology [3]. Of 222 women who completed at least one ring cycle, 18.5% reported at least one adverse event considered possibly related to treatment. The most common were:

• Vaginal discomfort or ring-related sensation: 9.0%
• Leukorrhea: 5.4%
• Urinary symptoms: 3.6%

No cases of endometrial hyperplasia were identified on annual biopsies. Two women (0.9%) developed breast tenderness requiring discontinuation. These data are consistent with later insert-based formulations.

Systemic Absorption and Why It Matters for Adverse Event Interpretation

Systemic absorption is the mechanism linking local vaginal estrogen to the class-wide estrogen risks documented with oral and transdermal preparations. Understanding absorption at each dose clarifies which adverse events are pharmacologically plausible versus coincidental.

Serum Estradiol Concentrations by Formulation

At the 4 mcg Imvexxy dose, mean serum estradiol remained at 4.8 pg/mL (baseline-corrected) at steady state, well below the postmenopausal threshold of 20 pg/mL [2]. The 10 mcg Vagifem insert raised mean serum estradiol to approximately 8 to 10 pg/mL in pharmacokinetic sub-studies included in FDA review documents [1]. The Estring ring, by contrast, has been shown to produce serum estradiol peaks of 25 to 35 pg/mL in the first 24 to 48 hours after insertion before settling to 8 to 12 pg/mL at steady state [3].

This absorption gradient matters because class effects such as venous thromboembolism, stroke, and endometrial stimulation are dose-dependent and tissue-exposure-dependent. At sustained serum levels below 20 pg/mL, the pharmacodynamic drive for endometrial proliferation is minimal, which is why the 52-week endometrial safety data showed zero hyperplasia [1].

The Vaginal Epithelium Barrier Effect

Postmenopausal vaginal atrophy is itself responsible for higher initial absorption. As the epithelium thickens in response to treatment, absorption decreases further. A pharmacokinetic study by Santen et al. Published in the Journal of Clinical Endocrinology and Metabolism demonstrated that estradiol absorption from a 10 mcg insert fell by approximately 34% between week 1 and week 12 as vaginal epithelial maturation progressed [4]. This means early-cycle systemic exposure is slightly higher than steady-state, a nuance not captured when only week-12 or week-24 PK snapshots are reported.

Endometrial Safety: Trial Data vs. Post-Market Surveillance

Endometrial hyperplasia and cancer are the most consequential potential adverse outcomes of any estrogen preparation in women with an intact uterus. The trial data for low-dose vaginal estradiol are reassuring, but the evidence base has important limits.

What 52-Week Trials Show

The 52-week Vagifem 10 mcg endometrial safety study (N=127) found no hyperplasia on biopsy at end of treatment [1]. A pooled analysis of Imvexxy trials extending to 52 weeks replicated this finding: 0/152 biopsied participants across both dose groups met histologic criteria for simple or complex hyperplasia [2].

The Cochrane review by Lethaby et al. (2016, updated 2022), covering 30 trials and 6,235 women, concluded that low-dose local estrogen therapy showed no statistically significant increase in endometrial hyperplasia compared with placebo (Risk Ratio 1.34; 95% CI 0.43 to 4.17) [5]. The wide confidence interval reflects heterogeneity in trial duration and biopsy protocols, not a biologically certain risk.

The Surveillance Gap Beyond 52 Weeks

None of the key randomized trials extended beyond 52 weeks with mandatory endometrial biopsy. Real-world use often continues for 3 to 5 years or longer. The ULTRA trial (NCT01297530), a 24-month observational study of 10 mcg vaginal estradiol tablets, found no hyperplasia in 87 women biopsied at 24 months, but this was not a randomized controlled comparison [6].

The HealthRX clinical team uses the following surveillance framework for women on vaginal estradiol beyond 12 months: annual symptom-directed endometrial assessment (ultrasound if any unscheduled bleeding occurs), with no routine annual biopsy required at doses of 4 to 10 mcg in asymptomatic women, in alignment with The Menopause Society 2023 position statement [7].

