577 were
1% and 18%. Obviously, retrospective popula-
analysis will looking for; they
tion
Hormone Replacement Therapy and Endometrial Cancer THE spectre of
a possible link between endoand metrial oestrogens continues to haunt the gynaecologist and the general practitioner. How should they now be dealing with a woman at or around the menopause whose symptoms are obviously due to oestrogen deprivation? Over a year has passed since publication of the five American studies’-5 purporting to show a cancer/oestrogen association, and much information, epidemiological, biological, and biochemical is to hand. Does a link exist between exogenous oestrogens and endometrial cancer? On existing evidence, yes. Is the link causal? We do not know. Yet on the assumption of causation some workers have already suggested avoidance of synthetic cestrogens (cestrone sulphate, in particular) in treatment. The main objections6- to the five studies are: that they were not double-blind and allocation to case or control groups was not randomised; that they were not prospective; and that the groups were unrepresentative of the female population. Other variables may have contributed to the association-use of cestrogens for perimenopausal bleeding caused by an unsuspected but pre-existing malign or pre-malign lesion; greater frequency of diagnostic examinations amongst hormone-treated women; and possible misinterpretation of oestrocancer
gen-induced hyperplasia
neoplasia. Probably, diagnosed earon in women lier oestrogen therapy-an assumption supported by a detailed histological review of the operative data7from one of the major studies.1 9S o of the oestrogen-treated group had a stage 0 or 1 lesion, as against 74% of the controls, while the corresponding figures for deeply invasive lesions endometrial
cancer
as
is detected and
1 Smith, D. C., Prentice, R., Thompson, D. J., Hermann, W. L. New Engl. J Med. 1975, 293, 1164. 2. Zeil, H. K, Finkle, W. D. ibid. p. 1167. 3 Quint, B.C. Am. J. Obstet. Gynec. 1975, 122, 498. 4 Mack,T. M., Pike, M. C., Henderson, B. E., Pfeffer, R. I., Gerkins, V. R., Arthur, M., Brown, S. E. New Engl. J. Med. 1976, 294, 1262. 5 Austin, D. Unpublished report presented before F.D.A. Advisory Committee, December, 1975; quoted by Gordon, G. S., Greenberg, B. G. Postgrad. Med. 1976,
59, 66.
Studd, J. ibid. p. 1144.
8 Greenblatt, R. B. in Consensus on Menopause Research (edited by P. A. Van Keep, R. B. Greenblatt,
and M.
Albeaux-Fernet); p. 59. Lancaster, Decennial (edited by
9 Vellios, F. in Genital and Mammary Pathology Sommers); p. 55. New York, 1975.
1976. S. C.
provide the answers we are only from prospective
studies with strict controls. Fragile though the epidemiological evidence may be, there is no escaping the fact that unopposed oestrogenic stimulation, either endogenous or exogenous, results in hyperplastic states of the endometrium which sometimes (in some 12%to 18%10 of cases) progress to invasive disease. Most such lesions can be reversed or destroyed by withdrawal of the oestrogens or administration of progestagenic or anti-oestrogenic agents. The prevention, detection, and treatment of such lesions is therefore of paramount importance. But in women taking oestrogens as postmenopausal replacement therapy the diagnosis of such hyperplastic lesions (glandular cystic, adenomatous, atypical hyperplasia, or carcinomain-situ) presents difficulties. Usually the dose of oestrogen will be so low that endometrial proliferation is slight and vaginal bleeding is absent; the Chelsea Hospital Group" have now shown that atypical hyperplasia can arise without symptoms with a low cyclical oestrogen dosage. If the dose of unopposed oestrogens is higher, then the endometrial stimulation will result in a form of irregular breakthrough bleeding; and this bleeding is hard to distinguish from that produced by an endometrial carcinoma: the gynaecologist has to prove that there is no carcinoma. Women given oestrogens with an exactly timed small dose of progestagens fare better since the withdrawal bleed is regular and deviation from regularity can be regarded as abnormal. Uncommonly the endometrium shows hyperplastic changes. How many women on replacement regimens acquire such lesions and what is the best method of surveillance? Probably as many as 1 in 3 women on continuous or cyclical oestrogen therapy
glandular cystic or adenomatous hyperplasia,l1-13 the rate being much lower in those taking a progestagen as well to induce regular endometrial shedding. Disturbing preliminary figures
get
emerge from the Chelsea Hospital Group investigation:" hyperplasia developed in 11 of 25 patients on cyclical high-dose oestrogen, half of them having the more sinister adenomatous type. All the lesions resolved on progestagen therapy. Surprisingly, 9 of 127 patients proved to have had hyperplastic endometrium before the start of replacement therapy, and 2 of these acquired an endometrial carcinoma
during replacement therapy. A number of diagnostic procedures could be used for screening such women, but large-scale studies 10. 11.
