Cervical cancer control in developing countries: memorandum from a WHO meeting.

Memoranda/Memorandums Cervical cancer control in developing countries: Memorandum from a WHO meeting* This memorandum summarizes the report of a WHO Consultation on the Control of Cervical Cancer in Developing Countries, held on 6-7 November 1994, in New Delhi, India. Evaluated was the current situation with regard to cervical cancer and the relevance of current practices in screening. New pragmatic approaches to cervical cancer were proposed that are relevant for developing countries; this includes empowerment of women to come forward, and visual inspection-"downstaging". Introduction Size of the problem Cervical cancer is the commonest form of cancer that affects women in virtually all developing countries and the second commonest form of cancer that affects women in the world. Globally, there are an estimated 450000 new cases each year, with 300000 deaths. If the undiagnosed, early cases are taken into account, the number of new cases each year would be 900000 worldwide. This is approximately the same as the total number of new cases of acquired immunodeficiency syndrome (AIDS) among males and females each year, but in contrast to AIDS, large numbers of women with cervical cancer are usually ignored. * This Memorandum is based on the report of a WHO Consulta- tion on the Control of Cervical Cancer in Developing Countries, held in New Delhi, on 6-7 November 1994. The participants were as follows: P. Blake, London, England; A. Beltram, Tlalpan, Mexico; K. Chaudhry, New Delhi, India; Z.M. Chirenge, Harare, Zimbabwe; H.S. Cronj6, South Africa; D.K. Das, Delhi, India; F. Geldenhuys, Johannesburg, South Africa; M.K. Hakama, Tampere, Finland; J.B. Halil, Kuala Lumpur, Malaysia; C.M. Jagdish, Bangalore, India; H.C. Kitchener (Rapporteur), Aberdeen, Scotland; S.P. Kohli, New Delhi, India; S. Krishnamurthy, Karnataka, India; I.N. Mittra, Bombay, India; P. Naud, Porto Alegre, Brazil; I. Oliver, London, England; A.E. Prasad, Bangalore, India; N. Rehman, Dhaka, Bangladesh; C.R. Ramachandran, New Delhi, India; A. Roxas, Manila, Philippines; L.H. Samarage, Galle, Sri Lanka; A. Sehgal, New Delhi, India; L. Shamsuddin, Dhaka, Bangladesh; V. Singh, New Delhi, India; R.P. Symonds (Rapporteu), Glasgow, Scotland; E. Vallikad, Bangalore, India; F. Welsch, Bethesda, MD, USA; Z.M. Zain, Kuala Lumpur, Malaysia. WHO Collaborating Centre: A.B. Miller (Chairman), Toronto. WHO Secretariat S. Nazeer (Secretary), J. Stjernsward (Co-chairman). Requests for reprints should be sent to Dr S. Nazeer, Division of Reproductive Health, World Health Organization, 1211 Geneva 27, Switzerland. Correspondence should be sent to Dr J. Stjernsward, Department of Community Medicine, Malmo Univer- sity Hospital, University of Lund, S 20502 Malmo, Sweden. A full French resume appears on pages 350-351. Reprint No. 5709 State of the art Cervical cancer is both preventable and curable, provided it is detected at an early stage. In developed countries 80% of cervical cancer cases detected are cured because of early detection. However, in developing countries 80% of cervical cancer cases are incurable at the time of detection, if they are detected at all. Five out of six women with cervical cancer live in developing countries, which possess only 5% of the global resources for cancer control. There are few, if any, cytology screening programmes with a coverage sufficient to have an impact in developing countries; standard radiotherapy is often not available. A realistic, pragmatic approach to cervical cancer control has to be sought for developing countries, coupled with the provision of curative therapy and palliative care and pain relief.a Over the years WHO has advocated "downstaging" (visual inspection of the cervix) as a more realistic ap- proach to active coverage in developing countriesbc than, e.g., cytology screening; however, its sensi- tivity and specificity remain to be evaluated in controlled studies. One of the purposes of the consultation was to initiate and coordinate such studies. Another approach that is highly relevant in developing countries is to educate women about the early warning signals of cervical cancer and to inform them that the disease is curable if diag- a National cancer control programmes: policies and managerial guidelines. Geneva, World Health Organization, 1995. b Stjernsward J et al. Plotting a new course for cervical cancer screening in developing countries. World health forum, 1987, 8: 42-45. c Stjernsward J. National training of radiotherapists in Sri Lanka and Zimbabwe: priorities and strategies for cancer control in developing countries. International joumal of radiation oncology, biology, physics, 1990, 19: 1275-1278. Bulletin of the World Health Organization, 1996, 74 (4): 345-351 © World Health Organization 1996 