BRITISH MEDICAL JOURNAL
846
benefit of treating even mild to moderate increases (90-109 mm Hg) in diastolic pressure. The cost of overhauling equipment and collecting and compiling results was supported by a grant from the Glasgow and West of Scotland Group of the Chest, Heart, and Stroke Association. The members of the study group are grateful to the Royal College of Physicians, Edinburgh, and the Royal College of Physicians and Surgeons of Glasgow for accommodating their meetings, and to Mrs Margaret Dick and the staff of the practices for their help. Professor V M Hawthorne convened the study group.
References
Joint Working Party of the Royal College of Physicians of London and the British Cardiac Society on Prevention of Coronary Heart Disease,
Journal of the Royal College of Physicians of London, 1976, 10, 3. 2Veterans Administration Co-operative Study Group on Anti-hypertensive Agents, J'ournal of the American Medical Association, 1970, 213, 1142. 3
Werko, L, in Pathophysiology and Management of Arterial Hypertension, ed G Berglund, L T4ansson, and L Werko, p 283. Molndal, Sweden, A B Hassle, 1975.
31 MARCH 1979
4Working Party on Functional Integration of Patient Care in Hyperteasion, in Proceedings of Annual Community Medicine Conference, Glasgow, 1975; p 19. Glasgow, Greater Glasgow Health Board, 1975. 5 Rose, G A, Lancet, 1965, 1, 673. 6 Cartwright, A, Patients and their Doctors, p 24. London, Routledge and Kegan Paul, 1967. 7Beevers, D G, et al, in Pathophysiology and Management of Arterial Hypertension, ed G Berglund, L Hansson, and L Werko, p 263. Molndal, Sweden, A B Hassle, 1975. 8Hawthorne, V M, et al, International3Journal of Epidemiology, 1972, 1, 370. 9 Hawthorne, V M, et al, British Medical_Journal, 1974, 3, 601. lOHawthorne, V M, Current Medical Research and Opinion, 1977, 5, suppl No 1, p 109. 1 Hawthorne, V M, et al, in Epidemiology and Control of Hypertension, Second International Symposium, p 542. New York, Stratton Intercontinental Medical Book Corp, 1974. 12 Tudor Hart, J, personal communication. "Waters, W E, British Medical_Journal, 1971, 1, 142. 14 Weiss, N S, New England Journal of Medicine, 1972, 287, 631. 15 Miall, W E, and Brennan, P J, Clinical Science and Molecular Medicine, 1975, 48, suppl No 2, p 165s. 16 Bryers, F, and Hawthorne, V M, Journal of Epidemiology and Community Health, 1978, 32, 171.
(Accepted 5 February 1979)
Value of cytology for detecting endometrial abnormalities in climacteric women receiving hormone replacement therapy J W W STUDD, MARGARET THOM, F DISCHE, M DRIVER, T WADE EVANS, D WILLIAMS British Medical3journal, 1979, 1, 846-848
Summary and conclusions Over six months 113 endometrial specimens from 110 menopausal women receiving hormone replacement. therapy were examined by two cytologists and two histopathologists. Specimens were obtained by aspiration with the Isaacs cell sampler immediately before Vabra suction curettage, both procedures being performed in the outpatient department without anaesthetic. The histologists agreed with each other on the classification of 85 specimens (75-2%) and the cytologists agreed on the classification of 44 (38-9%). In only 15 cases (13.3%) did all four observers agree. Of the three cases of cystic or adenomatous hyperplasia detected histologically, only one was diagnosed by cytology. Furthermore, both cases of adenocarcinoma escaped detection by cytology, though a third case-later confirmed-was suspected by one cytologist alone. Use of the Isaacs endometrial cell sampler cannot be advocated for routine screening of women with climacteric symptoms receiving hormone replacement therapy. Efforts should be made to establish the correct
King's College Hospital, London SE5 9RS J W W STUDD, MD, MRCOG, consultant gynaecologist MARGARET THOM, MB, MRCOG, senior registrar F DISCHE, MD, FRCPATH, consultant pathologist M DRIVER, MB, MRCPATH, consultant pathologist Birmingham and Midland Hospital for Women, Birmingham 11 T WADE EVANS, MD, MRCPATH, consultant pathologist D WILLIAMS, FIMLS, chief technician
dose and duration of treatment with an oestrogenprogestogen preparation in order to avoid over-stimulating the endometrium and the need for regular screening for endometrial abnormalities.
