Best Practice & Research Clinical Obstetrics and Gynaecology 26 (2012) 197 208
Contents lists available at SciVerse ScienceDirect
Best Practice & Research Clinical
Obstetrics and Gynaecology
journal homepage: www.elsevier.com/locate/bpobgyn
4
Alternative approaches to cervical cancer screening
for developing countries
a,* b
Thomas C. Wright Jr., MD , Louise Kuhn, PhD
a
Department of Pathology, College of Physicians and Surgeons of Columbia University, Room 16-404, P&S Building,
630 W. 168th Street, New York, NY 10032, USA
b
Mailman School of Public Health and Gertrude H. Sergievsky Center, NY, USA
Cervical cancer remains the most common cancer among women
Keywords:
living in developing countries, largely because of the failure either
human papillomavirus
to initiate or sustain effective cervical-cancer screening pro-
cervical cancer prevention
screening grammes. This potentially preventable and curable cancer
cytology
continues to cause high mortality among relatively young women
visual inspection with acetic acid
residing in low-resource countries. Cytology as a screening test,
cervical intraepithelial neoplasia
linked with a robust healthcare infrastructure, has significantly
affected cervical cancer prevention in countries that have had
sufficient resources to establish and sustain well-conducted pro-
grammes. The failure to establish such programmes has stimulated
a large body of research into alternative screening tests and
approaches to cervical-cancer prevention. Two of the most recent
research methods have been visual inspection with acetic acid and
molecular testing for high-risk types of human papillomavirus
deoxyribonucleic acid. Visual inspection with acetic acid has
shown a great deal of promise in cross-sectional studies; however,
in randomised-controlled trials, it has been shown to be signifi-
cantly less effective in reducing cervical cancer or its precursors.
The development of point-of-care human papillomavirus or other
highly sensitive tests for the prevention of cervical cancer is
imperative. It has also been clearly shown that linking testing or
screening to treatment (so-called  screen and treat ) without the
intervention of colposcopy or the need for sophisticated labora-
tories may potentially prevent cervical cancer in large numbers of
women.
Ó 2011 Published by Elsevier Ltd.
* Corresponding author. Tel.: þ212 305 3991; Fax: þ212 305 1295.
E-mail address: [e-mail] (T.C. Wright).
1521-6934/$  see front matter Ó 2011 Published by Elsevier Ltd.
doi:10.1016/j.bpobgyn.2011.11.004
198 T.C. Wright Jr., L. Kuhn / Best Practice & Research Clinical Obstetrics and Gynaecology 26 (2012) 197 208
Introduction
Worldwide, cervical cancer is the third most common cancer among women and the fourth leading
cause of cancer death in women worldwide.1 It is estimated that annually there are about 529,828 cases
and 275,125 deaths from cervical cancer globally (Table 1).2 The numbers of cases and deaths from
cervical cancer, however, underestimates its global effect, as cervical cancer occurs in relatively young
women. Each death from invasive cervical cancer occurring before the age of 70 years accounts for an
average of 17 potential years of life lost.3 Globally, the mortality : incidence ratio is 52%.4 About 3.4
million women-years of life before age 70 years are lost annually from cervical cancer worldwide.4
Marked global disparities exist in both the incidence of, and mortality from, cervical cancer.5 More
than 85% of the cases of cervical cancer occur in developing countries where it accounts for 13% of all
cancers in women (Table 1).2 Highest risk regions include Eastern and Western Africa, where the age-
2
standardised incidence rates are over 30 per 100,000 (Fig. 1). Other high-risk regions with age-
standardised incidence rates of over 20 per 100,000 include South Africa, South-Central Asia, South
America, Melanesia, Middle Africa, Central America and the Caribbean. Rates are lowest in Western
Asia, Australia and New Zealand, and Northern America, where the age-standardised incidence rates
are less than 6 per 100,000. India, the second most populous country in the world, accounts for 27% of
all cervical cancer deaths globally. In large part, these global disparities in the burden of cervical cancer
reflect differences in cervical cancer screening rates as well as rates of infection with high-risk types of
HPV. Few developing countries have organised and quality assured cervical cancer-prevention pro-
grammes that allow the detection of high-grade cervical cancer precursors (referred to as cervical
intraepithelial neoplasia (CIN) grade 2 and 3) and early invasive cervical cancers before the develop-
ment of advanced disease.
Requirements for a successful cytology-based screening program and barriers found in low-
resource settings
Since the introduction of cervical cytology (e.g. the Papanicolaou test) in the 1950s, the approach to
cervical cancer prevention in most developed countries has been to screen women for CIN 2,3 lesions
and early invasive cervical cancers using cervical cytology. Screening is followed by an evaluation using
colposcopy and cervical biopsy for women with cytological abnormalities.6 If a CIN 2,3 lesion or early
invasive cancer is detected on the cervical biopsy, treatment is provided. This approach works because
it takes on average more than 10 years for a CIN 2,3 lesion to progress to an invasive cervical cancer,
giving ample opportunity to detect and treat the precursor lesion and prevent the development of
invasive cancer.7 In developed countries that have established screening programmes, and in which
women are repetitively screened at relatively frequent intervals, cytology-based cervical cancer
prevention has proven to be the single most successful cancer prevention programme.
Unfortunately, although simple in concept, cytology-based prevention programmes have proven
extremely difficult to implement in low-resource settings. Bishop et al.8,9 described many of the
Table 1
Estimated numbers of cases and deaths from cervical cancer in 2008.
