9/13/2018 REVIEW VAGINAL ESTROGEN SAFETY AND Timeline LABELING: - - PDF document

9/13/2018 1

VAGINAL ESTROGEN SAFETY AND LABELING: STATE OF THE SCIENCE

Carolyn J. Crandall, MD, MS, FACP, NCMP Professor of Medicine David Geffen School of Medicine at the University of California, Los Angeles

REVIEW

• Timeline
• Class labeling
• Safety data

REVIEW

• Timeline
• Class labeling
• Safety data

Vaginal estrogen cream initial approval 1946 (conjugated equine estrogens) then 1984 (E2) 1946 cream

9/13/2018 2

Vaginal estradiol rings initial approval 1996 1946 cream 1996 ring Vaginal estradiol tablets (inserts) initial approval 2009 1946 cream 1996 ring 2009 insert (tablet)

VAGINAL DEHYDROEPIANDROSTERONE (DHEA)

Vaginal insert FDA-approved 2016 Estrogen is a metabolite of prasterone, so I include it today. 1946 cream 1996 ring 2009 insert (tablet) 2016 dehydroepiandrosterone Estradiol dose is 6X higher than Estring (which is 7.5 micrograms/day)

NOT OUR FOCUS TODAY

9/13/2018 3

REVIEW

• Timeline
• Class labeling
• Safety data

FIRST, THE FOREST VIEW

Vincent van Gogh, Trees and Undergrowth, 1887

CLASS LABELING

• Low-dose vaginal estrogen preparations approved by the U.S. Food

and Drug Administration carry the same boxed warning about health risks as the systemic formulations of estrogen alone and combination estrogen plus progestogen carry.

• This labeling is based on extrapolations of data from clinical trials of

systemic hormone therapy, which:

• involved substantially higher levels of systemic exposure, and
• were not based on evidence from clinical trials of vaginal estrogen.

(https://www.fda.gov/ucm/groups/fdagov-public/@fdagov-drugs- gen/documents/document/ucm075090.pdf FDA guidance)

SO….WHAT DO WE KNOW ABOUT SAFETY FROM VAGINAL TRIALS?

9/13/2018 4

REVIEW

• Timeline
• Class labeling
• Safety data

WHAT’S THE LEVEL OF EVIDENCE BEHIND THE LABELING?

Van Gogh, Undergrowth with Two Figures, 1890

EVIDENCE BEHIND THE LABEL: PRASTERONE

• WARNINGS AND PRECAUTIONS Based on: “Estrogen is a metabolite of
• prasterone. Prasterone has not been studied in women with a h/o

breast cancer. Use of exogenous estrogen is contraindicated in women with a suspected history of breast cancer.”

• Current or Past History of Breast Cancer.
• ADVERSE REACTIONS
• In four 12-week placebo-controlled RCTs, the most common

adverse reaction with an incidence ≥ 2% was vaginal discharge.

• In one 52-week open-label clinical trial, the most common adverse

reactions with an incidence ≥ 2 percent were vaginal discharge and abnormal Pap smear. www.fda.gov

NOW THE SAME EXERCISE FOR VAGINAL ESTRADIOL TABLETS

• … 17 steps long!

9/13/2018 5

EVIDENCE BEHIND THE LABEL: VAGINAL ESTRADIOL TABLET

Warnings:

EVIDENCE BEHIND THE LABEL: VAGINAL ESTRADIOL TABLET

Warnings:

• 1. There is an increased risk of endometrial cancer in a woman with a

uterus who uses unopposed estrogens.

• 2. In the Women’s Health Initiative estrogen-alone study, there was
• a. increased risk of stroke and DVT.
• b. Increased risk of probable dementia in postmenopausal women

65 years of age and older

• 3. The WHI estrogen plus progestin substudy found increased risks of:
• a. stroke, DVT, PE, MI, and breast cancer
• b. Probable dementia in women 65 years of age and older 

EVIDENCE BEHIND THE LABEL: VAGINAL ESTRADIOL TABLET

Warnings:BASED ON TRIALS OF ESTROGEN PILLS. DO VAGINAL ESTROGENS CAUSE SAME SERIOUS AES? NO CLINICAL TRIAL DATA EXIST.

