Abstract
During the present meeting, the participants reviewed the state of the art on screening for cervical cancer, and considered a number of promising new developments. Four key areas comprised a major focus of the discussions, programmatic issues, cervical cytology, alternative methods of screening and diagnosis and treatment of dysplasia. Worldwide, cervical cancer comprises approximately 12% of all cancers in women. It is the 2nd most common cancer in women worldwide and the most common in developing countries. The global estimates are for 452,000 new cases and more than 234,000 deaths from cervical cancer each year around 2000 (Ferlay et al., 1998; Parkin et al., 1999). Eighty percent of these cases are found in developing countries with only 5% of global cancer resources. Cervical cancer is both preventable and curable. Widespread comprehensive cervical cancer control programs have helped some developed countries to achieve up to an 80% reduction in incidence and mortality from cervical cancer. In developing countries, 60–80% of cases are seen in advanced stages (III and IV), if ever diagnosed, with a low probability of long-term survival (Sankaranarayanan et al., 1998a). In most developing countries, cytological screening is not currently available for population-based screening programs, and will not be possible for several decades. In a few countries where it is available, it is of suboptimal quality. Several studies are being conducted worldwide to evaluate alternative, feasible techniques for cervical cancer in comparison to cytology. Consideration of these studies comprised one of the main features of this consensus conference. The participants considered cervical screening projects and programs from Brazil, Chile, Finland, India, Iran, South Africa, Sri Lanka, Tunisia and Zimbabwe. These programs are in highly varied stages of evolution (from pilot studies to highly successful nationwide efforts) from areas of the world with major differences in incidence rates of cervical cancer and varied levels of economic development. The program in Finland is the model for organized programs of screening by cervical cytology worldwide. This program began in 1960 with 5 yearly screening for women age 30 to 59, invited through the national population register (Hakama and Louhivuori, 1988). Considerable attention was devoted to education of the population, fast feedback of screening results to women, a cost-effective system for referral of women with abnormalities, histological confirmation of diagnoses and continuous quality control (Syrjänen, 1995; Hristova and Hakama, 1997). Recent data show much greater efficacy following smears conducted in the organized program than opportunistic smears. The programs under development in Tunisia and Iran illustrate the relatively low risk of cervical cancer in Islamic populations. Nevertheless, in Tunisia, cervical cancer is the second most common cancer in women, therefore mass screening for women over the age of 35 is being developed. In Iran, a national program to screen by cytology all married females between 20 and 65 years every 3 years was proposed. This achieved a coverage of 61% in urban areas in 1997 but only 16% in rural areas. It is proposed to change the program to start at 25 or 30 years of age, and to screen every 5 years. Further development of organized approaches should help to reduce excessive expenditure on screening low-risk women, with little impact on the disease. In contrast, the experience in high incidence countries in Africa and South America confirms the need to ensure that those at risk are incorporated into programs. In many such settings, alternative approaches to screening must be evaluated if a major impact on the disease is to be achieved. In South Africa, where in some settings the sensitivity of the Pap smear is low, alternative approaches such as visual inspection with acetic acid (VIA), speculoscopy and human papilloma virus (HPV) DNA testing are being evaluated. Data from several studies suggest that most of these approaches have similar sensitivities to cytology. Similar data are being obtained in a study in Brazil. However, in South Africa, VIA has lower specificity than cytology. Further, in India and Zimbabwe, although studies have shown that VIA performed substantially better at identifying true disease than a Pap smear, the reverse was true for specificity (Chirenje et al., 1999; Sankaranarayanan et al., 1998b, Sankaranarayanan et al., 1999). This has substantial implications for treatment policies such as “see and treat” as there will be considerable overtreatment of women who in fact have a false positive finding. Combinations of screening tests or a simple immediately applied diagnostic test may help to resolve this. A once in a lifetime test would need to be of high sensitivity to be effective. Currently, an organized program involving 3 smears in a lifetime (starting at age 30) is being piloted as national policy in South Africa. Chile illustrates the difficulties in introducing organized programs. A national program was introduced in 1985, but initially appeared to have little impact. Therefore, concentration was devoted to the metropolitan area of Santiago. The lessons learnt there were extended to the rest of the country. Recently, an emerging impact in terms of declining incidence, improving proportions of cases diagnosed in stage I and declining mortality from the disease has been noted. Chile is, in practice, the only country in Latin America that has so far shown a decline in mortality from cancer of the cervix. Although there is every hope that a prophylactic vaccine against the oncogenic types of HPV will eradicate cervical cancer in the future, the likely time course for such a scenario to unfold will be prolonged. Once a “perfect” HPV prophylactic vaccine is synthesized, 3 years will be required for completion of all 3 phases of clinical trial. Another 2–4 years will be required to collect short-term