|
July 2005, Volume 27, No. 7
|
Original Articles
|
Prevalence of ASCUS and age as a predictor of outcome in Hong Kong Chinese womenHing-Ming Hung 孔慶明 HK Pract 2005;27:253-260 Summary
Objective: To estimate the prevalence of ASCUS in local Chinese
women and to test whether age is a significant predictor of outcome in women with
ASCUS.
Keywords: ASCUS, cervical dysplasia, age, colposcopy, cost-effectiveness 摘要
目的： 1.估計本地中國女性ASCUS的發病率；2.測試年齡是否可以做預測ASCUS預後的明確指標。
Introduction It has been well established that Pap smears have significantly reduced the incidence of squamous cell carcinomas of the uterine cervix in developed countries.1,2 However, cervical cancer is still one of the most common cancers among women worldwide and the leading cause of death due to cancer in many developing and third world countries.3,4 Early detection of the precursors of cervical cancer by cervical smears and their treatment are essential factors in the prevention of this type of cancer. 5 Minor cytologic abnormalities ASCUS (Atypical Squamous Cells of Undetermined Significance) are the most frequent abnormalities observed in smears performed in screening for cervical cancer.5 Many studies have demonstrated that ASCUS is a predictor of high-grade lesions,6-10 and therefore a good marker for detecting underlying squamous intraepithelial lesion. Moreover, ASCUS may be the only abnormality preceding a diagnosis of HSIL (High-grade Squamous Intraepithelial Lesion).11 In a prospective study of more than 46,000 women, Kinney et al12 reported that of all cases of biopsy-proven HSIL, ASCUS was the most frequently associated cytologic diagnosis, accounting for 38% of the cases. Follow-up of women with ASCUS are therefore very important in order not to miss treatable diseases. The association of ASCUS with patient's age has seldom been investigated.13 Some authors concluded that age did not seem to influence the probability of finding high-grade lesions in women with minor cytologic abnormalities.14-15 However, there are more studies which suggest that age is one of the important predictors of dysplasia in women with ASCUS16-23 and should be considered as an independent factor in any plan of management. In recent years it has become evident that the cost and morbidity associated with the detection and treatment of low-grade cervical lesions, including ASCUS, have escalated, probably with marginal benefit of cancer prevention.24-26 As a result of screening, many women are diagnosed as having equivocal cytologic abnormalities, including ASCUS.27 Managing this huge number of clients each year is a significant clinical and public health problem.28 Determining which women with cytologic abnormalities are at risk for significant cervical disease, performing appropriate diagnostic work-ups and treating cancer precursors present a major public health challenge.29 After an abnormal result is issued for a Pap smear, the clinician then needs to decide when (e.g. at which point on the diagnostic spectrum) further evaluation is indicated. If the threshold for such a work-up is low, significant lesions are less likely to be missed and more individuals will have their lesions detected at an early stage. However, at the same time, more women are likely to have unnecessary diagnostic tests for either false-positive results or very early lesions that do not require treatment. Clinicians who prefer immediate colposcopy argue that some women with ASCUS actually have HSIL or even invasive cancer.30-33 Clinicians who prefer to use repeat cytology argue that most women with ASCUS have either no lesion or LSIL (Low-grade Squamous Intraepithelial Lesion), which is likely to regress in the absence of any treatment.34 Although colposcopy is a low morbidity procedure, it is a costly component of cervical screening. Ideally, patients likely to have histological SIL, especially high-grade, would be referred, whereas those unlikely to have it would be followed up without colposcopy.35 Cervical cancer affects about 500 women and causes the death of 150 women (4.6 per 1,000,000 population) in Hong Kong each year. Despite a decrease in age standardized incidence, it is the 4th commonest malignancy among women older than 18 years and ranks 7th as a cause of cancer death in females in Hong Kong. The lifetime probability of developing the disease is 1 in 72 (1.4%).36 However, we cannot find local data on the prevalence of ASCUS in the Chinese population. According to the Guidelines on the management of an abnormal cervical smear published by the Hong Kong College of Obstetricians and Gynaecologists,37 the suggested follow-up action for women with ASCUS is to repeat Pap smear in 3-6 months and refer for colposcopy if the abnormality persists. They quoted that the 1996 Hong Kong colposcopy data showed 14% of cases with persistent ASCUS had CIN II-III (equivalent to HSIL). To combat this common cancer among women in Hong Kong, the Department of Health of the Hong Kong Government SAR has started a population-based cervical cancer screening programme in 2004. From a public health perspective this is a much welcome initiative from the government. However much needs to be done, including a consensus among the health service providers to offer an efficient, cost-effective