345 Memorandum nosed early enough.d It is an open question as to what will come first - effective cytology screening coverage, as is the case in developed countries, or a vaccine for primary prevention. For a cost-effective, basic prevention programme it would seem ra- tional to combine activities against sexually transmitted diseases (STDs), AIDS, and cervical cancer. Problems In developing countries the major problems asso- ciated with cervical cancer are the following: lack of knowledge among women about its symptoms; a fatalistic attitude towards cancer, in general, and lack of awareness about the possibility of a cure; shortage of health care facilities in rural areas, with often total lack of standard therapies; and male dominance and ignorance, combined with a low priority for women's health issues. Early detection and screening have been suc- cessful in reducing morbidity and mortality from cer- vical cancer in some developed countries, but not in others. Lack of effect is most often due to poor man- agement and implementation of inappropriate poli- cies with mainly young women being screened and insufficient coverage of older women.e In most developing countries a purposeful cov- erage of all women at risk through cytological screening will not be possible for decades to come, because of the paucity of economic and technically competent manpower resources and inadequate quality assurance for smear tests. Solution Empowering women with knowledge about cervi- cal cancer - its early warning signals, such as intermenstrual, postcoital or postmenopausal bleed- ing, foul discharge, its curability if diagnosed early- combined with the availability of adequate therapies could have a major impact. In developing countries the incidence of cervical cancer is often equal to mortality. Experience in Nordic countries shows that before the introduction of any formal cytology screening programme the proportion of advanced disease among invasive cancers decreased signifi- cantly when therapy became available and women had been made aware about the condition. Mortality decreased much more than that resulting from d Stjernsward J. Downstaging - strategies for cancer control programmes in developing countries; Cancer care, 1994, 1: 11. e Miller AB. Cervical cancer screening: managerial guidelines. Geneva, World Health Organization, 1992. the later introduction of cytology screening.d0 In developing countries a major effect on cervical cancer would therefore be expected using such an approach. Appropriate management and effective early detection and referral policies along with availability of standard therapy are vital for reducing morbidity and mortality from cervical cancer. International Study Group An International Study Group on the control of cer- vical cancer in developing countries has been set up to analyse and evaluate the feasibility and validity of various low-cost strategies for cervical cancer screen- ing in low resource settings compared with cytology. The group aims, in particular, to emphasize the ef- fect of health education and empowerment of women and of the clinical downstaging approach, which involves simple visual inspection of the cervix of asymptomatic women to detect cervical cancer at an earlier, curable stage. Most probably the study findings will be available long before a purposeful level of cytology screening coverage in the majority of developing countries. It is intended that the group will meet regularly to pursue its aims to develop protocols for improving early detection of cervical cancer in developing countries and hence increase the prospects of cure. Objectives The primary objectives of the study group are to compare the effect of health education, the empow- erment of women with knowledge about cervical cancer, and visual inspection (downstaging) versus cytology and to evaluate the relative contributions of each. The secondary objectives are shown below: - to accelerate improvements by coordinating efforts; - to evaluate scientifically health education materi- als for their effect; - to produce education models and a visual tumour atlas jointly, in order to avoid duplication;9 - to identify and support study participants; - to standardize the study protocols; I Porten J et al. Strategies for global control of cancer. Interna- tional journal of cancer, 1995, 60: 1-26. 9 Cervical cancer screening: technical guidelines, 2nd edit. Geneva, World Health Organization, 1996 (in press). 