ERT and CAD in Women

1
Estrogen Replacement Therapy and the Prevention
of Coronary Heart Disease in Women Friend or Foe?
David Parra, Pharm.D., BCPS Clinical Pharmacy
Specialist Department of Cardiology Veterans
Affairs Medical Center West Palm Beach, FL
2
Objectives

• Understand the magnitude of coronary heart
  disease (CHD) in postmenopausal women
• Explain the mechanisms behind estrogens proposed
  cardioprotective effect
• Review the observational data supporting the use
  of estrogen in preventing CHD
• Discuss the results of the recently completed
  Heart and Estrogen/progestin Replacement Study
  (HERS) and apply them to clinical practice

3
Cardiovascular Disease in Women

• One in two women will die of cardiovascular
  disease (CVD)
• if all forms of major CVD were eliminated life
  expectancy would increase by 10 years
• One in 26 women will die of breast cancer
• if all forms of cancer were eliminated life
  expectancy would increase by 3 years

1998 Heart and Stroke Statistical Update, AHA.
4
Cardiovascular Disease in Women

• 63 of women (48 of men) die suddenly from
  coronary heart disease
• 44 of women (27 of men) will die within one
  year after a heart attack

1998 Heart and Stroke Statistical Update, AHA.
5
Leading Causes of Death for All Females
505,440
256,844
Deaths in thousands
48,961
45,136
33,130
0
United States 1995 Mortality Data Adapted from
1998 Heart and Stroke Statistical Update, AHA.
6
Coronary Heart Disease Despite Advances, Still
the 1 Killer

• Percentage Breakdown of Deaths From
  Cardiovascular Diseases
• United States 1995 Mortality, Final Data

22 Other
1 Rheumatic Fever/Rheumatic Heart Disease
50 Coronary Heart Disease
1 Congenital Heart Defects
2 Atherosclerosis
4 Congestive Heart Failure
4 High Blood Pressure
16 Stroke
American Heart Association1998 Heart and Stroke
Facts Statistical Update
7
Cardiovascular Disease Mortality Trends
United States 1995 Mortality Data Adapted from
1998 Heart and Stroke Statistical Update, AHA.
8
Premenopausal Postmenopausal

• Decrease in HDL
• Increase in LDL, triglycerides, apolipoproteins B
  and A-I
• Increase in diastolic blood pressure

Menopause exerts a negative effect on CHD risk
9
Incidence of Coronary Heart Disease
Framingham Cohort
Adapted from Kannel et. al. American Heart
Journal. 1987114413-9.
10
The Framingham Heart StudyAnnual Rate of
Coronary Artery Disease in Women as a Function of
Age
Adapted from Castelli et al. Am J Obstet Gynecol
1988 158 1553-60.
11
Estimated Prevalence of CHD in Women by
Age United States 1988-91
25
18.2
20
13.6
15
Percent Female Population
11
7.9
10
5.3
3.8
3.1
5
0
20-29
30-39
40-49
50-59
60-69
70-79
gt 80
Ages
Adapted from 1998 Heart and Stroke Statistical
Update, AHA.
12
Early Outcome of Acute Myocardial Infarction -
ISIS-3
Adapted from Malacrida R et. al. N Eng J Med.
19983388-14.
13
Premenopausal Postmenopausal

• Loss of endogenous estradiol production
• Presumption estrogen has a role in premenopausal
  protection against CHD
• Conversely its loss has a role in postmenopausal
  risk

Estrogen replacement therapy (ERT) should
decrease this risk by maintaining metabolic
factors that affect CHD at premenopausal levels
14
Lipids/Atheroma
Hemostatic/Platelet
Homocysteine
Estrogens Cardioprotective
Mechanisms
Antioxidant
Nitrous Oxide
Inhibition of Constricting Agents
Calcium Channel Antagonism
Carbohydrate Metabolism
15
Female Life Cycle and Lipids
160
140
LDL-C
120
100
Mean values (mg/dL)
80
60
40
HDL-C
20
0
75-7
15-19
20-24
25-29
30-34
35-39
40-44
45-49
50-54
55-59
60-64
65-69
70-74
Age (years)
Adapted from Kannel et. al. American Heart
Journal. 1987114413-9.
16
Estrogen and Lipids

