© 1999 by Oxford University Press
Journal of the National Cancer Institute, Vol. 91, No. 3, 252-258,
February 3, 1999
© 1999 Oxford University Press
REPORTS |
Natural History of Dysplasia of the Uterine Cervix
Affiliation of authors: Faculty of Medicine, Department of Public Health Sciences, University of Toronto, Ontario, Canada.
Correspondence to: Philippa Holowaty, Ph.D., Durham Region Health Department, 1615 Dundas St., E., Suite 210, Whitby, Ontario L1N 2L1, Canada (e-mail: holowa_p{at}region.durham.on.ca).
 |
ABSTRACT |
|---|
 Top Abstract IntroductionMaterials and MethodsNote on TerminologyAnalysisResultsDiscussionReferences |
|---|
BACKGROUND: A historical cohort of Toronto (Ontario, Canada) women whose Pap smear histories were recorded at a major cytopathology laboratory provided the opportunity to study progression and regression of cervical dysplasia in an era (1962-1980) during which cervical squamous lesions were managed conservatively. METHODS: Actuarial and Cox's survival analyses were used to estimate the rates and relative risks of progression and regression of mild (cervical intraepithelial neoplasia 1 [CIN1]) and moderate (CIN2) dysplasias. In addition, more than 17 000 women with a history of Pap smears between 1970 and 1980 inclusive and who were diagnosed as having mild, moderate, or severe dysplasia were linked to the Ontario Cancer Registry for the outcome of any subsequent cervical cancers occurring through 1989. RESULTS: Both mild and moderate dysplasias were more likely to regress than to progress. The risk of progression from mild to severe dysplasia or worse was only 1% per year, but the risk of progression from moderate dysplasia was 16% within 2 years and 25% within 5 years. Most of the excess risk of cervical cancer for severe and moderate dysplasias occurred within 2 years of the initial dysplastic smear. After 2 years, in comparison with mild dysplasia, the relative risks for progression from severe or moderate dysplasia to cervical cancer in situ or worse was 4.2 (95% confidence interval [CI] = 3.0-5.7) and 2.5 (95% CI = 2.2-3.0), respectively. CONCLUSION: The risk of progression for moderate dysplasia was intermediate between the risks for mild and severe dysplasia; thus, the moderate category may represent a clinically useful distinction. The majority of untreated mild dysplasias were recorded as regressing to yield a normal smear within 2 years.
|
INTRODUCTION |
|---|
|
TopAbstractIntroductionMaterials and MethodsNote on TerminologyAnalysisResultsDiscussionReferences |
|---|
Understanding the natural history of cervical cancer is important for optimal management of its precursor lesions. Previous studies (1,2) of the natural history of precursor lesions, which followed up women with cervical dysplasia, were hampered by problems such as small numbers of study subjects, short follow-up times, inconsistent assessment of diagnostic labels, and failure to subclassify cases of dysplasia. A major limitation of current cohort studies in developed countries is that treatment following the diagnosis of cervical cancer precursors interrupts the natural history and makes it impossible to determine whether the lesions would have progressed or regressed if left untreated. However, it is clear from the number of cases of dysplasia diagnosed that there is a considerable excess of these lesions, such that the majority will not progress to invasive cancer, even without treatment (3).
The study reported here is a cohort study in which the Pap smear history of women is related to the risk of subsequent cervical cancer. In addition the Pap smear histories are used to estimate progression and regression rates of mild and moderate cervical dysplasia, prior to treatment, using life table analysis. This study makes use of a large and relatively unique historic cytology database and the excellent coverage and linkage facilities provided by the Ontario Cancer Registry (OCR). A preliminary analysis of the original dataset, prior to the linkage with the OCR, has been published (4).
|
MATERIALS AND METHODS |
|---|
|
TopAbstractIntroductionMaterials and MethodsNote on TerminologyAnalysisResultsDiscussionReferences |
|---|
The availability of a large cytology dataset provided a unique opportunity to study the natural history of a precancerous condition because the screening histories were collected in a time period (1962 through 1980) during which the management of cervical dysplasia in Ontario was relatively conservative, because of the care taken by the laboratory to make cervical cytology as reliable as possible, and because of the distinction made between different states of dysplasia. Furthermore, use of the OCR enabled identifying information from the cohort to be linked to cancer diagnosis data, to determine which of the women had developed cervical cancer during the screening period and beyond, through 1989. The OCR was also used to identify those women who had developed endometrial cancer, so that they could be censored at the time of removal of their uterus (and cervix). Censoring information for the study cohort was also obtained from two other databases made available via the OCR: Ontario mortality data for 1964-1989, obtained from the Registrar General's Office; and Ontario hospital discharges for hysterectomies, for the period from April 1, 1979, through March 31, 1989, obtained from the Hospital Medical Records Institute (HMRI). The three external databases all had province-wide coverage. Information on censoring due to treatment was available from the cervical cytology requisition forms.
Ethical approval for the study was received from both the University of Toronto and from The Ontario Cancer Research and Treatment Foundation (now Cancer Care Ontario).