• 1. There is an increased risk of endometrial cancer in a woman with a

uterus who uses unopposed estrogens.

• 2. In the Women’s Health Initiative estrogen-alone study, there was
• a. increased risk of stroke and DVT.
• b. Increased risk of probable dementia in postmenopausal women

65 years of age and older

• 3. The WHI estrogen plus progestin substudy found increased risks of:
• a. stroke, DVT, PE, MI, and breast cancer
• b. Probable dementia in women 65 years of age and older 

EVIDENCE BEHIND THE LABEL: VAGINAL ESTRADIOL TABLET

Contraindications

• 1. Undiagnosed abnormal genital bleeding
• 2. History of breast cancer (or suspected)
• 3. Known or suspected estrogen-dependent neoplasia
• 4. History of DVT or PE,
• 5. History of arterial thromboembolic disease, such as stroke or MI
• 6. Liver impairment or disease
• 7. Known protein C, protein S, or antithrombin deficiency, other

thrombophilic disorders

• 8. Known or suspected pregnancy



9/13/2018 6

EVIDENCE BEHIND THE LABEL: VAGINAL ESTRADIOL TABLET

Contraindications DITTO. NO DATA FROM VAGINAL ESTROGEN RCTS

• 1. Undiagnosed abnormal genital bleeding
• 2. History of breast cancer (or suspected)
• 3. Known or suspected estrogen-dependent neoplasia
• 4. History of DVT or PE,
• 5. History of arterial thromboembolic disease, such as stroke or MI
• 6. Liver impairment or disease
• 7. Known protein C, protein S, or antithrombin deficiency, other

thrombophilic disorders

• 8. Known or suspected pregnancy



EVIDENCE BEHIND THE LABEL: VAGINAL ESTRADIOL TABLET

• Warnings and precautions
• 1. Estrogens increase the risk of gallbladder disease
• 2. We have to monitor thyroid function if you are taking thyroid

replacement therapy

• 3. We need to check whether you are taking drugs that affect

estrogen drug metabolism (inducers and inhibitors of CYP3A4).

EVIDENCE BEHIND THE LABEL: VAGINAL ESTRADIOL TABLET

• Warnings and precautions THESE POINTS ARE FROM ORAL, NOT

VAGINAL, ESTROGEN STUDIES

• 1. Estrogens increase the risk of gallbladder disease
• 2. We have to monitor thyroid function if you are taking thyroid

replacement therapy

• 3. We need to check whether you are taking drugs that affect

estrogen drug metabolism (inducers and inhibitors of CYP3A4). In fact, it says “no drug-drug interaction studies have been performed for Vagifem”!

EVIDENCE BEHIND THE LABEL: VAGINAL ESTRADIOL TABLET

• The only adverse events information on the label that isn’t based on

WHI is:

• One pcRCT each for vaginal estradiol 10 mcg tablets (12-mo) and

25 mcg tablets (12-week)

9/13/2018 7

BY THE WAY

• Because of class labeling, the same warnings, precautions,

contraindications, drug interaction information is listed for:

• Premarin vaginal CEE cream
• Estring vaginal estradiol ring
• Estradiol cream

SO WHAT DO WE KNOW ABOUT VAGINAL ESTROGEN SAFETY FROM VAGINAL ESTROGEN TRIALS?

• Endometrial safety
• Serum levels

9/13/2018 8

SO WHAT DO WE KNOW ABOUT VAGINAL ESTROGEN SAFETY FROM VAGINAL ESTROGEN TRIALS?

• Endometrial safety
• GSM symptoms can’t be expected to resolve on their own. We

expect them to persist or worsen, so long-term endometrial safety data are of great interest.

