Publication of the International Union Against Cancer

Int. J. Cancer: 107, 337340 (2003)

2003 Wiley-Liss, Inc.

MINI REVIEW
CAN SCREENING FOR CERVICAL CANCER BE IMPROVED, ESPECIALLY IN
DEVELOPING COUNTRIES?
Anthony B. MILLER1*, Rengaswamy SANKARANARAYANAN2, F. Xavier BOSCH3 and Cecilia SEPULVEDA4
Division of Clinical Epidemiology, Deutsches Krebsforschungszentrum, Heidelberg, Germany
2
Unit of Descriptive Epidemiology, International Agency for Research on Cancer, Lyon, France
3
Catalan Institute of Oncology, Barcelona, Spain
4
Programme on Cancer Control, World Health Organization, Geneva, Switzerland

Worldwide, cervical cancer comprises approximately 12% of all

cancers in women.1 It is the second most common cancer in
women worldwide but still the commonest in many developing
countries. Cervical screening is acknowledged as currently the
most effective approach for cervical cancer control. However, in
many countries, including some middle-income developing countries, the existing programmes are failing to achieve a major
impact.2
The World Health Organization has recently released a comprehensive report on Cervical Screening in Developing Countries.2
The purpose of our review is to place this report in the context of
what we know of the effectiveness of cervical screening and the
ongoing research endeavours designed to evaluate new tests for
cervical cancer precursors, and major efforts in many countries to
improve programme organisation.
CYTOLOGY SCREENING

It is generally agreed that cytology screening for cancer of the

cervix has been effective in reducing the incidence and mortality
from the disease in many developed countries.3,4 The organised
programmes have shown the greatest effect, while using less
resources than the unorganised (opportunistic) programmes.4
There is general agreement that high-quality cytology is a highly
specic screening test, with estimates of the order of 98 99%.
There is less agreement on the sensitivity of the test; crosssectional studies have suggested sensitivity of the order of 50% in
some circumstances.57 However, studies that have been able to
assess sensitivity longitudinally have produced estimates of approximately 75%.8,9
There are several essential elements for successful cytology
screening.2 Critical is training of the relevant health care professionals, including smear takers (e.g., physicians, nurses, midwives), smear readers (cytotechnologists), cytopathologists, licensed colposcopists and programme managers, to ensure
adequate quality in the administration and assessment of the smear.
This should ensure adequately taken and xed smears, though
checks to make sure that adequate quality is obtained should be
built into the system, and retraining conducted for those who have
a high proportion of unsatisfactory smears (10%).10 Funding
should be such as to provide efcient and high-quality laboratory
services. These should preferably be centralised to ensure adequate
throughput through the laboratory and staff that can ensure quality
control of cytology reading. Communications should be exemplary, starting with a means to rapidly transport smears to the
laboratory and building in a mechanism to inform the women
screened of the results of the test in an understandable form,
coupled with a mechanism to ensure that women with an abnormal
test result attend for management and treatment and a mechanism
to follow up treated women.
Guidelines should be agreed upon regarding a number of policy
issues. These include a decision on the priority age group to be
screened, which in an unscreened population should initially be
women ages 3554 years to ensure maximum impact.11 However,
in many countries, it will often be found that previous policies

have resulted in a concentration on screening younger women,

often in association with family planning and maternal and child
health services, resulting in a diversion of resources from those
who need to be screened and very little impact on disease in the
population.2 Correcting such imbalances will be difcult and will
require many consultations and professional education on the
natural history of the disease. For those women with abnormalities,
the guidelines should include an accepted denition of an abnormality to be treated, i.e., high-grade lesions, because of the substantial tendency of low-grade lesions to regress spontaneously.12,13 It will also be crucial to make a decision on the frequency
of subsequent screens and develop a mechanism to invite women
with negative smears for subsequent smears, as this has a major
impact on costs for the health system.14,15
It follows that elements that interfere with the development of
successful cytology screening programmes include over-reliance
upon maternal and child health services for screening, opportunistic rather than organised screening and low coverage of the target
group.4,11 Setting too low a threshold for referral for colposcopy,
i.e., overtreating nonprogressive disease, will lead to reduced
cost-effectiveness.15
The major advantages of cytology screening are the considerable experience accrued worldwide in its use and that it is so far
the only established screening test for cervical cancer precursors
that has been shown to reduce the incidence and mortality of the
disease.3,4 However, cytology has limitations, and it is incompatible with some womens beliefs. It is important that women are not
coerced into screening, nor given an overoptimistic view of its
potential, as it is impossible to abolish mortality from the disease
with screening.15
New developments in cytology, such as liquid-based cytology
and automated reading, have advantages but are currently out of
reach of most programmes in developing countries.2
Research into means to improve programme efciency in middle-income countries is a high priority.
VISUAL INSPECTION WITH ACETIC ACID (VIA)