The study is based on the cervical cytology records of Cyto-Pathology Associates, which was established in Toronto in 1962, and rapidly became the largest laboratory serving Metropolitan Toronto. Cyto-Pathology Associates, directed by Dr. Don Thompson, was a leading proponent from the 1960s through the 1980s of Laboratory Proficiency Testing, standardization of cytologic diagnosis, education of medical personnel in smear-taking technique, and record keeping for epidemiologic studies. This laboratory established consistent recording of cervical abnormalities in accordance with Canadian national recommendations, which used the World Health Organization (WHO) classification (5-7). There was a follow-up system at the laboratory to identify the outcome for women with a diagnosis of mild dysplasia or worse. Follow-up information obtained by laboratory staff was entered on the appropriate cytology requisition form, but treatment information supplied by the physician was entered on the subsequent cytology requisition form at the time of a woman's next Pap smear.
Data collection has already been described in detail (4). In the early 1980s, 75 585 records were extracted from among the estimated 2 million records maintained by the laboratory for the period 1962 through 1980. The records were chosen in two ways: A reference sample of more than 20 000 records was systematically sampled (on the basis of the last two digits of the individual's Ontario Health Insurance Plan [OHIP] Number or of the laboratory account number if there was no OHIP number) to represent the complete file of women with Pap smear data held by the laboratory. An additional set of more than 50 000 records was selected, consisting of all women with any smears classified as minimal dysplasia or worse, excluding those already in the initial sample.
One record represented one woman, since, where possible, all smears for the same woman were kept together in the one record. The OHIP number and last name of the woman were used to further bring together all smears for one woman. This resulted in a file of 70 236 records (a total of 176 806 smears); 20 461 of these women were in the sample group and an additional 49 775 were selected because of diagnoses of minimal dysplasia or worse (28 867 of whom had mild dysplasia or worse). The term minimal dysplasia had been chosen by the laboratory to include a group of smears with insufficient cellular abnormality to be classed as definitive mild dysplasia. It is likely that many of these smears would be classified currently, in the 1990s, as atypical squamous or glandular cells of undertermined significance in the Bethesda classification (8).
For this study, internal probabilistic record linkage (9) was used to remove additional duplicates from the study file. As a result of the internal linkage, 4810 duplicate records were removed from the file, leaving a file of 65 426 records representing 65 426 women, 18 981 (29%) in the sample group and 46 445 selected because of a diagnosis of minimal dysplasia or worse. Identifying information, such as the woman's first name was missing for 0.8% (158) of the sample group and 0.5% (254) of the abnormal smear group, and these women were eliminated from the analysis because the records could not be linked to the external files. This left a total study file of 65 014 women, which consisted of 18 823 in the sample group and 46 191 in the abnormal smear group.
For the analysis that involved linkage of the cytology records to the records of the OCR to determine the presence of an outcome of cervical cancer, it had been hoped that women in the sample group whose screening history started with two normal smears could be used as the referent group. However, there were so few cervical cancer events in this group that the referent group had to be women with mild dysplasia. Women were therefore eligible if they had resided in Metropolitan Toronto, and had mild, moderate, or severe dysplasia not preceded by a more severe result or by a prior gynecologic cancer. In addition, for this analysis, women were included only if their screening history began between 1970 and 1980. This was because identifying information was more complete from 1970 onward. In all, 17 946 women who had diagnoses of mild, moderate, or severe dysplasia were eligible. However, 729 (4%) women were then excluded because their records did not have a full date of birth and thus were less likely to link successfully to the external files. Of the 17 217 women who remained, 12 244 had mild dysplasia, 5121 had moderate dysplasia, and 956 had severe dysplasia (total, 18 321 dysplasias). A total of 1020 women were seen in two different dysplastic states, and 42 women were seen in all three dysplastic states. The outcomes of interest were cervical cancer in situor worse and, in a separate analysis, invasive cervical cancer. Censoring was due to death, endometrial cancer (when the cervix would have been removed), hysterectomy, or other gynecologic treatment (except when cervical cancer was diagnosed within 1 year of the treatment) or due to termination of follow-up at the end of 1989. The outcome had to have occurred after the start date, which was the date of initial dysplasia. This left 12 058 mild, 4837 moderate, and 724 severe dysplasias. Exclusions at this stage (2% for mild, 6% for moderate, 24% for severe dysplasias) were mostly due to treatment that occurred on or before the start date. Women observed in more than one dysplastic state contributed to each of the dysplastic states in the analysis. The average lengths of follow-up (with 95% confidence intervals [CIs]) were 159 (95% CI = 158-160), 142 (95% CI = 140-144), and 86 (95% CI = 80-92) months for those with mild, moderate, and severe dysplasia, respectively. The maximum follow-up was 239 months. The average number of smears per study subject was 2.85 (95% CI = 2.82-2.88).
