SYMPOSIUM

National Institutes of Health (NIH) consensus development conferences bring together biomedical-research scientists, practicing physicians, consumers, and others as appropriate in an effort to reach agreement on the safety and efficacy of particular developments in medical technology. The report below summarizes conclusions reached at a consensus conference recently sponsored by the National Institute on Aging, assisted by the Office for Medical Applications of Research, NIH; it was prepared by the conference's chairman, panel members, three main speakers, and a consultant.

Estrogen Use and Postmenopausal Women: A National Institutes of Health Consensus Development Conference

A CONSENSUS DEVELOPMENT CONFERENCE On E s t r o g e n

Use and Postmenopausal Women was held at the National Institutes of Health, 13-14 September 1979, to explore the risks and benefits of estrogen use. Among the subjects addressed were the benefits of estrogen use, including treatment for menopausal symptoms and possible prevention of osteoporosis; hazards of estrogen use; relative risks and benefits of various types of estrogen therapy; economic issues; and indications for and contraindications to estrogen use. The consensus conclusions reached and presented here are based on the three position papers prepared for the conference, the response of the panel, and the general discussion by the audience, followed by an analysis by the panel and other invited conference participants. Efficacy of Estrogens

The group first reviewed the evidence for the efficacy of estrogens in treating specific conditions associated with menopause. It was accepted that estrogens are more effective than placebo in decreasing the frequency and severity, together or individually, of vasomotor symptoms (hot flashes and sweating). The questions that remain to be answered include whether vasomotor symptoms represent a homogeneous entity with a single cause and why some patients need much larger than average doses to control symptoms. There was general agreement that the decision of whether to initiate therapy should depend on the severity of symptoms and the patient's perceived need for relief and that the lowest effective dose should be used. The occurrence of hot flashes naturally declines over time, and therapy should not be unnecessarily prolonged. Estrogens are effective in overcoming the atrophy of the vaginal epithelium (wall) and the associated symp-

toms, which may include dryness, burning, itching, and pain during intercourse. The possible relation of urinary tract symptoms to estrogen lack must be more thoroughly investigated. Attempts to avoid systemic effects by treating vaginal symptoms with local application of estrogen-containing creams have been common, but there is evidence that the estrogens in these creams may be absorbed rapidly into the bloodstream. The biological consequences of this absorption are undetermined and should be studied. Present evidence does not justify the use of estrogens to treat primary psychological problems. Surveys have shown no established specific or temporal association of sleep patterns, mental performance, mood, or psychological state with menopause or estrogen deprivation. On the other hand, in preliminary intervention studies comparing estrogens to placebo, effects on sleep latency and rapid-eye-movement sleep have been noted. Some improvement in mental well-being in women receiving estrogens may be due to alleviation of physical symptoms. The group acknowledged the validity of three randomized trials indicating that exogenous estrogens can retard bone loss if given around the time of menopause. Except for dietary calcium, which appears to decrease bone loss to a lesser extent, other substances have not been shown to have such an effect. That this retardation of bone loss will prevent the ultimate development of osteoporosis and attendant fractures is a premise still lacking proof. Case-control studies not yet published but discussed at the meeting have associated estrogen use with a decreased risk of osteoporosis-related fractures. More data are definitely needed, however, before the efficacy of estrogens in preventing fractures can be established. A n inconsistency was noted: In the randomized trials, accelerated bone loss after estrogen use had been discontinued led to loss of any favorable effect on bone mass, whereas in the case-control studies, use of estrogens at any time in the past conferred some protection against fractures. Identification of patients at increased risk of osteoporosis is desirable because of the strong possibility of successful prophylaxis. One high-risk group in which to investigate possible benefits of estrogens consists of patients who already have developed osteoporosis and sustained fractures. Estrogen administration is a promising approach to prevention of the widespread problem of hip fracture. There is no convincing evidence that estrogens in customary doses increase the risk of thromboembolic phenomena, stroke, or heart disease in women who have undergone natural menopause. Although it was once hoped that estrogens would protect against heart disease in aging women, this effect has not yet been seen. One promising approach would be to devise a more physiologic mechanism for estrogen replacement. Because oral therapy results in the delivery of supraphysiologic concentrations of estrogens to the liver, it can exert an exaggerated effect on lipoprotein metabolism, blood coagulation, and other important processes. Symposium