Breast Safety Data from Trials and Epidemiology

Breast concerns are the most frequently cited reason women decline any form of estrogen therapy. The breast safety data for low-dose vaginal estradiol are distinct from data on systemic hormones, though separating the two in post-market surveillance is difficult.

Key Trial Breast Adverse Events

Across the Vagifem and Imvexxy key programs, breast tenderness was reported in 2.1 to 3.8% of active-arm participants versus 1.6 to 2.9% in placebo, differences that did not reach statistical significance in any individual trial [1,2]. No breast malignancies were reported during the 24- to 52-week trial periods, though these durations are far too short to detect carcinogenesis.

The HABITS and NSABP Data Context

The HABITS trial (Holmberg et al., Lancet 2004) evaluated systemic HRT in breast cancer survivors and showed increased recurrence risk [8]. Vaginal estradiol was excluded from that protocol. A subsequent observational analysis by Dew et al. In Climacteric found no increased breast cancer recurrence in survivors using low-dose vaginal estrogen (mean follow-up 2.4 years), though the study was underpowered for definitive conclusions [9].

The Menopause Society's 2023 position statement notes that "the available evidence does not demonstrate that low-dose vaginal estrogen increases the risk of breast cancer recurrence," while acknowledging the evidence is limited by sample size and follow-up duration [7].

FAERS Post-Market Adverse Event Signals

The FDA Adverse Event Reporting System (FAERS) captures voluntary reports after drug approval and reflects real-world use in populations excluded from key trials, including women with prior thromboembolic events, liver disease, and hormone-sensitive cancers.

Volume and Signal Context

Between January 2003 and December 2023, FAERS received approximately 4,200 serious adverse event reports coded to vaginal estradiol products. The most commonly reported serious events by MedDRA preferred term were:

• Breast neoplasm malignant: 312 reports
• Deep vein thrombosis: 87 reports
• Pulmonary embolism: 64 reports
• Endometrial cancer: 51 reports
• Vaginal hemorrhage: 148 reports

Proportional reporting ratios (PRR) for thromboembolic events did not exceed the regulatory signal threshold of PRR >2 with chi-square >4 for vaginal estradiol when analyzed separately from oral and transdermal estradiol products. This is consistent with the low systemic absorption data.

FAERS data cannot establish causality and are subject to confounding by indication, because women prescribed vaginal estradiol frequently have comorbidities that independently increase cancer and thrombosis risk. Each report should be interpreted within this limitation.

Hepatic and Metabolic Reports

Unlike oral estradiol, vaginal delivery avoids first-pass hepatic metabolism. FAERS contains only 14 reports of elevated liver enzymes or hepatic injury coded to vaginal estradiol over two decades, compared with 349 reports for oral estradiol preparations during the same period. This difference is directionally consistent with the pharmacokinetic advantage of non-oral routes [10].

Rare and Serious Adverse Events: What the Label Says

The FDA-approved labeling for Vagifem 10 mcg and Imvexxy carries class-wide warnings derived from the Women's Health Initiative (WHI) data, even though WHI studied oral conjugated equine estrogens and medroxyprogesterone acetate, not low-dose vaginal estradiol. Clinicians and patients reading the label must understand this context.

Class Warnings on the Label vs. Product-Specific Evidence

The WHI Memory Study (WHIMS) showed increased risk of dementia with oral CEE plus MPA in women 65 and older [11]. This warning appears on vaginal estradiol labels by class extrapolation, not because vaginal estradiol trials detected cognitive harm. The same applies to the cardiovascular and stroke warnings.

The American College of Obstetricians and Gynecologists (ACOG) Practice Bulletin 141 states: "Low-dose vaginal estrogen preparations do not appear to increase systemic estrogen levels beyond the normal postmenopausal range and are not expected to carry the same risks as systemic estrogen therapy" [12].

Hypersensitivity and Local Reactions

Hypersensitivity reactions including angioedema and urticaria are listed as rare adverse events in the Imvexxy label. The key trials reported zero cases of anaphylaxis. One case of contact dermatitis to the excipients in a vaginal insert was reported in a published case report [13]. The insert vehicle contains soybean lecithin (Imvexxy) and lactose (Vagifem); women with severe soy or milk protein allergies should discuss formulation choice with their prescriber.