results (i.e., prevention of HPV infection in vaccinated subjects). Information regarding the prevention of high-grade precancerous lesions (high-grade squamous intraepithelial neoplastic lesions or HSIL) in vaccinated individuals will not be available for 20–30 years after the completion of phase 3 clinical trials. Finally, acceptance and introduction of such a vaccine can be expected to take at least another 10 years once efficacy in prevention of HSIL has been established. It was noted in this context that the hepatitis B vaccine has been administered to less than 1% of children worldwide, despite the decades that have elapsed since the successful development of that vaccine. Several approaches to HPV DNA testing are available, including hybrid capture, PCR and in situ hybridization tests. Some are being piloted as alternatives to cervical cytology, or in terms of their clinical utility in the management of women with an abnormal Pap smear. The sensitivity of hybrid capture is very high for oncogenic types of HPV. Information regarding cervical cancer screening in women co-infected with human immunodeficiency virus (HIV) or other sexually transmitted diseases (STD) and HPV was also presented. Women with HIV infection tend to show high prevalence of both low-grade squamous intraepithelial neoplastic lesions (LSIL) and HSIL, and most of these test positive for infection with oncogenic types of HPV. Given what is known about the natural histories of cervical dysplasia and of HIV-induced disease, it is unclear whether screening for cervical cancer will improve health outcomes of women infected with HIV, particularly in light of the high rates of treatment failure for high-grade cervical disease in HIV-positive women. It was questioned whether cervical screening constitutes an effective allocation of scarce medical resources in areas of the world with high prevalence rates of HIV infection. Approaches are being made to improve the efficacy of cervical cytology. One promising approach is the liquid-based test (Thin Prep). In one evaluation in a developed country, a triage protocol involving a combination of the liquid-based test and an HPV DNA test has been evaluated. Attempts at automating cervical cytology reading are also encouraging in developed countries, in terms of reducing dependence on cytotechnicians and facilitating reduction of false negatives. However, there have been some difficulties in extending their use to developing countries, where maintenance of high technology approaches could prove to be difficult. The success of cervical cancer screening will depend on the program, from education to screening to acting on results, being acceptable to women themselves. Thus, the methods and activities of the program will have to respect the dignity, privacy and autonomy of women. It is more likely that this will be achieved if women and women's health and rights advocates are involved in the development of the program. The greatest burden of disease in terms of the rates of cervical cancer and the absolute number of women with cervical cancer occurs in poor countries and countries with economies in transition. Thus, mindful of this, minimal requirements are proposed as a priority to decreasing the number of women dying of cervical cancer internationally. The primary operational aim of the program must focus on achieving the highest possible coverage rate. To support this, indicators such as number of women screened, as opposed to number of Pap smears done, should be promoted. Implementation of programs should be undertaken when it is established that cervical cancer is an important disease within the country. The aim should be to set up rational national organized screening programs (Hakama et al., 1985). It is more costly, as it is less effective, to perform opportunistic screening than to choose a population-based approach. An incremental approach, involving a well running comprehensive demonstration program, including all the essential elements in an area of high prevalence, is advocated as a good method of starting. This should facilitate lobbying for national implementation and enable the organizers to learn from mistakes. An education program aimed to reach the target community. Education of service providers and, during basic and undergraduate training, of health professionals. Training of service providers of community education, smear taking and cytology reading and ongoing quality control of their work. Evaluation of people providing community education, smear taking and cytology reading. Creation of screening facilities. Designation of laboratories to provide cytological services, and a system to ensure quality control in these laboratories. Mechanisms for the processing of cytological smears and the return of results as rapidly as possible. Definition of referral mechanisms for patients. Creation and designation of treatment centers that can provide treatment for early lesions. Creation of a system for dealing with advanced disease, which in the worst case scenario would provide support and pain relief for women with untreatable disease. Creation of an information system that allows for evaluation of the program such that data can be analyzed at the local as well as the national level. Systems for rational implementation defined, including responsibilities for each level and equipment requirements at each level. Each of these elements is mutually dependant and essential to making an impact on cervical cancer incidence, morbidity and mortality. Together, they form the public health model of cervical screening (Fig. 1). Attention to each of these elements is essential prior to the initiation of a screening program. The notion of “first getting the house prepared and then inviting the guests” is a good guide for program planning. The public health model for cervical cancer screening. Community-based education is best done by people who have experience in this area and resources should