and evidence-based management strategy to handle the greatly increased number of women who are found to have ASCUS. If everyone with ASCUS is to be followed up with colposcopy, it is going to be a big financial burden to the community, whether it is to be paid by the clients or funded by government. It is hoped that this study will help us to know more of the prevalence of ASCUS in Hong Kong and the magnitude of its problem. The study is also expected to give evidence to support the use of age as a significant predictor of the outcome of ASCUS so that clinicians can use age as a criteria to decide whether prompt colposcpy and intervention is the best evidence-based recommendation. Methodology This study is a cross-sectional descriptive study, which is well suited for a study on the prevalence of a disease. It is divided into two parts. All Chinese women who had Pap smears done in Evangel Hospital in year 1999 were included in the first part of the study. The author conducted a retrospective review of the cytologic diagnosis of all the records. Pregnant and post-partum women, women with previous hysterectomy or ablative therapy to the cervix, women with history of histologically proven cervical dysplasia, or infection with human immuno-deficiency virus and Pap smear reported as unsatisfactory were excluded from the study. Evangel Hospital is a local community hospital with a strong emphasis and tradition in family medicine. Women who come for Pap smear are mainly self-referred. The hospital promotes preventive care. Therefore we can assume that the study sample by and large is quite representative of the Chinese female population in Hong Kong. The following variables were recorded for analysis: age at the time of Pap smear, result of the first Pap smear, follow-up method by repeat Pap smear or colposcopy, result of the follow-up. Result of the first Pap smear is the major outcome being studied. Frequencies of different types of results, viz., normal, ASCUS, LSIL, HSIL, and cancer were analysed using SPSS. The second part of the study included all Chinese women who were detected to have ASCUS in their Pap smears in the year 1998 to 2001 in Evangel Hospital. The same exclusion criteria used in the first part of the study were applied to this group. Women with ASCUS but lost to follow-up were excluded from the study because we could not ascertain their outcome. As such there is a selection bias (being a non-response bias) in the second part of the study. It was noted that 77% of women with ASCUS came purely for screening and did not have abnormal symptoms. The same variables used in the first part of the study were recorded for analysis. A few other variables, viz., marital status, parity, use of oral contraceptive pill or hormone replacement therapy, smoking status, record of sexually transmitted diseases and/ or multiple sexual partners, were also recorded for analysis. The major outcome being studied is the result of the follow-up, either by repeat Pap smear or by colposcopy. For women followed by cytology only, the outcome diagnosis was determined by the most recent Pap smear. For women who underwent colposcopic evaluation, the outcome diagnosis was the more significant of either the biopsy or Pap smear taken at the time of colposcopy. The cases were separated into two age groups, one group younger than or equal to 45 and the second group older than 45. The outcomes of the two groups were compared by the Chi-squared test and tested for significance of association with age. Linear regression was used to confirm the findings from the Chi-squared tests. Similarly, the Chi-squared test was used to test the significance of association of other variables with the outcome. Linear regression was used to confirm the findings from the Chi-squared tests. Univariate analysis of variance (ANOVA) was used to test the variability and interaction between the variables. Results In the first part of the study there were 2677 valid cases. The minimum age of the entire group was 18 years and the maximum age 84 years. The mean age was 42 (42.74) years, with standard deviation of 11.41. The prevalence of ASCUS in the present sample of 2677 clients was 2.4%. The ASCUS / SIL ratio, which is a useful laboratory quality indicator, of the present sample was 1.05 [63 / (34+26)]. In the second part of the study there were 165 cases of ASCUS, of which there were 139 valid cases with known results of follow-up. The minimum age of this second group (valid cases) was 22 and the maximum age was 72. The mean age was 43 (43.64) with standard deviation of 9.87. For easy understanding, in the following presentation of the result of analysis using Chi-squares tests, LSIL and HSIL are combined into a single category of SIL. From this analysis, women younger than or equal to 45 years were 2.1 times more likely to have SIL than those older than 45 years (p=0.037). This is statistically significant. Although the number of nulliparous women and the number of women using hormone (oral contraceptive pills or hormone replacement therapy) were small (14% and 7% respectively), Chi-squared tests were still done to test the association of outcome with these two variables. In both cases the results are not statistically significant. Analysis of Variance (ANOVA) was used to test for any significant interaction effect between these variables. Results show that the interaction between age and parity, and between age and use of hormone are both insignificant. 