• Serum levels

Cochrane systematic review 2016

(Lethaby et al Cochrane Database of Systematic Reviews 2016, Issue 8. Art. No.: CD001500) 4 studies of endometrial thickness included Cream vs. ring Cream vs. tablet

Cochrane systematic review 2016

(Lethaby et al Cochrane Database of Systematic Reviews 2016, Issue 8. Art. No.: CD001500) 4 studies of endometrial thickness included Cream vs. ring

• CEE cream may be associated with ↑

in endometrial thickness vs E2 ring (OR 0.36, 95% CI 0.14-0.94 ring vs. crm)

• This may be due to the higher doses
• f cream (Ayton Br J Obstet Gynaecol.

1996 used 0.625mg CEE nightly 3 wks

• n and 1 wk off vs. estradiol ring 5-10

mcg/d; Nachtigall Maturitas 1995 used 1.25mg CEE 3X/wk vs. E2 ring 10mcg/d; both 12-wk intervention)

• Thickness of >5 mm in 12% cream vs.

6% ring users. (Nachtigall 1995) Cream vs. tablet

• No difference in endometrial

thickness between vaginal E2 tablet and CEE cream 2X/wk (Manonai 2001 E2 tablet 25mcg once/wk or CEE cream 1gm = 0.625mg 2X/wk X 10 wks; Rioux 2000 E2 tablet 25mcg once/wk or CEE 2g = 1.25mg/d X 21 days f/b 1 wk

• ff for 6-month intervention).

BUT .…

• The 2016 Cochrane Systematic Review:
• Found that the longest intervention duration re endo thickness was
• nly 6 months! (Rioux et al Menopause 2000)
• Did not evaluate endometrial biopsy (EMB), hyperplasia, or cancer
• So, what are data from systematic EMB studies?

9/13/2018 9

LONGEST SYSTEMATIC EMB DATA

• tablet vs. crm –6 months
• estradiol tablet 25mcg once/wk or CEE cream 1.25 mg/d X 21

days f/b 1 wk off for 6-month intervention

• tablet: 1/49 proliferative endometrium;
• cream: 7/49 proliferative endometrium, 2/49 endometrial

hyperplasia) (Rioux et al Menopause 2000)

• Tablet without comparison group-12 months
• estradiol vaginal tablet 10mcg, 2 studies
• no hyperplasia or endometrial ca (Ulruch Climacteric 2010)
• 2 events of hyperplasia and carcinoma in 386 evaluable

biopsy samples (incidence rate 0.52% per year)(vs. background incidence of 0%-1%)(Simon Obstet Gynecol 2010)

LONGEST SYSTEMATIC EMB DATA

• Prasterone:
• E2 made in peripheral tissues from DHEA exerts action locally in

the same cells where synthesis takes place with only minimal release into the circulation. (Labrie et al Menopause 2009)

• Enzymes transform DHEA into estrogens are absent in the

endometrium (Portman et al Menopause 2015)

• Endometrial atrophy or inactive endometrium in all women

(Portman et al Menopause 2015, 52 wk; Archer et al Menopause 2015, 12 wk)

SO WHAT DO WE KNOW ABOUT VAGINAL ESTROGEN SAFETY FROM VAGINAL ESTROGEN TRIALS?

• Endometrial safety
• Serum levels—surrogate for systemic effects, particularly during

maintenance phase

SERUM LEVELS HEAD-TO-HEAD

Vaginal estradiol tablet

• vs. ring

Vaginal estradiol tablet

• vs. placebo

Vaginal estradiol tablet

• vs. CEE cream

(won’t discuss studies without a comparison group)

9/13/2018 10

SERUM LEVELS HEAD-TO-HEAD

Vaginal estradiol tablet

• vs. ring

Vaginal estradiol tablet

• vs. placebo

Vaginal estradiol tablet

• vs. CEE cream

(studies with a comparison group)

SERUM LEVELS: VAGINAL ESTRADIOL TABLET VS. CEE CREAM

• Frequency of serum E2 level outside postmenopausal range (>49

pg/mL) was significantly higher with vaginal cream vs. vaginal tablet, even at 6 months-