Introduction The widespread use of oestrogens for treating climacteric symptoms is restricted by the fear of endometrial carcinoma and the lack of a simple, painless method of monitoring endometrial changes. Regular histological examination of the endometrium has been advocated, which would necessitate a simple outpatient screening procedure. Formal dilatation and curettage under general anaesthesia is logistically impossible, and Vabra suction curettage may cause discomfort and fail to produce a satisfactory specimen in 18% of patients taking oestrogens.1 Hutton et at2 obtained satisfactory specimens in 90% of cases using the Isaacs cell sampler in routine gynaecological patients and suggested that the technique would be useful for screening postmenopausal women. We have evaluated this claim.
Patients and methods Over six months beginning in February'1978, 113 endometrial specimens were collected by aspiration with the Isaacs cell sampler from 110 menopausal women receiving various regimens of hormone replacement therapy. The technique was used immediately before Vabra suction curettage. Both procedures were performed in the outpatient department without anaesthetic. A bivalve speculum was inserted into the vagina and a single-toothed tenaculum applied to the anterior lip of the cervix. The assembled sampler was then introduced through the cervical os until the fundus was reached. After pushing the cervical stop against the cervix the plunger of the syringe was withdrawn to create a vacuum and the intrauterine cannula rotated into each cornu to obtain samples from the whole uterine cavity. The
BRITISH MEDICAL JOURNAL
847
31 MARCH 1979
cannula was then withdrawn and the aspirate transferred from the syringe and cannula to a single glass slide. This was placed immediately in fixative and later examined independently by two cytologists (MD and DW) without reference to hormone treatment. Vabra suction curettage was then performed on each patient, the specimens being fixed in formalin-glutaraldehyde and later examined independently by two histopathologists (FD and TWE). Each observer was asked to classify the specimens as atrophic, proliferative, secretory, cystic hyperplasia, adenomatous hyperplasia, atypical hyperplasia, adenocarcinoma, or unsatisfactory.
TABLE III-Agreement on classification of specimens when analysed according to basic categories of diagnosis Normal Histologists agreed (n = 85) Cytologists agreed (n = 44) Histologists and cytologists agreed (n= 15)
Cystic/ adenomatous Carcinoma Unsatisfactory hyperplasia
56
3
32
1
2
24
11
11
1
3
Results
Cytological and histological reports were available on all 113 specimens. The histologists agreed with each other on the classification of 85 specimens (75 20/o), half the cases of disagreement concerning their views on what constituted an unsatisfactory specimen. The cytologists agreed on the classification of 44 specimens (38-9%). In the case of only 15 specimens did all four observers agree; 10 of these specimens were classified as proliferative, one as secretory, one as adenomatous hyperplasia, and three as unsatisfactory. All four observers disagreed on the classification of one specimen. Table I gives the histological and cytological classification of specimens on which the respective observers were agreed. The agreed histological diagnosis was accepted by both cytologists in only 15 cases (table II). Of the 26 cases of proliferative endometrium agreed by the histologists, 10 were diagnosed by both cytologists. One cytologist agreed iis seven cases and neither cytologist agreed in the remaining nine. Similarly, out of 30 cases of secretory endometrium agreed by the two histologists, only one was detected by both cytologists. In 16 cases the secretory endometrium was diagnosed by one cytologist, and in 13 cases neither cytologist agreed that secretory changes were present. Only one of the two cases of adenomatous hyperplasia and neither of the two cases of adenocarcinoma was detected cytologically, although both cytologists agreed that the specimens from the patients with carcinoma were unsatisfactory because of blood and should be repeated, and that one slide showed atypical endometrial cells. When the findings of all four observers were analysed simply as normal, hyperplasia, carcinoma, and unsatisfactory unanimous agreement occurred in 11 of the 56 cases judged to be normal histologically, in only one of the three hyperplastic specimens, and in neither of the two specimens of carcinoma. The histologists reported 24 specimens as unsatisfactory and the cytologists only 11, but these were different specimens; only three specimens were reported as unsatisfactory by all four investigators (table III). Cytology proved to be superior in one case, in which adenocarcinoma cells were seen by one cytologist while the other three observers diagnosed secretory cells. Formal dilatation and curettage two weeks later disclosed adenocarcinoma.