6 Cooke, I. D. Br. med. J. 1976, i, 1209.
not
can come
Gusberg, S. Am. J.
Obstet. Gynec. 1976, 126, 535. Whitehead, M. I., McQueen, J., Beard, R. J., Minardi, J., Campbell, S. Acta obstet. gynoec. scand. (in the press). 12. Perm, J., Thomas, K. in Ageing and Oestrogens: Frontiers of Hormone Research; vol. II, p. 134. Basle, 1973. 13. Sturdee, D. W., Wade-Evans, T., Gustafson, R. C., Studd, J. W. W. Br. J. Obstet. Gynœc. (in the press).
578
of them in a poor light. One of the most has been the technique devised by popular GRAVLEE. Irrigation of the uterine cavity with a negative pressure jet-washer produces either gross tissue fragments suitable for histological assessment or cytological material which can be directly smeared or subjected to further filtration processes. CASEY14 reports that, without anaesthesia, the device is easily inserted into the cervix of about one in two symptomless women over 40. In about a quarter of the patients insertion was somewhat difficult, while in the remainder it was very difficult or impossible. A quarter of the patients show
some
reported moderate to severe discomfort during the procedure. In this study 98% of irrigation samples were suitable for histological and cytological interpretation. But although these favourable results have been reproduced by some workers, IS 16 there is increasing disillusionment with the Gravlee technique. ALFONS017 reviewed the data on 221 women in eight studies between 1970 and 1975 and found a 43% false-negative rate in the detection of endometrial adenomatous hyperplasia. The results for endometrial carcinoma were better, but again the washings were positive in only 233 (80%) of 289 18 women seen by twelve groups. TWIGGS and his co-workers in California employed uterine irrigation in 556 abnormal uterine or postmenopausal bleeders who presented as outpatients. Only 2 of 16 women with adenomatous hyperplasia had the diagnosis confirmed by washings while, out of 25 with adenocarcinoma proven at later curethad negative and 10 had unsatisfactory 4 tage, women
Such high false-positive and false-negative rates forced them to conclude that "in the clinical setting outside of the controlled environment of a formal study" this method was unreliable. What other screening techniques exist besides outright dilatation and curettage under general anaesthesia ? One possibility is outpatient curettage, either with the steel aspiration curette (Vabra) or with a small conventional curette. Steel aspiration curettage is not very uncomfortable (only 2 out of 50 women in one study19 described the pain as severe); satisfactory endometrial specimens are obtained and the accuracy seems to be far better than that of the uterine irrigation technique. 19 11 Passage of the 3 mm diameter metal cannula into the postmenopausal cervix seems to be
washings.
Casey, M. J., Madden, T. J. in Consensus on Menopause Research (edited by P. A. Van Keep, R. B. Greenblatt, and M. Albeaux-Fernet); p. 139. Lancaster, 1976. 15. Henderson, S. R., Roxburgh, D. R., Bobrow, L. G., Pollard, S. M., Greening, S. E. Br. J. Obstet. Gynœc. 1975, 82, 976. 16. Bibbo, M., Rice, A. M., Wied, G. Obstet. Gynec. 1974, 43, 253. 17. Alfonso, J. F. ibid. 1975, 46, 141. 18. Twiggs, L. B., Di Saia, P. J., Morrow, C. P., Townsend, D. E., Schwinn, C. E. J. Am. med. Ass. 1976, 235, 2748. 19. Cohen, C. J., Gusberg, S. B., Koffler, D. Gynec. Oncol. 1974, 2, 279. 20. Denis, R., Barnett, J. M., Forbes, S. E. Obstet. Gynec. 1973, 42, 672. 14.
