cedures treatment is by radiation or the appropriate radical surgery, depending on the extent of disease and the local facilities. If in situ carcinoma is found on cone biopsy, two possible situations may exist. One of these is that the cone biopsy has apparently removed the lesion completely. In this case it is appropriate to follow the patient with cytologic examination of smears from the canal and the portio vaginalis for an indefinite period. The second situation is that the cone biopsy has apparently not removed the lesion entirely. This may be handled by a repeat cone biopsy, if further pregnancies are desired, or by hysterectomy. While the foregoing has been the generally accepted management for the last 2 decades in Canada, and is still so today, two new factors have appeared that have influenced the management of the disease process. First is the fact that most of the cervical lesions are now being identified in the dysplastic phase. Second, they are being identified increasingly in younger women, many of whom have not yet borne children, so that preservation of reproductive function is an important factor. Because of these factors, colposcopy, which on earlier assessment in North America had been considered too cumbersome and inefficient for screening purposes, has been recently adapted to a role in the assessment and management of these early dysplastic lesions of the cervix. This is particularly applicable because conization, with its inherent morbidity, may be considered to
be an overly aggressive or excessive means of investigation and treatment for women who wish to bear children. Such disease can be assessed by colposcopically directed biopsy and then treated by cryosurgery or electrocautery techniques. It is also possible that colposcopy may make cone biopsy for diagnosis unnecessary in most cases. Conization would therefore become a predominantly therapeutic procedure designed to remove a lesion that has already been well defined by prior assessment. Several more years may be required for full evaluation of the role of colposcopy in the investigation and management of preclinical lesions of the cervix. However, the decreased morbidity, and the fact that colposcopically directed biopsies are taken in an outpatient facility rather than in an operating room, make it mandatory that the technique be fully assessed as an adjunct to screening programs. Colposcopy is currently being used in many centres in Canada but it is evident from experience to date that colposcopic investigation should only be performed in specially developed regional or local clinics that are properly equipped, and staffed by gynecologists well trained in colposcopy. It should not be performed in the random and relatively uncontrolled setting of individual medical offices.
2. WIED GL: Who should take the cytologic specimens? Acta Cytol (Baltimore) 9: 335, 1965 3. MCDONALD JR: Routine cervical screening of in-patients in a voluntary hospital. Acta Cytol
(Baltimore) 13: 455, 1969 4. CHRISTOPHERSON WM, PARKER JF: Control
of cervix cancer in women of low income in a community. Cancer 24: 64, 1969 5. FROST JK: Diagnostic accuracy of cervical
smears. Obstet Gynecol Surv 24: 893, 1969 6. ANDERSON WA, GUNN SA: A critical evaluation of the vaginal irrigation kit as a screening method for the detection of cancer of the cervix. Acta Cytol (Baltimore) 10: 149, 1966
7. REAGAN JW, UN F: An evaluation of the vaginal irrigation technique in the detection
of uterine cancer. Acta Cytol (Baltimore) 11: 374, 1967 8. PouLsaN RS: The feasibility of an automatic screening system for cytological specimens -
9. 10. 11.
12. 13.