• about 50% of women using cream vs. about 5% using tablet at 6

months (bar graph)

• But: estradiol tablet 25 mcg once/wk or CEE 1.25mg/d X 21 days

f/b 1 wk off for 6-month intervention (Rioux et al Menopause 2000) higher CEE dose than often used! (Cochrane 2016 systematic review and my updated searches) 

BUT EVEN WITH LOWER CEE DOSE…

• Vaginal estradiol tablet vs. CEE cream in another study:
• Mean plasma E2 level at 12 wks was higher in CEE grp (12.6 +/- 10.1

pg/mL) than vaginal tablet group (8.9 +/- 10.5 pg/mL) though not

• stat. sig. difference (total n = 48)
• (Manonai et al J. Obstet Gynaecol Res 2001, Estradiol tablet

25mcg once/wk or CEE cream 1gm = 0.625mg 2X/wk X 10 wks)

SERUM LEVELS HEAD-TO-HEAD

Vaginal estradiol tablet

• vs. ring

Vaginal estradiol tablet

• vs. placebo

Vaginal estradiol tablet

• vs. CEE cream

(won’t discuss studies without a comparison group)

9/13/2018 11

SERUM LEVELS: VAGINAL ESTRADIOL TABLET VS. PLACEBO

• No sig. difference in serum E2 level between vaginal estradiol

tablet and placebo at 12 months-

• (15.5 +/- 2.5 pg/mL for estradiol tablet vs. 13.8 +/- 1.9 placebo)

(Simunic et al Int J Gynaecol Obstet 2003; 25 mcg dose)

SERUM LEVELS HEAD-TO-HEAD

Vaginal estradiol tablet

• vs. ring

Vaginal estradiol tablet

• vs. placebo

Vaginal estradiol tablet

• vs. CEE cream

(won’t discuss studies without a comparison group)

SERUM LEVELS: VAGINAL ESTRADIOL TABLET VS RING

• Estring:
• serum estradiol level ↑d from 4.3+/- 6.0 pg/mL to 13.3 +/- 17.4

pg/mL at week 24, then fell to 5.4 +/- 5.2 pg/mL at week 48.

• Vagifem:
• serum estradiol level ↑d from 4.1+/- 9.0 pg/mL to 9.8 +/- 13.9 pg/mL

at week 24, same at week 48. (Weisberg et al Climacteric 2005 Vagifem 25 mcg 2X/wk vs. Estring 7.5mcg/d, ?p-value between preps) (Vagifem is also avail in 10 mcg dose)

PRASTERONE 6.5 MG (0.50%) OVER 12 WEEKS

• serum E2 concentration increased statistically significantly, but was

22% below the average normal postmenopausal value (3.26 versus 4.17 pg/ml) (Ke et al J Steroid Biochem Mol Biol 2015)

9/13/2018 12

AND WE HAVE “PRASTERONE-ISH”

• RCT postmenopausal women with breast or gynecologic cancer
• DHEA 3.25 vs. 6.5 mg/d (compounding pharmacy) vs. plain moisturizer
• Only the 6.5-mg/day DHEA dose demonstrated a statistically

significant increase in E2 level compared with moisturizer

• At 12 wks mean serum E2 level 4.0 +/- 2.8 pg/mL

(Barton et al Support Care Cancer 2018) (6.5 mg is same dose as prasterone)

TIDE OF PATIENT PREFERENCE

Vaginal estradiol tablet: rated more acceptable than cream (Rioux et al Menopause 2000- estradiol tablet 25mcg once/wk or CEE 1.25mg/d X 21 days f/b 1 wk off for 6-month intervention) Vaginal estradiol ring : rated more preferable than cream (Nachtigall Maturitas 1995- 1.25mg CEE 3X/wk

• vs. estradiol ring 10mcg/d)

THE DISCUSSION GOES LIKE THIS

9/13/2018 13

Point

• Vaginal estradiol has never

been shown to increase endometrial cancer risk.