easier than passage of the 4.5 mm tip of the irrigation device. Other methods for obtaining endometrial specimens have made use of swabs, sponges, and brushes; all seem to increase intrauterine pressure (with the theoretical risk of extrauterine contamination), are disliked by patients, and do not seem to obtain uniform endometrial samples. How then should the woman on hormone replacement therapy be screened? Since oestrogens probably stimulate the endometrium in doserelated fashion,11 periodic sampling of the endometrium seems essential: a year might be an acceptable interval between tests, but it should be less if irregular bleeding occurs. The Chelsea Hospital Group, who have undertaken one of the first prospective studies on the effect of exogenous oestrogens on the postmenopausal endometrium, suggest sampling at six-monthly intervals. Aspiration curettage would probably be sufficient for these tests. But is all this monitoring being too cautious? The incidence rate of endometrial carinoma in the U.S.A. seems not to have risen between the last two major national cancer surveys of 1948-49 and 1969-71,21 and indeed mortality in this time has fallen from 9.1 to 4/100 000. In the U.K. there has been no increase in mortality from endometrial cancer up to 1973 or in incidence up to 1970;22 the percentage of postmenopausal women who have been on oestrogen replacement is small, probably about 6% of those between 50 and 59.22 Existing data make a case for some form of surveillance, and this could be arranged in National Health Service clinics now providing hormone replacement. Maybe the economists would argue for screening only women at high risk of endometrial cancer. GUSBERG10 has re-emphasised the tendency of the diabetic, the obese, and the nullipara to get this disease. Giving a progestagen in sequential form might also be cost-effective, but the analysis will be difficult since uterine cancer is seldom fatal.
Reactivation of Hepatitis-B Virus WHATEVER their ancestral origins,1 viruses must have been exposed to intense evolutionary presfrom a hostile external environment which includes the internal environment of animal hosts which, during their own evolution, acquired ever more elaborate defence-mechanisms. The enormous variation in the patterns of host/virus interactions and the wide spectrum of clinical illness which results are perhaps visible reflections of these sures
21. 22. 1.
Cramer, D. W., Cutler, S. J., Christine, B. Gynec. Oncol. 1974, 2, 130. Doll, R., Kinlen, L. J., Skegg, D. C. G. Br. med. J. 1976, i, 1071. Almeida, J. D. Br. J. Hosp. Med. 1973, 10, 368.
1% and 18%. Obviously, retrospective popula-
analysis will looking for; they
tion
Hormone Replacement Therapy and Endometrial Cancer THE spectre of
a possible link between endoand metrial oestrogens continues to haunt the gynaecologist and the general practitioner. How should they now be dealing with a woman at or around the menopause whose symptoms are obviously due to oestrogen deprivation? Over a year has passed since publication of the five American studies’-5 purporting to show a cancer/oestrogen association, and much information, epidemiological, biological, and biochemical is to hand. Does a link exist between exogenous oestrogens and endometrial cancer? On existing evidence, yes. Is the link causal? We do not know. Yet on the assumption of causation some workers have already suggested avoidance of synthetic cestrogens (cestrone sulphate, in particular) in treatment. The main objections6- to the five studies are: that they were not double-blind and allocation to case or control groups was not randomised; that they were not prospective; and that the groups were unrepresentative of the female population. Other variables may have contributed to the association-use of cestrogens for perimenopausal bleeding caused by an unsuspected but pre-existing malign or pre-malign lesion; greater frequency of diagnostic examinations amongst hormone-treated women; and possible misinterpretation of oestrocancer
gen-induced hyperplasia
neoplasia. Probably, diagnosed earon in women lier oestrogen therapy-an assumption supported by a detailed histological review of the operative data7from one of the major studies.1 9S o of the oestrogen-treated group had a stage 0 or 1 lesion, as against 74% of the controls, while the corresponding figures for deeply invasive lesions endometrial
cancer
as
is detected and
1 Smith, D. C., Prentice, R., Thompson, D. J., Hermann, W. L. New Engl. J Med. 1975, 293, 1164. 2. Zeil, H. K, Finkle, W. D. ibid. p. 1167. 3 Quint, B.C. Am. J. Obstet. Gynec. 1975, 122, 498. 4 Mack,T. M., Pike, M. C., Henderson, B. E., Pfeffer, R. I., Gerkins, V. R., Arthur, M., Brown, S. E. New Engl. J. Med. 1976, 294, 1262. 5 Austin, D. Unpublished report presented before F.D.A. Advisory Committee, December, 1975; quoted by Gordon, G. S., Greenberg, B. G. Postgrad. Med. 1976,
59, 66.
