P1249945237VvHwx

1
Womens Health Initiative (WHI) Estrogen plus
Progestin Trial Design, Primary Results (JAMA
2002) 2003 Updates - Selected Outcomes
Marcia L. Stefanick, Ph.D. Associate Professor
of Medicine and of Gynecology and Obstetrics,
Stanford University PrincipaI Investigator,
Stanford Chair, WHI Steering Committee
2
WHI Hormone Trials Specific Aims

• To test whether Estrogen Progestin (EP) - or-
  Estrogen Only (E-Alone)
• reduces the risk of CHD or other CVD, e.g. Stroke
• increases the risk of Breast Cancer
• reduces the risk of Hip and other Fractures
• To determine the overall balance of health risks
  and benefits of EP and E-alone

3
WHI Hormone Program Design
CEE (Conjugated equine estrogens) 0.625 mg/d
YES
N 10,739
Premarin
Placebo
Hysterectomy
CEE 0.625 mg/d medroxyprogesterone acetate
(MPA) 2.5 mg/d

NO
N 16,608
Prempro
Placebo
Initially CEE only (N331), CEEMPA, or
Placebo
4
Outcomes Monitored by DSMB

• Cardiovascular disease
• Coronary Heart Disease (CHD)
• Strokes
• Pulmonary Emboli (PE)
• Invasive Breast Cancer
• Colorectal cancer
• Endometrial cancer
• Hip Fractures
• Deaths from other causes
• Global Index provides overall balance of
  benefits and risks earliest occurrence of CHD,
  stroke, PE, breast cancer, colorectal cancer,
  endometrial cancer, hip fracture or death from
  other causes

5
April 2000 WHI Data Safety Monitoring Board
(DSMB) requested that HT Participants be informed
that

• There was a small increase in the number of heart
  attacks, strokes, and blood clots in the lungs
  and legs in women receiving active hormones,
  compared to women taking placebo.
• After an average of 4 years of follow-up, there
  were more heart attacks, more strokes, and more
  PE and DVT, in active hormone group (EP
  E-Alone combined) vs placebo

June 2001 WHI DSMB required that all HT (EP and
E alone) Participants be informed that
6
May 2002 NHLBI accepted DSMB recommendation to
stop WHI Estrogen Progestin (EP) Trial

• After an average of 5.2 years
• Women in EP trial were told to stop study pills
  because the risks exceeded the benefits.
• Participants in the EP trial continue to be
  monitored, to determine how long risks or
  benefits persist, over time
• Women in E-Alone study were asked to continue
  study pills balance of benefits and risks is
  unclear.
• No increased risk of breast cancer had been seen
  in women taking estrogen only vs placebo by this
  time.
• E-Alone participants will continue to be closely
  monitored.

7
WHI Preliminary Results With CEE/MPA
Event nHR
CHD 1.29 Stroke 1.41 Breast Cancer 1.26 VTE
2.11 Colon Cancer 0.63 Hip Fracture 0.66 Total
Fracture 0.76 Death 0.98
Nominal 95 CI Adjusted 95 CI
Hazard Ratio
Writing Group for WHI Investigators JAMA 2002
288 321-333
8
Kaplan-Meier Estimates of Cumulative Hazards for
the Global Index The number of women at risk are
presented below the horizontal axis for each
treatment arm.
9
WHI EP Trial Annualized Event Rates
Placebo
E P
Risks
Neutral
60 50 40 30 20 10 0
Additional Events
Reduced Events
7
8
8
8
Number of Casesper 10,000 Women per Year
6
5
CHD
Stroke
Deaths
EndometrialCancer
BreastCancer
PE
ColorectalCancer
HipFracture
Statistically significant based on 95 nominal
CI on Hazard Ratios
Adapted from Writing Group for the Womens
Health Initiative. JAMA. 2002288321-333
10
EstrogenProgestin Study Results Summary

• 19 more health problems per 10,000 women
  assigned to EP compared to placebo
• Over 5 years, a net 1 per 100 women in EP group
  had a harmful outcome.
• Treatment with estrogen progestin for up to 5
  years is not beneficial to overall health.