Age at dysplasia was calculated from the date of dysplasia minus the date of birth. Other variables derived from the data available on the cytology forms included the following: oral contraceptive use (ever/never); parity; time period (dichotomized as earlier than 1975 and 1975 onward); number of dysplastic smears; and whether the initial dysplasia was prevalent (on the first smear), incident (preceded by normal smears), or progressive (progressing from a less severe state other than normal).
For the analysis of progression and regression of mild and moderate cervical dysplasia, women in the total study file were eligible if they were diagnosed with mild or moderate dysplasia between the years 1962 and 1980 inclusive, with no prior diagnosis that was more severe, and were resident in metropolitan Toronto. Of the total file, 25 486 women were eligible, but 5242 (21%) records were subsequently eliminated because they did not have a full date of birth; even the internal record linkage would have been questionable for these records. Of these eliminated records, 85% had a first cervical screen prior to 1970. Of the remaining 20 244 women, 14 795 had a diagnosis of mild dysplasia and 6414 had a diagnosis of moderate dysplasia (a total of 21 209 dysplasia diagnoses). Among these women, 974 (5%) had been diagnosed in both dysplasia states and contributed two records to the analysis. Records were censored at treatment or at the end of the screening history. Since the outcome had to occur after the start date, records were excluded if they had only one Pap smear result or if the outcome of interest or censoring occurred at the first dysplastic smear. This reduced the number of mild dysplasias to 7880 and moderate dysplasias to 4123.
|
Note on Terminology |
|---|
|
TopAbstractIntroductionMaterials and MethodsNote on TerminologyAnalysisResultsDiscussionReferences |
|---|
There are several different systems for classifying cervical cytology, all derived from the original Papanicolaou system (10). Until recently, the most commonly used classification scheme in Ontario was based on the WHO system (7). This system divides cervical dysplasia into mild, moderate, and severe dysplasia and has a separate category for cervical cancer in situ (CIS). In North America, the system of Richart (11) was commonly used until recently, and is still often used for histologic classification. This classifies lesions into three grades of cervical intraepithelial neoplasia (CIN). CIN1 corresponds to mild dysplasia, CIN2 to moderate dysplasia, and CIN3 to severe dysplasia and carcinoma in situ combined. The more recent Bethesda system (8) is designed to provide simplification of cytologic diagnoses. In this system, lesions with CIN1 are classified as low-grade squamous intraepithelial lesions (LSIL) and lesions with CIN2 or CIN3 are combined as high-grade squamous intraepithelial lesions. The latter two systems implicitly assume the same prognosis for the combined states, which makes it impossible to determine whether they have different progression and regression probabilities. For the present study, therefore, the original (WHO) terminology was retained. Comments included within the classification of moderate dysplasia in this laboratory ranged from mild to moderate to moderate/highly suspicious.
|
Analysis |
|---|
|
TopAbstractIntroductionMaterials and MethodsNote on TerminologyAnalysisResultsDiscussionReferences |
|---|
Initial analysis used the actuarial life table method to calculate progression or regression rates. Weighted regression was used to test for trend in annual conditional rates (12). Cox's regression (13) was used to estimate the relative risk of progression and regression for the different dysplastic states, adjusting for covariates, and changes in the relative risk with time were tested by use of a time-dependent covariate. The covariates themselves were also tested as risk factors for progression and regression, stratified by dysplastic state. Survival estimates were derived from Cox's model based on different degrees of dysplasia and specific values for covariates.
Some apparent progression from mild, moderate, or severe dysplasia to cervical cancer was probably due to an undercall of the initial smear. In particular, cancer diagnosed within a few months of a previous or noncancerous smear may have been present at the time of the smear. An arbitrary cutoff was set to differentiate incident from prevalent cancers. On examining the hazard of CIS or worse (i.e., including invasive cervical cancer) for severe dysplasias over the first 12 months from the initial severe smear, a sharp change in the slope of the line was found at 4 months. Records with outcomes occurring within 4 months of an initial dysplasia were therefore excluded from the analysis.
Women who had more than one dysplasia state contributed more than one record to the analysis. However, this only applied to 5%-7% of eligible women and is unlikely to have introduced a bias.
All tests of significance were two-sided. Approximately seven variables were tested as risk factors in 10 separate analyses. To reduce the possibility of associations appearing significant through chance alone, a Bonferroni correction was used (14) and only P values less than .00074 were considered significant. However, for the covariates apart from dysplastic state, the interest in this study was not so much in individual tests of significance as in seeing consistent patterns of association in the different analyses.
|
RESULTS |
|---|
|
TopAbstractIntroductionMaterials and MethodsNote on TerminologyAnalysisResultsDiscussionReferences |
|---|
The percentage of subjects undergoing treatment during a 2-year period from the time of their first dysplastic smear was 4% for those with mild dysplasia, 23% for those with moderate dysplasia, and 70% for those with severe dysplasia. Between the two time periods (from 1962 through 1974 and from 1975 through 1980), treatment of mild dysplasia increased from 3.6% to 5.3%, treatment of moderate dysplasia increased from 18.6% to 33.3%, and treatment of severe dysplasia increased from 69.1% to 75.9%. This confirmed our initial impression that the management of cervical dysplasia had been conservative at the time, particularly prior to 1975.
