Atherosclerosis and Lipoproteins |
From the Departments of Internal Medicine/Cardiology (D.M.H., B.E.P.), Neurology (W.A.R.), Endocrinology (T.Y.T.), and Surgical Sciences-General (Hypertension Center) (K.B.B.), Wake Forest University School of Medicine, Winston-Salem, NC; the University of Arizona/Phoenix VA Medical Center (E.A.B.); and Pfizer, Inc (D.B.M.), Groton, Conn.
| Abstract |
|---|
|
|
|---|
Key Words: droloxifene hormone replacement therapy women estrogen cardiovascular disease
| Introduction |
|---|
|
|
|---|
Selective estrogen receptor modulators (SERMs), like tamoxifen, are compounds with mixed estrogen agonistic/antagonistic effects. Other benzothiophene derivatives, such as raloxifene, and other structurally unrelated compounds, such as soy phytoestrogens and tibolone, have variable degrees of estrogen agonistic and antagonistic effects.5 Tamoxifen lowers cholesterol,6 and in trials in women with breast cancer, it was associated with 15% to 63% fewer cardiovascular events.7 However, tamoxifen also increases the risk for endometrial hyperplasia/carcinoma8 and is therefore not an ideal cardioprotective agent. Raloxifene is a related compound with tamoxifen-like LDL-lowering effects.9 However, the chemical structure of raloxifene differs considerably from that of tamoxifen, and the effects of raloxifene on atherosclerosis are uncertain.10 11
Droloxifene is a structural analogue of tamoxifen with tissue-specific estrogen agonistic/antagonistic effects.12 In ovariectomized rodents, it prevents bone loss and lowers cholesterol without causing uterine hypertrophy.13 14 Preliminary data suggest that it may be useful in advanced breast cancer.15 The cardiovascular effects of droloxifene in women are unknown. The purpose of the present study was to compare the effects of droloxifene with the effects of conventional estrogen replacement on measures of plasma lipids, coagulation and fibrinolytic factors, and brachial artery flow-mediated vasodilation in healthy postmenopausal women.
| Methods |
|---|
|
|
|---|
1 year or self-report of hysterectomy/oophorectomy with estradiol
concentrations <25 pg/mL and follicle-stimulating hormone
concentrations >40 IU/L. Women were ineligible with if they had a
fasting LDL cholesterol >200 mg/dL or a fasting
triglyceride level >400 mg/dL, if they were receiving
medical therapy for diabetes mellitus, or if their body mass index was
>35 or <18 kg/m2. Any woman with recurrent deep
venous thrombosis or who had taken anticoagulant therapy for
thromboembolic disease within 2 years was excluded. No women were
currently taking angiotensin-converting enzyme
inhibitors, anticoagulants (except aspirin), calcium
channel blockers, nitrates, or anticonvulsants. Subjects could not have
taken conjugated estrogen or another hormone replacement regimen within
3 months. We recruited 25 women, and they gave written informed consent to participate. One woman withdrew after the first treatment period because of an exacerbation of previously diagnosed chronic hepatitis. We report results from the 24 women who completed the trial. The study was approved by our institutions Clinical Research Practices Committee.
Study Design
Participants received oral conjugated equine estrogens (0.625
mg/d Premarin, Wyeth-Ayerst) or oral droloxifene (60 mg/d, Pfizer)
during 2 randomly ordered 6-week treatment periods, separated by a
4-week washout (Figure 1
). Investigators
and participants were blinded to treatment assignments. Subjects took 3
tablets and 1 capsule every morning. The tablets contained either
droloxifene (one 40-mg and two 10-mg tablets) or placebo, and the
capsule contained either conjugated equine estrogens or placebo. Each
subject received a 6-week supply of study medications for 1 regimen.
After completing the first treatment period and the washout, each woman
then received a 6-week pill supply for the other regimen. Compliance
was ascertained by pill count at the end of each treatment period.