• Decrease LDL 5-15
• Increase HDL 5-15
• Increase triglycerides 4-14
• /- lipoprotein (a)
• Effect dependent on route and formulation

17
Estrogens Mechanism of Action on Lipids

• Induction of LDL receptor formation
• Destruction of hepatic lipase
• 25-50 of beneficial effect on CHD

18
Platelet Effects

• Increased production of local prostacyclin (PGI2)
• enhancement of prostacyclin stabilizing factor
• Favorable prostacyclin/thromboxane balance
• increase in blood flow
• anti-aggregation effect

19
Peripheral Vascular Effects

• Estrogen receptors in blood vessels
• Estrogen increases in blood flow
• decrease in arterial impedance
• decreased vascular tone in uterine arteries
• increase in cerebrovascular blood flow
• Decreased anginal episodes and increases in
  treadmill times

20
Peripheral Vascular Effects

• Increased production of prostacyclin in
  endothelium
• Decreased thromboxane A2 synthesis by platelets
• Facilitate release or response to nitrous oxide
• Inhibit release or response to constrictor
  factors
• Calcium channel antagonist role

21
Direct Effect on Myocardium

• Estrogen receptors present in the heart and aorta
• ST segment changes induced by estrogen resemble
  those inducible by digoxin
• Evidence by echocardiogram of changes in stroke
  volume and mean acceleration
• Net result of possible positive inotropic effect

22
Hemostatic Effects

• Complex and variable
• decreased fibrinogen and antithrombin III
• increased factor VII and Protein C
• overall observational findings suggest a
  reduction in risk of thrombosis

23
Other Effects

• Carbohydrate metabolism
• increased insulin release and receptor
  sensitivity
• may be opposite at doses gt 1.25mg
• Antioxidant
• both estrogen and progesterone
• Body fat
• gynoid fat versus android fat

24
Progestin Effects on the Cardiovascular System

• Progestin product dependent
• Progesterone receptors present in blood vessels
• In-vitro negation of increased PGI2
• Attenuation of increased uterine blood flow
• Decreased estrogen receptor activity
• Blunting of estrogens effect on lipid profile

25
From Mechanistic to Observational Evidence
26
Estrogen Use and Risk of CHD
Hospital case-control
Population case-control
Prospective internal control
Cross-sectional
Prospective external control
All studies combined
Prospective internal control cross sectional
0.5
1
1.5
2
RR
Adapted from Stampfer et al. Prev Med 199120
47-63.
27
Relative Risk of CHD Among All Postmenopausal
Hormone Users
Hormone Use RR Major Coronary Disease
Never used 1.0 Currently used Estrogen 0.60
(0.43-0.83) Estrogen progesterone 0.39
(0.19-0.78)
Adjusted for multiple risk factors Adapted from
Grodstein et al. NEJM 1996335 453-61. Nurse
Health Study 1976-1994.
28
Change in RR with Estrogen Therapy with and
without Progesterone
Disease ERT ERT Progesterone
Osteoporosis 0.40 0.40 Endometrial
Cancer 6.0 1.0 Breast Cancer 1.37 1.60 Ischemic
Heart Disease 0.52 0.69 Stroke 0.50 0.67 Mortalit
y Change -328 -188 (per 100,000)
Adapted from Ross et al. Lancet 19814858-60.
29
Estimated Lifetime Probability of Selected Events
with ERT
Treated
Condition Not Treated () Assumed RR
CHD 46.1 0.65 34.2 Stroke 19.8 0.96 20.2 Hip
Fracture 15.3 0.75 12.7 Breast Cancer 10.2 1.25 13
.0 Endometrial CA 2.6 8.22 19.7 Life Expectancy
(y) 82.8 83.7
In women treated with ERT 15 years or more versus
those not treated. Adapted from Grady et al. Ann
Intern Med 1992 Petitti et al. Ann Epidemiol
1994.
30
Estimated Change in Mortality with Estrogen
Replacement Therapy
Condition RR Cumulative Change
Osteoporotic Fractures 0.4 - 563 per
100,000 Gallbladder Disease 1.5 2 per
100,000 Endometrial Cancer 2.0 63 per
100,000 Breast Cancer 1.1 187 per
100,000 Ischemic Heart Disease 0.5 - 5,250 per
100,000 Net Change - 5,561 per 100,000
Adapted from Henderson et al. Am J Obstet Gyn
1986154 1181-6.
31
Relative Risk of Death Among All Postmenopausal
Hormone Users
Hormone Use
Cause of Death Current Past Never
All Cause 0.63 (0.56-0.70) 1.03
(0.94-1.12) 1.0 CHD 0.47 (0.32-0.69) 0.99
(0.75-1.30) 1.0 Stroke 0.68 (0.39-1.16) 1.07
(0.68-1.69) 1.0 All Cancer 0.71 (0.62-0.81) 1.04
(0.92-1.17) 1.0 Breast Cancer 0.76
(0.56-1.02) 0.83 (0.63-1.09) 1.0
Adjusted for multiple risk factors Adapted from
Grodstein et al. NEJM 1997336 1769-75. Nurse
Health Study 1976-1994.
32
Summary of Observational Evidence