The technical and nancial constraints of implementing cytology-based screening programmes in developing countries have led
to the investigation of screening tests based on visual examination
of the uterine cervix. Among these tests, visual inspection with
35% acetic acid (VIA) seems to fulll the basic criteria of a
*Correspondence to: Division of Clinical Epidemiology, Deutsches
Krebsforschungszentrum, Heidelberg 69120, Germany.
Fax: 49-6221-42-2203. E-mail: [email protected]
Received 8 February 2003; Revised 13 May 2003; Accepted 20 May
2003
DOI 10.1002/ijc.11388

338

MILLER ET AL.

satisfactory screening test.2 VIA involves nonmagnied visualization of uterine cervix soaked with 35% acetic acid.
The results of assessments of test accuracy in cross-sectional
study settings indicate that the sensitivity of VIA to detect highgrade precancerous lesions ranged from 66 96% (median 84%);
the specicity varied from 64 98% (median 82%); the positive
predictive value ranged from 10 20% and the negative predictive
value ranged from 9297%.16 22 However, all reported studies,
except 2, suffered from verication bias.23 Despite different study
settings, providers, study protocols and denitions of positive tests,
the estimates of sensitivity of VIA have tended to cluster around a
mean of 76%. In most of the studies where cytology and VIA have
been provided under the same conditions, the sensitivity of VIA
was found to be similar to that of cytology, whereas the specicity
was consistently lower.2
A wide range of personnel ranging from doctors, nurses and
other allied health workers to nonmedical personnel have been
involved in the administration and reporting of results of VIA. The
most common form of reporting involved negative and positive
categories. The emerging consensus is that well-dened, demarcated, densely opaque acetowhite lesions located in the transformation zone (TZ) close to the squamocolumnar junction should
dene a positive VIA test. The criteria for a negative test included
one or more of the following: no acetowhite lesions, faint illdened translucent acetowhite lesions, endocervical polyps, nabothian cysts, dot-like acetowhite lesions and prominent squamocolumnar junction.2
The investigation of women with a positive VIA initially followed similar principles to those of cytology-positive women.
Now, in various settings, 5 options may be considered for women
testing positive on VIA:
Colposcopy with histologic sampling and treatment based on
the histologic nding;
Colposcopy with histologic sampling and treatment given on
the basis of the colposcopic diagnosis (with retrospective access to histologic diagnosis);
Colposcopy and treatment on the basis of the colposcopic
diagnosis;
Magnied visual inspection (VIAM) with histologic sampling
and immediate treatment with cryotherapy;
Immediate treatment with cryotherapy with diagnostic referral
(for colposcopy or biopsy) restricted to cases ineligible for
immediate treatment.

All of the above approaches are still being evaluated in terms of

safety, acceptability to women, feasibility and effectiveness in
eradicating pre-invasive cervical disease.
In most of the reported study settings, training in the administration and reporting of VIA has been carried out in sessions
lasting 3 days to 2 weeks, accompanied by written manuals. A
learning period has been recognised after the training sessions. In
the reported and unreported studies, the screen-positive rate among
newly trained screeners has ranged from 2535%, which later
decreased to 10 18% in most instances.2
The major limitations of VIA include: low specicity (generally
85%), which can lead to overinvestigation and overtreatment of
screen-positive women, and lack of standardised methods of quality control, training and competency evaluation. It is limited in its
ability to detect endocervical disease. The major strengths of VIA
include its simplicity and low cost, real-time availability of results
and potential for immediate linkage with investigations/treatment,
consistent estimates of accuracy, feasibility to be offered in lowresource settings and the possibility of rapid training of providers.
In recent model-based evaluations of the cost-effectiveness of VIA
compared to cytology and HPV DNA testing, a major advantage of
VIA has been the possibility of treatment (cryotherapy) in the
same session as an abnormality is detected, this obviating the need