a sur',ey, in Proueedings of the Fourth Tutonal of Clinical Cytology (Automation in Cytology), U of Chicago Pr, 1971, pp 134-82 WIED GL, BAHR GF: Automated Cell Identification and Cell Sorting, New York, Acad Pr, 1970 Proceedings of the international Conference on Automation of Uterine Cancer Cytology, U of Chicago Pr, 1975 (in press) WIED GL, Bieso M, BAHR GF, et al: Computerized recognition of uterine glandular cells. II. The application of a self-learning program. Acta Cytol (Baltimore) 13: 662, 1969 POuLSEN RS: Automated prescreening of cervical cytology specimens, thesis, McGill, Montreal, 1973 KuLcsAs DD: Survey of cytological facilities
in Canada. Can Med Assoc 1 94: 1228, 1966 14. ANDERSON GH, THOMPSON DW, KuLcsAs
DD: Survey of cytological facilities in Canada. Can Med Assoc I 101: 279, 1969 15. Canadian Society of Cytology: Survey of Cytology Facilities in Canada, 3rd survey, 1968-71 (unpublished data)
16. Idem: Survey of Cytology Facilities In Canada, 4th survey, 1972-75 (unpublished data) 17. COLLINs DN, KAUFMANN W: New York State computerized proficiency testing program in exfoliative cytology development. Acta Cytol (Baltimore) 15: 34, 1971 18. INHORN SL, CLARKE E: A state-wide proficiency testing program in cytology. Ibid, p 351 19. Renort on Proficiency Testing in Cytology Laboratories, Atlanta, GA, Center for Disease Control, Nov 1972 20. Ontario Council of Health: Cytological Serv-
ices in Ontario, 1973, Toronto et al: Quality control in cervical cytology. I Clin Pathol 27: 935, 1974
21. HuSAIN OAN, BUTLER ER, EVANS DMD,
References
22. Wian GL: Quality control standards for laboratories and cytotechnology registration.
1. Ama JE: Who should take the cytologic specimen? (C). Acta Cytol (Baltimore) 10: 66, 1966
Acta Cytol (Baltimore) 14: 557, 1970 23. Quality control in the cytology laboratory. Cytotechnol Bull 12: no 2, 1975
IV. Relations between screening programs for carcinoma of the cervix and other screening and preventive programs in Canada Screening for other malignant disease Other than cancer of the cervix, the most common malignant diseases that affect women in Canada are cancers of the breast, large bowel, ovary, lung and endometrium. For none of these is there a recognizable preclinical stage that is amenable to detection in a mass screening program. Thus, screening programs for these conditions have to be based on the detection of invasive cancers and not primarily of preinvasive disease, as is the case for cancer of the cervix. In addition, the age groups to
which screening programs for cancer of the cervix should be directed are 20 to 30 years younger than those for these other cancers. A further difficulty arises over the complexity and unreliability of some of the procedures required to detect these other common malignant diseases. As an example, one may consider cancer of the breast. In this condition, screening by mammography1 and clinical examination has been shown in one study to be efficacious in women aged 50 to 69, though the benefit was greatest in women aged 50 to 59. In
contrast, women aged 40 to 49 derived no benefit from the screening program. Even if the logistic difficulties of introducing screening by mammography and clinical examination in women aged 50 or more could be solved, it would be unrealistic to attempt to combine this with a screening program for cancer of the cervix, as the latter program should be based on women who are at least 30 years younger. In addition, the epidemiologic characteristics of women in the high-risk group for cervical cancer are completely different from those of women at high risk for breast cancer.
CMA JOURNAL/JUNE 5, 1976/VOL. 114 1031
For example, breast cancer is more frequent in single women or women who delay their first pregnancy until over the age of 25 or, particularly, 30.2 This is completely opposite to the situation for cancer of the cervix. Even in a condition such as carcinoma of the large bowel, where relatively simple tests for occult blood in the stool are currently being evaluated, it is unlikely, in view of the age distribution of individuals with these cancers, that screening tests would be initiated much before the age of 50. Similar considerations for other frequently occurring cancers in women lead to the conclusion that screening programs for cancer of the cervix should not be combined with programs for the detection of other cancers. It may be objected that the clinical examination that accompanies the taking of a smear is in itself a useful screen for ovarian and endometrial carcinoma. Apart from the age differential already discussed, no evidence has been seen in any program to support a beneficial effect of screening for cancer of the cervix on either the incidence or mortality of either of these conditions, which appear to be increasing in most technically advanced countries. Thus, in British Columbia and elsewhere the stage distribution for ovarian cancer has remained unchanged throughout the period when screening for carcinoma of the cervix has been in operation and there is no evidence of a consistent effect on mortality. Table I sets out the change in mortality Table I-Change In mortality from carcinoma of the ovary Mortality % increase Province 1960.62* 1970.72* (decrease) Newfoundland 13.92 16.81 20.8 Prince Edward Island 20.88 9.97 (52.3) Nova Scotia 16.19 16.88 4.3 New Brunswick 12.60 16.44 30.5 Quebec 12.75 12.81 0.5 Ontario 15.39 15.51 0.8 Manitoba 15.03 14.95 (0.5) Saskatchewan 12.18 10.06 (17.4) Alberta 13.98 13.34 (4.6) British Columbia 13.79 18.41 33.5 *Age.standardized rates per 100 000 women aged 35 to 64.