Relative Risk of Progression to CIS and Invasive Cervical Cancer
The number of cervical cancer outcome events detected in the course
of the study follow-up period and the corresponding rates are presented
in Table 1.
For severe dysplasia, most of those who
progressed to CIS or worse did so in the first 2 years. The relative
risk of progression to CIS or worse for moderate versus mild dysplasia
and severe versus mild dysplasia was tested for outcomes occurring from
greater than 4-24 months after a dysplastic smear and for outcomes
occurring beyond 2 years (Table 2). Up to 24 months
from an initial observation of dysplasia, moderate dysplasia progressed
eight times more often than did mild dysplasia, and severe dysplasia
progressed 22 times more often. From 2 years onward, the relative risks
of progression were at a constant, lower level. Overall, moderate
dysplasia was 2.5 times more likely to progress to CIS than was mild
dysplasia, and severe dysplasia was four times more likely to progress
than was mild dysplasia. Relative risk estimates were similar, but less
precise, when the outcome was invasive cervical cancer (Table 2). These
estimates were not changed by adjustment for covariates.
|
|
Progression and Regression of Dysplasia Based on Cervical Cytology
Actuarial life table analysis showed that at 2 years, 2.1 per 100
(95% CI = 1.7-2.4) women with mild dysplasia and 16.3 per 100
(95% CI = 15.0-17.7) women with moderate dysplasia had progressed
to severe dysplasia or worse (Table 3). Progression
from mild to severe dysplasia was at a constant low rate of 1% per
year for up to 10 years. Progression from moderate to severe dysplasia
(and from mild to moderate) was highest in the first 2 years, and then
annual conditional rates decreased linearly with time over the next 8
years to 1.0 per 100 women in the tenth year. At 2 years, 29% of
mild dysplasia was still present or had progressed. Progression from
mild to moderate or worse was 11% (95% CI = 10%-12%)
in 2 years. At 10 years, over half of the women with either mild or
moderate dysplasia had not progressed to a more severe lesion.
|
Mild dysplasia was more likely to regress to normal than was moderate dysplasia, but the rates were of the same order of magnitude (Table 3
).
By 5 years, the annual rate of regression for mild dysplasia had fallen
to the annual rate of regression for moderate dysplasia, for regression
to either one or two normal smears. There was a greater chance of
regression to a single normal (possibly followed by recurrence) than to
two normal smears. The covariates that were significantly associated with a modestly decreased risk of progression over the total time period beyond 4 months were incident dysplasia, age at least 50 years, and zero parity (15). These variables also increased the risk of regression to a first normal smear from mild or moderate dysplasia by approximately 40%. The time period for diagnosis of the initial dysplasia (1962 through 1974 versus 1975 through 1980) was not associated with altered risk of progression or regression, except for the transition from mild to moderate dysplasia, for which there was a 61% (95% CI = 42%-84%) increased risk associated with the later period. Ever/never use of oral contraceptives was not significantly associated with altered risk of progression or regression. The covariates made little difference to the estimates of relative risk of progression or regression for different degrees of dysplasia, but they inevitably affected the estimates of absolute risk derived from a Cox's proportional hazards model.
|
DISCUSSION |
|---|
|
TopAbstractIntroductionMaterials and MethodsNote on TerminologyAnalysisResultsDiscussionReferences |
|---|
It is not known how representative the study participants were of all Toronto women in the late 1960s and 1970s, although Cyto-Pathology Associates was the largest cytology laboratory in Toronto at the time. We investigated a random sample of more than 100 of the nearly 1000 physicians who used the laboratory during this time period. Twenty-six percent of the sample were specialists in obstetrics and gynecology at the time of data collection, and they contributed 60% of the smears.
A selection bias for women with different degrees of dysplasia was unlikely, but there were differences in follow-up. The average length of time between cytology smears was 17 months for mild dysplasia and 13 months for women with either moderate or severe dysplasia. These differences would increase the apparent risk of progression for severe and moderate dysplasia relative to mild dysplasia. The effectiveness of linkage to the external files (cancer, hysterectomies, and death) was unlikely to vary by grade of dysplasia and probably introduced no bias in the relative risks. Misclassification of cytology results is likely to have reduced the relative risks. Risk factors tested in this study (parity, age, number of dysplastic smears, and ever use of oral contraceptives) did not alter the relative risks of progression. However, factors, such as sexual activity, presence and type of human papillomavirus infection, and smoking, were not measured.
Censoring due to treatment was not available beyond the screening history, except for hysterectomies (from 1979 through 1989) or for hysterectomy related to endometrial cancer over the total time period. Therefore, rates of progression of dysplasia to CIS or worse may be underestimated. Observed deaths for the sample group were 36% less than expected, even for women for whom we had a full date of birth (standardized mortality ratio [SMR] = 0.64). However, the cohort of screened women may have been healthier than the general population of Ontario women on which the expected rates were based. In a simulation study, reducing the denominator according to the expected number of out-migrations (based on age, year, and out-of-province migration rates) only increased the SMR to 0.7.