to bring women back for diagnosis and treatment, with the associated costs and risk of failure to attend.24,25
Further research in addressing methods for improving specicity, quality control, tests to be used to follow up women who have
been treated and competency and evaluation of skills of screeners
and other health personnel involved in screening programs is
essential. The efcacy and cost-effectiveness of VIA-based population-screening programmes in reducing the incidence of and
mortality from cervical cancer is not known and remains to be
established, as do the long-term complications and safety of overtreatment in the context of a VIA screening programme.2
Further information from ongoing studies regarding VIAs longitudinal (programme) sensitivity, efcacy in reducing incidence/
mortality from cervical cancer, its cost-effectiveness and safety
will be useful in formulating public health policies to guide the
organisation of VIA-based, mass population-based screening programmes in developing countries and to reorganise programmes in
countries with currently ongoing inefcient cytology screening
programmes.
HPV TESTS IN CERVICAL CANCER SCREENING PROGRAMMES

Molecular and epidemiologic studies have unequivocally shown

that the vast majority of cervical cancer cases worldwide are
caused by persistent infections with some high-risk types of the
Human Papillomavirus family.26 From the point of view of dening preventive strategies, the HPV-attributable fraction should be
considered to be 100%.27
Current HPV-testing systems are able to detect the presence of
viral markers (HPV-DNA in exfoliated cervical cells) in close to
100% of invasive cervical cancer specimens, 7590% of precursor
lesions (LSIL/CIN1, CIN2/3, HSIL) and in some 50% of borderline cytology lesions (ASCUS).28,29
Commercially available, FDA approved testing systems can be
transferred to settings with some level of sophisticated technology.
Such laboratories can be found in all developed and many middleincome countries.
In triage studies (investigations of the minor abnormalities detected by cytology) and in screening studies when both cytology
and HPV tests are jointly performed, the cross-sectional sensitivity
of the HPV test to detect HSIL or more advanced lesions is at least
as good as cytology.30,31 In most studies, the reported sensitivity of
the HPV test is some 10% higher than cytology.
Triage studies, including large, randomised controlled trials,
have shown that reduction in the number of visits and referrals to
colposcopy/biopsy can be achieved with HPV tests.32,33
One of the strongest gains of the combination of HPV tests and
cytology lies in the very high negative predictive value (i.e.,
97%). Major savings to the health systems may derive from
substantially increasing the duration of the interval between
screens without losses in sensitivity for high-grade intraepithelial
lesions.2
The advantages of HPV tests compared to cytology are:
The objectivity of the test, resulting in very low inter- and
intra-observer variability;
The possibility of almost complete automation of the process.
This should ensure high throughput at a standard level of
quality;
Built-in quality-control procedures;
Opportunities for self-sampling for HPV DNA in some populations with limitations in health care facilities and manpower,
albeit with some loss of sensitivity;
The high sensitivity of the HPV DNA test to identify HSIL in
women ages 30 years;
Gains in effectiveness could be achieved by increasing the
length of the interval between screens and reducing the total
number of lifetime screens required.

339

CERVICAL SCREENING IN DEVELOPING COUNTRIES

The disadvantages of HPV DNA testing are:

Its relatively high costs compared to cytology and VIA;
Dependence on reagents currently produced by only a single
commercial manufacturer;
The requirement for a molecular diagnostic laboratory;
Its low specicity in younger women and populations with
signicant rates of HIV seropositivity;
Further, since HPV DNA testing, like cytology, is not a test
that provides results at the time of the visit or soon afterwards,
many of the traditional barriers to cytologic screening have not
been eliminated.