Discussion
Owing to adverse reports3-5 on the role of oestrogens used for climacteric symptoms in the pathogenesis of endometrial cancer our patients have been willing to undergo any screening procedure; thus they readily submit to yearly Vabra curettage, as recommended by some workers.6 A simpler screening method
would be invaluable and permit widespread prescription of oestrogens outside the specialist research clinic. Hutton et al2 produced encouraging results after investigating over 100 general gynaecological patients with the cell sampler. Satisfactory specimens were obtained in 90% of cases. In our series only 33 (75%) of the 44 specimens whose classification ,was agreed by both cytologists were -satisfactory for cytology. Nevertheless, patient acceptability was high, since the procedure is less painful than Vabra curettage, which many had experienced. Histology is currently used to assess the endometrium in patients receiving hormone replacement therapy, and we used the method as a baseline against which to evaluate cytology. At first sight disagreement between the two histopathologists on the classification of 28 specimens (25%) was disturbing, but in 14 cases (12%) this was explained by their views on what constituted an unsatisfactory specimen and could not be regarded as a fundamental difference of opinion. Cytological differentiation between atypical hyperplasia and adenocarcinoma is not possible, and in this series the method of cell sampling failed in one case of carcinoma; in the other case, however, abnormal cells were seen in the specimen and further investigation would have been requested had the procedure been for screening. A,third case of adenocarcinoma was suspected by one cytologist, although the other three observers diagnosed secretory endometrium. Cytology could not distinguish between proliferative and secretory endometrium, although these would have been adequately recognised as "normal." The failure to recognise hyperplastic specimens, however, was alarming. This type of change, which in minor degrees is difficult to distinguish histologically, is of particular concern to physicians prescribing oestrogens to postmenopausal women.7 Cystic hyperplasia has been found in lOo/< of women after a mean of 9-7 months' treatment with cyclical unopposed oral oestrogens' and in 32% of women after a mean of 15 8 months' treatment.4 It is therefore important that a screening procedure detects hyperplasia, and our study shows that the Isaacs cell sampler cannot do this. The samples we obtained were not of a consistently high enough standard to permit consistently reliable results, although we were careful to practise aspirating endometrial cells before the study and are confident that the specimens we obtained were the best possible. In view of our results we cannot advocate the use of the Isaacs endometrial cell sampler as a routine screening
TABLE I-Classification of specimens on which respective observers were agreed Atrophic
Histologists agreed (n = 85) Cytologists agreed (n = 44)
Proliferative
Secretory
26 28
30 4
Cystic
Adenomatous
Adenocarcinoma
Unsatisfactory
hyperplasia 1
hyperplasia 2 1
2
24 11
Adenocarcinoma
Unsatisfactory
TABLE II-Classification of specimens on which allfour observers were agreed or on which three were agreed (one cytologist disagreed) Atrophic All four observers agreed (n = 15) Three out of four observers agreed (n = 27)
Proliferative 10 7
Secretory 1 16
Cystic hyperplasia
Adenomatous
hyperplasia 1
3 4
BRITISH MEDICAL JOURNAL
848
procedure for climacteric women receiving hormone replacement therapy or to monitor the reversal of endometrial hyperplasia with oral progestogens.8 One cytologist felt that a confident diagnosis was possible only when a piece of endometrial tissue had been obtained, allowing a histological diagnosis to be made. We believe that cytology will not be a useful means of detecting endometrial abnormalities and that efforts should be directed towards establishing the correct dose and duration of treatment with an oestrogen-progestogen preparation to avoid overstimulating the endometrium. We thank Colgate-Palmolive Ltd for supplying Isaac samplers, and Ayerst Laboratories Ltd, Organon Laboratories Ltd, and Schering Chemicals Ltd for continued financial support of the menopause clinic at Dulwich Hospital.
31 MARCH 1979
References Sturdee, D W, et al, British Medical3Journal, 1978, 1, 1575.
Hutton, J D, et al, British Medical_Journal, 1978, 1, 947. 3Smith, D G, et al, New England Journal of Medicine, 1975, 293, 1164. 4Ziel, H K, and Finkle, W D, New England3Journal of Medicine, 1975, 293, 1167. Mack, T M, et al, New England Jrournal of Medicine, 1976, 294, 1262. 2
6 Whitehead, M I, and Campbell, S, in Proceedings of Second International Meeting on Endometrial Cancer and Related Topics, ed R W Taylor, M Brush, and R J King. London, Bailliere, Tindall and Cassell, 1978. 7Studd, J W W, Chakravarti, S, and Oram, D H, in Clinics in Obstetrics and Gynaecology, ed R B Greenblatt and J W Studd, vol 4, p 3. London,
Saunders, 1977. 8 Studd, J W W, Thom, Margaret, and White, P J, British
1978, 2, 1369.
