| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Letters to the Editor |
Wake Forest University School of Medicine Winston-Salem, North Carolina 27157
We are pleased that our colleagues at Lilly Research Laboratories have provided us with an opportunity to comment further on our report (1) that raloxifene failed to inhibit the progression of coronary artery atherosclerosis in surgically postmenopausal cynomolgus monkeys. In their letter to the editor (above), Bryant et al. state that they disagree with the "strong conclusions regarding the clinical relevance of these data...". However, we stand by our statement that "the data reported herein suggest that it [raloxifene treatment of estrogen deficiency and bone loss] would be at the expense of any protection against coronary artery atherosclerosis. Bryant and colleagues raised three areas of concern, and we shall deal with these separately.
Blood levels of 17ß-estradiol. We treated the monkeys with CEE based on a caloric adjustment of dose. The most commonly used dose of CEE for women is 0.625 mg per day, and we considered 1800 calories as an average daily caloric intake for women in the United States. The dose can then be calculated to be 0.000347 mg per calorie. Since the monkeys eat, on average, 120 calories per kilogram of body weight per day; they were given 0.042 mg CEE per kilogram of body weight per day. Assuming body weight is maintained, this type of caloric adjustment can account for differences in metabolic rate and physical activity both between species and among individuals. We found variable concentrations of plasma estradiol in the monkeys in this study, and we find that the plasma concentrations of estradiol following CEE treatment of postmenopausal women vary considerably based on a conflicting literature. We do point out in the Discussion of our paper, however, that the plasma concentrations of estradiol in the monkeys were closer to the plasma concentrations of women given a CEE dose of 1.25 mg per day. We also point out that on average, the estradiol concentrations in the CEE treated group were well below estradiol concentrations during the follicular phase of nonstressed premenopausal cynomolgus monkeys (about 240 pg/mL). Bryant et al. state that the premenopausal monkeys in this study had on average estradiol concentrations of 68 pg/mL. The sampling schedule for the premenopausal monkeys was done without knowledge of whether the monkeys were having normal menstrual cycles and were done at random with regard to phase of the cycle, thus these estradiol data are not readily interpretable relative to the CEE treated group. It is of interest, however, that the extent of coronary artery atherosclerosis in the premenopausal group was very similar to that of the CEE treated group (CEE group CAA = 0.035 mm2, premenopausal group CAA = 0.026 mm2). The premenopausal group had 79% less CAA than the untreated postmenopausal group compared to 72% less CAA in the CEE treated group. In comparing the coronary artery effects of CEE and raloxifene, arguments about whether the plasma estradiol concentrations of the CEE treated monkeys were more like women taking 1.25 mg vs. 0.625 mg of CEE per day, is at best a distraction to the primary finding that raloxifene had no clinically or statistically significant effect. If one considers the higher dose of raloxifene a fairer comparison, then it would appear from the point estimates that there was a trend toward worsened CAA as the raloxifene dose increased.
Bryant and colleagues place too much emphasis on the statistical analysis of the coronary artery plaque size in the CEE treated group stratified by tertile of estradiol concentrations. They state that the CEE treated monkeys with lower plasma estradiol concentrations had coronary artery plaque sizes that were statistically indistinguishable from the placebo treated postmenopausal control group. As we pointed out in both the Results and Discussion sections, the average plaque size of the tertile with the lowest estradiol concentration were 39% smaller than the placebo treated group. The second tertile had plaque sizes 52% smaller, and the highest estradiol tertile had plaque sizes 95% smaller than the placebo treated group. Not surprisingly, the plaque size estimates of the two lowest estradiol tertile groups are not statistically different than the untreated control group, likely because there were only seven or eight animals in each of these tertiles. Our intent in including these data in the publication was to provide evidence of a dose-response relationship between plasma estradiol concentrations and coronary artery atherosclerosis among individuals treated with a standard dose of CEE, suggesting that individual differences in metabolism of CEE resulting in different plasma concentrations of estradiol might impact atherogenesis. However, the results of this post hoc secondary analysis have no impact on the lack of effect of raloxifene on coronary artery atherosclerosis.