Rates of progression and regression of dysplasia within the screening history were considered to be more accurate than rates of progression to cervical cancer derived from the external linkages. If smears were sent to another laboratory during the screening history, failure to account for some censoring may have slightly reduced some rates. Changes of state were assumed to occur when they were observed. Delay in finding initial lesions (increasing apparent rates) would probably have been greater than the delay in finding the outcome (decreasing apparent rates). Increased rate of progression in the first 2 years, relative to subsequent years, indicated that there was undercall of the original smear, particularly since the initial increase in rates was greatest for progression to the next level of severity. This is the major reason for the recommendation of the Canadian Workshop Group that the first repeat smear should be at 6 months rather than at 1 year (16). However, it is also possible that regression or progression will occur mostly within the first 2 years. Some residual false negatives may have been left after 2 years, but the majority of outcomes occurring after 2 years were probably due to true progression.
Compared with the majority of previous studies, the present study was unusual in that it used survival analysis to compare risks of cervical cancer by degree of cervical dysplasia, particularly allowing for changes in relative risk with time. A study that followed up histologically confirmed CIN1, CIN2, and CIN3 lesions for a maximum of 4 years found a relative risk of progression to CIS of 6 for CIN2 relative to CIN1 (17). A study of 268 mild and 138 moderately displastic lesions found by cytology observed 27 times the progression in the moderate as opposed to mild lesions at 3 years and a relative risk of 13 at 5 years (18). Comparison of rates of progression to CIS for CIN2 (22%) versus CIN1 (11%) in a meta-analysis are discrepant with our study, but the rates of progression reported in the meta-analysis were not necessarily of progression to CIS as reported (2), but sometimes any progression beyond the original state.
Table 4 summarizes rates of progression and
regression of dysplasia in other studies. Rates of progression both
from mild and from moderate dysplasia to severe dysplasia or worse in
our study were lower than those observed in previous studies.
Reanalyzing our data to include progression to events beyond treatment
had very little effect on the results. Approximately half of the
eligible women were removed from the analysis because the outcome or
final screen was not later than the first screen. However, the excluded
women were found to have half the risk of developing CIS compared with
the women who were included in the analysis of progression and
regression rates. Rates of progression from mild to moderate dysplasia
for women with high-risk covariates (25% at 2 years and 45% at
5 years) were close to those reported by others. The rates of
progression to severe dysplasia seen in our study were particularly
lower than in studies that used colposcopically directed biopsies to
determine the outcome (19,20,25,26,28). It may be that in our
study, some progressions which later regressed, especially of smaller
lesions, were undetected through cytology (30).
|
Regression rates were greater than progression rates for mild and moderate cytology smears. Regression to two normal smears appeared to be much less frequent than regression to one normal. There was a greater chance to be seen to progress to one normal, but some regression to one normal may have been followed by progression, or been a false-normal smear, possibly due to sampling variability. The rates of regression from both mild and moderate dysplasia were reasonably similar to rates reported in other studies, especially after 5 years (Table 4
). These results support the conclusion that regression from
moderate dysplasia is less than from mild dysplasia, but not
substantially less (Table 3). A recent report (31)
from a
large, population-based study in Denmark, over the same time period as
our study, also found that regression outweighed progression, even for
combined, confirmed CIN lesions. Other studies (28,32-34) have associated older age with decreased risk of progression of cervical dysplasia. Diagnostic artefact was an unlikely explanation for this observation in our study, given the length of follow-up. Parity has been associated with increased risk of cervical cancer (35) and dysplasias that persist have a higher risk of progression (19-21,36).
In conclusion, the risk of progression to cervical cancer for moderate dysplasia was intermediate between the lower risk for mild dysplasia and the greater risk for severe dysplasia. The majority of cases of mild dysplasia will regress to normal cytology, as will approximately half of those with moderate dysplasia, and most of these regressions occur during the first 2 years after diagnosis of the dysplastic smear. There is, therefore, prognostic value in the classification of moderate dysplasia separate from severe dysplasia. The recommendation by the Canadian National Workshop on Cervical Screening (16) for cytologic surveillance rather than immediate referral to colposcopy for those with a cytologic diagnosis of mild dysplasia (CIN1 or LSIL) is supported by this study.
|
NOTES |
|---|
Supported in part by the Ontario Ministry of Health Research grant 04081 for funding the record linkage and analysis phases of this project, the National Cancer Institute of Canada for the initiation of the study and the entry of the cytopathology data, and Health Research Personnel Development Program fellowship 03037 (P. Holowaty).
We thank Dr. Don Thompson who was responsible for the very high quality of the original cytology data, Dr. Eric Holowaty and his staff who performed the external record linkages, and Andrew White who was invaluable as a computer and programming consultant for this project.
|
REFERENCES |
|---|
|
TopAbstractIntroductionMaterials and MethodsNote on TerminologyAnalysisResultsDiscussionReferences |
|---|
1 Chanen W. The CIN saga—the biological and clinical significance of cervical intrepithelial neoplasia. Aust N Z Obstet Gyneacol 1990;30:18-23.