Medical3Journal,
(Accepted 5 February 1979)
Comparison of salbutanmol given intravenously and by intermittent positive-pressure breathing in life-threatening asthma P BLOOMFIELD, J CARMICHAEL, G R PETRIE, N P JEWELL, G K CROMPTON British
Medical_Journal,
1979, 1, 848-850
Summary and conclusions A double-blind crossover trial was carried out during 22 episodes of life-threatening asthma in 19 patients to compare salbutamol given as a 500 ,ug intravenous injection and as a 0 5% solution administered by intermittent positive-pressure breathing (IPPB) for three minutes. Relief of pulsus paradoxus was significantly better after IPPB than the intravenous treatment. Both treatments significantly improved the peak expiratory flow rate. Salbutamol given intravenously produced a mean increase in heart rate of over 20 beats/min five minutes after treatment compared with the relief of tachycardia that occurred after administration by IPPB. Four patients had noticeable cardiovascular side effects after salbutamol given intravenously, but no such effects were noticed after administration by IPPB. Two patients withdrawn shortly after entry into the trial because of a worsening clinical condition had received intravenous salbutamol. It is concluded that salbutamol given by IPPB is better than that given by slow intravenous injection in severe acute asthma. Introduction Bronchodilators are essential in treating severe acute asthma in the critical initial period before corticosteroids have had time to Respiratory Unit, Northern General Hospital, Edinburgh EH5 2DQ P BLOOMFIELD, MRCP, research assistant J CARMICHAEL, MB, CHB, medical registrar G R PETRIE, MRCP, medical registrar (present appointment: senior registrar, Royal Infirmary, Glasgow) G K CROMPTON, FRCPE, consultant physician Medical Computing and Statistics Unit, University of Edinburgh Medical School, Edinburgh N P JEWELL, PHD, research fellow
act. Beta 2-stimulant drugs such as salbutamol or terbutaline may be given intravenously or as an aerosol via a nebuliser with or without the additional help of intermittent positive-pressure breathing (IPPB). The best route of administration in the severely ill patient is still controversial, and we designed this study to compare the efficacy of salbutamol given by IPPB and
by intravenous injection.
Patients and methods All patients who entered the trial had severe acute asthma and satisfied the criteria of having a heart rate of over 120 beats/min and a peak expiratory flow rate (PEFR) less than 20%// of the predicted normal. All patients gave their informed consent. The trial was a double-blind crossover comparison of two treatments given one hour apart, the order of treatment being randomly allocated. The treatments were (1) a 0 5%,' solution of salbutamol given for three minutes by IPPB, plus 5 ml of saline given by intravenous injection over three minutes; and (2) saline given for three minutes by IPPB, plus 500 ,ug salbutamol in 5 ml saline given by intravenous injection over three minutes. A patient-triggered Bennett ventilator with a tightly fitting face mask was used to deliver the nebulised solutions by IPPB. All patients also received 500 mg hydrocortisone intravenously and oxygen via 350o Ventimask. When the patients entered the trial we measured the PEFR with a Wright peak flow meter, using a low-reading paediatric meter when necessary, and recorded the best of three attempts. The heart rate was read from a Hewlett-Packard cardiac monitor and the respiratory rate recorded. These measurements were repeated five, 15, 30, 45, and 60 minutes after each treatment. Arterial blood-gas analysis was made on entry at least five minutes after the patient had been inspiring 350/ oxygen and repeated 60 minutes after each treatment. The degree of pulsus paradoxus was measured using a sphygmomanometer on entry and 60 minutes after each treatment. We studied 22 episodes of severe acute asthma in 19 patients aged 17-54 years (mean 27-35 years). Nine of the patients were receiving prednisolone or adrenocorticotrophic hormone and inhaled corticosteroids as regular maintenance treatment; three were taking inhaled beclomethasone alone; and two were taking sodium cromoglycate alone. All except one patient had been using a sympathomimetic aerosol before admission, and one had used a Bennett ventilator to administer salbutamol by IPPB at home some 90 minutes before admission. Two patients were withdrawn from the trial within 30 minutes after entry and given conventional treatment when their
846
benefit of treating even mild to moderate increases (90-109 mm Hg) in diastolic pressure. The cost of overhauling equipment and collecting and compiling results was supported by a grant from the Glasgow and West of Scotland Group of the Chest, Heart, and Stroke Association. The members of the study group are grateful to the Royal College of Physicians, Edinburgh, and the Royal College of Physicians and Surgeons of Glasgow for accommodating their meetings, and to Mrs Margaret Dick and the staff of the practices for their help. Professor V M Hawthorne convened the study group.