346 WHO Bulletin OMS. Vol 74 1996 Cervical cancer control in developing countries - to identify funding resources; - to provide epidemiology and biostatistical sup- port, which are essential for a conclusive study; and - to provide a forum for regular meetings to estab- lish protocols, exchange information, share expe- riences, and hold discussions in the presence of world experts. A model protocolh for a controlled randomized study was presented and agreed upon. It is designed to evaluate the applicability and cost-effectiveness of different approaches to the early detection of cervi- cal cancer in developing countries, especially health education and cytology versus downstaging, and to evaluate the relative contribution of each. Four study arms should be employed, with a sample size of at least 5000 women in each arm, targeting women aged 35-65 years. The intervention pattern in each arm is shown below. - Arm I: Control arm. - Arm II: Health education only (empowerment of women). - Arm III: Health education + visual inspection. - Arm IV: Health education + Pap-smears. Country experiences Outlined below is an analysis of the situation with regard to cervical cancer in selected countries. United Kingdom Until recently, the cervical screening programme in the United Kingdom had failed to have a significant impact on cervical cancer, largely because it com- prised nonsystematic, opportunistic screening with too few women aged over 35 years being screened. This situation has now been turned around, how- ever, by a government initiative that introduced computerized call/recall, tackled quality assurance in laboratories, and paid general practitioners for meeting coverage targets. An overall fall in incidence of 15% had been reported recently, but a rise in incidence was being seen among young women. Unless the programme, which now costs US$ 100 million per annum, results in a substantial reduction in deaths following these hPrevention and detection of cervical cancer in developing coun- tries: modelprotocol fora pilot study. Unpublished WHO document CAN 1995.1. initiatives, its credibility in terms of cost-effective- ness will be questionable. In considering the propos- als for under-resourced countries it is relevant to keep the United Kingdom experience in mind, and avoid a dogmatic noneffective cytology screening as the only relevant solution for the very different situ- ation that prevails in such countries. South Africa In South Africa as a whole, although the health infra- structure is good and the new government favours primary health care, the problems posed by distance, poor communication, and ignorance remain. While treatment and laboratory facilities were not prob- lems, screening and follow-up coverage of the black population certainly were. Cape Town. A protocol has been piloted in Cape Town, which is primarily designed to eliminate re- ferral problems. The protocol involves an educa- tional programme in the target area for 2-3 weeks backed up by a mobile clinic, where a trained nurse takes smears that are processed on the spot. If a high-grade cytological abnormality is found, colposcopy + acid solution and treatment by dia- thermy loop excision are performed. If a low-grade abnormality is found, a repeat smear is undertaken 6 months later. If the smear is normal, another is taken 10 years later (the national proposal is for smears at 30, 40, and 50 years of age). In the pilot project thus far, 86 abnormal smears were identified in 2600 screened women, the level that would be expected in the population involved. The project is currently be- ing supported by an independent development trust. Orange Free State. A parallel effort is being carried out in Orange Free State. Different methods of iden- tifying precancerous lesions of the cervix were first evaluated in terms of cost, low expertise required, immediate diagnosis, as well as sensitivity and specificity. Of these, affordability and immediate diagnosis were held to be particularly important. Cy- tology, cervicography, and a clinical test involving application of a solution of acetic acid to the cervix were evaluated; of these, the acetic acid test had a higher sensitivity than the other two (though it was much less specific), as well as being cheap and instan- taneous (Table 1). The primary objective is to screen 80% of the target population over a 3-year period, beginning in one district. The protocol involves primary screening by acetic acid testing at 30, 40, and 50 years of age. If the primary test is abnormal, cervicography (held to be more cost-effective than repeat cytology) will be performed, and if the results are abnormal, treat- WHO Bulletin OMS. Vol 74 1996 347 Memorandum Table 1: Comparative sensitivity and specificity of cy- tology, cervicography and the acetic acid test, Orange Free State Sensitivity (%) Specificity (%) Cytology 27 98 Cervicography 51 93 Acetic acid test 65 43 ment by colposcopy and diathermy loop excision will be carried out. Where resources were scarce, this could be scaled down to acetic acid testing, and if the results are abnormal, the cervix treated using cryosurgery, although its limited specificity (40-50%) would necessitate a large number of treatments. India