be allocated to develop and test appropriate and effective methods of education to encourage wide coverage and uptake of services. Messages to promote good uptake aimed at both men and women should be considered. Special attention must be paid to reaching older women, the target group. Skills required for each level of service provision (educators, smear takers, smear reading, treatment, information system) should be defined and methods of ensuring that this is adequately done and ongoing monitoring of the quality of services must be put in place. Categories of staff to provide each of these services will be determined by local resources. It is clear from pilot programs that training, motivation and monitoring of the quality of the staff, rather than their formal qualifications, are important. Taking Pap smears, by nurse midwives, nurses, doctors, lay women. Reading Pap smears, by cytotechnicians under supervision of cytopathologists (for further detail, see “Cervical Cytology” below). Treatment, predominantly by doctors, although if local circumstances dictate by well-trained and supervised technicians or nurses. Special attention must be paid to training and other interventions required to provide friendly and welcoming services. Repeating smears to make up for lack of quality of the first smear is very costly. Therefore, investing in systems to ensure high-quality smear taking and smear reading is highly recommended. These should be located in such a way as to ensure maximum coverage of the population. So far, integrated services have been proven to be very successful in many countries. However, locally appropriate solutions should always be the guiding principle. Age is the most important risk factor in cervical cancer and screening should aim to target high-risk women. A good guide would be to look at the peak incidence of cervical cancer and begin screening at 5 years prior to this age. In most countries, this will be at about 30–35 years of age and can continue to about 60 years of age. Local resources will determine how national programs are developed. It is recommended that women who have never had a Pap smear before and who are older than 60 years of age should have 1 Pap smear only. Data indicate that women who have had 4 normal smears and are about 50 years of age are at very low risk of developing cervical cancer and can exit the screening program at this point. Countries should aim for at least 1 Pap smear per woman in her lifetime. Once this has been achieved, the aim should be 10-yearly smears. Five-yearly screens would be better, but only when resources suffice (WHO, 1986). Local resources will determine this for each country. The laboratory should ensure feedback to service providers on the adequacy rate of the smears taken. A network of laboratories, with a center of excellence, coordinating and monitoring quality should be set up (for further detail, see “Cervical Cytology” below). Data at least to monitor: coverage, prevalence by age group and treatment and follow up should be collected. Access to appropriate treatment and follow up are fundamental parts of the program and mechanisms to ensure their provision at the local level and to monitor this at the national level must be set up. A patient-held record card may be beneficial in achieving good follow up, especially in highly mobile communities. The above recommendations should also be seen in the context that: Primary prevention for cervical cancer should involve interventions that support later age at first intercourse; men having fewer sexual partners; women having control over their sexual behavior, particularly their ability to have sex when they want and with whom they want; that violence against women and children is significantly decreased. Cervical cancer interventions will, therefore, greatly benefit from initiatives in other areas; notably acquired immunodeficiency syndrome (AIDS) and STD intervention programs and women empowerment interventions that address gender inequality. Screening by cervical cytology has proven to be most successful in reducing the incidence and mortality from cervical cancer (Miller et al., 1990). All other approaches to control of cancer of the cervix must be evaluated in relation to this established approach. However, only 5% of women in developing countries, where approximately 80% of new cases of cervical cancer occur, have received a cervical cytology (Papanicolaou-Pap) smear. Therefore, one of the key issues before the INCGC must be to evaluate whether cervical control in these countries is best achieved by increasing the capacity for high-quality cervical cytology in developing countries. It is already clear that unless resources are relatively unlimited, only an organized program of cervical cytology screening can achieve the expected reduction in incidence and mortality from cancer of the cervix (Miller et al., 1990). To achieve such an outcome, attention has to be paid to a number of managerial issues (Miller, 1992). The following recommendations on training, equipment, quality control and practice guidelines are proposed to enable countries to establish a comprehensive and organized cytology laboratory service. Cytopathologists, screeners, laboratory personnel and smear collectors must be trained via hands-on experience with the aid of available reference manuals. A pathologist with clear leadership potential should be selected. This pathologist will be responsible for the initiation, organization and operation of a successful cytology laboratory. This individual should be trained at an established National Center of Excellence in Cervical Cytology with the expectation of subsequently becoming the trainer of additional cytopathologists. The period of training should probably last 6 months. Training must include cytopathology, cytotechnology and laboratory management. The training institution selected should be linguistically compatible and prepared to establish long-term links with