2 Ostor AG. Natural history of cervical intraepithelial neoplasia: a critical review. Int J Gynecol Pathol 1993;12:186-92.[ISI][Medline]cancerlit;93216377
3 Boyes DA, Morrison B, Knox EG, Draper GJ, Miller AB. A cohort study of cervical cancer screening in British Columbia. Clin Invest Med 1982;5:1-29.[ISI][Medline]cancerlit;83002484
4 Narod SA, Thompson MJ, Wall C, Green LM, Miller AB. Dysplasia and the natural history of cervical cancer: early results of the Toronto Cohort Study. Eur J Cancer 1991;27:1411-6.cancerlit;92075373
5 Task force on cervical cancer screening programs. Conclusions and recommendations of the task force. Can Med Assoc J 1976;114:1003-33.[Medline]
6 Cervical cancer screening programs 1982. Ottawa; Health Services and Promotion Branch, Department of National Health and Welfare; 1982. p. 31.
7 Riotton G, Christopherson WM. Cytology of the female genital tract. In: World Health Organization, editor. International histological classification of tumors. Geneva (Switzerland): World Health Organization; 1973.
8 National Cancer Institute Workshop. The 1988 Bethesda System for reporting cervical/vaginal cytological diagnosis. JAMA 1989;262: 931-4.[CrossRef][ISI][Medline]cancerlit;92686453
9 Howe GR, Lindsay J. A generalized iterative record linkage computer system for use in medical follow-up studies. Comput Biomed Res 1981;14:327-40.[CrossRef][ISI][Medline]
10 Koss LG. The Papanicolaou test for cervical cancer detection. A triumph and a tragedy.JAMA 1989;261:737-43.[Abstract]cancerlit;89095119
11 Richart RM. The patient with an abnormal Pap smear—screening techniques and management. N Engl J Med 1980;302:332-4.[ISI][Medline]cancerlit;80078033
12 Fleiss TL. Statistical methods for rates and proportions. New York (NY): John Wiley & Sons, 1981.
13 Cox DR, Oakes D. Analysis of survival data. Monographs on statistics and applied probability. London (U.K.): Chapman & Hall; 1984. p. 201.
14 Snedecor GW, Cochran WG. Statistical methods. 7th ed. Ames (IA): Iowa State University Press; 1980.
15 Holowaty PH. The Natural history of cervical dysplasia. Ph.D. thesis. University of Toronto; 1996.
16 Miller AB, Anderson G, Brisson J, Laidlaw J, Le Pitre N, Malcolmson P, et al. Report of a National Workshop on Screening for Cancer of the Cervix. CMAJ 1991;145:1301-25.[Medline]cancerlit;92034499
17 Galvin GA, Jones HW, Te Linde RW. The significance of basal-cell hyperactivity in cervical biopsies. Am J Obstet Gynecol 1955;70:808-21.[ISI][Medline]
18 Luthra UK, Prabhakar AK, Seth P, Agarwal SS, Murthy NS, Bhatnagar P, et al. Natural history of precancerous and early cancerous lesions of the uterine cervix. Acta Cytol 1987;31:226-34.[ISI][Medline]cancerlit;87237302
19 Jones MH, Jenkins D, Cuzick J, Wolfendale MR, Jones JJ, Balogun-Lynch C, et al. Mild cervical dyskaryosis: safety of cytological surveillance. Lancet 1992;339:1440-3.[CrossRef][ISI][Medline]cancerlit;92292763
20
Flannelly G, Anderson D, Kitchener HC, Mann EM, Campbell
M, Walker F, et al. Management of women with mild and moderate cervical dyskariosis. BMJ 1994;308:1399-403.
21 Brown MS, Phillips GL Jr. Management of the mildly abnormal Pap smear: a conservative approach. Gynecol Oncol 1985;22:149-53.[CrossRef][ISI][Medline]cancerlit;86031504
22 Robertson JH, Woodend BE, Crozier EH, Hutchinson J. Risk of cervical cancer associated with mild dyskaryosis. BMJ 1988;297:18-21.cancerlit;88310152
23
Mikhail MS, Merkatz IR, Romney SL. Clinical usefulness of
computerized colposcopy: image analysis and conservative management of mild dysplasia. Obstet Gynecol 1992;80: 5-8.
24
Syrjanen K, Kataja V, Yliskoski M, Chang F, Syrjanen S,
Saarikoski S. Natural history of cervical human papillomavirus lesions does not substantiate the
biologic relevance of the Bethesda system. Obstet Gynecol 1992;79(5 Pt 1):675-82.