Overall mean compliance by pill count was 97.8%. When the second
6-week treatment period was finished, all subjects received a 12-day
supply of open-label medroxyprogesterone acetate
(10 mg/d Provera, Wyeth-Ayerst) to ensure the ablation of any residual
endometrial hyperplasia. A phone interview 3 weeks later documented no
adverse events or persistent vaginal bleeding.
|
Plasma Lipid and Hemostatic Measurements
Fasting plasma specimens for lipoprotein determinations were
obtained twice (on separate days) at the beginning and twice at the end
of each treatment period. Results were averaged to improve statistical
precision. Cholesterol and triglyceride
analyses were performed on a Technicon RA-1000
autoanalyzer as described in the Technicon technical manual for
cholesterol (SM4-0139A85) and for triglyceride
(SM4-0189K87, glycerol phosphase-oxidase blank method). The
cholesterol method is based on the enzymatic
cholesterol procedures of Allain et al16 and
Roeschlau et al17 and the peroxidase/4-aminophenazone
system of Trinder.18 The triglyceride method
was described by Fossati and Prencipe.19 Total glyceride
levels in plasma were quantified. HDL cholesterol was
measured by use of the heparin-manganese precipitation
procedure.20 21 LDL cholesterol was calculated
by using the Friedewald formula. Lp(a) measurements were made on a
COBAS FARA II centrifugal autoanalyzer.22
Fibrinolytic and coagulation elements were measured by MDS Clinical Trial Laboratories. Fibrinogen,23 plasminogen,24 tissue plasminogen activator (tPA),25 antithrombin III,26 27 and plasminogen activator inhibitor (PAI-1)28 were measured by using previously described techniques.
Measurement of Brachial Artery Flow-Mediated Dilation
Details of the procedure have been published
previously.29 Once the transducer position was established
over the left brachial artery, baseline images were obtained for 2
minutes. The brachial artery was continuously imaged for the 4 minutes
of blood pressure cuff occlusion and for 2 minutes immediately after
cuff release. Nitroglycerin was not administered to
avoid headaches and hypotension.
Ultrasound images were analyzed by using an automated
analysis system30 that determines changes in
brachial artery diameter for 2 minutes after flow stimulus (Figure 2
). Diameter is defined as the average
distance between the medial-adventitial boundaries for the segment of
interest. Baseline and maximum diameter and the area under the diameter
versus time curve (AUC) are automatically determined and stored in a
database for further analyses. Reproducibility was previously
determined to be R2=0. 80, 0.84, and
0.78 for percent change, absolute change, and AUC,
respectively.29
|
Statistical Analysis
Data are expressed as mean±SD. Repeated measures ANOVA models
found no evidence of an order effect for baseline or change in any
variable. Therefore, simple paired t tests were used to
compare baseline versus follow-up values for each treatment. ANOVA was
used to compare the effect of droloxifene versus estrogen within the
same subjects. ANCOVA models were used to determine whether
treatment-associated changes in brachial responses were independent of
treatment-associated changes in plasma lipids or hemostatic factors.
The primary outcome was prospectively defined as percent change in
brachial diameter after the flow stimulus. On the basis of previous
similar studies,29 sample size was estimated to
detect at least a 30% treatment effect with 80% power at the 2-sided
0.05 level of significance.
| Results |
|---|
|
|
|---|
|
|
Hemostatic Factors
Droloxifene and estrogen significantly lowered fibrinogen and
antithrombin III (P<0.01 for both factors and both
treatments, Table 1
). However, for fibrinogen, the
droloxifene-associated reduction was greater than with the reduction
with estrogen (17.8% versus 7.3%, P=0.004). In contrast,
droloxifene had no effect on fibrinolytic elements, whereas estrogen
significantly increased circulating plasminogen levels
(P<0.001) and lowered circulating levels of PAI-1
(P=0.008) and tPA (P=0.02).
Brachial Flow-Mediated Vasodilator Responses
Droloxifene produced a 36.4% increase in response to flow
stimulus (percent change, Figure 4
) and a
42.9% increase in absolute change in diameter (P<0.05 for
each, Table 2
), whereas estrogen produced
27.3% and 29.4% improvements, respectively (P=0.001 for
each). Two women showed no improvement with either regimen. Identical
patterns were observed when the vasodilator response was quantified by
using AUC. There were no statistical differences between any
improvements realized with estrogen and droloxifene.
