This document summarizes highlights from the 18th International Papillomavirus Conference held in Barcelona, Spain in July 2000. It discusses several key topics: 1) Epidemiological studies presented data on HPV prevalence by age and risk factors like oral contraceptive use. 2) Studies of HPV serology found correlations between antibody levels and sexual behavior or HPV infection status. 3) Research on potential cofactors in cervical cancer like chlamydia infection presented both supporting and non-supporting evidence for their role in carcinogenesis.

1. J Clin Pathol 2001;54:163–175 163
Leaders
Papillomavirus research update: highlights of the
Barcelona HPV 2000 international papillomavirus
conference
F X Bosch, T Rohan, A Schneider, I Frazer, H Pfister, X Castellsagué, S de Sanjosé,
V Moreno, L M Puig-Tintore, P G Smith, N Muñoz, H zur Hausen
Servei Abstract www.hpv2000.com and the numbers in paren-
d’Epidemiologia i The 18th international papillomavirus thesis in the text below refer to the abstract
Registre del Càncer, conference took place in Barcelona, Spain number on the website. Key publications for
Institut Català
d’Oncologia, Av Gran in July 2000. The HPV clinical workshop further reading are listed at the end of this
Via KM 2,7, E.08907 was jointly organised with the annual paper.1–19
L’Hospitalet de meeting of the Spanish Association of Cer- The opening session was given by PG Smith
Llobregat, Barcelona, vical Pathology and Colposcopy. The con- (London School of Hygiene and Tropical
Spain ference included 615 abstracts describing Medicine) who lectured on vaccination trials in
F X Bosch ongoing research in epidemiology, developing countries. Professor Smith stressed
X Castellsagué
S de Sanjosé diagnosis/screening, treatment/prognosis, that the next decade holds the promise of new
V Moreno immunology/human immunodeficiency vaccines against some of the major infectious
virus, vaccine development/trials, diseases aVecting mankind. Most of the
Department of transformation/progression, replication, research and deployment eVorts are, and will
Epidemiology and transcription/translation, viral protein continue to be, directed at vaccines against the
Social Medicine, functions, and viral and host interactions. major infectious diseases, such as malaria,
Albert Einstein
College of Medicine, This leader summarises the highlights tuberculosis, pneumonia, meningitis, and
New York NY–10461, presented at the conference (the full text AIDS. However, some will be directed at what
USA of the abstracts and lectures can be found have traditionally been called “chronic” dis-
T Rohan at www.hpv2000.com). Relevant material eases, an increasing number of which appear to
in Spanish can be found at www.aepcc. have an infectious aetiology. Hepatitis B vacci-
Department of org. nation should have a profound influence on the
Gynaecology,
(J Clin Pathol 2001;54:163–175) rates of chronic liver disease and liver cancer.
Friedrich Schiller
University D–07743, Similarly, the aetiological association between
Keywords: papillomavirus; epidemiology; immunology;
Jena, Germany biology; screening HPV and cervical cancer has become increas-
A Schneider ingly clear and vaccines that prevent or treat
these infections are on the near horizon.
Centre for The need for HPV vaccines is greatest in
The 18th international papillomavirus confer-
Immunology and
Cancer Research, ence took place in Barcelona, Spain in July areas of high incidence of cervical cancer and
University of 2000. The human papillomavirus (HPV) clini- especially in those settings where diagnostic
Queensland, 4102 cal workshop was jointly organised with the and treatment services are poor. In most of the
Brisbane, Australia annual meeting of the Spanish Association of industrialised countries eVective screening and
I Frazer Cervical Pathology and Colposcopy. The treatment programmes for early lesions have
conference included 615 abstracts describing the potential to reduce cervical cancer mor-
Institute of Virology,
University of Koln, ongoing research as follows: epidemiology tality rates to low levels, but even in these
D–50935 Koln, (127 abstracts), diagnosis/screening (129 countries it is often the case that those at high-
Germany abstracts), treatment/prognosis (39 abstracts), est risk of disease are least likely to enter
H Pfister immunology/human immunodeficiency virus screening programmes. In the developing
(HIV) (61 abstracts), vaccine development/ countries, the resources for screening pro-
Secció de Ginecologia grammes are sparse, as is access to eVective
trials (56 abstracts), transformation/progression
Oncològica ICGON,
Hospital Clínic (53 abstacts), replication (20 abstracts), treatment for any lesions identified. An eYca-
Universitari, E–08036 transcription/translation (30 abstracts), viral cious vaccine against HPV infections, espe-
Barcelona, Spain protein functions (40 abstracts), and viral and cially one that acted against established infec-
L M Puig-Tintore host interactions (60 abstracts). A special work- tions, as well as preventing new ones, could
shop devoted to presenting ongoing research in have a rapid and profound impact on one of the
continued over
developing countries was also organised and major cancers aZicting women globally.