Cost-benet analyses are underway. Modelling based upon results from South Africa suggests that VIA or HPV DNA tests may
offer attractive alternatives to cytology-based screening programmes.24
In countries with established cytology-based screening programmes, HPV tests are an alternative to repeat cytology in the
presence of abnormal cytology. Women who are HPV-negative
need not be rescreened for at least 5 years and possibly 10 years.
In countries without established cytology-based screening programmes, but with the necessary laboratory facilities, HPV tests
could be evaluated for primary screening. Appropriate trials are
strongly encouraged and are now underway in India.2
New tests for HPV are being developed to overcome the disadvantages of both conventional cervical cytology and current HPV
DNA tests. Ideal would be a test that indicates that an oncogenic
HPV virus has already enhanced genetic instability and rendered
infected cells susceptible to transformation, thereby facilitating the
development of cancer. An example is based upon the fact that
continuous expression of viral oncogenes E6 and E7 interferes
with normal cell cycle control by targeting p53 and pRB. This
phenomenon is a general event in persistent high-risk HPV infections but seems not to be induced by low-risk types. The functional
loss of pRB by binding of E7 protein in turn leads to an increase
in cell cycle regulating p16INK4a, possibly as a negative feedback
loop. Therefore, it should be possible to use the detection of
overexpression of this cell cycle regulatory protein as a surrogate
marker for essential steps in early cervical carcinogenesis.34 A
simple immunohistochemistry assay has been developed that detects p16 expression in both cell smears and tissue sections and is
under evaluation. Detection systems using microarray technology
are also under investigation. Alternative targets to HPV such as
telomerase gene expression and other host genetic targets alone or
in combination with HPV testing are being considered.2
PROGRAMME ORGANISATION

Central to the success of any screening programme is the functioning of that programme in its entirety. This is true whatever test
is chosen for screening. The requirements include the ability of a
programme to ensure high levels of coverage of the target population, to offer high-quality, caring services, to develop and monitor good referral systems that ensure good patient follow up and
to ensure that the patients receive appropriate, acceptable and
caring treatment in the context of informed consent.15
WHO recommends that cervical screening should be planned
within the context of national planning for cancer control.35 In

many countries some form of screening exists but will have to be

reorganised to achieve success. There needs to be the political will
to proceed, with support and funding from the Ministry of
Health.2,15 Screening has to be based on an adequate health infrastructure. There must be a dened target population and means to
identify and recruit into screening that population.15 The women in
this population will have to be educated about screening for
cervical cancer, and the health professionals who serve them may
need educating and retraining. As indicated above for cytology but
applicable for all tests, a dened referral system for women with an
abnormality and a mechanism to ensure women with an abnormality attend for diagnosis and treatment must be put in place.
Systems to manage the abnormalities and follow up those treated
will also be required, while the programme will require monitoring
and evaluation. Leadership, management skills, attention to linkages at all levels of the programme and budgeting skills are
essential.35 In many developing countries, lack of information
systems constitutes an important barrier to the effective control of
cervical cancer. A framework for creating such systems in developing countries has recently been described.36
In the long term, vaccination programmes against the relevant
oncogenic HPV types could become the mainstay of cervical
cancer control. However, development of these vaccines, though
proceeding apace, has to be followed by careful evaluation of their
safety and efcacy under eld conditions. Some univalent candidate vaccines are already under evaluation,37 but until they can be
established as effective and replaced by vaccines incorporating the
large majority of the prevalent oncogenic HPV types in each
region, they cannot substitute for screening. Further, if the hope
that therapeutic as distinct for prophylactic vaccines proves unfounded, it will take more than one generation before screening,
directed to those infected before vaccines became available, can be
withdrawn. So it seems probable that effective screening programmes for cervical cancer will be needed for many decades.
CONCLUSIONS

Cytology screening remains the standard for application in

middle-income resource settings. However, VIA holds substantial
promise, and provided this is conrmed in the ongoing studies and
the difculties associated with its lower specicity overcome, it
may replace cytology in lower-income settings. Tests for HPV
DNA have shown efcacy in the triage of equivocal diagnosis
(ASCUS in the Bethesda system) and, if their cost and the required
technology are made affordable, could eventually become the
preferred approach for screening in middle-income settings. All
tests, however, have to be applied within an organised setting,
preferably as a component of a National Cancer Control Programme.35
ACKNOWLEDGEMENTS

This article is dedicated to Prof. Dr. Harald zur Hausen on the

occasion of his retirement as head of Deutsches Krebsforschungszentrum, with gratitude and appreciation for 20 years of leadership and for his support of research into cervical cancer. We thank
all those who contributed to the discussions during the WHO
consultation on Cervical Cancer Screening in Developing Countries in March 2001 and others who have provided input from the
World Health Organization, on which this review is based.