25 Campion MJ, Singer A, Mitchell HS. Complacency in diagnosis of cervical cancer. Br Med J 1987;294:1337-9.cancerlit;87243348
26 Nasiell K, Roger V, Nasiell M. Behavior of mild cervical dysplasia during long-term follow-up. Obstet Gynecol 1986;67:665-9.[ISI][Medline]cancerlit;86175925
27 Carmichael JA, Maskens PD. Cervical dysplasia and human papillomavirus. Am J Obstet Gynecol 1989;160:916-8.[ISI][Medline]cancerlit;89225981
28
Nasiell K, Nasiell M, Vaclavinkova V. Behavior of moderate
cervical dysplasia during long-term follow-up. Obstet Gynecol 1983;61: 609-14.
29 Murthy NS, Sardana S, Narang N, Agarwal SS, Sharma S, Das DK. Biological behaviour of moderate dysplasia—a prospective study. Indian J Cancer 1996;33:24-30.[Medline]cancerlit;97216885
30 Giles JA, Hudson E, Crow J, Williams D, Walker P. Colposcopic assessment of the accuracy of cervical cytology screening. Br Med J 1988;296:1099-102.cancerlit;88241237
31 Bos AB, van Ballegooijen M, van Oortmarssen GJ, van Marle ME, Habbema JD, Lynge E. Non-progression of cervical intraepithelial neoplasia estimated from populationscreening data. Br J Cancer 1997;75: 124-30.[ISI][Medline]cancerlit;97153319
32 Kirby AJ, Spiegelhalter DJ, Day NE, Fenton L, Swanson K, Mann EM, et al. Conservative treatment of mild/moderate cervical dyskaryosis: long-term outcome. Lancet 1992;339:828-31.[CrossRef][ISI][Medline]
33 Montz FJ, Monk BJ, Fowler JM, Nguyen L. Natural history of the minimally abnormal Papanicolaou smear. Obstet Gynecol 1992;80(3 Pt 1):385-8.[ISI][Medline]
34 Stern E, Neely PM. Dysplasia of the uterine cervix: Incidence of regression, recurrence, and cancer. Cancer 1964;17:508-12.[CrossRef][ISI][Medline]
35
Brinton LA, Reeves WC, Brenes MM, Herrero R, de Britton
RC, Gaitan E, et al. Parity as a risk factor for cervical cancer. Am J Epidemiol 1989;130:486-96.
36 Hulka BS. Cytologic and histologic outcome following an atypical cervical smear. Am J Obstet Gynecol 1968;101:190-9.[ISI][Medline]
Manuscript received May 6, 1998; revised November 18, 1998; accepted December 3, 1998.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  M. Howard and A. Lytwyn The HPV vaccine: An analysis of the FUTURE II study Can Fam Physician, December 1, 2007; 53(12): 2157 - 2159. [Full Text] [PDF]
|
|
|
|
 |
|
 N. Van de Velde, M. Brisson, and M.-C. Boily Modeling Human Papillomavirus Vaccine Effectiveness: Quantifying the Impact of Parameter Uncertainty Am. J. Epidemiol., April 1, 2007; 165(7): 762 - 775. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Vinokurova, N. Wentzensen, J. Einenkel, R. Klaes, C. Ziegert, P. Melsheimer, H. Sartor, L.-C. Horn, M. Hockel, and M. von Knebel Doeberitz Clonal History of Papillomavirus-Induced Dysplasia in the Female Lower Genital Tract J Natl Cancer Inst, December 21, 2005; 97(24): 1816 - 1821. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. D. Blumenthal and L. Gaffikin Cervical Cancer Prevention: Making Programs More Appropriate and Pragmatic JAMA, November 2, 2005; 294(17): 2225 - 2228. [Full Text] [PDF]
|
|
|
|
 |
|
 W. J.M. Hrushesky, R. B. Sothern, W. J. Rietveld, J. Du Quiton, and M. E. Boon Season, Sun, Sex, and Cervical Cancer Cancer Epidemiol. Biomarkers Prev., August 1, 2005; 14(8): 1940 - 1947. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. L. Trimble, S. Piantadosi, P. Gravitt, B. Ronnett, E. Pizer, A. Elko, B. Wilgus, W. Yutzy, R. Daniel, K. Shah, et al. Spontaneous Regression of High-Grade Cervical Dysplasia: Effects of Human Papillomavirus Type and HLA Phenotype Clin. Cancer Res., July 1, 2005; 11(13): 4717 - 4723. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. M. Clifford, R. K. Rana, S. Franceschi, J. S. Smith, G. Gough, and J. M. Pimenta Human Papillomavirus Genotype Distribution in Low-Grade Cervical Lesions: Comparison by Geographic Region and with Cervical Cancer Cancer Epidemiol. Biomarkers Prev., May 1, 2005; 14(5): 1157 - 1164. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 N. B. Prasad, A. V. Biankin, N. Fukushima, A. Maitra, S. Dhara, A. G. Elkahloun, R. H. Hruban, M. Goggins, and S. D. Leach Gene Expression Profiles in Pancreatic Intraepithelial Neoplasia Reflect the Effects of Hedgehog Signaling on Pancreatic Ductal Epithelial Cells Cancer Res., March 1, 2005; 65(5): 1619 - 1626. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. J. Goldie, M. Kohli, D. Grima, M. C. Weinstein, T. C. Wright, F. X. Bosch, and E. Franco Projected Clinical Benefits and Cost-effectiveness of a Human Papillomavirus 16/18 Vaccine J Natl Cancer Inst, April 21, 2004; 96(8): 604 - 615. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. J. Goldie, J. J. Kim, and T. C. Wright Cost-Effectiveness of Human Papillomavirus DNA Testing for Cervical Cancer Screening in Women Aged 30 Years or More Obstet. Gynecol., April 1, 2004; 103(4): 619 - 631. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