11
Updated Results WHI Estrogen Progestin
Randomized, Placebo-controlled Clinical Trial

• Coronary Heart Disease (CHD)
• Manson J et al. N Engl J Med 2003 349 523-534
• Stroke
• Wassertheil-Smoller S et al. JAMA 2003 289
  2673-2684
• Breast Cancer (and Mammograms)
• Chlebowski, Rowan T et al. JAMA 2003 289
  3243-3253
• Fractures and Bone Mineral Density
• Cauley, Jane A. et al. JAMA 2003 290 1729-1738

12
CHD Update Manson JE, et al N Engl J Med 2003
349 523-34

• CHD outcomes through July 7, 2002
• mean 5.6 yr follow-up (vs. 5.2 in 1st report)
• 335 CHD cases (vs. 286 in first report)
• centrally-adjudicated by cardiologists
• Additional CHD endpoints
• angina, acute coronary syndrome, and CHF
• Analyses of subgroups defined by clinical
• characteristics and biomarkers

13
CHD Overall HR and HR Year of Follow-Up
CHD N (Annualized ) acute MI silent MI
CHD death (stratified, adjusted) EP 188 (0.39)
Placebo 147 (0.33) HR1.24 (1.00-1.54)
HR1.81
1.34
1.27
1.25
1.45
0.70
CHD Update Manson JE, et al N Engl J Med 2003
349 523-34
14
WHI EP Intermediate Outcomes at Year 1 (
change, EP minus Placebo)
CHD Update Manson JE, et al N Engl J Med 2003
349 523-34
15
WHI EP CHD Update Analyses by Subgroups

• Age (yrs) 50-59, 60-69, 70-79 p
  0.36
• Years since Menopause lt 10, 10-19, 20 p
  0.33
• Hot Flashes (in women aged 50-59) Yes, No
  p 0.16
• Hot Flashes Night Sweats (50-59) Yes, No p
  0.61
• BMI (kg/m2) lt 25.0, 25-29.9, 30 p
  0.60
• Race/Ethnic Group non-Hsp White, non-Hsp Black,
  Hispanic p 0.41
• Education Level H.S. or GED, gt
  H.S. p 0.86

CHD Update Manson JE, et al N Engl J Med 2003
349 523-34
16
WHI EP CHD Update Analyses by Subgroups

• LDL-C (mg/dl) lt 126, 126-155, gt155 p
  0.01
• HDL-C Triglycerides Total Cholesterol (p
  0.07) NS
• Lp(a) Fibrinogen Factor VIIIC
  C-reactive Protein NS
• Cigarette Smoking Never or former, Current p
  0.64
• Hypertension (140/90 or treated) No, Yes p
  0.49
• Diabetes No, Yes - treated, Yes - all cases p
  0.51
• Number of Risk Factors 0, 1-2, 3 (smoking,
  hypertension, p 0.96
• diabetes, high cholesterol, parental
  history-fatherlt55 yrs mother lt 65 yrs)
• Aspirin Use Yes, No p 0.71
• Statin Use Yes, No p 0.44

CHD Update Manson JE, et al N Engl J Med 2003
349 523-34
17
E P and Risk of CHD Additional Analyses of
Subgroups
adjusted for age and prior CHD
between subgroup and treatment variables include
s history of MI, CABG, PTCA, stroke, or transient
cerebral ischemia. includes history of MI,
CABG, or PTCA.
CHD Update Manson JE, et al N Engl J Med 2003
349 523-34
18
Clinical Outcomes (Annualized Percentage)
by Randomization Assignment
EstrogenProgestin Placebo Hazard Ratio
95 CI Nominal
Stroke 151 (0.31) 107 (0.24) 1.31 (1.02,1.08)
Ischemic 125 (0.26) 81 (0.18) 1.44 (1.09,
1.90) Hemorrhagic 18 (0.04) 20 (0.04) 0.82 (0
.43, 1.56)
Stroke Update Wassertheil-Smoller et al. JAMA
2003 2892673-2684
19
Summary and Implications of Stroke Data

• EP increases risk of ischemic stroke
• Excess strokes 7 per 10,000 women per year
• Excess risk is
• not explained by blood pressure increase
• apparent in hypertensives and normotensives
• apparent in all subgroups examined
• No significant interaction with any biomarker
  studied