from ovarian cancer in the period 196062 to 1970-72 in women aged 35 to 64. In the provinces of Newfoundland, New Brunswick and British Columbia there has been a substantial increase in mortality, while there has been a moderate increase in Nova Scotia. In Prince Edward Island and Saskatchewan there has been a substantial fall in mortality, while in Alberta there has been a moderate fall. In the remaining provinces - namely, Quebec, Ontario and Manitoba - there has been hardly any change in mortality. The increase in the incidence of endometrial carcinoma that has been seen in Saskatchewan is documented in part II of this report, while in Louisville, Kentucky, the mortality from endometrial carcinoma has been shown to have risen in spite of a screening program that has reduced mortality from cancer of the cervix. The task force is therefore of the opinion that routine annual physical examinations in asymptomatic women cannot be advocated as a means of lessening the mortality from ovarian or endometrial cancer. Such an argument should not be used to counter the recommendations in this report on frequency of repeat screening examinations in women found free of cytologic atypia. However, every screening program must be justified on its own merits. When individuals participating in a screening program have the risk factors relevant to screening for other conditions, screening for them might with economy be combined. Currently, however, for women in a cervical cancer screening program, this does not seem practical except possibly for hypertension and cancer of the breast when the women reach the relevant age. Screening for nonmalignant disease
Several programs of screening for nonmalignant disease have been suggested. Some of those that have been advocated in adults - for example, screening for diabetes - have not so far produced unequivocal evidence of benefit. Screening for hypertension is currently regarded as potentially beneficial if individuals found to have hypertension can be persuaded to comply
with the necessary medication regimen. However, as for the malignant diseases already discussed, the age group for whom screening for hypertension would be advocated is substantially older than the age group who would be sought for entry to a cervical cancer screening program. Even so, women in such a program who attain the age of risk for hypertension could have their blood pressure taken while they attend for their cervical smears. Screening programs for cancer of the cervix, however, can with benefit be combined with other health programs. The objective is to make use of these other programs to ensure that women are brought into the cervical cancer screening program, rather than the reverse. Such programs include family-planning schemes, pre- and postnatal services, venereal disease detection and treatment programs, and the examination of women admitted to penal institutions. Economic aspects of screening A detailed discussion of the economic aspects of screening for cancer of the cervix would be out of place in this report. It is estimated that implementation of the recommendations on frequency of screening made in this report would make it possible to screen all women at risk at a cost of only one quarter of the sum that would be required to screen them annually. The total number of examinations required to cover the whole population at risk should be less than the total number already being conducted in Canada. This is a reflection of the extent to which programs are presently being utilized for the annual rescreening of cytologically negative women rather than concentrating on bringing into the programs women who have never been screened. References 1. SHAPIRO 5, STRAX P, VENET L, et al: Changes in 5-year breast cancer mortality in a breast cancer screening programme. Proc Nati Cancer Confer 7: 663, 1973 2. MACMAHON B, COLE P, BROWN J: Etiology of human breast cancer; a review. / Nail
Cancer Inst 50: 21, 1973
3. CHRIsTOPHERsoN WM, MENDEZ WM, PARKER JE, et al: Carcinoma of the endometrium: a study of changing rates over a 15-year period. Cancer 27: 1005. 1971