It is important to emphasize that the Indian popula- tion is essentially rural, largely of low social class, many of whom are poorly informed in general and about health care in particular, especially women. Of India's population of approximately 900 million a total of 700 million cannot afford health care, and these are the very people who need to be targeted. In India, cytology lacks sensitivity owing to the high incidence of cervicitis, and it was held that the South African approach could be a useful model. Delhi. A team at the Indian Council for Medical Research (ICMR), Delhi, has undertaken cervical photography using a 300-mm lens. It is claimed that the photographs obtained are only slightly inferior to cervicography if a "ring"-type flash unit is used. Bangalore. In Bangalore, a study was conducted over a 3-year period using visual inspection by a nurse as the means of early detection. If the nurse found that the cervix was abnormal, the woman was transferred to the primary health care (PHC) system. The aim was to achieve a downshift in cervical cancer stage by activating or mobilizing the PHC system. Four interventions were instituted in four areas; the coverage of the target population was such that, of the entire targeted study population, 15.7% received education on cervical cancer and 7.8% had undergone visual inspection. The results demon- strated a very low population coverage and poor patient compliance. The number of cancers identi- fied was in proportion to the population screened, but no results are yet available on downshifting. The reasons for noneffectiveness or noncompliance were ascribed to under-resourcing and the unwillingness of women to avail themselves of the programmes, as well as their refusal to be examined. The factors in the PHC system that contributed to the poor coverage were lack of innovation and incentives, with the tasks of an already overworked PHC staff. The coverage was better at centres where the primary health care was being delivered by a local self-governing body rather than in the WHO/ ICMR areas, where the existing PHC centres had to be used, with insufficient incentives for the staff (of- ten the female medical officers) to collaborate ac- tively in the study. Nordic countries Swedish and Finnish data show that cervical cancer deaths had fallen in these countries very signifi- cantly prior to screening, presumably due to better health awareness among women, which led to earlier stage presentation and availability of free treatment. The subsequent contribution of screening to a fall in mortality from the disease was much less. The finding that professional and public education had such a marked effect on cervical cancer mortality has an important message for developing countries, and in such countries health awareness and education should be priorities. Prospective studies Below are outlined details of various prospective studies on cervical cancer screening in selected countries. Zimbabwe In Zimbabwe most of the population lives in a rural setting, with district hospitals in the cities and pri- mary health centres staffed by nurses all over the country. Most women live within lOkm of such a clinic. The study will involve recruiting 15 000 women, with 5000 in one of three arms; education interven- tion alone; education intervention and visual inspec- tion; and education intervention and cytology. Any women in the first arm with symptoms of cancer will be examined for abnormalities. Each arm will func- tion in a different part of Mashonaland Central prov- ince, an area containing a homogeneous population. It is planned to include also a control arm, i.e., with no health education intervention, in order to evalu- ate the effect of health education, which probably in itself could have a major impact. United Kingdom At the Royal London Hospital it is proposed to set up an integrated prostate and cervical cancer screen- WHO Bulletin OMS. Vol 74 1996348 Cervical cancer control in developing countries ing programme in association with the family plan- ning clinic and the school education programme on the common risk factors for both these diseases and AIDS. The aim is to tackle the contribution of males to cervical cancer. India In Bombay a study has been designed to answer whether downstaging can be achieved. The aim is to change significantly the current distribution of cervical cancer from 70% in stages 3 and 4 to a level of 30%. The study will also investi- gate downstaging of oral and breast cancer. The sample size will be 35000, randomized to each arm. In order to increase its cost-effectiveness, the study is linked to the expanded programme on immunization (EPI) in Bombay. A database of 1 million Bombay families currently exists and this will be used to identify a target population. Half the effort will be undertaken by dedicated health