• True, but we only have short-term

studies from placebo-controlled trials with systematic endometrial biopsy

• Vaginal E2 tablet 12 mo (Ulrich

2010 & Simon 2010, 10 mcg)

• Vaginal estradiol ring 12 wks

(Nachtigall 1995)

Counterpoint Point

• But vaginal estradiol use is

associated with lower serum estradiol levels than is oral estrogen use.

• True, but...
• It’s all about the dose!
• Case in point-Femring high dose ring is

approved to rx HF

• Vaginal estrogens do ↑ E2 level, even
• utside postmenopausal range, at higher

doses CEE (1.25mg/d 21 days on + 1wk

• ff)

(Rioux et al Menopause 2000)

Counterpoint Point

• But we have no RCT evidence

that vaginal estrogen preparations increase the risk of the adverse events mentioned

• n the black box warning.
• (breast cancer, stroke, DVT,

PE, dementia, MI)

• True, but:
• Absence of evidence ≠

absence of risk, and

• We’re unlikely to ever get that

evidence from an RCT.

Counterpoint

9/13/2018 14

SO SINCE WE’RE UNLIKELY TO GET LONG-TERM RCTS OF VAGINAL ESTROGEN……

• Is there information about breast cancer, stroke, DVT, PE, dementia, MI

from large prospective observational studies?

• We have one published large prospective U.S. study.

I acknowledge my outstanding collaborators.

OBJECTIVE

• To determine the association between the use of vaginal estrogen

and an index of overall risks and benefits: the global index

• Developed for the WHI Hormone Therapy Trials

GLOBAL INDEX EVENT (GIE)

• Defined as first occurrence of:
• coronary heart disease (CHD)-i.e. fatal or nonfatal MI
• stroke
• pulmonary embolism
• invasive breast cancer
• colorectal cancer
• endometrial cancer
• hip fracture
• death from any cause

9/13/2018 15

STUDY COHORT

• Participants of the Women’s Health Initiative Observational Study
• Recruited at 40 U.S. clinical centers
• 50-79 years at baseline
• Main study conducted 1993-2005
• Extension study 2005-2010

N = 93,676 WHI- Observational Study

Exclusions:

• current HT users (oral, transdermal

estrogen and/or progestogen) at baseline or during follow-up (N= 41,630)

• history of breast, ovarian, endometrial

cancer (N= 7,079)

• missing HT use data (N= 85)
• missing info hysterectomy status (N= 454)
• no follow-up data (N= 473)

N = 45,663 N = 1,207 Vaginal estrogen users N = 3,003 Vaginal estrogen users N = 14,133 Hysterectomy* N = 32,433 Intact uterus* Figure 1. Algorithm of Study Participants

*numbers don’t add up to 45,663 due to time-varying nature of

hysterectomy status: 903 change from no hysterectomy to hysterectomy and are counted in both cells (Crandall et al, Menopause 2017)

PRIMARY OUTCOME

• Time to first occurrence of GIE
• Assessed on:
• annual self-assessment questionnaires, and
• Confirmed by:
• medical record review by WHI physicians except:
• Venous thromboembolism self-reported, not medical record-

confirmed (Crandall et al, Menopause 2017)

PRIMARY PREDICTOR

• Use of vaginal estrogen:
• ascertained on self-assessment questionnaires
• baseline
• annually from years 3 through 8
• Ext 1 phase (year 4)
• Captured initiation at any time in follow-up (time-varying variable)

(Crandall et al, Menopause 2017)

9/13/2018 16

STATISTICAL ANALYSIS

• Cox proportional hazards models
• The end of the study follow-up period was defined as 2.5 years after

the last ascertainment of vaginal estrogen use. (Crandall et al, Menopause 2017)

COVARIATES

• Baseline:
• education
• age
• race/ethnicity
• body mass index
• diagnosis of diabetes
• physical activity (total MET-

hours/week)