REFERENCES

1.
2.
3.

Ferlay J, Bray F, Pisani P, Parkin DM. GLOBOCAN 2000. Cancer

incidence, mortality and prevalence worldwide. Version 1.0. IARC
Cancer Base no. 5. Lyon: IARC Press, 2001.
World Health Organization. Cervical cancer screening in developing
countries. Report of a WHO consultation. Geneva: World Health
Organization, 2003.
Hakama M, Chamberlain J, Day NE, Miller AB, Prorok PC. Evaluation of screening programmes for gynecological cancer. Br J Cancer
1985;52:669 73.

4.
5.

6.

Miller AB, Chamberlain J, Day NE, Hakama M, Prorok PC. Report on

a workshop of the UICC Project on evaluation of screening for cancer.
Int J Cancer 1990;46:7619.
Agency for Health Care Policy and Research. Evaluation of cervical
cytology. Technology Assessment Report No. 5. Rockville, MD:
Agency for Health Care Policy and Research, 1999. [Available at
http://www.ahcpr.gov]
Fahey MT, Irwig L, Macaskill P. Meta-analysis of Pap-test accuracy.
Am J Epidemiol 1995;141:680 9.

340
7.

8.
9.

10.
11.
12.
13.
14.

15.

16.
17.
18.
19.
20.

21.

22.
23.

MILLER ET AL.

Nanda K, McCrory DC, Myers ER, Bastian LA, Hasselblad V, Hickey

JD, Matchar DB. Accuracy of the Papanicolau test in screening for
and follow-up of cervical cytologic abnormalities: a systematic review. Ann Int Med 2000;132:810 9.
Boyes DA, Morrison B, Knox EG, Draper G, Miller AB. A cohort
study of cervical cancer screening in British Columbia. Clin Invest
Med 1982;5:129.
IARC Working Group on Cervical Cancer Screening. Summary chapter. In: Hakama M, Miller AB, Day NE, eds. Screening for cancer of
the uterine cervix. IARC Scientic Publications No. 76. Lyon: IARC,
1986. pp. 133 42.
Miller AB. Review: Quality assurance in screening strategies. Virus
Res 2002;89:2959.
Miller AB. Cervical cancer screening programmes. Managerial guidelines. Geneva: World Health Organization, 1992.
Ostor AG. Natural history of cervical intraepithelial neoplasia: a
critical review. Int J Gynecol Pathol 1993;12:186 92.
Holowaty P, Miller AB, Rohan T, To T. The natural history of
dysplasia of the uterine cervix. J Natl Cancer Inst 1999;91:252 8.
Miller AB, Anderson G, Brisson J, Laidlaw J, LePitre N, Malcolmson
P, Mirwaldt P, Stuart G, Sullivan W. Report of a National Workshop
on Screening for Cancer of the Cervix. Can Med Ass J 1991;145:
130125.
Miller AB, Nazeer S, Fonn S, Brandup-Lukanow A, Rehman R,
Cronje H, Sankaranarayanan R, Koroltchouk V, Syrjanen K, Singer
A, Onsrud M. Report on consensus conference on cervical cancer
screening and management. Int J Cancer 2000;86:440 7.
Cecchini S, Bonardi R, Mazzotta A, Grazzinni G, Iossa A, Ciatto S.
Testing cervicography and VIA as screening tests for cervical cancer.
Tumorigenesis 1993;79:225.
Megavand E, Denny L, Dehaeck K, Soeters R, Bloch B. Acetic acid
visualization of the cervix: an alternative to cytologic screening.
Obstet Gynecol 1996;88:383 6.
Londhe M, George SS, Seshadri L. Detection of CIN by naked eye
visualization after application of acetic acid. Ind J Cancer 1997;34:
88 91.
Denny L, Kuhn L, Pollack A, Wainwright H, Wright TC. Evaluation
of alternative methods of cervical cancer screening for resource-poor
settings. Cancer 2000;89:826 33.
Sankaranarayanan R, Wesley R, Somanathan T, Dhakad N, Shyamulakumary B, Amma NS, Parkin DM, Nair MK. Performance of visual
inspection after acetic acid application (VIA) in the detection of
cervical cancer precursors. Cancer 1998;83:2150 6.
Sankaranarayanan R, Shyamulakumary B, Wesley R, Amma NS,
Parkin DM, Nair MK. Visual inspection with acetic acid in the early
detection of cervical cancer and precursors. Int J Cancer 1999;80:
1613.
University of Zimbabwe/JHPIEGO Cervical Cancer Project. Visual
inspection with acetic acid for cervical-cancer screening: test qualities
in a primary-care setting. Lancet 1999;353:869 73.
Walter SD. Estimation of test sensitivity and specicity when disease
conrmation is limited to positive results. Epidemiology 1999;10:67
72.