Wassertheil-Smoller S et al. JAMA 2003 289
2673-2684
20
Updated Results WHI Estrogen Progestin
Randomized, Placebo-controlled Clinical Trial

• Gynecological Cancers
• Anderson GL S et al. JAMA 2003 290 1739-1748
• Colorectal Cancer
• Chlebowski RT. (Submitted to NEJM)
• Diabetes
• Margolis K et al. (Submitted to JAMA)
• Venous Thromboembolism (in preparation)

21
WHI EP Trial Findings for Gynecological Cancers
(Average 5.6 years of follow-up)

• Invasive Ovarian Cancer Risk (32 cases) HR
  1.58 CI 0.77-3.24
• Endometrial Cancer (58 cases)
• HR 0.81 CI 0.48-1.36
• No appreciable differences in distributions of
  tumor histology, stage, or grade for either.
• Cervical Cancer (13 cases)
• Data too limited.

Anderson GL S et al. JAMA 2003 290 1739-1748
22
Discontinuation and Drop-in Rates by
Randomization Assignment and Follow-up Time
Writing Group for WHI Investigators JAMA 2002
288 321-333
Percent
23
WHI EP Trial Primary EndpointsPercent Event
Rates Based on Analysis TypeFinal
centrally-adjudicated outcomes - 2003 (average
5.6 yrs of follow-up)
Intention to treat
80
60
Increase Over Placebo
40
31
20
24
24
0
CHD1
Stroke2
Breast Cancer3
1 Manson JE et al N Engl J Med 2003 349
523-534 2 Wassertheil-Smoller S et al JAMA
2003 289 2673-2684 3 Chlebowski RT et al.
JAMA 2003 289 3243-3253
24
WHI EP Trial Primary EndpointsPercent Event
Rates Based on Analysis TypeFinal
centrally-adjudicated outcomes - 2003 (average
5.6 yrs of follow-up)
Intention to treat
80
Compliant taking 80 study pills censored 6
mo. after becoming non adherent
60
Increase Over Placebo
40
31
20
24
24
0
CHD1
Stroke2
Breast Cancer3
1 Manson JE et al N Engl J Med 2003 349
523-534 HR 1.50 (1.14-1.97) 2
Wassertheil-Smoller S et al JAMA 2003 289
2673-2684 HR 1.50 (1.08-2.08) 3 Chlebowski
RT et al. JAMA 2003 289 3243-3253 HR 1.49
(1.13-1.96)
25
Updated Results WHI E P Trial

• Quality of Life
• Hays J et al. N Engl J Med 2003 3481839-1854
• Gynecological Symptoms (in preparation)
• Dementia
• Shumaker S et al JAMA 2003 289 2651-2662
• Global Cognitive Function
• Rapp S et al JAMA 2003 289 2663-2672

- - - - - - - - - - WHI Memory Study (WHIMS) -
- - - - - - - - EP Trial Participants (n4532)
aged 65 yrs (baseline)
26
WHIMS Frequencies of Probable Dementia and Mild
Cognitive Impairment Over 4.1 Years

• Dementia - 23 excess cases per 10,000 person-yrs
• MCI - No differences

Shumaker et al JAMA 2003 2892651-2662
27
Attributable Risk Summary

• Excess risk per 10,000 women per year on EP
• 8 more women with breast cancer
• 6 more women with CHD
• 7 more women with strokes
• 8 more women with PE
• Risk reduction per 10,000 women per year
• 6 fewer colorectal cancer
• 5 fewer hip fractures

Writing Group for WHI Investigators JAMA 2002
288 321-333 2003 UPDATES CHD (Manson)
Stroke (Wassertheil-Smoller) Breast Cancer
(Chlebowski) Hip Fractures (Cauley)
28
WHI EstrogenProgestin TrialImplications

• The overall risks of estrogenprogestin outweigh
  the benefits when taken to prevent chronic
  diseases in postmenopausal women.
• Estrogen progestin should not be initiated or
  continued for the primary prevention of CHD.
• Risk for CHD, stroke, PE and breast cancer must
  be weighed against benefit for fracture in
  selecting from available agents to prevent
  osteoporosis.