the new laboratory. Suitable training institutions may be identified through the Internet by posting to appropriate listserv groups. Training may be supported by national or international governmental or non-governmental organizations, if appropriate applications are made. Formal training should be delayed until the laboratory approaches a screening volume of 15,000 slides per annum. Cytotechnologist skills will atrophy unless maintained by this work level. During the interval, slides will be screened by the cytopathologist. Training should follow similar guidelines as for cytopathologists. The period of training should be approximately 6 months. The cytotechnologist will be expected to train additional cytotechnologists. The International Federation of Cytology has an international qualification for cytotechnicians. The examination could play a role in ensuring the quality of smear reading. A laboratory technician responsible for Pap staining can be trained by a visiting laboratory technician who is expert in establishing and maintaining Pap staining protocols. All smear collectors must undergo hands-on training. For physicians, this should be provided in medical school. In countries where such training was not previously provided, special courses will have to be mounted. For other health professionals, especially nurses, special courses will also be required. A senior nurse could be appointed to provide directly supervised training at the individual clinic level, and to act as a quality control supervisor thereafter. Smears should be taken with a modified wooden Ayre's spatula (and cytobrush, if resources are available). Microscopes and computers may be obtained through donations. Specifically, equipment contributions may be solicited through written requests to the corporate contributions department or to the public relations department at the national or international corporate headquarters of the manufacturers whose equipment is sold locally. Centralize the laboratory facilities as far as possible. Laboratories should use a computerized data collection system and be able to integrate enrollment information, Pap smear results, biopsy results and follow-up information. Laboratories must ensure timely delivery of Pap smear reports to smear collectors. Limit workload to no more than 100 slides per day per cytotechnologist (60 per day is preferable). To keep skills adequate, a minimum number of Pap smears to be done annually in each laboratory should be defined (20,000–30,000 per annum is one recommendation). Maintain an up-to-date laboratory procedures manual. The policy of random 10% re-screening of negative smears should be abandoned and, when possible, replaced by participating in an externally organized quality control scheme. Smear collectors should transport cervical smear slides to the laboratory as soon as possible, and ensure they arrive in satisfactory condition. They should receive constant feedback regarding specimen adequacy, with the objective of keeping inadequate smears to the minimum. Consistent poor performers may be identified and counseled by ongoing monitoring of performance by review of computer records. Effective action plans include retraining, reassignment or disciplinary action. Cytotechnologists should undergo periodic re-testing every 6–12 months with standardized slide sets. Consistent poor performers may be identified by ongoing performance audits of computer records (e.g., tracking atypical rate) and counseled. Effective action plans include retraining, reassignment or, ultimately, dismissal. Cytopathologists should undergo periodic re-testing every 6–12 months with standardized slide sets. Consistent poor performers may be identified by ongoing performance audits of computer records (e.g., tracking atypical squamous cells of uncertain significance-to-squamous intraepithelial neoplastic lesion ratio [ASCUS:SIL]) and counseled. Effective action plans include retraining, reassignment or disciplinary action. The initial goal of a cytology screening program should be to screen each woman between the ages of 30 and 55 at least once (WHO, 1986). Repeated screening at 5 to 10-year intervals should only be undertaken when resources allow. For purposes of standardization, the Bethesda System (National Cancer Institute Workshop, 1989) can be used for Pap smear classification. However, this system is designed for classification of cervical smears. It should not be expanded to include histology, as there are still prognostic distinctions to be made with the older classifications (Syrjänen et al., 1992; Holowaty et al., 1999). Alternate smear classification schemes may be used alongside the Bethesda system as appropriate to local experience and usage. Women with Pap smears reported as HSIL or atypical glandular cells of uncertain significance (AGUS) must be referred for further evaluation as soon as possible. For women with Pap smears reported as LSIL or ASCUS, there is no consensus regarding the most appropriate management. Whenever it is possible with local referral and follow-up mechanisms, cytology follow-up after 6 months is an appropriate option (Holowaty et al., 1999). It is unlikely that currently available automated and liquid-based cytology screening systems will be cost-effective in developing countries. The working group considered the currently available scientific evidence on the role of the following alternative methods to cervical cytology as primary screening tests in the early detection of high-grade cervical neoplastic lesions: downstaging, VIA, speculoscopy, cervicography and HPV DNA testing. The group discussed these approaches in detail. The data available on the performance of the tests to detect high-grade cervical lesions compared with that of cervical cytology were considered to arrive at the following consensus on the role of the above as primary screening approaches. The working group noted the lack of data on the extent of incidence or mortality reduction