Downloaded from https://annals.org by Karolinska Institute user on 01/18/2019

921

Adverse Effects

The evidence for adverse effects associated with postmenopausal estrogen use was reviewed. In the absence of exogenous estrogens, the incidence of endometrial cancer is about one case per 1000 postmenopausal women per year. This rate increases severalfold beginning after about 2 to 4 years of use of 0.625 or 1.25 mg of conjugated estrogens per day. Evidence was presented that the risk of endometrial cancer increases with the duration of use and declines after discontinuation. Estrogen use is most strongly associated with lesions of the lowest grade and earliest stage. Of interest is the relation in time of the number of estrogen prescriptions and the incidence of carcinoma of the endometrium: Both rose steadily until 1976 and then declined in parallel. Although the incidence of carcinoma of the endometrium rose, mortality from the disease did not increase. A considerable part of this discrepancy may be attributable to early detection and the high cure rate. Cystic hyperplasia of the endometrium, considered to be a premalignant condition, has been associated with unopposed estrogen, whether endogenous (as in anovulatory states) or exogenous. The cost-effectiveness of sampling the endometrium to screen for endometrial hyperplasia and cancer in completely asymptomatic women currently or potentially receiving estrogens is uncertain. Suction curettage is effective in evaluating the endometrium, and certainly the cause of any bleeding must be ascertained. A report was presented at the meeting that indicated uterine bleeding may sometimes be absent early in the course of endometrial cancer. Hence, prudence would suggest that even in the absence of bleeding, the endometrium should be sampled yearly before and during estrogen therapy. The use of progestins for several days of each estrogen treatment cycle has been shown to decrease the occurrence of endometrial hyperplasia and may also reduce the associated risk of developing cancer of the endometrium. Before the use of combined therapy becomes established, risks of the various progestins must be adequately evaluated. The association of estrogens and breast cancer in experimental animals is well known. Careful review of several well-conducted case-control studies has not found such a relation in humans. In two follow-up studies comparing estrogen users with the general population, differing associations were encountered. One study found an excess of cases in years 5 to 9 of estrogen use, but the other only after 15 years of its use. Incidence rates of breast cancer have not changed in parallel with those of estrogen use, as have those of endometrial carcinoma. Because of the high incidence and relatively poor prognosis of breast cancer, any possible association with estrogen use remains a concern. Some experimental data indicate that estrogens can induce the production of lithogenic bile; in one study a 2.5fold relative risk for the development of surgically confirmed gallbladder disease was observed.

922

Some concern was expressed about women who have undergone menopause many years in advance of the normally expected age. Although most of the participants felt intuitively that approximation of the normal physiologic state through hormone replacement therapy would be best, there are no carefully controlled studies comparing the risks and benefits in these circumstances. Support was voiced for conserving the ovaries of young women when possible. One area of general agreement was that patients should be given as much information as possible about the evidence for the effectiveness of estrogens in treating specific menopausal conditions and the risks of their use. Patients must be kept continually informed of new findings as they arise. With present knowledge, no general recommendation applicable to all postmenopausal women can be made. Conclusion

Clearly much additional information is needed. Specifically, we need systematic knowledge of the natural course of the menopause in the absence of hormonal therapy, of alternatives to estrogen use, of the optimal way to provide estrogens if they are to be used, and of all aspects of their beneficial and adverse effects. Special attention should be directed toward studies that can determine the proper use of estrogens in young women who have undergone oophorectomy 10 or more years before the natural time of menopause. N o general formulation on therapy can be given. Rather, each patient must base her decision on the relative values she assigns to relief of symptoms, to expectations for optimizing health and well-being, and to various risks sustained in the process. Socially and culturally based attitudes toward menopause may influence these values and should be further defined. ( K E N N E T H J. R Y A N , M.D., Boston Hospital for Women, Boston, Massachusetts;

Boston,

Massachusetts;

G E R A L D A. G L O W A C K I , M . D . ,

Franklin Square Hospital, Baltimore, Maryland; SAUL B. GUSBERG, M.D., Mt. Sinai School of Medicine, New York, New

York; E L I Z A B E T H D . J O N E S , M.S.W.,

en 's Hormone Information Service, Shelburne, H O W A R D L. J U D D , M.D., University

Wom-

Vermont;

of California,

Los

Angeles, California; W E L D O N G. KOLB, M.D., Memorial Hospital, Galveston County, Texas; STANLEY G. K O RENMAN, M.D., Veterans Administration Hospital, Sepulveda, California; A N N E M. SEIDEN, M.D., Cook County Hospital, Chicago, Illinois; N O E L S. WEISS, M.D., School of Public Health and Community Medicine, Seattle, Washington;

BARBARA S. H U L K A , M.D.,

University

of North Carolina, Chapel Hill, North Carolina; ISAAC SCHIFF, M.D., Boston Hospital for Women, Boston, Massachusetts; M I L T O N C. W E I N S T E I N , PH.D., Harvard

Uni-

versity, Cambridge, Massachusetts; and B. L A W R E N C E RIGGS, M.D., Mayo Medical School, Rochester, Minnesota)

December 1979 • Annals of Internal Medicine • Volume 91 • Number 6

Downloaded from https://annals.org by Karolinska Institute user on 01/18/2019

E L I Z A B E T H B A R R E T T - C O N N O R , M . D . , Uni-

versity of California at San Diego, La Jolla, California; D A N I E L D. F E D E R M A N , M.D., Harvard Medical School,

Estrogen use and postmenopausal women: a National Institutes of Health Consensus Development Conference.

The concensus conclusions reached at a concensus development conference on Estrogen Use and Postmenopausal Women in September 1979 are based on 3 posi...
464KB Sizes 0 Downloads 0 Views