24. Goldie SJ, Kuhn L, Denny L, Pollack A, Wright TC. Policy analysis
of cervical cancer screening strategies in low-resource settings: clinical benets and cost-effectiveness. JAMA 2001;285:310715.
25. Mandelblatt JS, Lawrence WF, Gafkin L, Limpahoyom KK, Lumbiganon P, Warakamin S, King J, Yi B, Ringers P, Blumenthal PD.
Costs and benets of different strategies to screen for cervical cancer
in less-developed countries. J Natl Cancer Inst 2002;94:1469 83.
26. IARC Monographs on the Evaluation of Carcinogenic Risks to Humans. Vol. 64. Human papillomaviruses. Lyon: IARC, 1995.
27. Walboomers JM, Jacobs MV, Manos MM, Bosch FX, Kummer JA,
Shah KV, Snijders PJ, Peto J, Meijer CJ, Munoz N. Human papillomavirus is a necessary cause of invasive cervical cancer worldwide.
J Pathol 1999;189:129.
28. Nobbenhuis MAE, Walboomers JMM, Helmerhorst TJM, Rosendaal
L, Remmink AJ, Risse EKJ, van der Linden H, Voorhorst FJ, Kenemans P, Meijer CJLM. Relation of human papillomavirus status to
cervical lesions and consequences for cervical-cancer screening: a
prospective study. Lancet 1999;324:20 5.
29. Clavel C, Masure M, Bory JP, Putaud I, Mangeonjean C, Lorenzato
M, Nazeyrollas P, Gabriel R, Quereux C, Birembaut P. Human
papillomavirus testing in primary screening for the detection of highgrade cervical lesions: a study of 7932 women. Br J Cancer 2001;84:
1616 23.
30. Manos MM, Kinney WK, Hurley LB, Sherman ME, Shieh-Ngan J,
Kurman RJ, Ransley JE, Fetterman BJ, Hartinger JS, McIntosh KN.
Identifying women with cervical neoplasia: using human papillomavirus DNA testing for equivocal Papanicolaou results. JAMA 1999;
281:160510.
31. Solomon D, Schiffman M, Tarone R. Comparison of three management strategies for patients with atypical squamous cells of undetermined signicance: baseline results from a randomized trial. J Natl
Cancer Inst 2001;93:2939.
32. Cuzick J, Beverley E, Ho L, Terry G, Sapper H, Mielzynska I, Lorincz
A, Chan W-K, Krausz T, Soutter P. HPV testing in primary screening
of older women. Br J Cancer 1999;81:554 8.
33. Zielinski GD, Snijders PJF, Rozendaal L, Voorhorst FJ, van der
Linden HC, Runsink AP, de Schipper FA, Meijer CJLM. HPV precedes abnormal cytology in women developing cancer and signals
false negative smears. Br J Cancer, 2001;85:398 404.
34. Klaes R, Friedrich T, Spitkovsky D, Ridder R, Rudy W, Petry U,
Dallenbach-Herweg G, Schmidt D, von Knebel Doeberitz M. Overexpression of p16INK4a as specic marker for dysplastic and neoplastic epithelial cells of the cervix uteri. Int J Cancer 2001;92:276
84.
35. National Cancer Control Programmes. Policies and managerial guidelines, 2nd ed. Geneva: World Health Organization, 2002.
36. Marrett LD, Robles S, Ashbury FD, Green B, Goel V, Luciani S. A
proposal for cervical screening information systems in developing
countries. Int J Cancer 2002;102:2939.
37. Koutsky et al. A controlled trial of a human papillomavirus type 16
vaccine. N Engl J Med 2002;347:164551.