associated with the above approaches and the lack of formal cost-effectiveness analyses. Downstaging involves naked eye speculum examination of the uterine cervix (without acetic acid application) under adequate light by health workers who designate the findings as normal (test negative) or abnormal (test positive). Those with an abnormal finding need to be further investigated, which would entail cytology (in settings where limited cytology facilities are available) or examination by a gynecologist (where no cytology is available). Historically, this is the first ever alternative approach proposed to be evaluated in low resource settings (Miller, 1992; Stjernsward et al., 1987). Data from cross-sectional studies in India indicate that the test would result in referral of 40–70% of women for cytology or evaluation by gynecologists (Singh et al., 1992; Nene et al., 1996; Sakaranarayanan et al., 1997; Wesley et al., 1997). This method was not intended for the detection of pre-invasive disease but for the early diagnosis of invasive cervical cancer. However, the approach has a low sensitivity, and particularly, low specificity to detect invasive disease. Further, the lack of, or inadequacy of, treatment facilities for management of invasive cervical cancer in settings where downstaging is intended to be used precludes its role as a primary screening test in low resource settings. This approach was, however, largely the forerunner for further developments in visual inspection-based approaches. This test involves visual examination of the uterine cervix after the application of 3–5% acetic acid. It has been consistently demonstrated to have a sensitivity equivalent to that of cytological screening to detect high-grade lesions (Chirenje et al., 1999; Sankaranarayanan et al., 1998b, Sankaranarayanan et al., 1999). Therefore, this test should be considered as a primary screening method in settings where quality cervical cytology is not available. However, the specificity is lower than that of cervical cytology. Further evaluation is required to identify ways to improve the specificity and to reduce the number of women who suffer from adverse outcomes when VIA is coupled with immediate treatment. In addition, quality control issues in the application of the test need to be addressed. Such evaluation can be carried out in “demonstration projects” (or “follow-up”/cohort studies) or in randomized controlled intervention trials. This test involves visual examination, in a dark room, of the 3–5% acetic acid-impregnated cervix to detect aceto-white areas, with illumination provided by a chemiluminescent light source in the upper blade of the vaginal speculum. Sensitivity and specificity appear to be comparable to that of VIA. However, in view of the additional resources needed, it is an unlikely option as a primary screening test in developing countries. This test involves examination of magnified photographic documentation of the acetic acid-impregnated cervix. Sensitivity is lower than that of cytology and VIA to detect high-grade lesions, although specificity is comparable to that of cytology. As a primary screening test, it is an unlikely option in developing countries. HPV DNA testing with molecular methods for the high-risk HPV types has been shown to have similar or higher sensitivity to detect high-grade disease than that of cytology. However, the specificity is lower than that of cervical cytology, as the prevalence of HPV DNA positivity in women without cervical neoplasia varies markedly and may be particularly high in young women. Currently, the cost of the test is a barrier to its wider use, especially in developing countries. If the cost issue can be solved, HPV DNA testing should be considered as a primary screening method, especially in women over the age of 35, but further evaluation is required to investigate its role in high HPV prevalence populations. Evaluation can be carried out in routine programs or in randomized controlled intervention trials. The working group has not expressed any opinion on different possible combinations of screening tests in view of insufficient data. In any research involving alternative approaches, the working group recommends high-grade disease as the cut-off point for defining disease. Further evaluation in terms of incidence and/or mortality reduction from cervical cancer as well as cost-effectiveness of the recommended alternate primary screening approaches need to be carried out. Although programmatic issues with alternate screening approaches are assumed to be the same as those for cervical cytology, further research may provide additional leads with reference to the alternate approaches. A number of new devices are being evaluated. One recently introduced, portable, real-time device in the form of a hand-held piece attached to a small console offers an instant diagnosis of cervical neoplastic disease. It relies on the fact that cervical tissue has characteristic electrical and optical properties that can be measured with their subsequent cataloging by means of an algorithm. There are preliminary indications that the sensitivity is comparable to cytology but further evaluation is needed. Screening. Referral from such screening programs to a gynecology center. Within the referral center, colposcopy should be offered by trained personnel, with adequate pathological back-up. It is recognized that in less than ideal circumstances, colposcopy will not be available. Within these referral centers, treatment should be given in either the outpatient or office/ambulatory center or in an operating room. After treatment, an efficient follow-up mechanism or system should be in place, run by either medical or paramedical personnel. It should be possible to refer women to a non-hospital center for long-term follow-up which should, in the case of high-grade premalignant lesions, be between 5 and 10 years. The efficiency
Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
