HORMONOTHERAPIE POSTMENOPAUSIQUE GRANDES ETUDES EPIDEMIOLOGIQUES: ASPECTS CARDIO-VASCULAIRES

1
HORMONOTHERAPIE POSTMENOPAUSIQUE GRANDES ETUDES
EPIDEMIOLOGIQUESASPECTS CARDIO-VASCULAIRES

• Ulysse GASPARD
• Service de Gynécologie-Obstétrique
• CHU SAR TILMAN
• Université de Liège - Belgique

2
WOMENS HEALTH INITIATIVE

• 16,608 healthy postmenopausal women (EP)
• 10,739 healthy surgically women (E)
• Age 50 79 years
• CEE 0.625mg (E)/ MPA 2.5mg (EP)
• Duration 5.2 years (EP) / 6.8 years (E) (planned
  8.5 years)
• Primary benefit CHD events
• Primary adverse event breast cancer
• Writing Group for the Womens Health Initiative
  Investigators, JAMA 2002288321-33
• The Womens Health Initiative Steering Committee.
  JAMA 20042911701-12

3
AIM AND RESULTS OF THE WHI STUDIES

• WHI is a study on prevention of cardiovascular
  disease in elderly women by hormonal treatment
  and NOT on HRT in menopausal women.
• (Marcia Stefanick, Principal Investigator,
  Florence 24/04/2004)
• Giving one high dose specific HRT regimen to
  older women who do NOT require HRT results in
  neither harm nor benefit for 99 of them.
• (John Stevenson, Amsterdam 03/10/2004)
• WHI did not study the appropriate population in
  the appropriate time period to establish that HT
  does not exert a primary preventive effect on the
  risk of CHD.
• (Leon Speroff, Maturitas, 2004483-4)

4
I. HRT AND RISK OF VTE OBSERVATIONAL STUDIES

• Meta-analysis of new observational
  epidemiological studies (after 1996) from Daly,
  Grodstein, Jick, Perez Gutthann and Varas groups
• RR according to type of current HRT regimen
• Overall RR for the FIRST year of treatment and
  thereafter
• ? 1 yr of HRT 4.2 (2.3 - 7.6)
• gt 1 yr 2.2 (1.3 - 3.8)
• Dose response relationship doubling of risk in
  CEE users lt 0.6 vs
• gt 0.6 (Daly and Jick)

Any unopposed opposed
oral transdermal RR
2.6(1.6-4.2) 2.5(1.3-4.8) 3.1(1.6-5.8)
2.8(1.6-4.8) 2.1(1.0-4.7)
Castellsague J et al,
Drug Safety 199818117
5
HRT AND RISK OF VTE RANDOMIZED THERAPEUTIC
TRIALS (RCTs)

• WHI study HR 2.11 (1.58 - 2.82) - (3.29if
  compliance adj.)
• (EP) (idem PE or DVT adj or not)
• 5.2 yrs FU - 2xmore 1st year of use
  ("healthy survivors ?)
• HERS study HR 2.7 (1.4 - 5.0) p 0.003
• 4.1 yrs FU - 2x more 2 first years of use,
• mainly for PE rather than DVT
• Risk was decreased with Aspirin or Statins !
  HR 0.5
• (0.2 0.9)
• Tamoxifen and Raloxifene RCTs RR ? 2.3
• (Fisher 1998 Cummings 1999)
• BERAL's meta-analysis of RCTs RR 2.2 (1.5 -
  3.2)

Beral V et al, Lancet
2002360942
6
Venous Thromboembolic disease outcomes in
WHI-E-only arm
Variable E-only group n5310 Placebo group n5429 Adjusted hazard ratio Adjusted 95 CI
n (annualized ) n (annualized )
VTE (all) Deep vein Thromb Pulm. Embolism BY AGE 50-59 60-69 70-79 101 (0.28) 77 (0.21) 48 (0.13) 18 (0.15) 49 (0.31) 34 (0.40) 78 (0.21) 54 (0.15) 37 (0.10) 15 (0.13) 39 (0.23) 24 (0.28) 1.33 1.47 1.34 1.22 1.31 1.44 0.86-2.08 0.87-2.47() 0.70-2.55 0.62-2.42 0.86-2.00 0.86-2.44()
() DVT significant only for nominal CI 95 p
0.03 () P value for interaction NS (0.39)
WHI Steering Committee JAMA
20042911701
7
RISK OF VTE DURING USE OF TRANSDERMAL HRT

