Endometrial findings in asymptomatic postmenopausal women on exogenous estrogens—A preliminary report

GYNECOLOGIC

ONCOLOGY

11,

56-63 (1981)

Endometrial Findings in Asymptomatic Postmenopausal Women on Exogenous Estrogens-A Preliminary Report’ DONALD A. SIMSEN, M.D., FACOG; STEVEN R. SHIRTS, M.D.; FRED M. HOWARD, M.D.; JERRY SIMS, M.D., JAY M. HILL, M.D., FACOG Department

of Obstetrics

and Gynecology, Fitzsimons Aurora, Colorado 80045

Army Medical

Center,

Received August 7, 1980 Multiple retrospective studies have been done implicating estrogen replacement therapy as a factor in the increasing rise of endometrial cancer. This prospective study consisted of 94 asymptomatic postmenopausal females on exogenous estrogen. Duration of therapy ranged from I through 5 and greater than 5 years. All these patients were subjected to endometrial biopsies. We encountered a total of 71 negative biopsies (75.5%) and 23 positive biopsies (24.4%) ranging from cystic hyperplasia, through frank adenocarcinoma. All positive biopsies underwent formal diagnostic curettage. Histologic findings on the dilation and curettage specimens correlated with the original biopsy findings. All histologic slides were reviewed by an independent pathologist without knowledge of history or previous grading. Risk factors in the Positive biopsy group were present in only 17%. Therefore, it is not possible to predict the patient on estrogen who is at a higher risk for developing future hyperplasia, and in order to avoid missing early precursor lesions in endometrial carcinoma all patients must be biopsied. Significant cancer precursor lesions did not appear until after the patients had been on at least 3 years duration of unopposed estrogen replacement therapy. We conclude that postmenopausal patients on estrogen therapy at high risk can not be identified by risk factors atone, and as a minimum their endometrial status should be evaluated at least every 3 years.

INTRODUCTION

This is a longitudinal study of the effects of estrogen on endometrial histology. Biopsies taken from postmenopausal females who are asymptomatic and give no history of postmenopausal bleeding have shown that almost 25% of those receiving estrogen replacement therapy have abnormal endometrium as shown by biopsy. The fact that endometrial hyperplasia and adenocarcinoma can exist concomitantly in the human endometrium was first described by Allen in 1900[ 11. Carcinogenicity of estrogen in laboratory animals was first demonstrated by Cook and Dodds in 1933 [2]. Gusberg [3] in 1947 demonstrated that adenocarcinoma of the endometrium was preceded in a significant number of cases by adenomatous or atypical adenomatous hyperplasia. These and other authors have implicated estrogens either exogenous or endogenous as playing a significant role in this process in the genetically predisposed patient. The risk factors include obesity, * The opinions asserted in this article are those of the authors and they are not to be interpreted as official views of the Department of Defense or the Department of the Army. 0090~8258/81/010056-08$01.00/O Copyright All IiEhts

0 1981 by Academic Press, inc. of reproduction in any form reserved.

56

POSTMENOPAUSAL

WOMEN

ON EXOGENOUS

ESTROGENS

57

infertility, diabetes, hypertension, delayed menopause, estrogen-secreting tumors of the ovary, and polycystic ovarian syndrome, all conditions accompanied by anovulation or continued unopposed estrogenic stimulation of the endometrium. Until two decades ago, carcinoma of the cervix was encountered in a 2:l ratio over endometrial cancer. Today the trend is toward an almost complete reversal. This can not be explained totally on the decrease in number of invasive cervical cancers because of earlier diagnosis as a result of increased cytologic screening techniques. In 1976, Weiss ef al. [4] reported an increasing incidence of endometrial cancer in the United States; from 1960 to 1974 there was a 6% increase in New York state with 82% of the increase from 1969to 1974. Norway and England [5] reported an increasing incidence also but at a lower rate. One explanation as proposed by Gusberg [S] attributed this to the marked increase of estrogen users in the early 1960samong postmenopausal females. The 15year period from the delay until the present peak can be correlated with the interval frequently used in laboratory animals exposed to carcinogens which usually takes one-fifth of their life span to appear. As a result of these observations, a multitude of retrospective studies were published. Initially Ziel [63, Smith [7], and Mack [8] in 1975 and 1976 reviewed in a retrospective fashion their cases of adenocarcinoma with a history of estrogen replacement therapy in postmenopausal females. These initial reports were followed by numerous retrospective studies, concluding that estrogen replacement therapy in postmenopausal females led to an increased risk of a low of 2.8- to a high of 1I .8- fold [9] increase for estrogen users. The marked disparity in the true risk of taking estrogens was explained on the basis of different population groups, faulty control groups, inaccurate statistical formulas used, failure of using the same histologic criteria, and utilization of various pathologists. These pathologists themselves disagreed on criteria for differentiating between severe atypical hyperplasia and invasive adenocarcinoma. All of these reports led to a great deal of confusion among practicing gynecologists and a severe degree of confusion in the consumer-the postmenopausal American female. The confusion was so significant that it led Greenblatt [lo] in November 1977 to state that “estrogens provide a first-line defense against osteoporosis and denying postmenopausal women hormone replacement therapy will result in more collapsed vertebrae and fractured hips. The increased risk of death in the latter is 35%, far outweighing the risk of a few nonlethal endometrial cancers. The physician concerned by the many articles and editorials alleging an estrogen-cancer relationship, will now shun hormone therapy for the menopausal women in need of it. He thus avoids the possible consequences of estrogen therapy and detaches himself from any blame should a gynecic tumor arise. Brave New World. Surely, to do no harm and benign neglect are not mutually exclusive.” From all of the above, we arrived at the following conclusions: (1) In a certain group of genetically predisposed postmenopausal females, long-term administration of unopposed estrogens increases the risk of the development of future endometrial carcinoma (dosage and length of administration are not yet precisely established but definitely significant). (2) Do the risks outweigh the benefits? (3) Frank carcinoma (like cervical carcinoma) is preceded by precursor lesions in 20 to as high as 37% [ 1I] depending upon the author. Since the previous reports