Studd, J. ibid. p. 1144.
8 Greenblatt, R. B. in Consensus on Menopause Research (edited by P. A. Van Keep, R. B. Greenblatt,
and M.
Albeaux-Fernet); p. 59. Lancaster, Decennial (edited by
9 Vellios, F. in Genital and Mammary Pathology Sommers); p. 55. New York, 1975.
1976. S. C.
provide the answers we are only from prospective
studies with strict controls. Fragile though the epidemiological evidence may be, there is no escaping the fact that unopposed oestrogenic stimulation, either endogenous or exogenous, results in hyperplastic states of the endometrium which sometimes (in some 12%to 18%10 of cases) progress to invasive disease. Most such lesions can be reversed or destroyed by withdrawal of the oestrogens or administration of progestagenic or anti-oestrogenic agents. The prevention, detection, and treatment of such lesions is therefore of paramount importance. But in women taking oestrogens as postmenopausal replacement therapy the diagnosis of such hyperplastic lesions (glandular cystic, adenomatous, atypical hyperplasia, or carcinomain-situ) presents difficulties. Usually the dose of oestrogen will be so low that endometrial proliferation is slight and vaginal bleeding is absent; the Chelsea Hospital Group" have now shown that atypical hyperplasia can arise without symptoms with a low cyclical oestrogen dosage. If the dose of unopposed oestrogens is higher, then the endometrial stimulation will result in a form of irregular breakthrough bleeding; and this bleeding is hard to distinguish from that produced by an endometrial carcinoma: the gynaecologist has to prove that there is no carcinoma. Women given oestrogens with an exactly timed small dose of progestagens fare better since the withdrawal bleed is regular and deviation from regularity can be regarded as abnormal. Uncommonly the endometrium shows hyperplastic changes. How many women on replacement regimens acquire such lesions and what is the best method of surveillance? Probably as many as 1 in 3 women on continuous or cyclical oestrogen therapy
glandular cystic or adenomatous hyperplasia,l1-13 the rate being much lower in those taking a progestagen as well to induce regular endometrial shedding. Disturbing preliminary figures
get
emerge from the Chelsea Hospital Group investigation:" hyperplasia developed in 11 of 25 patients on cyclical high-dose oestrogen, half of them having the more sinister adenomatous type. All the lesions resolved on progestagen therapy. Surprisingly, 9 of 127 patients proved to have had hyperplastic endometrium before the start of replacement therapy, and 2 of these acquired an endometrial carcinoma
during replacement therapy. A number of diagnostic procedures could be used for screening such women, but large-scale studies 10. 11.
6 Cooke, I. D. Br. med. J. 1976, i, 1209.
not
can come
Gusberg, S. Am. J.
Obstet. Gynec. 1976, 126, 535. Whitehead, M. I., McQueen, J., Beard, R. J., Minardi, J., Campbell, S. Acta obstet. gynoec. scand. (in the press). 12. Perm, J., Thomas, K. in Ageing and Oestrogens: Frontiers of Hormone Research; vol. II, p. 134. Basle, 1973. 13. Sturdee, D. W., Wade-Evans, T., Gustafson, R. C., Studd, J. W. W. Br. J. Obstet. Gynœc. (in the press).
578
of them in a poor light. One of the most has been the technique devised by popular GRAVLEE. Irrigation of the uterine cavity with a negative pressure jet-washer produces either gross tissue fragments suitable for histological assessment or cytological material which can be directly smeared or subjected to further filtration processes. CASEY14 reports that, without anaesthesia, the device is easily inserted into the cervix of about one in two symptomless women over 40. In about a quarter of the patients insertion was somewhat difficult, while in the remainder it was very difficult or impossible. A quarter of the patients show
some
reported moderate to severe discomfort during the procedure. In this study 98% of irrigation samples were suitable for histological and cytological interpretation. But although these favourable results have been reproduced by some workers, IS 16 there is increasing disillusionment with the Gravlee technique. ALFONS017 reviewed the data on 221 women in eight studies between 1970 and 1975 and found a 43% false-negative rate in the detection of endometrial adenomatous hyperplasia. The results for endometrial carcinoma were better, but again the washings were positive in only 233 (80%) of 289 18 women seen by twelve groups. TWIGGS and his co-workers in California employed uterine irrigation in 556 abnormal uterine or postmenopausal bleeders who presented as outpatients. Only 2 of 16 women with adenomatous hyperplasia had the diagnosis confirmed by washings while, out of 25 with adenocarcinoma proven at later curethad negative and 10 had unsatisfactory 4 tage, women
Such high false-positive and false-negative rates forced them to conclude that "in the clinical setting outside of the controlled environment of a formal study" this method was unreliable. What other screening techniques exist besides outright dilatation and curettage under general anaesthesia ? One possibility is outpatient curettage, either with the steel aspiration curette (Vabra) or with a small conventional curette. Steel aspiration curettage is not very uncomfortable (only 2 out of 50 women in one study19 described the pain as severe); satisfactory endometrial specimens are obtained and the accuracy seems to be far better than that of the uterine irrigation technique. 19 11 Passage of the 3 mm diameter metal cannula into the postmenopausal cervix seems to be
washings.