• Oral E (P) ? ? in AT, PC, PS
• ? in APCR (APTT and ETP based), F
  VIII, IX, XI
• ? tPA, PAI1
• ?FPA and F12
• Dose-related effects
• Sidelmann JJ et al, BJOG 2003110541
• Oger E et al, ATVB 2003231671
• Transdermal E (P) no change
• Scarabin PY et al, ATVB 1997173071
• Scarabin PY et al, Lancet 2003362428

8
DIFFERENTIAL ASSOCIATION OF ORAL AND TRANSDERMAL
E-RT WITH VTE (ESTHER STUDY)

• Oral ERT ? blood coagulation and ? risk of VTE
• Transdermal ERT little effect on coag ? VTE
• ? Multicentre, hospital-based, case control study
  of PMW 155 documented idiopathic VTE (92 PE
  63 DVT) vs 381 controls matched for centre,
  age, time of recruitment
• ? Adj O.R. oral-ERT vs no use 3.5 (1.8 - 6.8)
• TTS-ERT vs no use 0.9 (0.5 - 1.6)
• oral-ERT vs TTS-ERT estimated risk 4.0 (1.9 -
  8.3)
• ? Transdermal safer than oral estrogen for VTE
  risk

Scarabin P.Y. et al,
Lancet 2003362428
9
II. STROKE AND HRT USE NURSE'S HEALTH STUDY AND
STROKE

• Observational cohort study of 70.533 PMW followed
  up from 1976 to 1996 767 strokes identified
• Risk of stroke in HRT users vs non users
• CEE ? 0.625mg/d RR 1.35 (1.08-1.68)
• CEE ? 1.25mg/d RR 1.63 (1.18-2.26)
• CEE progestin RR 1.45 (1.10-1.92)
• No increase in risk if CEE 0.3mg/d

Grodstein F et al, Ann Intern
Med 2000133933
10
WHI (EP) UPDATE ON STROKE IN PM WOMEN

• - n 8506 CEE MPA 8102 placebo 50 - 79y.o.
  F.U. 5.6yrs
• a) Overall strokes n 151 (1.8) treated and 107
  (1.3) placebo 31
• Overall HR 1.31 (1.08-2.08)
• ischemic (80) 1.44 (1.09 - 1.90)
• hemorrhagic (15) 0.82 (0.43 - 1.56)
• ? Risk of ischemic stroke only is increased
  in all age subgroups.
  Risk factors are not modified by EE MPA
• b) Case-control study 140 strokes (cases) vs
  513 controls assessment of biomarkers of
  inflammation, thrombosis, lipids increased risk
  of stroke NOT modified by EP intake
• ? Significant increase in thrombotic stroke if
  CEEMPA reinforce initial results of WHI (JAMA
  2002).

Wassertheil-Smoller S -
JAMA 20032892673
11
STROKE OUTCOMES IN WHI E-ONLY ARM
Outcomes E-only group n5310 Placebo group n5429 Adjusted hazard ratio Adjusted 95 CI
n (annualized ) n (annualized )
Stroke all Fatal Non fatal 158 (0.44) 15 (0.04) 114 (0.32) 118 (0.32) 14 (0.04 85 (0.23) 1.39 1.13 1.39 0.97-1.99 () 0.38-3.36 0.91-2.12
() Excess risk of stroke only significant for
nominal CI 95 (44 strokes in users vs 32 placebo
per 10.000 WY Z - 2.72. Monitoring boundary
was set at Z - 2.69
WHI Steering
Committee, JAMA 20042911701
12
WHI STROKE RISK
HR
HR
Stroke (E)
Stroke (EP)
5-10 10-15
Age (years)
Years postmenopause
Wassertheil-Smoller et al. JAMA
20032892673-84 The Womens Health Initiative
Steering Committee. JAMA 20042911701-12