Region Cases Deaths
World 529,828 275,125
More developed regions 76,000 32,000
Less developed regions 453,000 242,000
WHO Africa Region 75,000 50,000
WHO Americas Region 80,000 36,000
European Union 31,000 13,000
United States 11,000 3,000
China 75,000 33,000
India 134,000 72,000
WHO, World Health Organization. Adapted from Ferlay et al. Estimates of worldwide burden of cancer
in 2008: GLOBOCAN 2008. Int J Cancer 2010; 127: 2893 2917. By permission of Oxford Uniersity
Press.2
T.C. Wright Jr., L. Kuhn / Best Practice & Research Clinical Obstetrics and Gynaecology 26 (2012) 197 208 199
Fig. 1. Estimated cumulative incidence risk of cervical cancer up to the age of 74 Years. Ferlay et al. Estimates of worldwide burden of
cancer in 2008: GLOBOCAN 2008. Int J Cancer 2010; 127: 2893 2917. By permission of Oxford University Press.
challenges faced in implementing and sustaining cytology-based cervical-cancer prevention pro-
grammes in low-resource settings. Typically, the major barrier to cervical cancer prevention is not the
cost of the screening test, which is relatively inexpensive; instead it is the cost and complexity of
providing the infrastructure required for the screening programme. Not only must quality-assured
cytology laboratories be established, but the healthcare systems must also put into place the infra-
structure and capacity to track women with abnormal cytology as well as to provide quality assured
colposcopy, pathology, and treatment services. Each of these associated services has its own highly
specific infrastructure requirements. Cervical cytology is a good example of how complex a seemingly
simple test can be. Obtaining and evaluating the cytology specimen first requires that patients and
clinicians be made aware that screening is available and beneficial. Supplies need to be made available
where screening is carried out, and providers must be trained in speculum exams and obtaining the
specimens. Once the cytology specimen is collected, it has to be transported to the cytology laboratory
and stained. It then has to be screened by a trained cytotechnologist, whose training in most developed
countries takes about 2 years. Trained pathologists also need to be available who can review specimens
classified as abnormal by the cytotechnologists and make a final diagnosis. Maintaining a high-quality
cytology laboratory requires rigorous quality control mechanisms to assure that positive smears are not
inadvertently missed and that negative smears are correctly diagnosed as being negative. These
controls, in and of themselves, require extensive training, follow up, and oversight. Finally, once
a cytology result is available, the patient must be made aware of the result and whether or not col-
poscopy is needed. This requires an infrastructure for patient tracking and follow up. This example
serves to simply illustrate the complex training and infrastructure requirements for obtaining and
evaluating the Pap smear. The other components of an effective cervical cancer-prevention programme
include maintaining registries of women who have been screened; providing invitations to attend
screening; providing colposcopic services for evaluating women with abnormal screening test results;
treating women with preinvasive lesions; and finally providing medical, surgical, and terminal care for
women with invasive cervical cancer identified through the screening process. All have their own
infrastructure requirements.
Competing public health priorities are a major barrier to developing the infrastructure and capacity
described above, which is needed for cytology-based cervical cancer prevention programmes. For
example, in sub-Saharan Africa, other conditions such as communicable diseases including human
immunodeficiency virus (HIV) and acquired immune deficiency syndrome (AIDS), tuberculosis and
malaria, combined with maternal or perinatal complications, accounted for 70% of deaths in women in
1995.10 Despite the fact that the global AIDS epidemic seems to have stabilised and the annual number
of new HIV infections has been steadily declining since the late 1990s, the effort required to handle the
200 T.C. Wright Jr., L. Kuhn / Best Practice & Research Clinical Obstetrics and Gynaecology 26 (2012) 197 208
HIV and AIDS epidemic in sub-Saharan Africa has drained limited healthcare resources. It was esti-
mated that, in 2009, 5% of all adults aged 15 49 years living in sub-Saharan Africa were infected with
HIV and required ongoing and expensive care.11 High rates of communicable diseases are compounded
by widespread poverty in many developing countries. The United Nations Children s Fund estimates
that almost 50% of the developing world s population (2.5 billion people) lack improved sanitation
facilities, and over 884 million people still use unsafe drinking water sources.12 Primary healthcare
facilities, which are where preventative healthcare services, including cervical cancer screening, should
be situated, are poorly resourced and over-burdened in most developing countries. Finally, an urban
rural bias exists in health-care service provision in developing countries, which means that large
segments of the population lack access to health care. For example, in sub-Saharan Africa 87% of the
urban population has access to a health care, but over 50% of the region s population is non-urban and
lives more than 10 km from a primary healthcare facility.13
It should also be stressed that even if a developing country found the resources to establish a high-
quality cytology laboratory and had the capacity for follow up and treatment of abnormal results, the
programme would be only moderately effective in reducing death from invasive cervical cancer if
women were screened only once or twice in their lifetime. This is because the sensitivity of a single
cervical cytology is relatively low. Meta-analyses of the performance of single cervical cytology find
a sensitivity of 49 51% for CIN 2 and 3 and cancer (CIN2þ).14,15 Although these meta-analyses can been
criticised because they include a number of older studies, recent large screening studies that have used
contemporary high-quality cytology laboratories have confirmed that the sensitivity of a single cervical
cytology for CIN2þ is only 53 57%.16 18
Alternative approaches to cervical cancer prevention
Because cytology-based cervical cancer prevention programmes have proven so difficult to estab-
lish in low-resource settings, over the past 15 years there has been considerable interest in the global
health community in developing new strategies for cervical cancer prevention specifically for low-
resource settings. These include changing the screening test that is used as well as simplifying how
we evaluate and treat screen-positive women.
Alternative screening tests
Visual inspection with acetic acid
Two alternatives to cervical cytology have been evaluated in large studies in developing countries.