Casey, M. J., Madden, T. J. in Consensus on Menopause Research (edited by P. A. Van Keep, R. B. Greenblatt, and M. Albeaux-Fernet); p. 139. Lancaster, 1976. 15. Henderson, S. R., Roxburgh, D. R., Bobrow, L. G., Pollard, S. M., Greening, S. E. Br. J. Obstet. Gynœc. 1975, 82, 976. 16. Bibbo, M., Rice, A. M., Wied, G. Obstet. Gynec. 1974, 43, 253. 17. Alfonso, J. F. ibid. 1975, 46, 141. 18. Twiggs, L. B., Di Saia, P. J., Morrow, C. P., Townsend, D. E., Schwinn, C. E. J. Am. med. Ass. 1976, 235, 2748. 19. Cohen, C. J., Gusberg, S. B., Koffler, D. Gynec. Oncol. 1974, 2, 279. 20. Denis, R., Barnett, J. M., Forbes, S. E. Obstet. Gynec. 1973, 42, 672. 14.
easier than passage of the 4.5 mm tip of the irrigation device. Other methods for obtaining endometrial specimens have made use of swabs, sponges, and brushes; all seem to increase intrauterine pressure (with the theoretical risk of extrauterine contamination), are disliked by patients, and do not seem to obtain uniform endometrial samples. How then should the woman on hormone replacement therapy be screened? Since oestrogens probably stimulate the endometrium in doserelated fashion,11 periodic sampling of the endometrium seems essential: a year might be an acceptable interval between tests, but it should be less if irregular bleeding occurs. The Chelsea Hospital Group, who have undertaken one of the first prospective studies on the effect of exogenous oestrogens on the postmenopausal endometrium, suggest sampling at six-monthly intervals. Aspiration curettage would probably be sufficient for these tests. But is all this monitoring being too cautious? The incidence rate of endometrial carinoma in the U.S.A. seems not to have risen between the last two major national cancer surveys of 1948-49 and 1969-71,21 and indeed mortality in this time has fallen from 9.1 to 4/100 000. In the U.K. there has been no increase in mortality from endometrial cancer up to 1973 or in incidence up to 1970;22 the percentage of postmenopausal women who have been on oestrogen replacement is small, probably about 6% of those between 50 and 59.22 Existing data make a case for some form of surveillance, and this could be arranged in National Health Service clinics now providing hormone replacement. Maybe the economists would argue for screening only women at high risk of endometrial cancer. GUSBERG10 has re-emphasised the tendency of the diabetic, the obese, and the nullipara to get this disease. Giving a progestagen in sequential form might also be cost-effective, but the analysis will be difficult since uterine cancer is seldom fatal.
Reactivation of Hepatitis-B Virus WHATEVER their ancestral origins,1 viruses must have been exposed to intense evolutionary presfrom a hostile external environment which includes the internal environment of animal hosts which, during their own evolution, acquired ever more elaborate defence-mechanisms. The enormous variation in the patterns of host/virus interactions and the wide spectrum of clinical illness which results are perhaps visible reflections of these sures
21. 22. 1.
Cramer, D. W., Cutler, S. J., Christine, B. Gynec. Oncol. 1974, 2, 130. Doll, R., Kinlen, L. J., Skegg, D. C. G. Br. med. J. 1976, i, 1071. Almeida, J. D. Br. J. Hosp. Med. 1973, 10, 368.