(Courtesy J.C. Stevenson, Nov 2004)
13
WOMEN'S ESTROGEN FOR STROKE TRIAL (WEST STUDY)

• 664 PMW (mean age 71 years) with TIA or stroke
• RCT E2 1mg/d vs placebo x 2-8 years
• Overall RR E2 1.1 (0.8-1.4) for death or new
  stroke
• ? E2 does not reduce mortality or recurrence of
  stroke in PMW with cerebrovascular disease. E2
  not indicated for secondary prevention of
  cerebrovascular disease

Viscoli CM et al,
NEJM 20013451243
14
III. CHD/MI AND HRT USE

• MI incidence 50-54yrs 50/100.000WY
• 60-64yrs 200/100,000WY
• Ratio M/F 55-65yrs 3.3 1.0
• Possible reasons for ? risk with age ?
  atherosclerosis ? classical risk factors ?
  markers of inflammation, impaired thrombotic and
  rheological variables
• Smoking impact calculated in PM women (studies
  from Grodstein, Sidney, Rosenberg and Psaty)
• overall RR 3.3 (2.9-3.7)
• Hypertension impact overall RR 2.6 (2.3-2.9)

Lowe GDO, Maturitas 2004
Farley T et al, Maturitas 2004
15
NURSESHEALTH STUDY RISK OF CHD

• Observational cohort study of 70533
  postmenopausal women
• Follow-up 1976 to 1996
• 1258 non fatal MI or CHD death
• RR in current users of HRT 0.61 (0.52-0.71)
• Evidence of primary prevention of CHD (MI) by E
  P
• or E alone vs former or never use ()
• Recent Danish Nurses Obs Study no beneficial
  effect of HT
• () Similar results for secondary prevention in
  that study

Grodstein F et
al, Ann Intern Med 2000133933
Lokkegaard E et al, BMJ 2003326426
16
CHD OUTCOMES IN WHI (E P)
- Primary efficacy outcome of the trial non fatal MI or death due to CHD at 5.6yrs - Primary efficacy outcome of the trial non fatal MI or death due to CHD at 5.6yrs - Primary efficacy outcome of the trial non fatal MI or death due to CHD at 5.6yrs - Primary efficacy outcome of the trial non fatal MI or death due to CHD at 5.6yrs - Primary efficacy outcome of the trial non fatal MI or death due to CHD at 5.6yrs
Outcomes EP Group n8506 Placebo group n8102 Adjusted Hazard Ratio Adjusted CI 95 ()
n (annualized ) n (annualized )
All CHD Fatal CHD 188 (0.39) 39 (0.08) 147 (0.33) 34 (0.08) 1.24 1.10 0.97-1.60 (NS) 0.65-1.89 (NS)
() Adjusted for age, coronary events/therapy and
for sequential monitoring
Manson JE et
al NEJM 2003349523
17
WOMENS HEALTH INITIATIVE
CHD events

• increased CHD events in early years
• ? increased thrombogenesis / adverse remodelling
• age skewed to older women
• ? oestrogen dose too high
• ? effect of MPA (E-alone arm had early harm)
• ? healthy survivor effect