58

SIMSEN

ET AL.

linking estrogen with endometrial cancers were based on retrospective studies we decided upon a prospective study by sampling asymptomatic postmenopausal females on exogenous therapy in an effort to see if we could determine the incidence of precursor lesions, to see if we could identify the high-risk patient who requires more frequent monitoring, and finally what the correct interval for monitoring (endometrial sampling) should be. Such a prospective study was done by Buckman et al. [12]. Unfortunately, 59.6% of those patients with benign endometrial biopsies and 50.7% of those patients with abnormal endometrial biopsies were also taking progestogens in conjunction with exogenous estrogen therapy. All of our patients reported were only on estrogen therapy. None were being withdrawn with progestogens. MATERIALS AND METHODS Between June 15, 1979 and March 10, 1980, 94 postmenopausal women on estrogens presenting themselves for annual pelvic and breast examination and refill prescriptions for continued use of their estrogens were entered into the study. Age and predisposing risk factors (weight, blood pressure, diabetes, parity, and family history of carcinoma) were recorded. We also recorded length and dosage of estrogenic therapy. Any history of bleeding removed the patient from this study group. In addition to the questionnaire, the patients were subjected to office endometrial biopsy utilizing the Novak curette with an attached lo-cc syringe for suction. A four-quadrant biopsy of the uterine walls was performed and the tissue samples obtained were placed in 10% formalin and submitted for histologic examination. In 7 out of 94 cases (7.4%) this was unsuccessful because of cervical stenosis, and these patients underwent formal dilatation and curettage. After histologic evaluation, we divided these patients into two groups, negative and positive, by histologic report. Seventy-one reports (75.5%) were from patients whose biopsies were reported as secretory, proliferative, or atrophic endometrium. We also had 23 patients (24.5%) with abnormal biopsies reported as simple, cystic, adenomatous, atypical adenomatous hyperplasia, or frank adenocarcinema. These 23 patients with hyperplasia or adenocarcinoma had a dilatation and curettage under anesthesia in the operating room. The endometrial specimens were reviewed with the biopsy specimens and we found either no lesions remaining or a lesion similar to the office specimen. In the positive biopsy group, estrogens were stopped and curettage was performed between 10 and 28 days following office biopsy. In no patient did the endometrial specimen show a lesion more progressive than the office biopsy specimen. The 7 patients on which an office biopsy could not be performed because of stenosis all fell into the benign group. During the same period we also encountered 29 postmenopausal patients on estrogen with postmenopausal bleeding. We treated these patients the same way as the asymptomatic group. In this symptomatic group of 29 patients, there were 16 patients with benign biopsies (55.2%) and 13 patients with positive results (48.8%) (Tables 1 and 2). The latter also underwent dilatation and curettage. As in the asymptomatic group, those who underwent curettage showed either no remaining endometrium or a lesion similar to the office biopsy. There were no

75.5 24.4

14.9 6.4 2.1 1.1

100

71

23

14

6

2 1

94

Negative

Positive Simple or cystic hyperplasia Adenomatous hyperplasia Atypical adenomatous hyperplasia Adenocarcinoma

Percentage

No.

Diagnosis

Overall total

TABLE I

1 0 30

-

0

3

4

26

No.

56.5 61.0

56.0

60.3

58.9

53.7

Mean age

50 -

-

21

17

37

%

64

1 1

6

11

19

45

No.

50 100

100

79

83

63

%

24

0 0

0

4

4

20

No.