These are visual inspection with acetic acid (VIA) and molecular testing for high-risk types of human
papillomavirus (HPV). Visual inspection with acetic acid is the simplest, and the low unit cost and the
cost of the test itself makes it the least inexpensive screening test. Visual inspection with acetic acid is
based on the principle that most CIN 2,3 lesions are acetowhite (i.e. they develop a white colour when
vinegar or 5% acetic acid is applied to the cervical epithelium that harbors preinvasive lesions). This
means that mid-level providers can screen for CIN 2 and 3 lesions by conducting a speculum exami-
nation to visualise the cervix, applying vinegar to the cervix, and looking for areas of acetowhitening.
Normal cervical epithelium is pink, but most CIN 2 and 3 lesions will be white. The clinician conducting
the test then classifies a woman as screen positive (she has an acetowhite lesion); screen negative, (she
has no acetowhite lesions); or as suspicious for invasive cervical cancer if a necrotic, ulcerated, or
exophytic cervical lesion is found. Although this sounds quite simple, it is actually more complicated
than it appears. First, acetowhitening is a relatively non-specific cervical finding. Many women without
CIN 2 and 3 lesions have acetowhitening of the cervix owing to areas of immature squamous meta-
plasia or reparative conditions. This can lead to a considerable degree of over-referral and over-
treatment of women who do not have CIN2þ lesions. To minimise the amount of over-referral and
over-treatment of women with CIN2þ, over the years a number of criteria have been developed to try
and improve the specificity of VIA, and comprehensive training guides have been developed for mid-
level providers.19 21 Training typically takes 1 2 weeks and is competency based. The recommended
courses for providers allow the participants to learn by doing, and focuses on specific knowledge,
attitudes, and skills needed to perform VIA in low-resource settings. At the end of the training,
T.C. Wright Jr., L. Kuhn / Best Practice & Research Clinical Obstetrics and Gynaecology 26 (2012) 197 208 201
a competency-based performance evaluation is used to assess the participants performance in
VIA.20,21
The performance of VIA has been evaluated in numerous cross-sectional and prospective studies, as
well as in a meta-analysis. A recent large meta-analysis of VIA studies published through 2010 divided
the studies based on patient population (e.g. asymptomatic women undergoing screening versus
symptomatic women) and whether or not all women underwent a gold standard of colposcopy and
histology.22 A total of 26 studies enrolled only asymptomatic women and all screened women
underwent colposcopy and histology. In these studies the pooled sensitivity of VIA for CIN2þ was 80%
and the specificity was 92%, with a positive predictive value (PPV) of 10% and a negative predictive
value (NPV) of 99%.
When interpreting the results of the VIA studies, it is important to recognise that most studies were
large screening studies conducted in low-resource settings and many used local pathologists to
diagnose the cervical biopsies. In addition, many of the pathologists were not blinded to the colpo-
scopic findings. This can lead to quite significant problems in measuring the performance of a screening
test, as a combination of colposcopy and cervical biopsy is an imperfect gold standard. It is now rec-
ognised that a single colposcopy examination misses about one-third of CIN 2 and 3 lesions.23,24 In
addition, the pathological interpretation of cervical biopsies is highly subjective, and studies have
shown poor inter-observer and intra-observer reproducibilty of a diagnosis of CIN 2,3.25,26 When an
imperfect gold standard is used to determine disease status, significant bias can be introduced into
measures of a test s performance.27 The direction of the bias is determined by whether or not the
imperfect gold standard and the test being evaluated tend to err in the same patients. If a test and the
imperfect standard have a tendency to err in the same patients, the test s performance characteristics
can be significantly overestimated.28 Previously, we have shown that that classification errors occur-
ring with VIA correlate with histopathologic classification errors.29 What happens is that women with
areas of immature squamous metaplasia are classified as VIA positive, as these areas are frequently
acetowhite. When colposcopy is carried out, these areas of immature squamous metaplasia are
frequently biopsied, again because they are acetowhite. Unfortunately, as pathologists can also find it
difficult to differentiate between immature squamous metaplasia and CIN 2 and 3 on the biopsy, some
of the biopsies from areas of immature squamous metaplasia are misclassified as CIN 2 and 3.7 The
correlation between the VIA result and the erroneous biopsy result would increase the sensitivity of
VIA and decrease the sensitivity of cytology and HPV deoxyribonucleic acid (DNA) testing.29 In another
study,30 it was shown that the sensitivity of VIA was overestimated by 20% when colposcopy-directed,
rather than random cervical biopsies were used as the gold standard.30
A high degree of variability is found in different studies that have measured the sensitivity of VIA. For
example, in one study of 11 separate screening studies conducted by the International Agency for Research
in Cancer, the sensitivity of VIA for CIN 2 and 3 varied from 56.1% and 58% in Kolkata and Mumbai,
respectively, to 93.9% and 90.3% in Burkina Faso and Guinea, respectively.31 In this same study, specificity of
VIA for CIN 2 and 3 varied from 75.1% and 76.6% in Jaipur and Congo, respectively, to 93.2% and 93.8% in
Guinea and Niger, respectively.31 This high degree of variation occurred despite the fact that all of the
providers conducting VIA underwent a similar intensive training course. In two different South African
screening studies in which providers who underwent similar intensive training and were supervised by
the same clinician carried out VIA in both, the sensitivity of VIA for CIN 2þ was found to be 55% and 73%.32,33
Although it is difficult to know what the real performance will be of VIA when used in a developing
country screening programme, considerable evidence currently shows that VIA should be considered
as a reasonable alternative to cervical cytology. The advantages of VIA compared with cytology are that
it seems to be somewhat more sensitive (about 50 70%) sensitivity for VIA compared with about 50%
for cytology; it is a point-of-care test, which means that women can be immediately informed if they
are screening test positive; and it does not require a cytopathology laboratory. The disadvantages of VIA
compared with cytology are that it is much more difficult to provide quality control with VIA than it is
with cytology, as rescreening is not easily done; it requires that a considerable number of providers be
trained in VIA, whereas cytology requires only a limited number of cytotechnologists; its performance
declines substantially in women aged 40 years and older34,35; and its specificity is lower than that of
cytology. The variability in the performance of VIA observed in different studies, however, does not
seem to be any greater than that of cervical cytology.18
202 T.C. Wright Jr., L. Kuhn / Best Practice & Research Clinical Obstetrics and Gynaecology 26 (2012) 197 208
High-risk human papillomavirus testing
It is now accepted that infection with one of 14 oncogenic HPV genotypes (referred to as  high-risk
genotypes) is required for the development of CIN 2 and 3 and invasive cervical cancer.36,37 This
underlying causal association has spurred considerable interest in using high-risk HPV (hrHPV) testing
to improved cervical-cancer screening, and has resulted in commercially available sensitive molecular
tests to detect hrHPV genotypes in clinical specimens.16 A number of large clinical trials and cost-
effectiveness analyses have firmly established the clinical utility of using hrHPV testing to improve
the sensitivity of cervical screening in women aged 30 years and over undergoing routine screening.