trend p0.02
Manson et al. N Engl J Med 2003 349 523-34
(Courtesy J.C. Stevenson, Nov 2004)
18
CHD OUTCOMES IN WHI E-ONLY ARM
- Primary efficacy outcomenon fatal MI or death due to CHD at 6.8 years - Primary efficacy outcomenon fatal MI or death due to CHD at 6.8 years - Primary efficacy outcomenon fatal MI or death due to CHD at 6.8 years - Primary efficacy outcomenon fatal MI or death due to CHD at 6.8 years - Primary efficacy outcomenon fatal MI or death due to CHD at 6.8 years
Outcomes E-only group n5310 Placebo group n5429 Adjusted hazard ratio Adjusted 95 CI
n (annualized ) n (annualized )
All CHD Fatal CHD 177 (0.49) 54 (0.15) 199 (0.54) 59 (0.16) 0.91 0.94 0.72-1.15 0.54-1.63
HR changes with time of study Year 1 1.16 Trend with time Year 2 1.20 significant reduction Year 3 0.89 of risk (P 0.02) Year 4 0.79 Year 5 1.28 Year 6 1.24 Year 7 0.42 HR changes with time of study Year 1 1.16 Trend with time Year 2 1.20 significant reduction Year 3 0.89 of risk (P 0.02) Year 4 0.79 Year 5 1.28 Year 6 1.24 Year 7 0.42 HR changes with time of study Year 1 1.16 Trend with time Year 2 1.20 significant reduction Year 3 0.89 of risk (P 0.02) Year 4 0.79 Year 5 1.28 Year 6 1.24 Year 7 0.42 HR changes with time of study Year 1 1.16 Trend with time Year 2 1.20 significant reduction Year 3 0.89 of risk (P 0.02) Year 4 0.79 Year 5 1.28 Year 6 1.24 Year 7 0.42 HR changes with time of study Year 1 1.16 Trend with time Year 2 1.20 significant reduction Year 3 0.89 of risk (P 0.02) Year 4 0.79 Year 5 1.28 Year 6 1.24 Year 7 0.42
WHI Steering Committee, JAMA 20042911701-1712
19
WHI CHD RISK
HR
HR
CHD events (E)
CHD events (EP)
10-19
age (years)
years postmenopause
Writing Group for the Womens Health Initiative
Investigators. JAMA 2002 288 321-33
The Womens Health Initiative Steering Committee.
JAMA 2004 291 1701-12
(Courtesy J.C. Stevenson, Nov 2004)
20
HERS (I) TRIAL

• 2763 women (1380 EP vs 1383 placebo)
• mean age 66.7 years
• gt6 months from cardiac event
• conjugated equine oestrogens 0.625 mg MPA 2.5
  mg
• event rate 3.3 (estimated 5)
• mean follow-up 4.1 years (estimated 4.75 years)
• no overall benefit seen
• RH 0.99 (0.82 1.14)

CHD events
trend p0.009
(Courtesy J.C. Stevenson, Nov 2004)
Hulley et al. J Am Med Assoc 1998 260 605-13
21
(No Transcript)
22
ARE HRTs EFFECTS ON CHD AGE-DEPENDENT ?

• Nijmegen study later age at menopause ? CV
  mortality
• (De Kleijn MJJ et al, Am J Epidemiol
  2002155339)
• Brachial artery flow-mediated dilation
• ? in healthy and young HRT users w/o CHD
• T in older women and HRT users with CHD
• (Herrington D et al, ATVB 2001211955)
• Clarkson non-human primates
• Age, stage of atherosclerosis, ? artery wall
  ER,
• ? plaque inflammation suppress efficacy of HRT
  for
• decreasing plaque area.
• ? Clarksons model Window of opportunity for
  HRT benefit 6 years postmenopause
• (Clarkson, Gynaec Forum 2004911)
• E ? MMPs. Statins ? MMPs
• - WHI (EP)Statins HRT HR for CHD 0,99HRT
  and no statins HR1.27 (Manson JE et al, NEJM
  2003349523)

23
Relation of Age Distribution in WHI to Stage of
Progression of Coronary Artery Atherosclerosis
70
10 50-54
20 55-59
0
45 60-69
25 70-79
yrs
Clarkson, 2002.
24
Timing of HT and Prevention of Atherothrombosis
Hypothetical Pathogenic Sequence
No HRT
MMP-9
Fibrous Cap
Fibrous Cap
Fibrous Cap
Adventitia Media Internal Elastic Lamina