References
Joint Working Party of the Royal College of Physicians of London and the British Cardiac Society on Prevention of Coronary Heart Disease,
Journal of the Royal College of Physicians of London, 1976, 10, 3. 2Veterans Administration Co-operative Study Group on Anti-hypertensive Agents, J'ournal of the American Medical Association, 1970, 213, 1142. 3
Werko, L, in Pathophysiology and Management of Arterial Hypertension, ed G Berglund, L T4ansson, and L Werko, p 283. Molndal, Sweden, A B Hassle, 1975.
31 MARCH 1979
4Working Party on Functional Integration of Patient Care in Hyperteasion, in Proceedings of Annual Community Medicine Conference, Glasgow, 1975; p 19. Glasgow, Greater Glasgow Health Board, 1975. 5 Rose, G A, Lancet, 1965, 1, 673. 6 Cartwright, A, Patients and their Doctors, p 24. London, Routledge and Kegan Paul, 1967. 7Beevers, D G, et al, in Pathophysiology and Management of Arterial Hypertension, ed G Berglund, L Hansson, and L Werko, p 263. Molndal, Sweden, A B Hassle, 1975. 8Hawthorne, V M, et al, International3Journal of Epidemiology, 1972, 1, 370. 9 Hawthorne, V M, et al, British Medical_Journal, 1974, 3, 601. lOHawthorne, V M, Current Medical Research and Opinion, 1977, 5, suppl No 1, p 109. 1 Hawthorne, V M, et al, in Epidemiology and Control of Hypertension, Second International Symposium, p 542. New York, Stratton Intercontinental Medical Book Corp, 1974. 12 Tudor Hart, J, personal communication. "Waters, W E, British Medical_Journal, 1971, 1, 142. 14 Weiss, N S, New England Journal of Medicine, 1972, 287, 631. 15 Miall, W E, and Brennan, P J, Clinical Science and Molecular Medicine, 1975, 48, suppl No 2, p 165s. 16 Bryers, F, and Hawthorne, V M, Journal of Epidemiology and Community Health, 1978, 32, 171.
(Accepted 5 February 1979)
Value of cytology for detecting endometrial abnormalities in climacteric women receiving hormone replacement therapy J W W STUDD, MARGARET THOM, F DISCHE, M DRIVER, T WADE EVANS, D WILLIAMS British Medical3journal, 1979, 1, 846-848
Summary and conclusions Over six months 113 endometrial specimens from 110 menopausal women receiving hormone replacement. therapy were examined by two cytologists and two histopathologists. Specimens were obtained by aspiration with the Isaacs cell sampler immediately before Vabra suction curettage, both procedures being performed in the outpatient department without anaesthetic. The histologists agreed with each other on the classification of 85 specimens (75-2%) and the cytologists agreed on the classification of 44 (38-9%). In only 15 cases (13.3%) did all four observers agree. Of the three cases of cystic or adenomatous hyperplasia detected histologically, only one was diagnosed by cytology. Furthermore, both cases of adenocarcinoma escaped detection by cytology, though a third case-later confirmed-was suspected by one cytologist alone. Use of the Isaacs endometrial cell sampler cannot be advocated for routine screening of women with climacteric symptoms receiving hormone replacement therapy. Efforts should be made to establish the correct
King's College Hospital, London SE5 9RS J W W STUDD, MD, MRCOG, consultant gynaecologist MARGARET THOM, MB, MRCOG, senior registrar F DISCHE, MD, FRCPATH, consultant pathologist M DRIVER, MB, MRCPATH, consultant pathologist Birmingham and Midland Hospital for Women, Birmingham 11 T WADE EVANS, MD, MRCPATH, consultant pathologist D WILLIAMS, FIMLS, chief technician
dose and duration of treatment with an oestrogenprogestogen preparation in order to avoid over-stimulating the endometrium and the need for regular screening for endometrial abnormalities.