Correspondence to: included 138 abstracts. The conference also Before new vaccines are introduced into
Dr Bosch featured 80 invited lectures. The highlights pre- widespread use it is essential that their safety
x.bosch@ico.scs.es sented at the conference were selected and are and eYcacy should be properly established.
Accepted for publication briefly reported below. The full text of the This will require large randomised controlled
6 December 2000 abstracts and lectures can be found at trials, usually double blind, preceded by
www.jclinpath.com

2. 164 Bosch, Rohan, Schneider, et al
Department of appropriate smaller scale studies (preclinical HPV-16, 182; HPV-18, 231; HPV-45, 148; and
Infectious and Tropical and phase I/II trials) to establish immuno- HPV-31, 71.5. The study updated the conclu-
Diseases, London genicity and to detect any major adverse sions of the HPV IARC monograph providing
School of Hygiene and
Tropical Medicine, eVects. Initial trials may be conducted in the strongest evidence to support the identifi-
London WC1E 7HT, developed countries, where the candidate vac- cation of HPV types 16, 18, 31, 33, 35, 45, 51,
UK cines are most likely to be produced, at least 52, 58, and 59 as human carcinogens.
P G Smith initially. But even if they are found to be safe Most of the epidemiology presentations
and eYcacious in these settings it cannot be were on genital HPV infections and cervical
Unit of Field and assumed that the same eYcacy will hold in neoplasia. One descriptive epidemiological
Intervention Studies, other geographical and ecological situations. report of interest was from Lazcano-Ponce et
International Agency Further large scale trials will be required in
for Research on al (100) who, in studying a population based
Cancer, F–69372 developing countries, where the main market sample of 1340 Mexican women, observed
Lyon, France for the vaccines is likely to be, in public health two peaks of HPV DNA prevalence, one in
N Muñoz terms if not economically. women aged < 25 years, and the other in
Special methods need to be developed to women over the age of 65 years, with the low-
Deutsche ensure that individuals included in trials, many est prevalence being seen in women aged
Krebsforschungs- of whom may be illiterate, can be followed up 35–44 years. In all age groups, high risk HPV
zentrum, D–69120 and the endpoints of interest ascertained, often
Heidelberg, Germany types predominated. These data are intriguing
many years after an individual has entered into because they seem to contradict the previously
H zur Hausen
the trial. Furthermore, in such situations, widely accepted notion; namely, that HPV
major ethical issues need to be dealt with. The prevalence declines with age. They are,
level of medical care that must be provided for however, in agreement with data from a recent
those in a trial has been the subject of heated JNCI report expanded by an abstract by Her-
debate. rero et al (054) from the Costa Rica study, and
Conducting trials of new “expensive” vac- with those from some of the studies included
cines in developing countries has been consid-
in his international prevalence survey reported
ered unethical because it is unlikely that those
at this meeting. The interpretation of the peak
countries will be able to aVord to include a
in prevalence in older women is not clear, but
vaccine that is shown to be eYcacious in the
could represent either a cohort eVect or reac-
routine vaccination programme. An issue that
tivation of latent infection. Evidence for this
must be discussed before a trial starts is the
extent to which an eYcacious vaccine will be being a cohort eVect was provided by Peto et al
made available after a trial. At the very least (308), who showed a positive correlation
those in the “control” group should be oVered between the HPV prevalence by birth cohort
the vaccine, if they are still likely to benefit. and the corresponding lifetime risk of dying
However, many consider the responsibility of from cervical cancer in the UK.