Multiple studies have convincingly shown that hrHPV testing is more sensitive and more reproducible
than cervical cytology.17,18,38 40 Moreover, because of the high negative predictive value of hrHPV
testing, women who are hrHPV negative do not need to be rescreened before 6 years.36,41 Over the past
several years, evidence has grown that HPV DNA testing should be considered as an adjunct to cervical
cytology when screening, and also as a true replacement for cervical cytology. Several large screening
trials conducted in Europe and the USA have shown that combining cervical cytology with hrHPV
testing adds little compared with using hrHPV testing alone.16,18,40,41 Recently, an Italian screening
study40 showed a lower occurrence of invasive cervical cancer in women screened using hrHPV testing
compared with cytology alone. Currently, several European countries are beginning to transition from
a cytology-based screening programme to a hrHPV-based one. A potential advantage of hrHPV testing
over other screening methods for developing countries is that testing can be done on self-collected
vaginal swabs.42 In 2007, Petignat et al.43 conducted a systematic review of studies that compared
the prevalence of HPV infection detected with cervical versus self-collected vaginal swabs and
concluded that the two methods produced comparable results.43 More recently, Gravitt et al.44 con-
ducted another systematic review and reached the same conclusion.44 In our South African experience,
however, we have uniformily found that hrHPV testing of self-collected samples has both a lower
sensitivity and lower specificity for CIN 2 and 3 compared with clinician-collected samples.42,45 In
these studies, the sensitivity of hrHPV testing using a self-collected specimen is roughly equivalent to
that of cervical cytology.
The greatest issues in hrHPV testing for developing countries are the unit cost of the test, the
sophistication of the laboratories that are needed for carrying out current versions of the test, and the
fact that a point-of-care hrHPV test is not yet available. The current generation of commercially
available hrHPV assays use various amplification methods for detecting either hrHPV DNA or RNA.
Globally, the most widely used hrHPV test is the Hybrid CaptureÒ 2 HPV Assay (HC2) (Qiagen, Gai-
thersburg, MD, USA). Hybrid CaptureÒ 2 HPV uses signal amplification methodology to detect 13 hrHPV
genotypes (16, 18, 31, 33, 35, 39, 45, 51, 52, 56, 58, 59, and 68) as a pooled mixture. It was developed
over several years and, during development, it underwent clinical validation in large clinical trials in
developed and developing countries. As a result of these trials, the sensitivity of HC2 was adjusted to
provide a reasonable trade-off between sensitivity for CIN3þ lesions and positivity in women without
cervical disease. The latter is extremely important in reducing unnecessary or redundant follow-up
procedures for women who are hrHPV positive without cervical disease.46,47 Therefore, the perfor-
mance of HC2 in a screening setting is widely viewed by experts in Europe and North America as the
benchmark by which other hrHPV tests should be judged.48 It has been proposed that any other
candidate hrHPV test should have a sensitivity for CIN2þ of 90% that of HC2 as assessed by a non-
inferiority score test.48 Furthermore, the candidate hrHPV test should have at least 98% of the speci-
ficity of HC2 as assessed by a non-inferiority test. Recently, several genomic amplification hrHPV tests
have become commercially available. These tests use either the polymerase chain reaction or other
techniques to amplify DNA or RNA. Provided the tests have been clinically validated in a sufficiently
large general screening population, and been shown to have a sensitivity and specificity comparable to
that of HC2, there is no reason that these newer tests should not perform reasonably for cervical cancer
screening.
A new hrHPV test that is modelled on HC2 has been developed in collaboration between Qiagen (the
company that makes HC2), the Bill and Melinda Gates Foundation, and PATH (a NGO based in Seattle,
WA). This test that is called careHPVÔ, and is specifically designed for use in low-resource settings. It
does not require reagent refrigeration, can be carried out in a clinic setting by a healthcare worker with
minimal laboratory training, and takes only 2.5 h or less to complete.49 In 2008, careHPVÔ was directly
T.C. Wright Jr., L. Kuhn / Best Practice & Research Clinical Obstetrics and Gynaecology 26 (2012) 197 208 203
compared with HC2 and was shown to have a somewhat lower sensitivity for CIN2þ than HC2 (90% v
97%). The specificity of the two tests, however, was equivalent (84% v 86%).50 It is expected that
careHPVÔ will be made available to countries introducing nationwide HPV-based screening pro-
grammes at a much lower cost than other hrHPV tests. When this occurs, it will overcome one of the
major barriers to the introduction of HPV-based cervical cancer screening programmes in low-resource
settings. careHPVÔ is expected to become available in 2012.49
Alternative approaches to managing screen-positive women
The approach used in most developed countries to manage screen-positive women is to initially
carry out colposcopy and obtain cervical biopsies of all lesions. If the biopsies are found to have CIN2þ,
the woman undergoes treatment.51 Treatment for CIN 2 and 3 is typically loop electrosurgical excision
procedure or large loop excision of the transformation zone, whereas treatment for early invasive
cancers is generally surgical. Many developing countries do not have the capacity to carry out colpo-
scopy on screen-positive women. Moreover, many have a limited number of pathology laboratories and
pathologists to process and evaluate cervical biopsies. In rural settings, the need to track and recall
screen-positive women to provide them with their test results is also a barrier. Therefore, an alternative
approach to managing screen-positive women is needed if large-scale screening programmes are going
to be expanded into developing countries. These alternative approaches to managing screen-positive
women in developing countries have to take into account two competing priorities. One is assuring
that the cervical cancer prevention programme is as effective as possible in reducing death from
invasive cervical cancer. The other is to minimise over-treatment of women without CIN2þ lesions.