Introduction The widespread use of oestrogens for treating climacteric symptoms is restricted by the fear of endometrial carcinoma and the lack of a simple, painless method of monitoring endometrial changes. Regular histological examination of the endometrium has been advocated, which would necessitate a simple outpatient screening procedure. Formal dilatation and curettage under general anaesthesia is logistically impossible, and Vabra suction curettage may cause discomfort and fail to produce a satisfactory specimen in 18% of patients taking oestrogens.1 Hutton et at2 obtained satisfactory specimens in 90% of cases using the Isaacs cell sampler in routine gynaecological patients and suggested that the technique would be useful for screening postmenopausal women. We have evaluated this claim.
Patients and methods Over six months beginning in February'1978, 113 endometrial specimens were collected by aspiration with the Isaacs cell sampler from 110 menopausal women receiving various regimens of hormone replacement therapy. The technique was used immediately before Vabra suction curettage. Both procedures were performed in the outpatient department without anaesthetic. A bivalve speculum was inserted into the vagina and a single-toothed tenaculum applied to the anterior lip of the cervix. The assembled sampler was then introduced through the cervical os until the fundus was reached. After pushing the cervical stop against the cervix the plunger of the syringe was withdrawn to create a vacuum and the intrauterine cannula rotated into each cornu to obtain samples from the whole uterine cavity. The
BRITISH MEDICAL JOURNAL
847
31 MARCH 1979
cannula was then withdrawn and the aspirate transferred from the syringe and cannula to a single glass slide. This was placed immediately in fixative and later examined independently by two cytologists (MD and DW) without reference to hormone treatment. Vabra suction curettage was then performed on each patient, the specimens being fixed in formalin-glutaraldehyde and later examined independently by two histopathologists (FD and TWE). Each observer was asked to classify the specimens as atrophic, proliferative, secretory, cystic hyperplasia, adenomatous hyperplasia, atypical hyperplasia, adenocarcinoma, or unsatisfactory.
TABLE III-Agreement on classification of specimens when analysed according to basic categories of diagnosis Normal Histologists agreed (n = 85) Cytologists agreed (n = 44) Histologists and cytologists agreed (n= 15)
Cystic/ adenomatous Carcinoma Unsatisfactory hyperplasia
56
3
32
1
2
24
11
11
1
3
Results
Cytological and histological reports were available on all 113 specimens. The histologists agreed with each other on the classification of 85 specimens (75 20/o), half the cases of disagreement concerning their views on what constituted an unsatisfactory specimen. The cytologists agreed on the classification of 44 specimens (38-9%). In the case of only 15 specimens did all four observers agree; 10 of these specimens were classified as proliferative, one as secretory, one as adenomatous hyperplasia, and three as unsatisfactory. All four observers disagreed on the classification of one specimen. Table I gives the histological and cytological classification of specimens on which the respective observers were agreed. The agreed histological diagnosis was accepted by both cytologists in only 15 cases (table II). Of the 26 cases of proliferative endometrium agreed by the histologists, 10 were diagnosed by both cytologists. One cytologist agreed iis seven cases and neither cytologist agreed in the remaining nine. Similarly, out of 30 cases of secretory endometrium agreed by the two histologists, only one was detected by both cytologists. In 16 cases the secretory endometrium was diagnosed by one cytologist, and in 13 cases neither cytologist agreed that secretory changes were present. Only one of the two cases of adenomatous hyperplasia and neither of the two cases of adenocarcinoma was detected cytologically, although both cytologists agreed that the specimens from the patients with carcinoma were unsatisfactory because of blood and should be repeated, and that one slide showed atypical endometrial cells. When the findings of all four observers were analysed simply as normal, hyperplasia, carcinoma, and unsatisfactory unanimous agreement occurred in 11 of the 56 cases judged to be normal histologically, in only one of the three hyperplastic specimens, and in neither of the two specimens of carcinoma. The histologists reported 24 specimens as unsatisfactory and the cytologists only 11, but these were different specimens; only three specimens were reported as unsatisfactory by all four investigators (table III). Cytology proved to be superior in one case, in which adenocarcinoma cells were seen by one cytologist while the other three observers diagnosed secretory cells. Formal dilatation and curettage two weeks later disclosed adenocarcinoma.
