Effects of Conjugated Equine Estrogen in Postmenopausal Women with Hysterectomy The Women’s Health Initiative Randomized Controlled Trial JAMA 2004;291:
Postmenopausal Hormone Use Observational studies suggested –30%– 50% reduction in cardiac risk –8% – 30% increase in breast cancer –Strong benefit for osteoporosis In 1980’s – 90’s –Increasing use of hormone therapy for long-term cardiovascular disease prevention
WHI Hormone Trial Specific Aims Primary Endpoint: To test whether CEE reduces the incidence of CHD and other CVD –Definition of CHD: Nonfatal MI plus CHD death Primary Safety Endpoint: To assess whether CEE increases the risk of breast cancer Secondary Endpoints: To test whether CEE reduces the incidence of hip fractures and all osteoporosis-related fractures separately
Women’s Health Initiative Timeline
WHI Hormone Trial Design Hysterectomy CEE mg/d + medroxyprogesterone acetate (MPA) 2.5 mg/d E Alone N = 10,739 YES NO Placebo Conjugated equine estrogens (CEE) mg/d Placebo E + P N=16,608
Eligibility Age years at first screen Postmenopausal Likely to reside in clinic area for at least 3 years Providing written informed consent
Excluded if: Severely underweight Severe hypertension (>200/105) Severe menopausal symptoms Substance abuse, mental illness, dementia History of: –Cancer in last 10 years –Melanoma any time –Stroke or heart attack in last 6 months –Previous fractures treated with hormones Any condition likely to limit survival < 3 yrs So these were generally healthy women and this was a primary prevention trial
Follow-up Methods 6 weeks after randomization: phone contact –Assess symptoms –Reinforce adherence Every 6 months: phone or clinic contact –Assess adherence –Assess symptoms –Determine if outcomes had occurred Annually: clinic visit –Mammograms + clinical breast exams –Assess adherence –Assess symptoms –Determine if outcomes had occurred
Study Medication Use Intolerable symptoms e.g., breast tenderness –Reduced number of days of treatment –No unblinding for management Major event possibly related to hormones –Study meds stopped Temporarily for fracture, immobilization, surgery Permanently for breast CA, DVT/PE, malignant melanoma, meningioma, TG > 1000 mg/dl, if participant’s health provider prescribed open label hormones Resumption up to participant’s health provider for MI or stroke –No unblinding for discontinuation of medications
Ascertainment of Outcomes Self report of diagnosis/hospitalization Medical records obtained Local WHI physician adjudicated (coded) events (blinded to treatment assignment) Central adjudication (blinded to treatment assignment) –CHD –Stroke –VTE –Cancer –Hip fractures
Events during the Hormone Trial April 2000 – Participants notified of increase in CV events during first 2 years July 2001 – Participants informed risk does not disappear after 2 years July 2002 – –E+P trial intervention terminated due to increased breast cancer risk and overall risks exceeding benefits; Principal results published (JAMA 2002;288: ). –E alone trial is continued. February 2004 – NIH terminates E-alone intervention due to increased stroke risk and absence of CHD benefit March 1, 2004 – E- alone participants stop study pills
Baseline Characteristics of women in the WHI Estrogen alone trial
Age at entry CEE N = 5310 Placebo N = (30.8)1673 (30.8) (45.0)2465 (45.4) (24.2)1291 (23.8)
Age of E-alone participants at entry Percent
Race/Ethnicity Percent
Age at hysterectomy CEE N = 5310 Placebo N = 5429 < (39.8)2149 (39.8) (43.2)2275 (42.2) (9.5)566 (10.5) > (7.6)404 (7.5)