Unfortunately, at the current time, none of the current approaches are suitable for low-resource
settings that can fulfill both priorities. Therefore, public health authorities have to decide which
priority is most important in a given setting: preventing as many deaths as possible from cervical
cancer or minimising over-treatment of women without CIN2þ lesions.
The simplest approach would be to use the most sensitive screening test possible and ablate the
transformation zone in all women in whom the screening test is positive. Ablation of the trans-
formation zone is typically achieved using cryotherapy, and is generally considered to be quite safe and
to have only minor complications.52 54 This approach is referred to as  screen-and-treat .55 The
advantages of  screen-and-treat are that it produces the greatest reduction in CIN2þ. The reduction in
CIN 2 and 3 to be expected from a  screen-and-treat programme is simply the (sensitivity of the
screening test) x (success rate of the ablative method in eliminating CIN 2,3).56 Therefore, if a highly
sensitive screening test such as hrHPV testing is used that has a sensitivity of 90% for CIN 2 and 3
lesions, and this test is coupled with cryotherapy to ablate the transformation zone in all screen-
positive women (which eliminates about 80% of CIN 2 and 3 lesions), the overall reduction in CIN 2
and 3 achieved with the programme would be about 72%. Another advantage of  screen-and-treat is
that, if a point-of-care screening test is used, the entire process can be completed in a single visit and
there is no need to establish a track and recall system for patients. The major disadvantage of  screen-
and-treat is that, depending on the specificity of the screening test, 5 20% of all women without CIN 2
and 3 lesions will undergo unnecessary treatment.
In order to reduce the over-treatment of women without CIN 2 and 3, many demonstration projects
are using either colposcopy or, if hrHPV is used as the screening test, VIA to identify a subset of screen-
positive women for treatment who are at highest risk for having CIN2þ. We have previously referred to
this approach as a  two-stage screening approach .56 The expected reduction in CIN 2 and 3 to be
expected from a  two-stage screening programme that incorporates an intermediate triage step to
select a subset of screen-positive women for treatment is (sensitivity of the screening test) x (sensi-
tivity of the triage method for detecting CIN 2 and 3) x (success rate of the ablative method in elim-
inating CIN 2 and 3). Therefore, if hrHPV testing is used to screen and then hrHPV positive women then
undergo VIA, which has a sensitivity of 60% for CIN 2 and 3 to select a subset of women for cryotherapy,
the overall reduction in CIN 2 and 3 of the programme would be only 43%. Thus, the reduction in CIN 2
and 3 achieved with a  two-stage screening programme is greatly reduced compared with a  screen-
and-treat programme. The  two-stage screening approach does reduce unnecessary treatments. It is
important for policy makers to recognise that the sensitivity of colposcopy is considerably lower than
204 T.C. Wright Jr., L. Kuhn / Best Practice & Research Clinical Obstetrics and Gynaecology 26 (2012) 197 208
previously thought.23,24 Most studies have found that a single colposcopic examination misses about
one-third of CIN 2 and 3 lesions.23,24,57 Therefore, colposcopy seems to offer little advantage over VIA as
a triage approach for screen-positive women.
Outcomes of large screening trials
Over the past decade, a number of screening trials evaluating a variety of different approaches have
been conducted in low-resource settings. The results of these trials have frequently been conflicting,
which emphases how hard it is to conduct large-scale screening programmes in low-resource settings
and the difficulty of monitoring the effectiveness of screening using end points other than cancer
incidence and mortality. The results of just three of the many trials are provided in this section. One
was a randomised trial of  screen-and-treat carried out in South Africa, the other two were
community-cluster, randomised-controlled trials of  two-stage screening carried out in rural
India.34,57,58
The earliest of the Indian trials had 57 study clusters randomised to a single round of VIA carried out
by trained nurses and 57 study clusters randomised to a control group (no intervention).34 Women
aged 30 59 years were eligible for screening in the VIA arm. In the screening arm, women who were
VIA positive underwent colposcopy and directed biopsies. On the basis of the colposcopic appearance
of the cervix, women with a low-grade or high-grade CIN that involved less than three quadrants of the
cervix and did not extend into the endocervical canal were offered immediate cryotherapy, which was
carried out by nurses. Women with more advanced lesions were referred for a loop electrosurgical
excision procedure. In the clusters randomised to screening, 31,343 (63.6% of all eligible women)
underwent screening. In the VIA arm, 9.9% of all screened women screened were VIA positive and
offered treatment. During 6 years of follow up (2000 2006), cervical cancer incidence seen in the
clusters randomised to screening reduced by 25% compared with the control arm.34
More recently Sankaranarayanan et al.58 reported on the other cluster-randomised Indian screening
trial.58 This trial included 131,746 women aged 30 59 years of age who were assigned to four groups:
cytology screening, VIA screening, hrHPV screening, and a control group (no screening). As in the other
Indian trial, women with positive screening tests ( ASCUS or hrHPV or VIA positive) were referred to
colposcopy and were offered immediate cryotherapy if a low-grade or high-grade CIN suitable for
cryotherapy was found on colposcopy. During follow-up, a significant reduction in the incidence of
stage II or higher cervical cancer as well as death from cervical cancer was found in the hrHPV
screening arm compared with the control arm. No reduction in either cancer incidence or death from
cancer was found in either the VIA or cytology screening arms (Table 2).