the sponsors goes beyond this to the wider In work that has some bearing on the
community. It is common now to incorporate a Bethesda system, Peto et al (179) presented
cost–benefit analysis into field trials of new data from the Manchester cohort showing that
interventions. the prevalence of cervical intraepithelial neo-
Most existing vaccines that are in widespread plasia 3 (CIN3) increased linearly with increas-
use in developing countries are directed at ing time since last normal smear (for intervals
infections usually acquired in infancy or child- up to six years), suggesting therefore that the
hood. However, it seems unlikely that HPV true incidence of new disease can be estimated
vaccines will be directed at this age group, at reliably in a screened cohort and that CIN3
least initially. The target group, for preventive cases persist for at least this interval. The
vaccines, is likely to comprise girls or young prevalence of other abnormalities, including
women shortly before the onset of sexual activ- notably CIN2, was independent of the screen-
ity. This age group is more diYcult to recruit ing interval and essentially unrelated to HPV in
into trials, especially because in many develop- a previous normal smear, suggesting that these
ing countries girls of this age are unlikely to be abnormalities are part of the normal morpho-
in school. Those planning HIV vaccine trials logical spectrum of transient HPV infection. If
will have similar diYculties and the ground- correct, this would suggest that the inclusion of
work already done in preparing for HIV trials CIN2 in the high grade squamous intraepithe-
may be of great value for the planning of HPV lial lesion (HSIL) category of the Bethesda
vaccine trials. Indeed, a successful HPV classification system might be both a clinical
vaccine looks to be closer than an HIV vaccine. and scientific error, a conclusion supported by
some recent cohort data from Toronto.
Epidemiology HPV serology raised some attention at the
The results of combined analyses of the Inter- meeting. Two population based studies by
national Agency for Research on Cancer Herrero et al, one in Argentina (165) and one
(IARC) based multicentric case control study in Thailand (167), examined correlates of type
were presented (Muñoz et al, 053). The studies specific virus-like particles (VLPs). Both stud-
included 2288 cases and 2513 controls from ies were similar in design and both showed that
seven countries and used advanced polymerase the prevalence of antibodies against VLPs of
chain reaction (PCR) technology for HPV HPV types 16, 18, 31, 33, and 58 increased
detection and typing. The combined odds ratio with increasing number of sexual partners.
(OR) for viral detection was 83.3 (95% They also showed a strong correlation between
confidence interval (CI) 54.9 to 105.3) and the the detection of antibodies against a specific
ORs for the most relevant viral types were: HPV type and the presence of the same type in
www.jclinpath.com

3. Highlights of HPV 2000 165
the cervix, and that there is considerable cross- COFACTORS IN CERVICAL CARCINOGENESIS
reactivity. Antibodies were rare in virgins. In a Cofactors continue to be of interest in relation
study of anti-VLP-16 responses in young to HPV related cervical neoplasia because not
women, Ho et al (058) showed that those with all HPV positive women develop cervical
persistent IgG antibodies to HPV-16 VLP were cancer. In this regard, Moreno et al (061), using
half as likely to acquire HPV of any type, data from the IARC multicentric case control
suggesting that persistent antibody may cross study, showed strong evidence for the role of
protect against infection with other HPV types. OCs as a late stage factor in HPV induced car-
Thomann et al (157), in the Ludwig-McGill cinogenesis. According to this study, women
cohort, showed that IgM seropositivity at who had used OCs for six or more years and
enrolment was predictive of the risk of acquisi- were found to be positive for a high risk HPV
tion of incident HPV-16 infection, whereas IgG type had a four times greater risk of developing
seropositivity was more consistently associated squamous cell cervical cancer than women
with HPV DNA who were not exposed to OCs.