AVAILABLE TO MEDICAL FELLOWS Fellows of the Royal College of Physicians and Surgeons of Canada in the medical specialties who are interested in receiving the Canadian Journal of Surgery can do so without charge. Send your request for a free subscription to: Division of fellowship affairs Royal College of Physicians and Surgeons of Canada 74 Stanley Ave. Ottawa, ON KIM 1P4 1032 CMA JOURNAL/JUNE 5, 1976/VOL. 114
be an overly aggressive or excessive means of investigation and treatment for women who wish to bear children. Such disease can be assessed by colposcopically directed biopsy and then treated by cryosurgery or electrocautery techniques. It is also possible that colposcopy may make cone biopsy for diagnosis unnecessary in most cases. Conization would therefore become a predominantly therapeutic procedure designed to remove a lesion that has already been well defined by prior assessment. Several more years may be required for full evaluation of the role of colposcopy in the investigation and management of preclinical lesions of the cervix. However, the decreased morbidity, and the fact that colposcopically directed biopsies are taken in an outpatient facility rather than in an operating room, make it mandatory that the technique be fully assessed as an adjunct to screening programs. Colposcopy is currently being used in many centres in Canada but it is evident from experience to date that colposcopic investigation should only be performed in specially developed regional or local clinics that are properly equipped, and staffed by gynecologists well trained in colposcopy. It should not be performed in the random and relatively uncontrolled setting of individual medical offices.
2. WIED GL: Who should take the cytologic specimens? Acta Cytol (Baltimore) 9: 335, 1965 3. MCDONALD JR: Routine cervical screening of in-patients in a voluntary hospital. Acta Cytol
(Baltimore) 13: 455, 1969 4. CHRISTOPHERSON WM, PARKER JF: Control
of cervix cancer in women of low income in a community. Cancer 24: 64, 1969 5. FROST JK: Diagnostic accuracy of cervical
smears. Obstet Gynecol Surv 24: 893, 1969 6. ANDERSON WA, GUNN SA: A critical evaluation of the vaginal irrigation kit as a screening method for the detection of cancer of the cervix. Acta Cytol (Baltimore) 10: 149, 1966
7. REAGAN JW, UN F: An evaluation of the vaginal irrigation technique in the detection
of uterine cancer. Acta Cytol (Baltimore) 11: 374, 1967 8. PouLsaN RS: The feasibility of an automatic screening system for cytological specimens -
9. 10. 11.
12. 13.
a sur',ey, in Proueedings of the Fourth Tutonal of Clinical Cytology (Automation in Cytology), U of Chicago Pr, 1971, pp 134-82 WIED GL, BAHR GF: Automated Cell Identification and Cell Sorting, New York, Acad Pr, 1970 Proceedings of the international Conference on Automation of Uterine Cancer Cytology, U of Chicago Pr, 1975 (in press) WIED GL, Bieso M, BAHR GF, et al: Computerized recognition of uterine glandular cells. II. The application of a self-learning program. Acta Cytol (Baltimore) 13: 662, 1969 POuLSEN RS: Automated prescreening of cervical cytology specimens, thesis, McGill, Montreal, 1973 KuLcsAs DD: Survey of cytological facilities
in Canada. Can Med Assoc 1 94: 1228, 1966 14. ANDERSON GH, THOMPSON DW, KuLcsAs
DD: Survey of cytological facilities in Canada. Can Med Assoc I 101: 279, 1969 15. Canadian Society of Cytology: Survey of Cytology Facilities in Canada, 3rd survey, 1968-71 (unpublished data)
16. Idem: Survey of Cytology Facilities In Canada, 4th survey, 1972-75 (unpublished data) 17. COLLINs DN, KAUFMANN W: New York State computerized proficiency testing program in exfoliative cytology development. Acta Cytol (Baltimore) 15: 34, 1971 18. INHORN SL, CLARKE E: A state-wide proficiency testing program in cytology. Ibid, p 351 19. Renort on Proficiency Testing in Cytology Laboratories, Atlanta, GA, Center for Disease Control, Nov 1972 20. Ontario Council of Health: Cytological Serv-
ices in Ontario, 1973, Toronto et al: Quality control in cervical cytology. I Clin Pathol 27: 935, 1974