Bilateral Oophorectomy CEE N = 5310 Placebo N = 5429 Bilateral Oophorectomy 1938 (39.5)2111 (42.0)
History of Hormone Use CEE N = 5310 Placebo N = 5429 Never2769 (52.2)2770 (51.1) Past1871 (35.2)1948 (35.9) Current (3 month wash-out before randomization) 669 (12.6)708 (13.0)
Duration of prior hormone use CEE N = 5310 Placebo N = 5429 < 5 yrs1352 (53.2)1412 (53.1) 5-10 yrs469 (18.5)515 (19.4) > 10 yrs720 (28.3)732 (27.5)
Body Mass Index CEE N = 5310 Placebo N = 5429 < (21.0)1096 (20.3) (34.0)1912 (35.5) > (45.0)2383 (44.2) Mean BMI30.1
Smoking Status CEE N = 5310 Placebo N = 5429 Never2739 (51.9)2705 (50.4) Past1986 (37.8)2089 (38.9) Current542 (10.3)571 (10.6)
Medical History CEE N = 5310 Placebo N = 5429 Prior MI * 165 (3.1)172 (3.2) CABG/PTCA * 120 (2.3)114 (2.1) Angina308 (5.8)306 (5.7) Stroke76 (1.4)92 (1.7) DVT/PE87 (1.6)84 (1.5) * 441 (4.1%) women with “hard” CHD events prior to enrollment (> 6mos)
CHD Risk Status at Entry CEE N = 5310 Placebo N = 5429 Diabetes treatment410 (7.7)411 (7.6) Hypertension (Rx or > 140/90) 2386 (48.0)2387 (47.4) Elevated Cholesterol (on Rx) 694 (14.5)766 (15.9) Statin Use394 (7.4)427 (7.9) Aspirin Use1030 (19.4)1069 (19.7)
Baseline Characteristics of E-alone Participants (N=10,739) by Randomization Assignment NMeanSDNMeanSD Age (yrs) at screening BMI (kg/m 2 ) Systolic BP (mm Hg) Diastolic BP (mm Hg) CEE Placebo
Ethnicity White Black Hispanic American Indian Asian/Pacific Islander Unknown CEEPlacebo N%N%N%N% Baseline Characteristics of E-alone Participants (N=10,739) by Randomization Assignment
Treated diabetes Treated for hypertension or BP > 140/90 High cholesterol requiring pills Statin use at baseline Aspirin (>80mg) use at baseline CEEPlacebo N%N%N%N% Baseline Characteristics of E-alone Participants (N=10,739) by Randomization Assignment
History of MI History of angina History of CABG/PTCA History of stroke History of DVT or PE CEEPlacebo N%N%N%N% Baseline Characteristics of E-alone Participants (N=10,739) by Randomization Assignment
Hormone use Never Past Current Duration of prior hormone use (years) < CEEPlacebo N%N%N%N% Baseline Characteristics of E-alone Participants (N=10,739) by Randomization Assignment
Parity Never pregnant / no term pregnancy >1 term pregnancy Age at first birth < CEEPlacebo N%N%N%N% Baseline Characteristics of E-alone Participants (N=10,739) by Randomization Assignment
Female relative had breast cancer CEEPlacebo N%N%N%N% Baseline Characteristics of E-alone Participants (N=10,739) by Randomization Assignment
Fracture at age Falls in last 12 months or more CEEPlacebo N%N%N%N% Baseline Characteristics of E-alone Participants (N=10,739) by Randomization Assignment
Cumulative Drop-out and Drop-in Rates by Randomization Assignment and Follow-up Time
Intermediate Outcomes
Lipid Levels (8.6% subsample) Percent change, 1 yr
Clinical Outcomes
CEE and Coronary Heart Disease Events (Annualized %) by randomization assignment CHD † 177(0.49%)199(0.54%)0.91(0.75,1.12)(0.72,1.15) CHD Death54(0.15%)59(0.16%)0.94(0.65,1.36)(0.54,1.63) Non-fatal MI132(0.37%)153(0.41%)0.89(0.70,1.12)(0.63,1.26) * Adjusted for multiple comparisons across time (OBF procedures). A Bonferroni adjustment for 6 outcomes was applied to all outcomes other than CHD, Breast Cancer and the global Index. † CHD includes acute MI requiring hospitalization, silent MI determined from serial electrocardiograms and coronary deaths. There were 14 silent MIs. 95% CI CEEPlaceboHazard RatioNominalAdjusted *
Kaplan-Meier Estimates of Cumulative Hazards for CHD CHD HR, % nCI,
CHD events by years since randomization YearCEEPlaceboHR P-value for trend with time, 0.02.