• hypertension
• Gail breast cancer risk score
• smoking
• income
• alcohol use (servings/week)
• Before study baseline history of:
• estrogen use
• cancer (other than breast,

endometrial, or ovarian)

• cardiovascular disease
• deep vein thrombosis or

pulmonary embolism

• fracture after age 55

(Crandall et al, Menopause 2017)

RESULTS: VAGINAL ESTROGEN USE

Median duration of vaginal estrogen use Median total duration f/u Users at baseline 3 yrs before enrollment Initiated during f/u 2 yrs 7.2 yrs Among vaginal estrogen users, on average, use encompassed 40% of the time during follow-up. (Crandall et al, Menopause 2017) Baseline characteristics No Vaginal est. N = 41,563 Vaginal est. N=4,100 Age baseline, years <60 26.2% 21.5% 60-69 44.1% 46.6% ≥70 29.8% 31.9% Race (Nonwhite n = 8933) White (not Hispanic) 79.8% 89.2% Education High school or less 25.5% 19.0% College or more 37.8% 46.3% BMI, kg/m2 <25 35.9% 45.5% 25-<30 34.2% 35.4% ≥30 29.9% 19.1% Current smoking 7.4% 3.6% (Crandall et al, Menopause 2017)

9/13/2018 17

Baseline characteristics, contd (no sig. diff. bet. users and nonusers) No Vaginal Est. N = 41,563 Vaginal est. N=4,100 Stroke ever 1.9% 1.5% Myocardial infarction ever 2.9% 2.0% History of revascularization (CABG or PTCA) 2.2% 1.6% History of cardiovascular disease (past MI, stroke, or revascularizaton) 5.6% 4.1% History of VTE (deep vein thrombosis

• r pulmonary

embolism) 5.1% 5.8% (Crandall et al, Menopause 2017)

Adjusted for age, education, past estrogen use, history of cancer before study baseline, history of cardiovascular disease before study baseline, history of deep vein thrombosis or pulmonary embolism before study baseline, race/ethnicity, baseline body mass index, baseline diagnosis of diabetes, baseline physical activity (total MET-hours/week), hypertension, Gail breast cancer risk score, fracture after age 55 prior to study enrollment, smoking, income, and alcohol use (servings/week). Model for

• verall analytic sample also includes hysterectomy status.

(Crandall et al, Menopause 2017) VE use

N events (rate)

No VE use

N events (rate)

HR (95%CI) GIE Overall 153 (14.0) 5906 (21.0) 0.76 (0.64, 0.91) Intact Uterus 96 (12.0) 4110 (19.9) 0.68 (0.55, 0.86) Hysterectomy 57 (19.7) 1796 (24.1) 0.94 (0.70, 1.26) Adjusted for numerous covariates

Adjusted for age, education, past estrogen use, history of cancer before study baseline, history of cardiovascular disease before study baseline, history of deep vein thrombosis or pulmonary embolism before study baseline, race/ethnicity, baseline body mass index, baseline diagnosis of diabetes, baseline physical activity (total MET-hours/week), hypertension, Gail breast cancer risk score, fracture after age 55 prior to study enrollment, smoking, income, and alcohol use (servings/week). Model for

• verall analytic sample also includes hysterectomy status.

(Crandall et al, Menopause 2017) VE use

N events (rate)

No VE use

N events (rate)

HR (95%CI) CHD Overall 20 (1.8) 1311 (4.5) 0.52 (0.31, 0.85) Intact Uterus 10 (1.2) 865 (4.0) 0.39 (0.19, 0.78) Hysterectomy 10 (3.4) 446 (5.6) 0.74 (0.37, 1.50) Stroke Overall 20 (1.8) 902 (3.1) 0.78 (0.49, 1.24) Intact Uterus 9 (1.1) 619 (2.9) 0.54 (0.28, 1.04) Hysterectomy 11 (3.7) 283 (3.6) 1.27 (0.67, 2.40) PE/DVT Overall 12 (1.1) 600 (2.0) 0.68 (0.36, 1.28) Intact Uterus 9 (1.1) 385 (1.8) 0.72 (0.34, 1.53) Hysterectomy 3 (1.0) 215 (2.7) 0.57 (0.18, 1.81)