• Meta-analyses
• relative risk of coronary heart disease is
  0.50-0.65 with ERT
• addition of progesterone does not appear to
  attenuate the effect of ERT
• ERT estimated to save 5,250 lives per 100,000
  users

33
Confounding Factors

• Effects of progesterone
• Study biases
• Limitations of meta-analyses

34
Confounding Factors-progesterone

• Early epidemiologic studies usually in women on
  estrogen replacement therapy only
• Progesterone formulations and their effects
• 19-nortestosterone derivates (levonorgesterol)
• 17-hydroxyprogesterone derivates
  (medroxyprogesterone)
• micronized progesterone

35
Confounding Factors-healthy user

• Many studies demonstrate estrogen users
• higher socioeconomic status
• better educated
• younger
• thinner
• more likely to drink alcohol

36
Confounding Factors-compliance

• Long term hormone replacement therapy (HRT) users
  are compliers
• More likely to take ASA, MVI, and exercise
• Patients who comply with therapy (even placebo)
  have reduction in mortality

37
Confounding Factors-compliance

• Beta-blocker Heart Attack Trial

Beta-blocker Placebo Compliance
Mortality Relative Risk Mortality Relative
Risk lt75 8.7 19.0 gt75 4.5 0.53 6.8 0.36
Crude relative risk of mortality in women
participating in the trial
Petitte DB Ann Epidemiol 4 115-118, 1994
38
Confounding Factors-compliance

• Coronary Drug Project

• Clofibrate Placebo
• Compliance Mortality Relative Risk
  Mortality Relative Risk
• lt80 22.5 25.8
• gt80 15.7 0.70 16.4 0.64
• Adjusted 5-year mortality rate and relative risk
  in high-compliance group compared with
  low-compliance group

Petitte DB Ann Epidemiol 4 115-118, 1994
39
Confounding Factors-compliance

• Relative risk of CHD is 0.5-0.65 with ERT
• Relative risk of mortality in some studies is
  0.36-0.64 in patients compliant with placebo
• Estrogen users are by definition compliant

Suggestive that compliance bias, in theory, could
account for most of the benefit of ERT seen in CHD
40
Confounding Factors-surveillance

• Medical care
• sought on a more regular basis by HRT users
• risk factors identified earlier
• more likely to have preventative health screens

41
Confounding Factors-healthy survivor

• Breast Cancer Detection Demonstration Project
  Follow-up Study (Sturgeon et al.)
• current ERT users had best survival
• recent past users had highest all cause mortality
  (greater than those who never used ERT)

42
Confounding Factors meta-analyses

• Unable to remove biases from observational
  studies
• Not predictive 35 of the time
• LeLorier et al. NEJM 337(8) 536-542, 1997
• Example-beta carotene
• supporting observational and mechanistic studies
• lack of benefit and potential harm in RCTs

43
Estrogen Replacement Therapy and CHD
No Benefit?
Benefit?
44
The Answer?

• Need for trials to confirm these findings that
  are
• prospective
• randomized
• double-blind
• placebo-controlled

45
HERS Study

• Heart and Estrogen/progestin Replacement Study
• 2,763 women with CHD, mean age 67
• 0.625mg conjugated estrogens 2.5mg
  medroxyprogesterone (PremproTM) daily vs. placebo
• Average follow-up was 4.1 years