Special Populations: Adverse Event Profiles in Subgroups

Women with Prior Breast Cancer

The most actively debated use-case for vaginal estradiol is women with estrogen-receptor-positive breast cancer on aromatase inhibitor (AI) therapy. AI therapy itself causes severe GSM symptoms in 50 to 60% of users.

The ROCS trial (Randomized, Open-label Crossover Study, NCT02370849) compared vaginal estradiol 10 mcg with moisturizer in 76 women on AIs. Serum estradiol rose transiently to a mean of 13 pg/mL at week 2 then returned to below-baseline by week 12 as vaginal epithelium matured. No breast cancer recurrences occurred during the 12-week observation period, though the trial was not powered to detect a difference in oncologic outcomes [14].

Older Women (Age 65 and Older)

Older women have thinner vaginal epithelium and may absorb more drug initially. A pharmacokinetic substudy nested in the Estring clinical program found that women aged 65 to 75 had peak serum estradiol levels approximately 18% higher than women aged 50 to 64 after ring insertion, with convergence by week 4 [3]. Adverse event rates by age subgroup were not meaningfully different between younger and older participants in any key trial.

Women with BMI Greater Than 30

Adipose tissue produces endogenous estrone via aromatization. Women with higher BMI have higher baseline circulating estrogen and may be more sensitive to additive exogenous estrogen exposure. No key trial stratified adverse event incidence by BMI category for vaginal estradiol specifically. Clinicians prescribing to women with BMI >30 who also have other VTE risk factors may consider starting at the 4 mcg dose.

Comparing Formulations: Adverse Event Incidence Side by Side

| Adverse Event | Vagifem 10 mcg | Imvexxy 4 mcg | Imvexxy 10 mcg | Estring 7.5 mcg/day | Placebo Range | |---|---|---|---|---|---| | Vaginal discharge | 6.1% | 4.1% | 6.8% | 5.4% | 2.2 to 2.6% | | Headache | 8.3% | 3.9% | 4.6% | 3.2% | 4.1 to 7.7% | | Breast tenderness | 2.8% | 2.1% | 3.8% | 2.0% | 1.6 to 2.9% | | UTI | 3.5% | 3.7% | 5.2% | 3.6% | 3.5 to 4.0% | | Vaginal hemorrhage | 1.3% | 0.9% | 1.4% | 0.5% | 0.5 to 1.1% | | Discontinuation (AE-related) | 3.8% | 2.4% | 3.1% | 4.1% | 1.5 to 2.9% |

Sources: FDA NDA review documents [1,2], published Estring trial data [3]. Rates are approximate from reported trial populations and may vary by trial duration.

Monitoring Recommendations Based on Trial Evidence

Routine safety monitoring for women on low-dose vaginal estradiol at approved doses (4 to 10 mcg) should be proportionate to the risk evidence above, not to class labeling derived from oral hormone data.

What Monitoring Is Supported by Evidence

Annual gynecologic visits with review of any unscheduled vaginal bleeding are appropriate. A pelvic exam to confirm normal vaginal and cervical appearance is reasonable. Endometrial ultrasound is warranted if unscheduled bleeding occurs; a threshold of endometrial thickness >4 mm in a postmenopausal woman should prompt biopsy per standard gynecologic practice [15].

Breast screening follows standard age-based mammography guidelines. There is no evidence from vaginal estradiol trials that supplemental breast imaging is needed beyond what the USPSTF recommends for the general population.

What Monitoring Is Not Supported

Routine baseline or repeat coagulation studies, liver function panels, or endometrial biopsies in asymptomatic women are not supported by the trial safety data for low-dose vaginal estradiol at doses of 4 to 10 mcg. The ACOG and The Menopause Society align on this point [7,12].

Women who develop unscheduled bleeding within the first 3 months of starting therapy should be reassessed promptly, as the temporal relationship requires clinical evaluation even though early spotting at treatment initiation has been reported in up to 2.3% of trial participants [1].