Discussion
Owing to adverse reports3-5 on the role of oestrogens used for climacteric symptoms in the pathogenesis of endometrial cancer our patients have been willing to undergo any screening procedure; thus they readily submit to yearly Vabra curettage, as recommended by some workers.6 A simpler screening method
would be invaluable and permit widespread prescription of oestrogens outside the specialist research clinic. Hutton et al2 produced encouraging results after investigating over 100 general gynaecological patients with the cell sampler. Satisfactory specimens were obtained in 90% of cases. In our series only 33 (75%) of the 44 specimens whose classification ,was agreed by both cytologists were -satisfactory for cytology. Nevertheless, patient acceptability was high, since the procedure is less painful than Vabra curettage, which many had experienced. Histology is currently used to assess the endometrium in patients receiving hormone replacement therapy, and we used the method as a baseline against which to evaluate cytology. At first sight disagreement between the two histopathologists on the classification of 28 specimens (25%) was disturbing, but in 14 cases (12%) this was explained by their views on what constituted an unsatisfactory specimen and could not be regarded as a fundamental difference of opinion. Cytological differentiation between atypical hyperplasia and adenocarcinoma is not possible, and in this series the method of cell sampling failed in one case of carcinoma; in the other case, however, abnormal cells were seen in the specimen and further investigation would have been requested had the procedure been for screening. A,third case of adenocarcinoma was suspected by one cytologist, although the other three observers diagnosed secretory endometrium. Cytology could not distinguish between proliferative and secretory endometrium, although these would have been adequately recognised as "normal." The failure to recognise hyperplastic specimens, however, was alarming. This type of change, which in minor degrees is difficult to distinguish histologically, is of particular concern to physicians prescribing oestrogens to postmenopausal women.7 Cystic hyperplasia has been found in lOo/< of women after a mean of 9-7 months' treatment with cyclical unopposed oral oestrogens' and in 32% of women after a mean of 15 8 months' treatment.4 It is therefore important that a screening procedure detects hyperplasia, and our study shows that the Isaacs cell sampler cannot do this. The samples we obtained were not of a consistently high enough standard to permit consistently reliable results, although we were careful to practise aspirating endometrial cells before the study and are confident that the specimens we obtained were the best possible. In view of our results we cannot advocate the use of the Isaacs endometrial cell sampler as a routine screening
TABLE I-Classification of specimens on which respective observers were agreed Atrophic
Histologists agreed (n = 85) Cytologists agreed (n = 44)
Proliferative
Secretory
26 28
30 4
Cystic
Adenomatous
Adenocarcinoma
Unsatisfactory
hyperplasia 1
hyperplasia 2 1
2
24 11
Adenocarcinoma
Unsatisfactory
TABLE II-Classification of specimens on which allfour observers were agreed or on which three were agreed (one cytologist disagreed) Atrophic All four observers agreed (n = 15) Three out of four observers agreed (n = 27)
Proliferative 10 7
Secretory 1 16
Cystic hyperplasia
Adenomatous
hyperplasia 1
3 4
BRITISH MEDICAL JOURNAL
848
procedure for climacteric women receiving hormone replacement therapy or to monitor the reversal of endometrial hyperplasia with oral progestogens.8 One cytologist felt that a confident diagnosis was possible only when a piece of endometrial tissue had been obtained, allowing a histological diagnosis to be made. We believe that cytology will not be a useful means of detecting endometrial abnormalities and that efforts should be directed towards establishing the correct dose and duration of treatment with an oestrogen-progestogen preparation to avoid overstimulating the endometrium. We thank Colgate-Palmolive Ltd for supplying Isaac samplers, and Ayerst Laboratories Ltd, Organon Laboratories Ltd, and Schering Chemicals Ltd for continued financial support of the menopause clinic at Dulwich Hospital.
31 MARCH 1979
References Sturdee, D W, et al, British Medical3Journal, 1978, 1, 1575.
Hutton, J D, et al, British Medical_Journal, 1978, 1, 947. 3Smith, D G, et al, New England Journal of Medicine, 1975, 293, 1164. 4Ziel, H K, and Finkle, W D, New England3Journal of Medicine, 1975, 293, 1167. Mack, T M, et al, New England Jrournal of Medicine, 1976, 294, 1262. 2
6 Whitehead, M I, and Campbell, S, in Proceedings of Second International Meeting on Endometrial Cancer and Related Topics, ed R W Taylor, M Brush, and R J King. London, Bailliere, Tindall and Cassell, 1978. 7Studd, J W W, Chakravarti, S, and Oram, D H, in Clinics in Obstetrics and Gynaecology, ed R B Greenblatt and J W Studd, vol 4, p 3. London,
Saunders, 1977. 8 Studd, J W W, Thom, Margaret, and White, P J, British
1978, 2, 1369.