21. HuSAIN OAN, BUTLER ER, EVANS DMD,
References
22. Wian GL: Quality control standards for laboratories and cytotechnology registration.
1. Ama JE: Who should take the cytologic specimen? (C). Acta Cytol (Baltimore) 10: 66, 1966
Acta Cytol (Baltimore) 14: 557, 1970 23. Quality control in the cytology laboratory. Cytotechnol Bull 12: no 2, 1975
IV. Relations between screening programs for carcinoma of the cervix and other screening and preventive programs in Canada Screening for other malignant disease Other than cancer of the cervix, the most common malignant diseases that affect women in Canada are cancers of the breast, large bowel, ovary, lung and endometrium. For none of these is there a recognizable preclinical stage that is amenable to detection in a mass screening program. Thus, screening programs for these conditions have to be based on the detection of invasive cancers and not primarily of preinvasive disease, as is the case for cancer of the cervix. In addition, the age groups to
which screening programs for cancer of the cervix should be directed are 20 to 30 years younger than those for these other cancers. A further difficulty arises over the complexity and unreliability of some of the procedures required to detect these other common malignant diseases. As an example, one may consider cancer of the breast. In this condition, screening by mammography1 and clinical examination has been shown in one study to be efficacious in women aged 50 to 69, though the benefit was greatest in women aged 50 to 59. In
contrast, women aged 40 to 49 derived no benefit from the screening program. Even if the logistic difficulties of introducing screening by mammography and clinical examination in women aged 50 or more could be solved, it would be unrealistic to attempt to combine this with a screening program for cancer of the cervix, as the latter program should be based on women who are at least 30 years younger. In addition, the epidemiologic characteristics of women in the high-risk group for cervical cancer are completely different from those of women at high risk for breast cancer.
CMA JOURNAL/JUNE 5, 1976/VOL. 114 1031
For example, breast cancer is more frequent in single women or women who delay their first pregnancy until over the age of 25 or, particularly, 30.2 This is completely opposite to the situation for cancer of the cervix. Even in a condition such as carcinoma of the large bowel, where relatively simple tests for occult blood in the stool are currently being evaluated, it is unlikely, in view of the age distribution of individuals with these cancers, that screening tests would be initiated much before the age of 50. Similar considerations for other frequently occurring cancers in women lead to the conclusion that screening programs for cancer of the cervix should not be combined with programs for the detection of other cancers. It may be objected that the clinical examination that accompanies the taking of a smear is in itself a useful screen for ovarian and endometrial carcinoma. Apart from the age differential already discussed, no evidence has been seen in any program to support a beneficial effect of screening for cancer of the cervix on either the incidence or mortality of either of these conditions, which appear to be increasing in most technically advanced countries. Thus, in British Columbia and elsewhere the stage distribution for ovarian cancer has remained unchanged throughout the period when screening for carcinoma of the cervix has been in operation and there is no evidence of a consistent effect on mortality. Table I sets out the change in mortality Table I-Change In mortality from carcinoma of the ovary Mortality % increase Province 1960.62* 1970.72* (decrease) Newfoundland 13.92 16.81 20.8 Prince Edward Island 20.88 9.97 (52.3) Nova Scotia 16.19 16.88 4.3 New Brunswick 12.60 16.44 30.5 Quebec 12.75 12.81 0.5 Ontario 15.39 15.51 0.8 Manitoba 15.03 14.95 (0.5) Saskatchewan 12.18 10.06 (17.4) Alberta 13.98 13.34 (4.6) British Columbia 13.79 18.41 33.5 *Age.standardized rates per 100 000 women aged 35 to 64.