Adjusted for age, education, past estrogen use, history of cancer before study baseline, history of cardiovascular disease before study baseline, history of deep vein thrombosis or pulmonary embolism before study baseline, race/ethnicity, baseline body mass index, baseline diagnosis of diabetes, baseline physical activity (total MET-hours/week), hypertension, Gail breast cancer risk score, fracture after age 55 prior to study enrollment, smoking, income, and alcohol use (servings/week). Model for

• verall analytic sample also includes hysterectomy status.

(Crandall et al, Menopause 2017) VE use

N events (rate)

No VE use

N events (rate)

HR (95%CI) Breast Cancer Overall 40 (3.6) 1185 (4.1) 0.91 (0.64, 1.29) Intact Uterus 26 (3.2) 858 (4.0) 0.79 (0.51, 1.22) Hysterectomy 14 (4.7) 327 (4.2) 1.23 (0.68, 2.21) Colorectal Cancer Overall 13 (1.2) 486 (1.6) 0.67 (0.35, 1.31) Intact Uterus 9 (1.1) 342 (1.6) 0.63 (0.28, 1.41) Hysterectomy 4 (1.3) 144 (1.8) 0.82 (0.26, 2.62) Endometrial Cancer Overall NA Intact Uterus 11 (1.3) 222 (1.0) 1.47 (0.75, 2.90) Hysterectomy NA

9/13/2018 18

Adjusted for age, education, past estrogen use, history of cancer before study baseline, history of cardiovascular disease before study baseline, history of deep vein thrombosis or pulmonary embolism before study baseline, race/ethnicity, baseline body mass index, baseline diagnosis of diabetes, baseline physical activity (total MET-hours/week), hypertension, Gail breast cancer risk score, fracture after age 55 prior to study enrollment, smoking, income, and alcohol use (servings/week). Model for

• verall analytic sample also includes hysterectomy status.

(Crandall et al, Menopause 2017) VE use

N events (rate)

No VE use

N events (rate)

HR (95%CI) Hip Fracture Overall 10 (0.9) 582 (2.0) 0.40 (0.19, 0.85) Intact Uterus 6 (0.7) 409 (1.9) 0.40 (0.16, 0.96) Hysterectomy 4 (1.3) 173 (2.2) 0.42 (0.10, 1.69) Death Overall 59 (5.3) 2592 (8.7) 0.78 (0.58, 1.04) Intact Uterus 32 (3.9) 1683 (7.7) 0.62 (0.41, 0.93) Hysterectomy 27 (9.0) 909 (11.3) 1.04 (0.68, 1.58)

SUMMARY

• In this large prospective cohort study, compared to nonusers of

vaginal estrogen, postmenopausal women who used vaginal estrogen had similar risks of:

• invasive breast cancer
• stroke
• colorectal cancer
• endometrial cancer, and
• venous thromboembolism
• We did not find evidence for elevated risk of CHD or death in vaginal

estrogen users compared with non-users.

THESE FINDINGS SHOULD:

• Provide reassurance to women and their health care providers

regarding the safety of vaginal estrogen and

• Help to inform menopausal hormone therapy clinical decision-

making.

LIMITATIONS

• Although we adjusted for numerous potential confounders, there

could have been residual confounding in this observational study.

• Information was not collected regarding specific vaginal estrogen:
• Dose, or
• type (cream, ring, or tablet)
• Use of the vaginal estradiol ring was only assessed at the end of the

follow-up period in the Extension Study, so users are not included in this analysis.

9/13/2018 19

STRENGTHS OF OUR STUDY

• Prospective data collection
• Large sample size
• Detailed information regarding multiple relevant confounders
• Medical record confirmation of all GIE events except venous

thromboembolism

• Ours was the first large prospective study to evaluate the overall

health risks and benefits of vaginal estrogen therapy in U.S. women.