25.5

-

-

28

17

28

%

One or more precipitating factors

PATIENTS ON EXOGENOUS ESTROGENS

Duration of therapy Wars)

CHARACTERISTICS AND DIAGNOSES IN 94 ASYM~OMATIC

: UJ m z 8 fi z,

57 0 8

z

2 0 $ E

z c

m3 3

B Y

0 0

6

0 0

29

20.7

7

Overall total

24.1

13

Positive Simple or cystic hyperplasia Adenomatous hyperplasia Atypical adenomatous hyperplasia Adenocarcinoma

100

55.2 44.8

16

Negative

Percentage

No.

Diagnosis

-

57.3

55.0

56.1

48.3

Mean age

10

0

0 -

57

30

30

%

4

4

6

No.

Wears)

Duration of therapy

19

6

3

9

10

No.

-

100

43

70

63

%

12

0 0

3

3

6

6

No.

One or more precipitating factors

TABLE 2 CHARACTERISTICS AND DIAGNOSESIN 29 SYMPTOMATICPATIENTSON EXOGENOUSESTROGENS

41.4

50

43

46

37.5

%

POSTMENOPAUSAL

WOMEN

ON EXOGENOUS

ESTROGENS

61

curettage specimens more progressive than the office biopsy. Before subjecting the data to analysis, all positive biopsies and curettage specimens were reviewed by an independent pathologist not familiar with the cases and provided with no history. In three patients there were differences of opinion histologically and these were reviewed by Dr. S. G. Silverberg, Pathologist, University of Colorado Medical Center. RESULTS AS recorded in Table 1, 94 postmenopausal patients on replacement therapy were evaluated by endometrial biopsy or curettage for evaluation of their endometrial status. Of the 94 asymptomatic patients, 71 (75.5%) had negative biopsies. Their mean age was 53.7 years; their duration of therapy was less than 3 years in 26 (37%) and 3 years or greater in 45 (63%). Twenty patients (28%) had one or more risk factors. There were 23 (24.4%) patients with abnormal biopsies confirmed by endometrial curettage. Their mean age was 58.9 or 5.2 years oJ&r than the benign group. The duration of therapy was less than 3 years in 4 (17%) and 3 years or greater in 19 (83%) compared to 63% or over 3 years in the benign group. In the group of 23 patients evaluated for one or more risk factors, only 4 were at risk. Of the histologic findings, there were 14 (14%) with simple or cystic hyperplasia, 6 (6.4%) with adenomatous hyperplasia, 2 (2.1%) with atypical adenomatous hyperplasia, and 1 (1.1%) with well-differentiated adenocarcinoma. If one accepts that adenomatous and atypical adenomatous hyperplasia are significant precursor lesions, plus the one case of adenocarcinoma we discovered, the incidence of significant lesions in asymptomatic women was approximately 10%. Of importance was that none of these 9 patients had any of the standard risk factors and only one had been on conjugated estrogens less than 3 years. These patients therefore cannot be selectively identified for endometrial sampling based on risk factors alone. Of interest was that one of our patients with adenocarcinoma was positive on an office biopsy, but uterine curettage revealed only benign endometrium. However, after hysterectomy had been accomplished there was superficial invasion to the inner one-third of the myometrium. This was a welldifferentiated lesion. In our 29 symptomatic patients, 16 (55.2%) had benign biopsies and 13 (44.8%) were abnormal. There was a 7.8-year difference in median ages-48.3 benign and 56.1 abnormal. Duration of therapy was similar in both groups as were risk factors. Of significance, however, is that of 13 positive biopsies there were 6 with adenomatous hyperplasia (20.7%), all of which were on conjugated estrogens 3 years or longer. We tabulated dosage of the estrogen utilized as well as the regimen of usage and could find no statistical differences in either group. DISCUSSION We believe that exogenous estrogens do play a role in the development of endometrial cancer. Despite the opinion of some authors who believe the benefits of replacement estrogen therapy far outweigh the risk; over 3,000 deaths per year occur in the United States from endometrial carcinoma. The American Cancer

62

SIMSEN

ET AL.