workers with the remainder being carried out by EPI staff. The protocol involves randomization to one of two arms: either no intervention or visual inspection of the mouth, breast, and cervix annually for 5 years. The outcomes will be changes in disease stage and mortality. This is viewed as a study defining the ef- fectiveness of downstaging. Bangladesh In Dhaka, it is believed that efforts on cervical can- cer screening should revolve around gynaecologists and it is intended to compare the findings from visual inspection alone versus those from a non- intervention control group. Philippines In the Philippines, cervical cytology is too expensive for a nationwide programme and an alternative strategy was required. The Ministry of Health is pro- ducing a proposal to obtain funding from the World Bank based on a modification of the Zimbabwe pro- tocol discussed above. A study design of three arms is planned: 15 000 women per arm to evaluate visual inspection versus education, versus a combination of both these plus cervical cytology. The outcomes will include improvements in knowledge, coverage, and rates of detection of cervical cancer and cervical intraepithelial neoplasia. Brazil A planned study in southern Brazil will compare visual inspection plus cervical cytology versus visual inspection alone, with 6000 women in each arm. Women would be screened every 3 years with a 5- year follow-up. Malaysia In Malaysia, cervical cancer control has been made a priority under the national cancer control pro- gramme. A protocol similar to that of Zimbabwe will be implemented in Kelantan State and a cervical cancer register established. It is intended to study a total of 15 000 women aged 35-59 years. A fulltime position for 3 years for an epidemiologist-biostatistician has been created for this purpose. Recommendations * Cervical cancer needs to be recognized as a major but soluble, health problem among women in most developing countries. * Empowerment of women through health educa- tion about cervical cancer, its signs and symptoms, and its curability could have a major impact. There is a great need to enhance women's awareness of their own reproductive health in order to make them come forward willingly for early detection tests and therapy. * Appropriate management and effective early de- tection and referral policies along with availability of standard therapy are vital to reduce the morbidity and mortality from cervical cancer. * Controlled studies to evaluate ways of most effectively empowering the women are urgently needed. * In developing countries, downstaging cervical can- cer by visual inspection and the possibility of achiev- ing coverage with quality controlled cytology screening need to be evaluated. * Early detection should not be carried out in isola- tion, and other clinical aspects should also be ad- dressed simultaneously, such as the availability of treatment for any STDs found, treatable cancers, and incurable cases. Thus, clear policies, including referral systems, for therapy and palliative care should also be established. * For an effective primary prevention approach, it seems rational to combine efforts against STDs, AIDS, and cervical cancer. WHO Bulletin OMS. Vol 74 1996 349 Memorandum Resume Lutte contre le cancer du col uterin dans les pays en developpement: Memorandum d'une reunion de l'OMS Le cancer du col uterin est la forme la plus fr6- quente de cancer f6minin dans la quasi-totalit6 des pays en developpement et, a I'echelle mondiale, la duxieme forme de cancer chez la femme. On denombre chaque annee dans le monde environ 450 000 nouveaux cas et 300000 d6ces dus a ce cancer. Si on tient egalement compte des cas pr6coces non diagnostiques, on arrive a au moins 900000 nouveaux cas par an. Cette affection est a la fois 6vitable et curable, a condition d'etre d6tect6e a un stade precoce. Dans les pays d6veloppes, 80% des cas d6tectes sont gu6ris grace au d6pistage pr6coce; en re- vanche, dans les pays en developpement, 80% des cas ne sont pas detect6s ou sont d6ja incurables au moment de la d6tection. 11 est necessaire de parvenis a une approche realiste de la lutte contre le cancer du col uterin dans les pays en developpement, approche qui doit etre assortie de la fourniture d'un traitement curatif, de soins palliatifs et d'un traitement de la douleur. L'OMS encourage le depistage pr6coce simplifi6, par inspection visuelle du col uterin, comme alterna- tive au depistage cytologique pour la recherche active des cas dans les pays en d6veloppement. La sensibilite et la sp6cificite de cette approche doivent encore etre evaluees lors d'etudes contr6l6es; I'un des buts de la Consultation 6tait de mettre en place et de coordonner de telles etudes. L'education des femmes concernant les signes d'alerte precoces du cancer du col et la possibilite de gu6ris la maladie si elle