|
|
Correlation Between Vasodilator Responses and Other
Outcomes
To determine whether treatment-associated improvements in brachial
vasodilator responses were attributable to changes in plasma lipids or
hemostatic factors, we examined the correlations between changes in
brachial responses and other variables and developed ANCOVA models
of treatment effects after adjusting for changes in plasma lipoproteins
and hemostatic factors. There were no significant correlations between
treatment-associated changes in any variables, and the
statistically significant effects of each regimen on brachial artery
flow-mediated responses remained.
| Discussion |
|---|
|
|
|---|
Lipid Effects of SERMs
Observational studies and clinical trials have demonstrated modest
LDL-lowering effects of tamoxifen in normal postmenopausal
women31 and in women treated for breast
cancer,32 sometimes accompanied by estrogen-like increases
in HDL and triglycerides.32 Clinical studies
of raloxifene have demonstrated 5% to 14% reductions in LDL
cholesterol with little or no effect on HDL
cholesterol and
triglycerides.33
In ovariectomized rats (2.5 to 10 mg · kg-1 · d-1), droloxifene caused 65% to 70% reductions in total cholesterol.13 The present study confirms a hypolipidemic effect in healthy postmenopausal women, and its magnitude (17% reduction in LDL cholesterol) is similar to that with tamoxifen and raloxifene. Like raloxifene, droloxifene also leaves plasma levels of HDL cholesterol or triglycerides unchanged. The modest LDL-lowering effect of droloxifene and other SERMs, and the absence of an HDL-elevating effect, may limit the lipid-mediated benefit of SERMs on cardiovascular risk. However, the lack of effect on triglyceride levels could be a benefit of SERMs versus estrogen, because elevated triglyceride levels are an important independent cardiovascular risk factor, especially in women.34
The present study reveals a significant estrogen agonistic effect of droloxifene on Lp(a). Previous studies of tamoxifen35 and raloxifene33 have also shown an Lp(a)-lowering effect in healthy normal women.36 Lp(a) levels are related to angiographically defined coronary atherosclerosis, risk for myocardial infarction, and cerebrovascular disease.37 Elevated levels of Lp(a) may lead to atherosclerotic disease by interfering with LDL catabolism, promoting vascular smooth muscle cell proliferation, and attenuating plasmin formation.37 Thus, the ability of estrogen and certain SERMs, including droloxifene, to lower Lp(a) could be an important cardiovascular benefit.
Effects of SERMs on Hemostasis
Like estrogen replacement,38 39
tamoxifen40 and raloxifene33 lower fibrinogen
by 10% to 20%, an effect that we confirmed for droloxifene.
Fibrinogen levels are independently associated with coronary
disease risk,41 42 perhaps because of its role in the
coagulation cascade and acute coronary thrombosis. However, the
Bezafibrate Infarction Prevention Study showed no effect of bezafibrate
on cardiovascular mortality despite an 11.8% reduction
in fibrinogen.43 Furthermore, lowering fibrinogen levels
via drugs may not have the same effect as naturally low levels. van
Baal et al44 observed an estrogen-associated 12%
reduction in clottable fibrinogen but no change in fibrinogen antigen,
suggesting that estrogen may act on fibrinogen clotting, not
synthesis.
However, in some, but not all, studies, estrogen replacement,38 45 tamoxifen,46 47 and raloxifene33 also reduced antithrombin III levels, potentially a prothrombotic effect.48 Only limited data are available concerning the effects of estrogen or SERMs on the activation of the coagulation cascade. Caine et al49 reported significant increases, and Walsh et al33 reported nonsignificant trends toward higher levels of prothrombin fragments 1 and 2 and fibrinopeptide A with estrogen replacement therapy. In the present study, droloxifene and estrogen lowered antithrombin III levels by 7% and 10%, respectively, but also lowered fibrinogen 18% and 7%, respectively. On the basis of these limited data, the net effect on risk for arterial thrombosis is unknown.