Recently, we completed 3 years follow up of our South African, three-arm randomised  screen-and-
treat trial. In this study, 6,637 women aged 35 65 years of age were randomised to receive either
hrHPV-based  screen-and-treat , VIA-based  screen-and-treat , or to be in a control (no intervention for
6 months) arm. After 6 months follow up, all women in the three arms underwent colposcopy and
cervical biopsy. After 36 months follow up there was a sustained, significant, 3.6-fold decrease in the
cumulative prevalence of CIN2þ in the hrHPV arm (1.5%; 95% CI. 0.9-2.2) compared with either the
control (5.6%; 95% CI 4.6 to 6.7%) or VIA arm (3.8%; 95% CI 2.9 to 4.7%) (Fig. 2).57 New cases of CIN 2 and 3
(identified at 12 months or over) were less common among the hrHPV arm (0.3%, 95% CI 0.05 to 1.02)
than control (1.0%, 95% CI 0.5 to 1.7) or VIA arms (1.3%, 95% CI 0.8 to 2.1). In this study, the risk benefit
Table 2
Incidence of stage II or greater cervical cancer and cancer mortality
Randomised arm
hrHPV Cytology VIA Control
Incidence of stage II or greater cervical cancera 14.5 23.2 32.2 33.1
Death from cervical cancera 12.7 21.5 20.9 25.8
a
Rate per 100,000 person years. hrHPV, high-risk human papillomavirus; VIA, visual inspection with acetic acid. Adapted
from reference 58.
T.C. Wright Jr., L. Kuhn / Best Practice & Research Clinical Obstetrics and Gynaecology 26 (2012) 197 208 205
Fig. 2. Cumulative detection of cervical intraepithelial neoplasia up to 36 months of follow-up in women enrolled in South African
 screen-and-treat study. Published with permission from reference 57. CIN, cervical intraepithelial neoplasia.
analysis of  screen-and-treat was quite favourable. Cryotherapy was, in general, well tolerated and,
although minor complaints such as bleeding an discharge were common after cryotherapy, significant
complications were quite rare.33 Out of 949 women treated with cryotherapy, there was only a single
significant complication, which was bleeding from the cervix of an immunosuppressed woman
infected with HIV. Moreover, 99% of participants stated they would recommend this type of  screen-
and-treat programme to their friends and family.33 A similar low complication rate and high accep-
tance rate of  screen-and-treat was found in a large trial in Thailand.55
Conclusion
Progress is clearly being made in developing alternative cervical cancer screening approaches that
are appropriate for developing countries. Advances have been made in how we screen and how those
tested positive after screening are evaluated and treated. Randomised-controlled trials carried out in
developing countries clearly show reductions in CIN2þ prevalence and incidence, with both VIA-based
and hrHPV-based screening programmes. Perhaps, more importantly, a recent Indian community
cluster-randomisation trial has, for the first time, shown that a single round of hrHPV-based screening
206 T.C. Wright Jr., L. Kuhn / Best Practice & Research Clinical Obstetrics and Gynaecology 26 (2012) 197 208
can reduce the incidence of, and death from, invasive cervical cancer.58 In these trials, the benefits of
hrHPV-based programmes seem to be significantly greater than the benefits of VIA-based programmes,
and a top priority should be the development and widespread introduction of inexpensive, point-of-
care hrHPV diagnostic tests.
Practice points
Cervical cancer remains the most common cancer cause of death among women in devel-
oping countries.
Establishing national cervical cancer screening programmes in developing countries has
failed as a result of the complexity of the infrastructure required and the intensity of
competing health needs.
Most people living in developing countries still do not have access to basic necessities such as
clean water.
Cervical cytology carried out once has a sensitivity of around 50%; its success in reducing
cervical cancer in developed countries was a result of the test being used repetitively over
time
The two most studied alternatives to cytology in the past 15 years are VIA and testing hrHPV
testing with a test known as HC2.
The advantages of VIA are that it is a low-cost, point-of-care test, using low technology;
however, it has a poor positive predictive value, leading to significant over-treatment or over-
referral of women with positive tests. In addition, establishing reliable quality control is
difficult.
hR-HPV DNA testing has been shown in numerous studies to have a high sensitivity (over 90%
in most studies) but a lower specificity than cytology.
hR-HPV DNA testing combined with immediate ablation of the transformation zone using
cryotherapy has been shown in large randomised-controlled trials in India and South Africa
to significantly reduce the incidence of cervical cancer precursors and cervical cancer.
Research agenda
The development of reliable point-of-care molecular tests (i.e. HC2) that allows women to be
 screened and treated at a single visit is urgently needed.
Test qualities that need attention to reduce over-treatment, over-referral, or both, for col-
poscopy include higher specificity and positive predictive value.
Recognising the multiple barriers to establishing cervical cancer screening programmes in
poor countries, approaches to preventing cancer should focus on the tests used and the
programmatic research to maximise implementation.
Establishing mechanisms for monitoring and evaluating cervical cancer screening pro-
grammes in developing countries is critical and should be linked to the development of
cancer registries and national cancer-control programmes.
Appropriate advocacy programmes that alert health authorities to the public health signifi-
cance of cervical cancer need to be developed.