Plaque
Necrotic Core
Necrotic Core
Fatty Streak/Plaque
HRT Early Continued
HRT
Mural Thrombus
HRT Late
HRT
3545 yrs
4555 yrs
gt 65 yrs
5565 yrs
Clarkson, 2002.
25
Hormone Therapy and Plaque Reduction
A. Early Intervention
B. Late Intervention
0.40 0.30 0.20 0.10 0
PN.S.
70 Plt0.05
Coronary Artery Plaque Size (mm2)
Placebo CEE Baseline Placebo
CEE CEE MPA
Surgical Menopausal Monkeys Rx with Conjugated
Estrogen With and Without MPA.
Clarkson 2002. INT. J. FERTIL. 4761
26
DIFFERENCES BETWEEN OBSERVATIONAL AND CLINICAL
TRIAL PARTICIPANTS
Observational Clinical trials
Menopause symptoms Age started HRT Time since menopause Stage of atherosclerosis Most users symptoms 45-55 years 0 1 year Fatty streaks, plaques Few with symptoms 63 (WHI) 67 (HERS) gt 70 (WEST) ? 14 (WHI) ? 18 (HERS) gt 20 (WEST) Advanced, unstable plaques
Adapted from Clarkson
TB et al, Gynaec Forum 2004911
27
Some problems with interpretation of RCTs in
perimenopausal women

• WHI was ten-fold underpowered to show
  cardioprotection of women starting HRT during
  menopausal transition
• (Naftolin, Fertil Steril, 2004)
• Importance of timing of intervention peri or
  postmenopausal the 6-year window
• (Clarkson data !)
• E-only arm of WHI RR of CHD 0.91 vs 1.24 (EP
  arm)
• Past E use in E-only arm was 35 (young age) vs
  15 use in EP arm
• 50-59 yrs HR 0.56 CEE alone vs placebo
• 60-69 yrs 0.92
• 70-79 yrs 0.94

Anderson G, 2003
28
CONCLUSIONS HT AND CVD (I)

• Confirmation of increased risk of venous
  thromboembolism and thrombotic stroke in
  postmenopausal women particularly if risk
  factors accidents begin early (1-2 y) E dose and
  route strongly implied in VTE potentially in
  stroke
• CHD no primary prevention in older women by EP
  and potentially E-alone cardio-prevention if lt
  10 years postmenop not ruled out. No secondary
  prevention demonstrated. E-dose implied ??

Stevenson JC, Pract CV risk man
200317 Herrington
DM et al, Atherosclerosis 2003166203
29
CONCLUSIONS HT AND CVD (II)

• For risk of CHD we need to study a young
  postmenopausal healthy population in order to
  establish that MT does or does not exert a
  primary preventive effect on CHD (Speroff L,
  Maturitas 2004483-4)
• The comparative study of Scarabin et al (Lancet
  2003362428) on E2 oral vs transdermal has to be
  expanded/initiated not only in case of VTE but
  also for stroke and CHD
• The estrogen dose and route, and the progestin
  type dose and regimen should be carefully
  selected (or different regimens compared) in
  these new randomized trials

30
MOTOR CARS
Lamborghini
Both are motor cars. .but performance
differs! Compare with HRT!
(Courtesy
J.C. Stevenson, Nov. 2004)
31
ESTIMATING THE RISK-BENEFIT BALANCE FROM WHI
STUDIES

• Global index of health benefit vs risk created by
  Data Safety and Monitoring Board
• Adverse outcome of HRT (EP) demonstrated
• HR 1.15(1.03-1.28)excess risk in 19
  women/10,000WY (lt 0.2)
• Adverse outcome not found in WHI E-only arm
• HR 1.01(0.91-1.12)benefit in 2 women/10,000WY
  (lt 0.05)
• ?This global index does not encompass ALL
  clinical outcomes recorded in the WHI studies !
• ? - This global index was not previously
  validated
• - Not really considered as playing a role in
  understanding how hormones should be used