In contrast, droloxifene had no estrogen agonistic effect on fibrinolytic factors. In the present and other studies,50 estrogen replacement raised plasminogen and lowered PAI-1 levels. This may occur directly via PAI-1 gene transcription51 or indirectly through an angiotensin-converting enzymeinhibiting effect of estrogen,52 which lowers angiotensin II, a known potent stimulus for PAI-1 synthesis.53 Estrogen also lowers tPA levels; however, because most tPA is complexed to PAI-1, this reduction in tPA likely reflects lower PAI-1 levels and not a reduced capacity for plasminogen activation. Thus, by increasing plasminogen and enhancing its conversion to plasmin, estrogen could promote fibrinolysis. If these estrogenic effects on the fibrinolytic system are important modulators of heart disease risk, they are apparently not shared with droloxifene and other tamoxifen analogues.33
Effects of SERMs on Endothelial Function
The favorable effects of acute and chronic estrogen therapy on
endothelium-dependent vasodilator responses are well
documented in monkeys54 and postmenopausal
women.29 55 56 In vitro, estrogen causes
endothelial cell release of NO.57 Clinical
studies have confirmed that the favorable effect of estrogen on
brachial flow-mediated dilation is accompanied by increased NO
production and release58 and can be inhibited by
N
-monomethyl-L-arginine,
an NO synthase inhibitor.59
Far fewer data are available regarding SERMs and
endothelium-dependent vasodilator responses. In in
vitro studies, tamoxifen and raloxifene have vasorelaxant effects
similar to those of 17ß-estradiol in comparable
doses.60 61 These effects are inhibited by
endothelial denudation and by coadministration of
N
-nitro-L-arginine
methyl ester, another competitive NO synthase inhibitor,
and by ICI 182,780, an estrogen receptor-
antagonist. In
contrast, in ovariectomized atherosclerotic monkeys, tamoxifen had no
effects on coronary vasomotor responses to acetylcholine,
whereas conjugated estrogen produced significant
improvements.62
In the present study, 60 mg droloxifene daily improved flow-stimulated brachial dilation in healthy women as much as 0.625 mg conjugated estrogens daily. Numerous studies have documented that flow-stimulated vasodilation is an endothelium-dependent response mediated in part by NO production.63 The roughly 30% improvement in brachial response with droloxifene and estrogen is comparable to improvements with estrogen in studies using a variety of techniques.29 56 The fact that our subjects were healthy, the treatment period was brief, and there was a lack of correlation between vasodilation and plasma lipids suggests a direct effect on the vessel wall, not an effect secondary to lipid-mediated changes in atherosclerosis.
A randomized, blind, crossover study is an efficient means to observe the effects of multiple therapies within the same subjects. Nonetheless, the relatively few subjects hampers statistical inferences concerning subtle differences between the 2 hormone regimens. In addition, on the basis of this short-term study, we cannot say whether the effects of droloxifene will persist, increase, or decrease over time. Finally, we did not comprehensively evaluate the effects of droloxifene on hemostasis or vasomotor tone. For example, we did not evaluate the effects of droloxifene on endothelium-independent vasodilation or metabolism of other endothelial modulators of vascular tone, eg, endothelin or prostaglandins. In addition, we did not determine whether droloxifene or estrogen altered the magnitude of the flow stimulus that was due to distal hyperemia, although several other investigators64 65 have failed to observe such an effect.
In summary, droloxifene has estrogen agonistic properties on LDL and Lp(a) metabolism, certain coagulation factors, and endothelium-dependent vasodilation. This is the first evidence in women of an estrogen agonistic effect on endothelium-directed vasodilation with a SERM. However, droloxifene had no estrogen agonistic activity on HDL or triglyceride metabolism or on the expression of key elements of the fibrinolytic cascade. It remains unknown whether the selective profile of cardiovascular effects from this or other SERMs will translate into fewer cardiovascular events. Moreover, preliminary data (Pfizer Central Research, unpublished data, 2000) suggest that droloxifene, like raloxifene66 and estrogen,4 67 68 69 may also increase the risk for venous thromboembolic events. Thus, the efficacy and safety of droloxifene and other SERMs should be evaluated in clinical trials before recommendations can be made concerning their use for the prevention of heart disease.9
| Acknowledgments |
|---|
| Footnotes |
|---|
Received December 28, 1999; accepted February 21, 2000.