References
1. Jemal A, Bray F, Center MM et al. Global cancer statistics. CA Cancer J Clin 2011; 61: 69 90.
2. Ferlay J, Shin HR, Bray F et al. Estimates of worldwide burden of cancer in 2008: GLOBOCAN 2008. Int J Cancer 2010; 127:
2893 2917.
3. Yang BH, Bray FI, Parkin DM et al. Cervical cancer as a priority for prevention in different world regions: an evaluation
using years of life lost. Int J Cancer 2004; 109: 418 424.
T.C. Wright Jr., L. Kuhn / Best Practice & Research Clinical Obstetrics and Gynaecology 26 (2012) 197 208 207
4. WHO. Burden of cervical cancer globally. WHO/ICO Information Center on HPV and Cervical Cancer, 2011.
5. Castellsague X, de Sanjose S, Aguado T et al. HPV and cervical cancer in the world 2007 report. Vaccine 2007: 25.
6. Wright Jr. TC. Cervical cancer screening in the 21st century: is it time to retire the PAP smear? Clin Obstet Gynecol 2007; 50:
313 323.
7. Wright TC, Ronnett BM, Kurman RJ et al. Precancerous lesions of the cervix. In Kurman RJ, Ellenson LH & Ronnet BM (eds.).
Blaustein s Pathology of the Female genital tract. 6th edn. New York: Springer-Verlag, 2011, pp. 193 252.
8. Bishop A, Sherris JD & Tsu V. Cervical dysplasia treatment in developing countries: a situation analysis. Seattle, WA: PATH, 1995.
*9. Bishop A, Wells E, Sherris J, Tsu V et al. Cervical cancer: evolving prevention strategies for developing countries. Repro-
ductive Health Matters 1995; 6: 60 71.
10. Denny L, Quinn M & Sankaranarayanan R. Chapter 8: Screening for cervical cancer in developing countries. Vaccine 2006;
24(Suppl. 3). S3/71 S3/77.
11. UNAIDS Report on the Global AIDS Epidemic 2010. 2010; http://www.unaids.org/documents/20101123_GlobalReport_
Chap2_em.pdf [last accessed 16.11.11].
12. UNICEF. Water, sanitation and hygeine 2011, http://www.unicef.org/wash/ [last accessed 16.11.11].
13. Medicine CtSFMaMiS-SAIo. In her lifetime: female morbidity and mortality in sub-Saharian Africa. Washington, DC: National
Academy Press, 1996.
14. Fahey MT, Irwig L & Macaskill P. Meta-analysis of Pap test accuracy. Am J Epidemiol 1995; 141: 680 689.
15. Nanda K, McCrory DC, Myers ER et al. Accuracy of the Papanicolaou test in screening for and follow-up of cervical cytologic
abnormalities: a systematic review. Ann Intern Med 2000; 132: 810 819.
16. Castle PE, Stoler MH, Wright Jr. TC et al. Performance of carcinogenic human papillomavirus (HPV) testing and HPV16 or
HPV18 genotyping for cervical cancer screening of women aged 25 years and older: a subanalysis of the ATHENA study.
Lancet Oncol 2011; 12: 880 890.
17. Mayrand MH, Duarte-Franco E, Rodrigues I et al. Human papillomavirus DNA versus Papanicolaou screening tests for
cervical cancer. N Engl J Med 2007; 357: 1579 1588.
18. Cuzick J, Clavel C, Petry KU et al. Overview of the European and North American studies on HPV testing in primary cervical
cancer screening. Int J Cancer 2006; 119: 1095 1101.
19. Sankaranarayanan R & Wesley R. A practical manual on visual screening for cervical neoplasia. Thiruvananthapuram, India:
Regional Cancer Centre: International Agency for Research on Cancer, 2001.
20. Sellors JW, Camacho CK, Winkler J et al. Teaching visual inspection of the cervix with acetic acid (VIA). Curriculum document.
Seattle, Washington: PATH, 2003.
21. JHPIEGO. Cervical cancer prevention course notebook for trainers: Global draft. Baltimore, Maryland: JHPIEGO, 2001.
*22. Sauvaget C, Fayette JM, Muwonge R et al. Accuracy of visual inspection with acetic acid for cervical cancer screening. Int J
Gynaecol Obstet 2011; 113: 14 24.
23. Cox JT, Schiffman M & Solomon D. Prospective follow-up suggests similar risk of subsequent cervical intraepithelial
neoplasia grade 2 or 3 among women with cervical intraepithelial neoplasia grade 1 or negative colposcopy and directed
biopsy. Am J Obstet Gynecol 2003; 188: 1406 1412.
24. Pretorius RG, Zhang WH, Belinson JL et al. Colposcopically directed biopsy, random cervical biopsy, and endocervical
curettage in the diagnosis of cervical intraepithelial neoplasia II or worse. Am J Obstet Gynecol 2004; 191: 430 444.
25. Stoler MH & Schiffman M. Interobserver reproducibility of cervical cytologic and histologic interpretations: realistic
estimates from the ASCUS-LSIL Triage Study. JAMA 2001; 285: 1500 1505.
26. Galgano MT, Castle PE, Atkins KA et al. Using biomarkers as objective standards in the diagnosis of cervical biopsies. Am J
Surg Pathol 2010; 34: 1077 1087.
27. Valenstein PN. Evaluating diagnostic tests with imperfect standards. Am J Clin Pathol 1990; 93: 252 258.
28. Valenstein PN & Emancipator K. Sensitivity, specificity, and reproducibility of four measures of laboratory turnaround
time. Am J Clin Pathol 1989; 91: 452 457.
29. Zhang Q, Kuhn L, Denny LA et al. Impact of utilizing p16INK4A immunohistochemistry on estimated performance of three
cervical cancer screening tests. Int J Cancer 2007; 120: 351 356.