HOW DO THOSE RESULTS COMPARE WITH THOSE OF PRIOR STUDIES?

GENERALLY CONSISTENT…

• with the (few) prospective observational studies:
• CHD (Finnish cohort Mikkola et al Human Reprod 2016)
• stroke (Finnish cohort Mikkola et al Human Reprod 2016; Danish

cohort Lokkegaard et al Stroke 2017)

• breast cancer (Finnish cohort Lyytinen et al Obstet Gynecol 2006)
• hip fracture (Swedish case-control Michaelsson et al Bone 2002)

among vaginal estrogen users.

AND GENERALLY CONSISTENT…

• Endometrial cancer studies
• Systematic EMB, two 12-month studies:
• estradiol vaginal tablets10 mcg
• no  in endometrial hyperplasia or cancer

(Ulrich et al Climacteric 2010 open-label; Simon et al Obstet Gynecol 2010 RCT includes Ulrich data & analyzed active rx arms)

• Study of Morch et al is the apparent exception? Let’s look closer…. 

9/13/2018 20

MORCH AND COLLEAGUES STUDY

• Large Danish registry study (1995-2009, 9 million person-years of follow-

up) that did not exclude users of systemic hormone therapy found two-fold increased endometrial cancer risk in users of vaginal estrogen

• vs. never users of HT
• prep not specified
• The differences in results between the Danish study and our recent

study might be related to:

• the fact that we censored data from vaginal estrogen users if they

initiated oral or transdermal progestogen, or

• differing doses or formulations of vaginal estrogen preparations in

the two studies (Morch et al Int J Cancer 2016)

• On May 29, 2018, the FDA rejected a proposal to modify package

labelling of low-dose vaginal estrogen products to accurately reflect evidence-based information for low-dose vaginal estrogen products approved for treating symptoms of vulvovaginal atrophy. (Docket No. FDA-2016-P-1246

ACKNOWLEDGEMENTS

• Rowan T. Chlebowski, MD, PhD
• Marcia L. Stefanick, PhD
• Dorothy S. Lane, MD, MPH
• Jan Shifren, MD
• Chu Chen, PhD
• Andrew Kaunitz, MD
• Jane A. Cauley, DrPH
• JoAnn Manson, MD DrPH
• Kathleen M. Hovey, MS
• Christopher A. Andrews, PhD
• The WHI program is funded by

the National Heart, Lung, and Blood Institute, National Institutes

• f Health, U.S. Department of

Health and Human Services through contracts:

• HHSN268201100046C,

HHSN268201100001C, HHSN268201100002C, HHSN268201100003C, HHSN268201100004C, HHSN271201100004C

BACK TO THE FOREST VIEW

Vincent van Gogh, Trees and Undergrowth, 1887

9/13/2018 21

SUMMARY

• Class labeling is based on extrapolations of data from clinical trials of

systemic hormone therapy, which:

• involved substantially higher levels of systemic exposure, and
• were not based on evidence from clinical trials of vaginal estrogen.
• But, we don’t have clinical trial data specific to vaginal estrogens in

relation to CHD, stroke, etc.

• GSM sx are likely long-term, but…
• Endometrial biopsy data show short-term safety
• 12 mo. for vaginal E2 tablet, 6 mo. for CEE cream, 12 wks. for E2 ring.

SUMMARY, CONTD

• Serum estradiol levels increase:
• But generally still in postmenopausal range
• outside postmenopausal range with CEE cream more often

than with vaginal E2 tablet, could be dose-related

• vaginal E2 25 mcg tablet = placebo, higher than Estring, but both
• postmeno. range during maintenance
• Women prefer the vaginal ring and the vaginal tablet to CEE cream
• In our prospective WHI observational study:
• the risks of CHD, invasive breast cancer, stroke, colorectal cancer,

endometrial cancer, VTE, and death are no higher in vaginal estrogen users compared with nonusers.

BUT…

• There’s still more to learn.