69 cases (49.6%) did not have their smoking status recorded in their case notes. There was no data of sexually transmitted diseases and only in two cases were multiple sexual partners being mentioned. Because of this lack of data, no attempt was made to analyse the association of smoking, sexually transmitted disease, and multiple sexual partners with the outcomes and their interaction with age of the patients. Discussion The prevalence of ASCUS in Chinese women of the present sample was 2.4%. This figure is within the range of the prevalence of 1.2% to 6% as reviewed in other similar studies.30,38-40 Our figure is on the lower side of the range. The ASCUS / SIL ratio of the present study was 1.05 which is also low compared with other studies.24,41 A low ASCUS/SIL ratio means that the cytopathologist(s) who reported the Pap smears tended to be conservative and avoided overuse of ASCUS in reporting. This may partially account for the low prevalence rate in the present study. In line with the results of most of the other similar studies,16-22 our analysis also confirmed that age was a significant predictor of dysplasia in women with ASCUS. In previous studies different age ranges were used for analysis. Some used 35, some used 45, and some used 50 as the cut-off. This study used 45 as an arbitrary cut-off, mainly because of the age distribution in our study sample. If we used 35 as the cut-off, there were only 27 cases in our sample and the number was too small for analysis. As there is no published data on the prevalence and outcome of women with ASCUS in the local Chinese female population, the present study with a large sample size of 2677 gave an evidence-based estimation of the prevalence of ASCUS in local Chinese women. Although the rate of 2.4% is not alarming, with the launching of the population-based cervical cancer screening programme in Hong Kong, a large number of women will be diagnosed to have ASCUS on their Pap smears. An efficient, cost-effective and evidence-based management strategy is needed to allow the clinicians and health service providers to advise these women on management options. The information is also very important to the policy makers for them to plan wisely on the proper use of our limited resources. The results indicated that older women (in this study 45 was used as the arbitrary cut-off) with ASCUS were less likely to have SIL than the younger age group. If they can accept the expectant approach, we can offer them the choice to repeat the Pap smear after three months instead of immediate colposcopy. Only when the repeat Pap smear shows persistent ASCUS or higher-grade lesions, then one needs to recommend colposcopy. Of course, the clinician needs to take into consideration the clinical history, presence of risk factors, pattern of practices and patient preference.42 One may be concerned with the actual risk of having SIL in the older age group if colposcopy is deferred. In this study the risk of having SIL in the age group older than 45 was as high as 31.5%. However, only one among the 17 cases with SIL has HSIL and cervical dysplasia is a very slow process. It should not cause real harm if the patient is comfortable with the expectant approach. The more important consideration in the older age group is the identification of specific risk factors, such as smokers and multiple sexual partners. These women with risk factors should be advised to consider immediate colposcopy instead of repeating cytology at a later date. When considering resource implications, the main difference between cytologic follow-up and immediate referral is the cost of colposcopy. There is a significant cost saving if colposcopy is not required. However this apparent saving must be interpreted in the light of two considerations. The first concerns the accuracy of the conventional Pap smear screening. There is an irreducible false negative rate associated with cervical cytology that is created by errors in sampling, screening, and interpretation.43 Pap smear will not detect all cases of dysplasia. There is the possibility that some of the women being followed with repeat cytology and reported as having normal Pap smear might have an unrecognized lesion. However, continued cytologic surveillance should identify those women with significant disease.8,42 Therefore one should not be satisfied with a single normal Pap smear on cytologic follow-up for women with ASCUS and there should be further cytologic surveillance. As recommended by the National Cancer Institute Working Group, the surveillance of women with ASCUS has to be carried out with smears repeated every four to six months for two years.24 The second and more worrisome consideration is the fact that 17.5% of the women with ASCUS in the study sample did not return for follow-up. Being a private hospital, clients have to pay for the service and this may account for the high default rate. In a previous similar study the defaulted rate was as high as 30%.42 To improve patient compliance the