Society in published data for 1975 stated new invasive cervical cancers reported were approximately 19,000 and new endometrial carcinomas were 28,000. They also stated that regarding the cervix there were 21,000 preinvasive lesions which were almost 100% curable. These lesions were detected by increased usage of cytologic screening and increased public awareness that such a test was available and exceedingly accurate as well as the increased usage of colposcopy. The conclusion of most of the articles on endometrial carcinoma and exogenous estrogens state that estrogens should be administered only after careful analysis on an individual patient basis, that the benefits should outweigh the risks, and that careful or annual monitoring of the endometrium be a necessity. The retrospective reports agree fairly well that a 3-year period is usually present before development of endometrial carcinoma and that this increases in a linear fashion to 10 years where it levels off and becomes stable. Several papers state that precursor lesions can be reversed by either stopping the drug altogether or by withdrawal with a progestational agent. Our study shows that 24.4% of women on unopposed estrogens have abnormal endometrial histology and are asymptomatic. About 10% have significant precursor lesions of carcinoma without bleeding. Of 94 asymptomatic patients only 17% have one or more of the commonly described risk factors. More important, none of the 9 patients with significant precursor lesions had any of the risk factors. As stated by Mack [81, “The non-obese parous women usually considered to be a low risk are those seemingly most endangered by estrogens.” We believe that patients on estrogen therapy deserve the same amount of care and scrutiny of the endometrium now practiced with the cervix. Since no significant lesions were discovered before 3 years of therapy, a safe practice to follow would be to do a pretreatment endometrial biopsy, and, if normal, follow with a repeat biopsy in 3 years. As to further recommendations following this 3-year biopsy, we are currently continuing to follow the same group of patients. Our population is a stable, military retired community eligible for medical benefits. We plan to continue follow-up at annual intervals and hopefully we will be able to make a further recommendation with respect to timing of future biopsies while on estrogen therapy. Annual endomtrial biopsy does not appear to be justified. After a 3-year period, it may be indicated based on future data. If we as gynecologists, through increased use of Papanicolaou tests and colposcopy, can detect precursor lesions before invasion in 50% of more patients with cervical disease by exerting the same effort and attention to the endometrium as we do to the cervix, we certainly should be able to improve the current 63% 5-year survival rate. We cannot ignore the asymptomatic patient on estrogens if we are to achieve this goal. We believe outpatient office biopsy is adequate for staging endometrial carcinoma if uterine sounding for depth of uterine cavity plus a separate endocervical scraping to rule out cervical involvement are accomplished. ACKNOWLEDGMENTS

We wish to at this time express our thanks to Kenneth Guenther, M.D., FACOG, who by his diligence and efforts provided us with the majority of our biopsy cases submitted and without whose help this paper could not have been

POSTMENOPAUSAL WOMEN ON EXOGENOUS ESTROGENS

63

written. We also wish to express our sincere thanks to Dr. E. Stewart Taylor for his helpful suggestions and interest in continuing this study, and his time and effort in helping us put the data together in a more meaningful manner allowing us to arrive at our conclusions. REFERENCES 1. Cullen, T. S. The pathology of diseases of the endometrium, Appleton, New York, pp. 353-408 (J9of8.

2. Cook, J. W., and Dodds, E. C. Sex hormones and cancer producing compounds, Nature (London) 131, 205 (1933). 3. Gusberg, S. B. Precursors ofcorpus carcinoma estrogens and adenomatous hyperplasia,Amer. J. Obstet. Gynecol. 54, 905 (1947). 4. Weiss, S., Szekely, D. R., and Austin, D. F. Increasing incidence of endometrial cancer in the United States, N. Engl. J. Med. 294, 1259-1262 (1976). 5. Gusberg, S. B. The individual at high risk for endometrial carcinoma, Amer. J. Obstet. Gynecol. 126, 535-542 (1976). 6. Ziel, H. K., and Finkle, W. D. lncresed risk of endometrial carcinoma among users of conjugated estrogens, N. Engl. J. Med. 293, 1167-l 170 (1975). 7. Smith, D. C., Prentice, R., Thompson, D. J., et al. Association of exogenous estrogen and endometrial carcinoma, N. Engl. .I. Med. 293, 1164(1975). 8. Mack, T. M., Pike, M. C., Henderson, B. E., et al., Estrogens and endometrial cancer in a retirement community, N. Engl. J. Med. 294, 1262-1267 (1976). 9. Jelovsek, F. R., Hammond, C. B., Woodard, B. H., et al., Risk of exogenous therapy and endometrial cancer, Amer. .I. Obstet. Gynecol. 137, 85-91 (1980). 10. Greenblatt, R. B. Estrogens and endomettial cancer-gross exaggeration or fact? Geriatrics 32, 60-72 (1977). 11. Sherman, A. J., and Brown, S. The precursors of endometrial carcinoma, Amer. J. Obstet. Gynecol. 135, 947-956 (1979). 12. Buchman, M. D., Kramer, E., and Feldman, G. B. Aspiration curettage for asymptomatic patients receiving estrogen, Obstet. Gynecol. 51, 339-341 (1978). 13. Studd, J. W., Thorn, M., Disch, F., et al. Value of cytology for detecting endometrial abnormalities in climacteric women receiving hormone replacement therapy, Erit. Med. J. 1(6167), 846-848 (1979). 14. Kottmeier, H. L., Ed Annual report on results oftreatment, Radiumhemmet, Stockholm, Vol. 16 (1977).