Risk of CHD events by prior disease CEEPlaceboHR95% CI Prior MI or revascularization (N=441) No prior disease
CEE and Stroke Events (Annualized %) By randomization assignment Stroke158(0.44%)118(0.32%)1.39(1.10,1.77)(0.97,1.99) Fatal stoke15(0.04%)14(0.04%)1.13(0.54,2.34)(0.38,3.36) Non-fatal stroke114(0.32%)85(0.23%)1.39(1.05,1.84)(0.91,2.12) 95% CI CEEPlaceboHazard RatioNominalAdjusted
Kaplan-Meier Estimates of Cumulative Hazards for Stroke Stroke HR, % nCI,
Risk of stroke by prior disease CEEPlaceboHR95% CI Prior stroke (N=168) No prior stroke
CEE and Venous Thromboembolic Events (Annualized %) By randomization assignment VTE † 101(0.28%)78(0.21%)1.33(0.99,1.79)(0.86,2.08) DVT † 77(0.21%)54(0.15%)1.47(1.04,2.08)(0.87,2.47) PE † 48(0.13%)37(0.10%)1.34(0.87,2.06)(0.70,2.55) † VTE, venous thromboembolic disease; DVT, deep vein thrombosis; PE, pulmonary embolism 95% CI CEEPlaceboHazard RatioNominalAdjusted
Kaplan-Meier Estimates of Cumulative Hazards for PE PE HR, % nCI
CEE and Cardiovascular Disease Events (Annualized %) By randomization assignment Total CVD811(2.25%)746(2.01%)1.12(1.01,1.24)(0.97,1.30) 95% CI CEEPlaceboHazard RatioNominalAdjusted
CEE and Cancer Incidence (Annualized %) By randomization assignment Invasive breast94(0.26%)124(0.33%)0.77(0.59,1.01)(0.57,1.06) cancer Colorectal cancer61(0.17%)58(0.16%)1.08(0.75,1.55)(0.63,1.86) Total cancer372(1.03%)408(1.10%)0.93(0.81,1.07)(0.75,1.15) 95% CI CEEPlaceboHazard RatioNominalAdjusted
Kaplan-Meier Estimates of Cumulative Hazards for Breast Cancer Invasive Breast Cancer HR, % nCI,
Kaplan-Meier Estimates of Cumulative Hazards for Colorectal Cancer Colorectal Cancer HR, % nCI,
CEE and Fracture Events (Annualized %) By randomization assignment Hip fracture38(0.11%)64(0.17%)0.61(0.41,0.91)(0.33,1.11) Vertebral fracture39(0.11%)64(0.17%)0.62(0.42,0.93)(0.34,1.13) Total fracture503(1.39%)724(1.95%)0.70(0.63,0.79)(0.59,0.83) 95% CI CEEPlaceboHazard RatioNominalAdjusted
Kaplan-Meier Estimates of Cumulative Hazards for Hip Fracture Hip Fracture HR, % nCI,
CEE Summary Measures (Annualized %) By randomization assignment Death from other193(0.53%)185(0.50%)1.08(0.88,1.32)(0.79,1.46) causes Total death291(0.81%)289(0.78%)1.04(0.88,1.22)(0.81,1.32) † Global index is the first event for each participant from among the following types: CHD; stroke; PE; breast cancer; colorectal cancer; hip fracture; and death from other causes. Global index † 692(1.92%)705(1.90%)1.01(0.91,1.12)(0.89,1.14) 95% CI CEEPlaceboHazard RatioNominalAdjusted
Kaplan-Meier Estimates of Cumulative Hazards for Death Death HR, % nCI,
Kaplan-Meier Estimates of Cumulative Hazards for Global Index Global Index HR, % nCI,
Causes of Death (Annualized Percentages) by Randomization Assignment CEEPlacebo Number randomized Mean follow-up time (months) Total deaths291(0.81%)289(0.78%) Adjudicated deaths278(0.77%)272(0.73%) Cardiovascular93(0.26%)95(0.26%) Breast cancer4(0.01%)8(0.02%) Other cancer110(0.30%)118(0.32%) Other known cause51(0.14%)38(0.10%) Unknown cause20(0.06%)13(0.04%)
CEE Effects on Cardiovascular Outcomes by Age
CEE Effects on Cancer Outcomes by Age
CEE Effects on Hip Fracture by Age
Summary Measures of CEE Effects by Age
Sensitivity Analyses HR (compliers)HR (int. to treat) Stroke PE Total Mortality CHD Breast CA Colorectal CA Hip fracture Higher risk; Lower risk
Women’s Health Initiative Trial of E–Alone Summary
Absolute risk differences per 10,000 person/years with CEE 12 more strokes (p =.007) 6 fewer hip fractures (p =.01) –56 fewer osteoporotic fractures at any site (p <.001) 7 more VTE events (ns) 7 fewer breast cancers (ns) 5 fewer CHD events (ns) No difference in colorectal cancer No difference in total mortality No difference in pre-defined global index
WHI E-Alone: summary 1 st large-scale, long-term, randomized, double- blind, placebo-controlled trial to test unopposed estrogen on rates of chronic diseases finds that CEE –Increases risk of stroke –Reduces hip and other fractures –Does not significantly affect CHD rates –Does not increase risk of breast cancer risk
Implications CEE provides no clear benefit for chronic disease prevention Overall findings support current FDA recommendations to use postmenopausal hormone therapy for severe symptoms, at the lowest effective dose, for the shortest time. Women should be counseled about stroke risk, but no increased risk of heart disease or breast cancer for 6.8 years of use.