set-up of cervical screening registries and recall systems is essential to a safe implementation of recommending repeat cytology to follow-up women with ASCUS in low-risk groups. Age should not be the only factor in considering referral for colposcopy in women with ASCUS. The American College of Obstetricians and Gynecologists suggests that women with ASCUS and with the presence of high risk factors, such as human papillomavirus infection, cigarette smoking, or multiple sexual partners, should be referred for colposcopy.44 In this study, some of these risk factors have been included as variables. However there were insufficient data on smoking status and multiple sexual partners to allow for meaningful analysis. Other risk factors have been mentioned in literature45 to be associated with developing cervical cancer. Coexistence of these risk factors with ASCUS should call for an early colposcopic referral. Married women are at greater risk than singles. In this study all the clients were married or already having a sexual relationship. Multiparity carries a higher risk than nulliparity. In this study, parity was not found to be a significant predictor of the outcome of ASCUS. However, it should be noted that there were only twenty nulliparous women (14.4% of the sample) compared with the multiparous women in this study sample, and we should not make inference from this small sample size. Use of hormone is also considered a risk factor. In this study, again we could not find a significant association between use of hormone and outcome of ASCUS. However there is a similar limitation in that only ten women (7.2% of the sample) were taking hormone and we should not make inference from this small sample size. The use of ASCUS as a diagnostic category remains a source of debate and controversy. It has been argued that the use of ASCUS is an incomplete evaluation of Pap smear and should be replaced by a more definitive diagnosis.46 Uncertainty in the ASCUS category eventually may be resolved by the use of intermediate triage tests to identify women at high-risk for high-grade lesions and carcinoma. Studies evaluating the usefulness of new modalities in the clarification of ASCUS are ongoing. These new modalities include HPV DNA testing, cervicography and ancillary new technologies for Pap smear collection and screening.11,27,29,37,48-49 From a public health perspective we may not be able to do the triage tests for every woman with ASCUS because of the financial constraint; however, if the tests are used for the older age group, the clinicians can be more certain that the high-risk group has been excluded and recommend repeat cytology to the low-risk group. Conclusion This study has succeeded in collecting a large sample of local Chinese women, and most of them presented themselves for routine Pap smears. The prevalence of ASCUS was found to be 2.4%. The ASCUS / SIL ratio was 1.05, which suggests that the cytopathologists have been quite conservative in their interpretation of Pap smears. Therefore, the prevalence rate may be under-estimated. The study also confirmed that age is a significant predictor of dysplasia in women with ASCUS. Our analysis suggests that women who are above 45 years of age presenting with ASCUS have a sufficiently low rate of SIL, and in particular HSIL(only 1 in 54 in our study sample), that immediate colposcopy might not be necessary. Exclusion of these women from immediate colposcopy might save significant costs. When we recommend repeat Pap smears for the follow-up of women with ASCUS in the age group older than 45 and who are not at high-risk for high-grade lesions, there are three concerns which need to be addressed to ensure the safety of this approach.
Addressing these issues through public and physician education is necessary to maximize the efficiency of cervical cytology screening programmes, and to ensure the proper management of women diagnosed of having ASCUS. With these understandings family physicians should be able to take on a strong role in the management of women with ASCUS. After consideration of risk factors we can be confident to recommend an expectant approach to women with ASCUS who are above 45 years of age. If they are not over anxious, we can ask them to come back for repeat Pap smear after 3 months. For women with ASCUS who are under 45 years of age, we should encourage them to consider an early colposcopy. As there are inherent limitations with the retrospective cohort design of this study, it is recommended to plan a prospective study so that the association of ASCUS with risk factors, such as smoking, multiple sexual partners, presence of human papillomavirus, in the Chinese female population can be better ascertained. The significance of parity, use of hormone and menopause can also be studied in greater depth. If the unfavourable factor of lost to follow-up cannot be eliminated, the demographic characteristics of the "defaulters" can be compared with those with known outcomes to see whether the results can be generalized to all women with ASCUS. Key messages
Hing-Ming Hung, MBBS(HK), MPH(CUHK)
Hospital Superintendent, Evangel Hospital. Correspondence to : Dr Hing-Ming Hung, Evangel Hospital, 222 Argyle Street, Kowloon.
References
|
|