*30. Pretorius RG, Bao YP, Belinson JL et al. Inappropriate gold standard bias in cervical cancer screening studies. Int J Cancer
2007; 121: 2218 2224.
31. Sankaranarayanan R, Chatterji R, Shastri SS et al. Accuracy of human papillomavirus testing in primary screening of
cervical neoplasia: results from a multicenter study in India. Int J Cancer 2004; 112: 341 347.
32. Denny L, Kuhn L, Pollack A et al. Direct visual inspection for cervical cancer screening: an analysis of factors influencing
test performance. Cancer 2002; 94: 1699 1707.
*33. Denny L, Kuhn L, De Souza M et al. Screen-and-treat approaches for cervical cancer prevention in low-resource settings:
a randomized controlled trial. JAMA 2005; 294: 2173 2181.
*34. Sankaranarayanan R, Esmy PO, Rajkumar R et al. Effect of visual screening on cervical cancer incidence and mortality in
Tamil Nadu, India: a cluster-randomised trial. Lancet 2007; 370: 398 406.
35. Li N, Shi JF, Franceschi S et al. Different cervical cancer screening approaches in a Chinese multicentre study. Br J Cancer
2009; 100: 532 537.
36. Schiffman M, Wentzensen N, Wacholder S et al. Human papillomavirus testing in the prevention of cervical cancer. J Natl
Cancer Inst 2011; 103: 368 383.
37. IARC. Human papillomaviruses. Lyon: IARC, 2007.
*38. Cuzick J, Arbyn M, Sankaranarayanan R et al. Overview of human papillomavirus-based and other novel options for
cervical cancer screening in developed and developing countries. Vaccine 2008; 26(Suppl. 10): K29 K41.
39. Naucler P, Ryd W, Tornberg S et al. Human papillomavirus and Papanicolaou tests to screen for cervical cancer. N Engl J Med
2007; 357: 1589 1597.
40. Ronco G, Giorgi-Rossi P, Carozzi F et al. Efficacy of human papillomavirus testing for the detection of invasive cervical
cancers and cervical intraepithelial neoplasia: a randomised controlled trial. Lancet Oncol 2010; 11: 249 257.
41. Dillner J, Rebolj M, Birembaut P et al. Long term predictive values of cytology and human papillomavirus testing in cervical
cancer screening: joint European cohort study. BMJ 2008; 337: a1754.
208 T.C. Wright Jr., L. Kuhn / Best Practice & Research Clinical Obstetrics and Gynaecology 26 (2012) 197 208
42. Wright Jr. TC, Denny L, Kuhn L et al. HPV DNA testing of self-collected vaginal samples compared with cytologic screening
to detect cervical cancer. JAMA 2000; 283: 81 86.
43. Petignat P, Faltin DL, Bruchim I et al. Are self collected samples comparable to physician-collected cervical specimens for
HPV DNA testing? Gynecol Oncol 2007; 105: 530 535.
*44. Gravitt PE, Belinson JL, Salmeron J et al. Looking ahead: a case for HPV testing of self-sampled vaginal specimens as
a cervical cancer screening strategy. Int J Cancer 2011; 129: 517 527.
*45. Taylor S, Wang C, Wright TC et al. A comparison of human papillomaviirus testing of clinician-collected and self-collected
samples during follow-up after screen-and-treat. Int J Cancer 2011; 129: 879 886.
46. Stoler MH, Castle PE, Solomon D et al. The expanded use of HPV testing in gynecologic practice per ASCCP-guided
management requires the use of well-validated assays. Am J Clin Pathol 2007; 127: 335 337.
47. Kinney W, Stoler MH & Castle PE. Special commentary: patient safety and the next generation of HPV DNA tests. Am J Clin
Pathol 2010; 134: 193 199.
48. Meijer CJ, Berkhof J, Castle PE et al. Guidelines for human papillomavirus DNA test requirements for primary cervical
cancer screening in women 30 years and older. Int J Cancer 2009; 124: 516 520.
49. Qiagen. Expanding access to cervical cancer screening 2009, http://wwwqiagencom/about/whoweare/qiagencares/the-
carehpv-testpdf [accessed 04.09.11].
50. Qiao YL, Sellors JW, Eder PS et al. A new HPV-DNA test for cervical-cancer screening in developing regions: a cross-
sectional study of clinical accuracy in rural China. Lancet Oncol 2008; 9: 929 936.
51. Cox T. Management of cervical intraepithelial neoplasia. Lancet 1999; 353: 857 858.
52. Kalliala I, Nieminen P, Dyba T et al. Cancer free survival after CIN treatment: comparisons of treatment methods and
histology. Gynecol Oncol 2007; 105: 228 233.
53. Ferenczy A. Comparison of cryo- and carbon dioxide laser therapy for cervical intraepithelial neoplasia. Obstet Gynecol
1985; 66: 793 798.
54. Castro W, Gage JC, Gaffikin L et al. Effectiveness, safety, and acceptability of cryotherapy: a systematic literature review.
Seattle, WA: PATH, 2003.
55. Gaffikin L, Blumenthal PD, Emerson M et al. Safety, acceptability, and feasibility of a single-visit approach to cervical-
cancer prevention in rural Thailand: a demonstration project. Lancet 2003; 361: 814 820.
56. Denny L, Kuhn L, Risi L et al. Two-stage cervical cancer screening: an alternative for resource-poor settings. Am J Obstet
Gynecol 2000; 183: 383 388.
*57. Denny L, Kuhn L, Hu CC et al. Human papillomavirus-based cervical cancer prevention: long-term results of a randomized
screening trial. J Natl Cancer Inst 2010; 102: 1557 1567.
*58. Sankaranarayanan R, Nene BM, Shastri SS et al. HPV screening for cervical cancer in rural India. N Engl J Med 2009; 360:
1385 1394.