est diagnostiquee suffisament t6t est 6galement du plus grand int6ret dans les pays en developpement. L'autonomisation des femmes grace a l'in- formation sur le cancer du col uterin-portant sur les signes d'alerte precoces comme les saigne- ments intermenstruels, postcoitaux ou postmeno- pausiques et les 6coulements vaginaux malodorants, et sur la possibilite de guerir cette moladie si elle est diagnostiqu6e suffisamment t6t-assocfe6 a la possibilite de disposer de traitements appropri6s, pourrait avoir un impact majeur sur la maladie. Dans les pays en developpement, l'incidence du cancer du 6gale souvent sa mortalit6. En revanche, dans les pays nordiques, avant meme l'adoption de tout programme officiel de d6pistage cytologique, la proportion de cas avanc6s parmi les cancers invasifs a baiss6 de fa9on importante, plus encore qu'apres I'adoption du d6pistage, lorsqu'un traite- ment est devenu disponible et que les femmes ont re,u une information sur la maladie. Dans les pays en d6veloppement, on pourrait par cons6quent attendre d'une telle approche un important effet sur les cancers du col. Un groupe d'6tude international sur la lutte contre le cancer du col uterin dans les pays en developpement a et6 constitu6 en vue d'analyser et d'evaluer la faisabilit6 et la validit6 de differentes strat6gies a faible coOt de d6pistage du cancer du col dans un contexte de faibles ressources, par rapport au depistage cytologique. Le groupe cher- chait en particulier a souligner l'effet de I'education sanitaire et de I'autonomisation des femmes, et celui de l'approche consistant an un depistage pre- coce simplifie, avec une simple inspection visuelle du col ut6rin chez les femmes asymptomatiques, afin de depister le cancer du col a un stade precoce, et curable. 11 est probable que dans la plupart des pays en d6veloppement, les r6sultats de l'etude seront disponibles bien avant l'introduction du depistage cytologique a grande echelle. Un protocole type pour les 6tudes randomisees contr6lees a ete prepare et approuve par les mem- bres du groupe. Ce protocole vise a evaluer I'appli- cabilite et le rapport coOt/efficacite de differentes approches de la detection pr6coce du cancer du col uterin dans les pays en d6veloppement, en particulier l'education sanitaire et le d6pistage cyto- logique, par rapport au d6pistage precoce simplifi6, en vue d'evaluer leur contribution relative. L'etude compatera quatre branches, avec des 6chantillons d'au moins 5000 femmes par branche, et sera axee sur les femmes sexuellement actives ag6es de 35 a 65 ans, comme indique ci-dessous. Branche 1: Lutte Branche I1: Education sanitaire seule (autonomisation de la femme). Branche IlI: Education sanitaire + inspection visuelle. Branche IV: Education sanitaire + frottis et test de Papanicolaou. Le groupe d'etude a formul6 les recommandations suivantes: * Le cancer du col uterin doit etre reconnu en tant que probleme de sante majeur mais soign- able chez les femmes de la plupart des pays en developpement. * L'autonomisation des femmes, grace a l'6duca- tion sanitaire concernant la maladie, ses symp- t6mes et ses possibilites de traitement, pourrait avoir un impact majeur sur les cancers du col. 11 est indispensable d'ameliorer l'information des femmes dans le domains de la sant6 en matibre de WHO Bulletin OMS. Vol 74 1996350 Cervical cancer control in developing countries reproduction afin de les encourager a se pr6senter pour des tests de d6pistage pr6coce et a suc6re un traitement. * Une prise en charge appropriee, et une politique efficace de d6tection pr6coce et d'orientation, as- sorties de la possibilite de disposer d'un traitement standard, sont indispensables pour reduire la mor- bidite et la mortalite dues au cancer du col uterin. * 11 est urgent d'entreprendre des 6tudes contr6lees afin d'6valuer le moyen le plus efficace de mobiliser les femmes contre la maladie. * Dans les pays en d6veloppement, il est neces- saire d'evaluer par des etudes contr6l6es le depistage precoce simplifi6 du cancer du col par inspection visuelle et les moyens de parvenir a une bonne couverture par un d6pistage cytologique de qualit6 suffisamment contr6l6e. * Le d6pistage pr6coce ne doit pas etre effectue isol6ment. D'autres aspects cliniques doivent etre envisag6s simultanement, par exemple la possibilite de traitement de toute maladie sexuellement trans- missible observ6e, des cancers pouvant etre soignes et de la prise en charge des cas incurables. 11 faut par cons6quent 6tablir 6galement des politiques claires, comportant des systemes d'orientation, pour les soins therapeutiques et palliatifs. * Pour une approche de prevention primaire efficace, il convient d'associer les efforts de litte contre les maladies sexuellement transmissibles, le syndrome d'immunod6ficience acquise (SIDA) et le cancer du col ut6rin. WHO Bulletin OMS. Vol 74 1996 351