| References |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
H.-L. Noh, K. Okajima, J. D. Molkentin, S. Homma, and I. J. Goldberg Acute lipoprotein lipase deletion in adult mice leads to dyslipidemia and cardiac dysfunction Am J Physiol Endocrinol Metab, October 1, 2006; 291(4): E755 - E760. [Abstract] [Full Text] [PDF] |
||||
![]() |
T. Pelzer, V. Jazbutyte, K. Hu, S. Segerer, M. Nahrendorf, P. Nordbeck, A. W. Bonz, J. Muck, K.-H. Fritzemeier, C. Hegele-Hartung, et al. The estrogen receptor-{alpha} agonist 16{alpha}-LE2 inhibits cardiac hypertrophy and improves hemodynamic function in estrogen-deficient spontaneously hypertensive rats Cardiovasc Res, September 1, 2005; 67(4): 604 - 612. [Abstract] [Full Text] [PDF] |
||||
![]() |
M. L. Bots, J. Westerink, T. J. Rabelink, and E. J.P. de Koning Assessment of flow-mediated vasodilatation (FMD) of the brachial artery: effects of technical aspects of the FMD measurement on the FMD response Eur. Heart J., February 2, 2005; 26(4): 363 - 368. [Abstract] [Full Text] [PDF] |
||||
![]() |
J. Widder, T. Pelzer, C. von Poser-Klein, K. Hu, V. Jazbutyte, K.-H. Fritzemeier, C. Hegele-Hartung, L. Neyses, and J. Bauersachs Improvement of Endothelial Dysfunction by Selective Estrogen Receptor-{alpha} Stimulation in Ovariectomized SHR Hypertension, November 1, 2003; 42(5): 991 - 996. [Abstract] [Full Text] [PDF] |
||||
![]() |
N. Colacurci, D. Manzella, F. Fornaro, M. Carbonella, and G. Paolisso Endothelial Function and Menopause: Effects of Raloxifene Administration J. Clin. Endocrinol. Metab., May 1, 2003; 88(5): 2135 - 2140. [Abstract] [Full Text] [PDF] |
||||
![]() |
T. Simon, P. Boutouyrie, J.M. Simon, B. Laloux, C. Tournigand, A.I. Tropeano, S. Laurent, and P. Jaillon Influence of Tamoxifen on Carotid Intima-Media Thickness in Postmenopausal Women Circulation, December 3, 2002; 106(23): 2925 - 2929. [Abstract] [Full Text] [PDF] |
||||
![]() |
D. M. Herrington, E. Vittinghoff, F. Lin, J. Fong, F. Harris, D. Hunninghake, V. Bittner, H. G. Schrott, R. S. Blumenthal, R. Levy, et al. Statin Therapy, Cardiovascular Events, and Total Mortality in the Heart and Estrogen/Progestin Replacement Study (HERS) Circulation, June 25, 2002; 105(25): 2962 - 2967. [Abstract] [Full Text] [PDF] |
||||
![]() |
N. J. Brown, A. Abbas, D. Byrne, J. A. Schoenhard, and D. E. Vaughan Comparative Effects of Estrogen and Angiotensin-Converting Enzyme Inhibition on Plasminogen Activator Inhibitor-1 in Healthy Postmenopausal Women Circulation, January 22, 2002; 105(3): 304 - 309. [Abstract] [Full Text] [PDF] |
||||
![]() |
B. Andersson, G. Johannsson, G. Holm, B.-A. Bengtsson, A. Sashegyi, I. Pavo, T. Mason, and P. W. Anderson Raloxifene Does Not Affect Insulin Sensitivity or Glycemic Control in Postmenopausal Women with Type 2 Diabetes Mellitus: A Randomized Clinical Trial J. Clin. Endocrinol. Metab., January 1, 2002; 87(1): 122 - 128. [Abstract] [Full Text] [PDF] |
||||
![]() |
D. M. Herrington, M. A. Espeland, J. R. Crouse III, J. Robertson, W. A. Riley, M. A. McBurnie, and G. L. Burke Estrogen Replacement and Brachial Artery Flow-Mediated Vasodilation in Older Women Arterioscler. Thromb. Vasc. Biol., December 1, 2001; 21(12): 1955 - 1961. [Abstract] [Full Text] [PDF] |
||||
![]() |
M. Cushman, J. P. Costantino, R. P. Tracy, K. Song, L. Buckley, J. D. Roberts, and D. N. Krag Tamoxifen and Cardiac Risk Factors in Healthy Women : Suggestion of an Anti-inflammatory Effect Arterioscler. Thromb. Vasc. Biol., February 1, 2001; 21(2): 255 - 261. [Abstract] [Full Text] [PDF] |
||||
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||