Stevenson JC, 2004 Anderson G,
2003
32
Global index of ALL clinical outcomes of WHI
studies
? Comparison of all events (excess or reduction) per 10,000 WY in WHI studies ? Comparison of all events (excess or reduction) per 10,000 WY in WHI studies ? Comparison of all events (excess or reduction) per 10,000 WY in WHI studies ? Comparison of all events (excess or reduction) per 10,000 WY in WHI studies
WHI EP Study WHI EP Study WHI E-only Study WHI E-only Study
RISKS All CVD 25 All cancers 3 BENEFITS All fractures 44 Deaths 1 RISKS All CVD 24 Deaths 3 BENEFITS All fractures 56 All cancers 7
Overall Balance 17 overall benefit in 8 women per 10,000 WY Overall Balance 17 overall benefit in 8 women per 10,000 WY Overall Balance 36 overall benefit in 22 women per 10,000 WY Overall Balance 36 overall benefit in 22 women per 10,000 WY

Stevenson J.C., 2004
33
HT AND VTE IN POSTMENOPAUSAL WOMEN SUMMARY 1

• RR 2-3 (OBS and RCT studies agree)
• RR not dependent of age or time since menopause
• 1st year of HT (E or EP) 2 x more VTE
• E-dose related E-route probably related
• RR ?? if HT risk factor (eg FV Leiden)
• ? HT or ET increased risk of VTE in PMW
• No primary or secondary prevention
• ? Explanation Prothrombotic effect of E ?
  (dose/route)

34
HRT AND STROKE IN POSTMENOPAUSAL WOMEN SUMMARY 2

• RR 1.31-1.45 (OBS and RCT studies agree)
• RR not dependent of age or time since menopause
• Incidence increases from 2nd year of HT
  onwards
• E-dose related thrombotic stroke only
• 2nd Prevention (WEST study) RR 1.1
• ? HT or ET increased risk of thrombotic stroke
  in PMW
• No primary or secondary prevention
• ? Explanation Prothrombotic effect of E ?
  (dose)

35
HT AND CHD IN POSTMENOPAUSAL WOMEN SUMMARY 3

• OBS and RCT studies disagree (healthy user bias
  in OBS a.o.)
• OBS studies (eg NHS) RR 0.61 (I and II
  Prevention)
• RCT studies - Older women one high dose regimen
  of
• CEE MPA
• - RR dependent of age and
  time since menopause
• EP RR 1.24 only significant 1st year
  (1.8)
• RR 0.89 if lt 10 years postmenop
• RR 1.71 if gt 20 years postmenop
• E-only RR 0.91
• RR 0.56 if PMW 50-59 years of age
• Secondary Prevention (HERS I II) RR 0.99
• RR 1st year 1.5
• ?HT or ET slightly increased risk of CHD in
  older women. Probability of LOWER risk in early
  postmenop (primary prevention). No secondary
  prevention.
• ? Explanation Prothrombotic and Proinflammatory
  effects of E, ? in older women (unstable plaques)
  ?

36
Effect of ET on coronary atherosclerosis in
monkeys timing of initiation
Thomas B.
Clarkson,Gynaecology ForumVol.9,n3, 2004
37
JoAnn E. Manson, N Engl J Med
2003349523-34
38
HERS II TRIAL

• HERS I
• Higher risk of CHD events during the first year
  in EP users
• Decreased risk during years 3 to 5 (significant
  trend)
• ? HERS II
• Unblinded follow up for 2.7 additional years
  (93 of those surviving)
• HERS-I overall RH 0.99 (0.81-1.22)
• HERS-II overall RH 1.00 (0.77-1.29)
• HERS-I II overall RH 0.99 (0.84-1.17)
• (similar results after adjustment for
  confounders and use of statins)
• CONCLUSION
• HERS-I early harm prothrombotic effect of
  EP ?
• late benefit improvement due to HRT
  or healthy survivor effect ?
• HERS-II benefit did not persist at 6.8years
• ? EP did not reduce risk of recurrent CHD in PM
  women with CHD

Grady D et al, JAMA 200228849
39
Association Française pour l'Etude de la Ménopause
1979-2004 XXVes Journées