Sensitivity of the model. While our Lilly colleagues may
have viewed assessment of coronary artery atherosclerosis development
as a secondary endpoint, it was the reason that the leadership of the
Comparative Medicine Clinical Research Center was willing to commit its
resources to a long-term study. The general approach to our evaluation
of raloxifene has been comparable to our studies of estrogen effects in
the cynomolgus model over the past 20 years. Some of those observations
relating to estrogen’s effects on coronary artery atherosclerosis are
summarized in Table 1
.
|
, we believe that we have made a case that comparable studies with
similar group sizes and similar lengths of time have been able to
detect estrogen effects on coronary artery atherogenesis that are of
clinical/pathologic significance. In this study we found an effect of
CEE on coronary artery atherosclerosis that was similar in magnitude to
a previous study (7). A determinant of statistical power is the
"effect size." If one considers the point estimates in this study
to be unbiased estimates of the effect size on coronary artery
atherosclerosis, then in this study we saw about a 7% smaller plaque
size in the lower dose (clinical dose) of raloxifene compared to the
untreated group and about a 16% larger plaque size with the higher
raloxifene dose compared with the untreated group. While it is possible
to determine the sample size required to be able to detect these
differences, we believe that a 7% reduction in plaque size is of
limited clinical importance, especially when one considers that the
point estimate for plaque size for the lowest tertile of estrogen
concentrations was 39% smaller. Since the submission of our manuscript in July 1997, we have gained some insight into the potential biological mechanisms underlying these outcomes. A comparison between the results of studies of CEE and the selective estrogen receptor modifiers (SERMs), tamoxifen and raloxifene, on atherosclerosis development in ovariectomized monkeys have provided some interesting results. CEE treatment resulted in a 70–80% inhibition of coronary artery atherosclerosis (based on two separate trials) (1, 7); tamoxifen was intermediate between no treatment and the CEE effect with an inhibition of about 50% (8), while raloxifene had no beneficial effect on the development of coronary artery atherosclerosis (1).
These data need to be considered in light of the potential mechanisms
underlying estrogens anti-atherosclerotic properties as well as new
information regarding estrogen receptors. Several lines of evidence
from both human trials and animal studies suggest that the
anti-atherogenic effects of estrogen may depend in large part on direct
effects of estrogens on the artery (9, 10). Arteries are known to
contain estrogen receptors, and recent work has shown that messenger
RNA for both the classical estrogen receptor (ER
) and the recently
described ERß are expressed in coronary arteries of female cynomolgus
monkeys (11). ERß has properties that distinguish it from ER, in
that transcription from an AP-1 site is regulated by the SERMs
tamoxifen and raloxifene in opposite directions for ER and ERß
(12). Thus the intermediate effects of tamoxifen and the lack of an
effect of raloxifene on atherosclerosis progression may relate in part
to the presence of both receptors in macaque coronary arteries. The
issue is further complicated as it appears that the liver primarily
expresses the classical ER (ER) in both rats (12) and monkeys (T.C.
Register, personal communication), suggesting that alterations in
plasma cholesterol may not tell the full story regarding potential
outcomes of a particular drug regimen.
Comparability of the results relative to historical experience with the model. Questions were raised about studies conducted by our colleague Dr. J. Koudy Williams on coronary artery reactivity in this study. No vascular reactivity data were included in the published manuscript because it was agreed upon by the scientists at Lilly Research Laboratories and Dr. Williams that the results were uninterpretable. The difficulties in those studies have not been clearly identified, but in no way would indicate that the monkeys’ responses to these treatments were unusual given that the plasma lipid concentrations, effects on bone density, and effects on the endometrium mimicked those seen in studies with women.
Comments on generalizations. Given that our monkey model shares greater than 90% homology in DNA with human beings, we feel comfortable with our generalizations. Regarding the "tenuous relationship between atheroma size and cardiovascular events," Bryant et al. confuse the difference in primary vs. secondary effects on coronary artery atherosclerosis. Our monkey study focused on primary prevention of coronary artery atherosclerosis. It is highly likely that estrogen treatment can improve plaque stability of established atherosclerotic plaques; however, that may be of somewhat less relevance in a treatment recommended for women during the early stages of menopause, before the development of significant coronary artery atherosclerosis.
Footnotes
Received April 30, 1998. Address correspondence to: Thomas B. Clarkson, Department of Comparative Medicine, Wake Forest University School of Medicine, Medical Center Boulevard, Winston-Salem, North Carolina 27157.
References
and ERß at
AP1 sites. Science. 277:1508–1510. and ß. Endocrinology. 138:863–870.
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