from ovarian cancer in the period 196062 to 1970-72 in women aged 35 to 64. In the provinces of Newfoundland, New Brunswick and British Columbia there has been a substantial increase in mortality, while there has been a moderate increase in Nova Scotia. In Prince Edward Island and Saskatchewan there has been a substantial fall in mortality, while in Alberta there has been a moderate fall. In the remaining provinces - namely, Quebec, Ontario and Manitoba - there has been hardly any change in mortality. The increase in the incidence of endometrial carcinoma that has been seen in Saskatchewan is documented in part II of this report, while in Louisville, Kentucky, the mortality from endometrial carcinoma has been shown to have risen in spite of a screening program that has reduced mortality from cancer of the cervix. The task force is therefore of the opinion that routine annual physical examinations in asymptomatic women cannot be advocated as a means of lessening the mortality from ovarian or endometrial cancer. Such an argument should not be used to counter the recommendations in this report on frequency of repeat screening examinations in women found free of cytologic atypia. However, every screening program must be justified on its own merits. When individuals participating in a screening program have the risk factors relevant to screening for other conditions, screening for them might with economy be combined. Currently, however, for women in a cervical cancer screening program, this does not seem practical except possibly for hypertension and cancer of the breast when the women reach the relevant age. Screening for nonmalignant disease
Several programs of screening for nonmalignant disease have been suggested. Some of those that have been advocated in adults - for example, screening for diabetes - have not so far produced unequivocal evidence of benefit. Screening for hypertension is currently regarded as potentially beneficial if individuals found to have hypertension can be persuaded to comply
with the necessary medication regimen. However, as for the malignant diseases already discussed, the age group for whom screening for hypertension would be advocated is substantially older than the age group who would be sought for entry to a cervical cancer screening program. Even so, women in such a program who attain the age of risk for hypertension could have their blood pressure taken while they attend for their cervical smears. Screening programs for cancer of the cervix, however, can with benefit be combined with other health programs. The objective is to make use of these other programs to ensure that women are brought into the cervical cancer screening program, rather than the reverse. Such programs include family-planning schemes, pre- and postnatal services, venereal disease detection and treatment programs, and the examination of women admitted to penal institutions. Economic aspects of screening A detailed discussion of the economic aspects of screening for cancer of the cervix would be out of place in this report. It is estimated that implementation of the recommendations on frequency of screening made in this report would make it possible to screen all women at risk at a cost of only one quarter of the sum that would be required to screen them annually. The total number of examinations required to cover the whole population at risk should be less than the total number already being conducted in Canada. This is a reflection of the extent to which programs are presently being utilized for the annual rescreening of cytologically negative women rather than concentrating on bringing into the programs women who have never been screened. References 1. SHAPIRO 5, STRAX P, VENET L, et al: Changes in 5-year breast cancer mortality in a breast cancer screening programme. Proc Nati Cancer Confer 7: 663, 1973 2. MACMAHON B, COLE P, BROWN J: Etiology of human breast cancer; a review. / Nail
Cancer Inst 50: 21, 1973
3. CHRIsTOPHERsoN WM, MENDEZ WM, PARKER JE, et al: Carcinoma of the endometrium: a study of changing rates over a 15-year period. Cancer 27: 1005. 1971
AVAILABLE TO MEDICAL FELLOWS Fellows of the Royal College of Physicians and Surgeons of Canada in the medical specialties who are interested in receiving the Canadian Journal of Surgery can do so without charge. Send your request for a free subscription to: Division of fellowship affairs Royal College of Physicians and Surgeons of Canada 74 Stanley Ave. Ottawa, ON KIM 1P4 1032 CMA JOURNAL/JUNE 5, 1976/VOL. 114