N. F. Schlecht, R. W. Platt, A. Negassa, E. Duarte-Franco, T. E. Rohan, A. Ferenczy, L. L. Villa, and E. L. Franco Modeling the Time Dependence of the Association between Human Papillomavirus Infection and Cervical Cancer Precursor Lesions Am. J. Epidemiol., November 1, 2003; 158(9): 878 - 886. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. F. Sawaya, K. J. McConnell, S. L. Kulasingam, H. W. Lawson, K. Kerlikowske, J. Melnikow, N. C. Lee, G. Gildengorin, E. R. Myers, and A. E. Washington Risk of Cervical Cancer Associated with Extending the Interval between Cervical-Cancer Screenings N. Engl. J. Med., October 16, 2003; 349(16): 1501 - 1509. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Goodman and D. C. Wilbur Case 32-2003 - A 37-Year-Old Woman with Atypical Squamous Cells on a Papanicolaou Smear N. Engl. J. Med., October 16, 2003; 349(16): 1555 - 1564. [Full Text] [PDF]
|
|
|
|
|
|
D. Cha, K. Khosrotehrani, Y. Kim, H. Stroh, D. W. Bianchi, and K. L. Johnson Cervical Cancer and Microchimerism Obstet. Gynecol., October 1, 2003; 102(4): 774 - 781. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
N. F. Schlecht, R. W. Platt, E. Duarte-Franco, M. C. Costa, J. P. Sobrinho, J. C. M. Prado, A. Ferenczy, T. E. Rohan, L. L. Villa, and E. L. Franco Human Papillomavirus Infection and Time to Progression and Regression of Cervical Intraepithelial Neoplasia J Natl Cancer Inst, September 3, 2003; 95(17): 1336 - 1343. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. J. Weissenborn, A. M. Funke, M. Hellmich, P. Mallmann, P. G. Fuchs, H. J. Pfister, and U. Wieland Oncogenic Human Papillomavirus DNA Loads in Human Immunodeficiency Virus-Positive Women with High-Grade Cervical Lesions Are Strongly Elevated J. Clin. Microbiol., June 1, 2003; 41(6): 2763 - 2767. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A E Raffle, B Alden, M Quinn, P J Babb, and M T Brett Outcomes of screening to prevent cancer: analysis of cumulative incidence of cervical abnormality and modelling of cases and deaths prevented BMJ, April 26, 2003; 326(7395): 901 - 901. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E. M. Burd Human Papillomavirus and Cervical Cancer Clin. Microbiol. Rev., January 1, 2003; 16(1): 1 - 17. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. Saslow, C. D. Runowicz, D. Solomon, A.-B. Moscicki, R. A. Smith, H. J. Eyre, and C. Cohen American Cancer Society Guideline for the Early Detection of Cervical Neoplasia and Cancer CA Cancer J Clin, November 1, 2002; 52(6): 342 - 362. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. S. Mandelblatt, W. F. Lawrence, L. Gaffikin, K. K. Limpahayom, P. Lumbiganon, S. Warakamin, J. King, B. Yi, P. Ringers, and P. D. Blumenthal Costs and Benefits of Different Strategies to Screen for Cervical Cancer in Less-Developed Countries J Natl Cancer Inst, October 2, 2002; 94(19): 1469 - 1483. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. S. Mandelblatt, W. F. Lawrence, S. M. Womack, D. Jacobson, B. Yi, Y.-t. Hwang, K. Gold, J. Barter, and K. Shah Benefits and Costs of Using HPV Testing to Screen for Cervical Cancer JAMA, May 8, 2002; 287(18): 2372 - 2381. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. W. Sellors, J. Jeronimo, R. Sankaranarayanan, T. C. Wright, M. Howard, and P. D. Blumenthal Assessment of the Cervix After Acetic Acid Wash: Inter-Rater Agreement Using Photographs Obstet. Gynecol., April 1, 2002; 99(4): 635 - 640. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. J. Goldie, L. Kuhn, L. Denny, A. Pollack, and T. C. Wright Policy Analysis of Cervical Cancer Screening Strategies in Low-Resource Settings: Clinical Benefits and Cost-effectiveness JAMA, June 27, 2001; 285(24): 3107 - 3115. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E. L. Franco, E. Duarte-Franco, and A. Ferenczy Cervical cancer: epidemiology, prevention and the role of human pa |