Medical3Journal,
(Accepted 5 February 1979)
Comparison of salbutanmol given intravenously and by intermittent positive-pressure breathing in life-threatening asthma P BLOOMFIELD, J CARMICHAEL, G R PETRIE, N P JEWELL, G K CROMPTON British
Medical_Journal,
1979, 1, 848-850
Summary and conclusions A double-blind crossover trial was carried out during 22 episodes of life-threatening asthma in 19 patients to compare salbutamol given as a 500 ,ug intravenous injection and as a 0 5% solution administered by intermittent positive-pressure breathing (IPPB) for three minutes. Relief of pulsus paradoxus was significantly better after IPPB than the intravenous treatment. Both treatments significantly improved the peak expiratory flow rate. Salbutamol given intravenously produced a mean increase in heart rate of over 20 beats/min five minutes after treatment compared with the relief of tachycardia that occurred after administration by IPPB. Four patients had noticeable cardiovascular side effects after salbutamol given intravenously, but no such effects were noticed after administration by IPPB. Two patients withdrawn shortly after entry into the trial because of a worsening clinical condition had received intravenous salbutamol. It is concluded that salbutamol given by IPPB is better than that given by slow intravenous injection in severe acute asthma. Introduction Bronchodilators are essential in treating severe acute asthma in the critical initial period before corticosteroids have had time to Respiratory Unit, Northern General Hospital, Edinburgh EH5 2DQ P BLOOMFIELD, MRCP, research assistant J CARMICHAEL, MB, CHB, medical registrar G R PETRIE, MRCP, medical registrar (present appointment: senior registrar, Royal Infirmary, Glasgow) G K CROMPTON, FRCPE, consultant physician Medical Computing and Statistics Unit, University of Edinburgh Medical School, Edinburgh N P JEWELL, PHD, research fellow
act. Beta 2-stimulant drugs such as salbutamol or terbutaline may be given intravenously or as an aerosol via a nebuliser with or without the additional help of intermittent positive-pressure breathing (IPPB). The best route of administration in the severely ill patient is still controversial, and we designed this study to compare the efficacy of salbutamol given by IPPB and
by intravenous injection.
Patients and methods All patients who entered the trial had severe acute asthma and satisfied the criteria of having a heart rate of over 120 beats/min and a peak expiratory flow rate (PEFR) less than 20%// of the predicted normal. All patients gave their informed consent. The trial was a double-blind crossover comparison of two treatments given one hour apart, the order of treatment being randomly allocated. The treatments were (1) a 0 5%,' solution of salbutamol given for three minutes by IPPB, plus 5 ml of saline given by intravenous injection over three minutes; and (2) saline given for three minutes by IPPB, plus 500 ,ug salbutamol in 5 ml saline given by intravenous injection over three minutes. A patient-triggered Bennett ventilator with a tightly fitting face mask was used to deliver the nebulised solutions by IPPB. All patients also received 500 mg hydrocortisone intravenously and oxygen via 350o Ventimask. When the patients entered the trial we measured the PEFR with a Wright peak flow meter, using a low-reading paediatric meter when necessary, and recorded the best of three attempts. The heart rate was read from a Hewlett-Packard cardiac monitor and the respiratory rate recorded. These measurements were repeated five, 15, 30, 45, and 60 minutes after each treatment. Arterial blood-gas analysis was made on entry at least five minutes after the patient had been inspiring 350/ oxygen and repeated 60 minutes after each treatment. The degree of pulsus paradoxus was measured using a sphygmomanometer on entry and 60 minutes after each treatment. We studied 22 episodes of severe acute asthma in 19 patients aged 17-54 years (mean 27-35 years). Nine of the patients were receiving prednisolone or adrenocorticotrophic hormone and inhaled corticosteroids as regular maintenance treatment; three were taking inhaled beclomethasone alone; and two were taking sodium cromoglycate alone. All except one patient had been using a sympathomimetic aerosol before admission, and one had used a Bennett ventilator to administer salbutamol by IPPB at home some 90 minutes before admission. Two patients were withdrawn from the trial within 30 minutes after entry and given conventional treatment when their