JAMA 1998280605-613,650-652
46
Endpoints

• Primary outcome was nonfatal MI or CHD death
• Secondary outcomes included total mortality,
  cancer death, breast cancer, endometrial cancer,
  DVT, PE, fractures, gallbladder disease

Heart and Estrogen/progestin Replacement Study.
JAMA 1998280605-613,650-652
47
Statistical Power

• Able to detect a 24 intention-to-treat effect
• 90 power, 2-tailed alpha of 0.05
• However lower than expected
• event rate (3.3 versus 5)
• treatment duration (4.1 years versus 4.75)
• compliance rate (75 and 81 at 3 years)
• Offset by
• 18 more participants than planned

Heart and Estrogen/progestin Replacement Study.
JAMA 1998280605-613,650-652
48
Demographics

• No differences (pgt0.05) between groups in
• Age, Education
• CHD risk factors (LDL 145 /- 37 mg/dL)
• CHD manifestations
• Medication use
• aspirin (78)
• b-blockers (33)
• lipid lowering medications (45-47)
• calcium channel blockers (55)

Heart and Estrogen/progestin Replacement Study.
JAMA 1998280605-613,650-652
49
Results-primary endpoint

• HRT Placebo
• First Year
• Nonfatal MI 42 (3) 29 (2)
• CHD Death 17 11
• End of study
• Combined 172 176

plt0.05 (95 CI, 1.01-2.29) pgt 0.05 (95 CI,
0.87-1.75)
Heart and Estrogen/progestin Replacement Study.
JAMA 1998280605-613,650-652
50
Results-primary endpoint

• Risk for CHD Death or nonfatal MI with HRT
• year one 52
• year two no difference
• year three 13
• years 4-5 23

p0.009 for trend in log relative hazard
Heart and Estrogen/progestin Replacement Study.
JAMA 1998280605-613,650-652
51
Results-primary endpoint

• As-treated analysis
• 80 compliant by pill count
• no difference between groups
• relative hazard 0.87 (95 CI, 0.67-1.11)
• Subgroup analyses
• no differential effects in 86 subgroups

Heart and Estrogen/progestin Replacement Study.
JAMA 1998280605-613,650-652
52
Results-other endpoints

• Lipids
• 11 reduction in LDL (125mg/dL versus 140mg/dL)
• 10 higher HDL (54mg/dL versus 49mg/dL)
• 8 higher triglycerides (181mg/dL versus
  170mg/dL)
• Total Mortality
• no differences (95 CI, 0.84-1.38)

Heart and Estrogen/progestin Replacement Study.
JAMA 1998280605-613,650-652
53
Results-other endpoints

• Venous thromboembolic events with HRT
• 34 versus 12, RH 2.89 (95 CI, 1.50-5.88)
• Gallbladder disease
• 84 versus 62, RH 1.38 (95 CI, 1.00-1.92)

Heart and Estrogen/progestin Replacement Study.
JAMA 1998280605-613,650-652
54
Explanation?

• Older population
• Established CHD
• Estrogen progesterone
• Effects over time
• random variation?
• early prothrombotic, arrhythmic, ischemic
  effects?
• later change in underlying atherosclerosis?

55
HERS Study-Key Points

• Continuous HRT in women with CHD did not reduce
  cardiovascular outcomes at 4.1 years
• HRT should not be started specifically for the
  secondary prevention of CHD
• Women already receiving HRT with CHD should not
  necessarily change therapy
• Increased risk of thromboembolic events and
  gallbladder disease

56
Coumadin Aspirin Reinfarction Substudy

• Incidence of post-MI cardiac events at one year
  in 1,853 postmenopausal women
• prior/current (n 411)
• new (n 126)
• never ( n 1,316)

Unstable Angina (RR) Death or MI
(RR) Prior/current 1.16 (p 0.25) 0.75 (p
0.14) New 1.96 (p lt 0.0001) 0.39 (p lt 0.02)
Alexander K., ACC 1999 Abstracts.
57
Unanswered questions

• Unopposed ERT
• Micronized or other forms of progesterone
• Cyclical versus continuous
• Lag effect
• Primary prevention
• Womens Health Initiative