with prevalent or persistent oncogenic HPV
OCs did not modify the chances of being HPV
infection. IgM seropositivity had no diagnostic
positive among controls. This finding might
value with respect to the detection of HPV
have clinical implications. Two studies sug-
infection or cervical lesions. gested that Chlamydia trachomatis infection
In an intriguing report, Lewis et al (145) might increase the risk of invasive squamous
described the use of capture enzyme linked cervical cancer. Wallin et al (113) presented
immunosorbent assay (ELISA) to study anti- results from a prospective study in Sweden in
bodies against HPV types 6/11 and 16 in saliva, which C trachomatis detected in cervical smears
oral mucosal transudates, and serum. IgG anti- collected on average 5.6 years before diagnosis
bodies to HPV types 6/11 and 16 were detected or termination of follow up was associated with
less frequently in the oral samples than in a 17-fold increase in risk of invasive cervical
serum, possibly because of suboptimal collec- cancer after adjustment for HPV status, and
tion of oral specimens or because antibody Smith et al (071) found a positive association
concentrations in the oral specimens are truly between C trachomatis seropositivity and inva-
lower. Nevertheless, HPV IgG values in oral sive squamous cervical cancer risk in HPV
specimens correlated with seropositivity, and if positive subjects in case control studies in Bra-
this approach can be optimised, it is conceiv- zil and the Philippines, with the ORs increasing
able that oral samples could be useful in epide- with increasing C trachomatis antibody titres. In
miological studies as a non-invasive alternative contrast, van den Brule et al (063) found no
to serological specimens. association between genital C trachomatis infec-
Some data were presented on infections with tion and risk of subsequent HSIL in HPV
multiple HPV types, which is of interest in positive women, but a possible association with
relation to vaccine development. For example, the risk of abnormal squamous cells of indeter-
in the Ludwig-McGill cohort, Rousseau et al mined significance/low grade squamous in-
(137) showed that multiple infections were traepithelial lesion (ASCUS/LSIL), and Mo-
relatively common, occurring in about 19% of scicki et al (075) did not show an association
women each visit. Furthermore, assuming between C trachomatis (and other sexually
independence of types, she observed an excess transmitted diseases) and the risk of HSIL. If,
of joint cumulative positivity of types involving indeed, C trachomatis does influence the devel-
HPV types 16, 31, and 53, suggesting that opment of cervical neoplasia, the mechanism
individual type infections are not independent. by which it does so is not clear, but might
To a large extent, risk factors for single and involve the induction of a chronic inflamma-
multiple infections were similar, although age tory process. In relation to this, Castle et al
and number of sexual partners were more (070) presented evidence from the Costa Rican
strongly associated with multiple infections cohort suggesting that cervical inflammation
was positively associated with the risk of the
than with single infections. In addition, Slavin-
progression of oncogenic HPV to HSIL or car-
sky et al (146) showed that HIV positive
cinoma. They also found that bacterial vagino-
women were more likely to be infected with
sis was inversely associated with the risk of
multiple HPV types than HIV negative women. progression. In terms of other possible cofac-
Bosch et al (064) presented data from the tors, Smith et al (082) also found a positive
IARC multicentre case control study of HPV association between herpes simplex virus 2
and cervical adenocarcinoma, the largest of (HSV-2) serology and cervical cancer risk
such studies to date. He showed a very strong overall, and in HPV positive subjects in the
positive association between HPV infection Philippine component but not the Brazilian
and risk (OR, 68.7; 95% CI, 36.2 to 130.5) and component of their Brazilo-Philippine study.
similarly high ORs for individual HPV types. One phenomenon of interest in relation to
The report also showed that risk factors for risk of cervical cancer is the so called “male”
adenocarcinoma are largely the same as those factor. Essentially, this is taken to mean that the
for squamous cell carcinoma, the exception sexual behaviour of men (for example, the
being an apparent inverse association between number of sexual partners that they have had)
parity and risk. An observation of interest, con- can influence the risk of cervical cancer in their
sistent with the results of squamous cell carci- female sexual partners. In this regard, Bleeker
nomas, is the likely interaction between long et al (140) presented data showing that 80% of
term use of oral contraceptives (OCs) and cer- the male sexual partners of women with CIN
vical adenocarcinoma. had penile lesions, of which a substantial
www.jclinpath.com

4. 166 Bosch, Rohan, Schneider, et al
proportion were infected with HPV, and they detection methods (usually nested PCR) indi-
speculated that penile lesions in sexual partners cates that these lingering viruses replicate at
of women with CIN are probably productive very low levels, which helps them to escape
and that they might play an important role in from immune detection. However, the detec-
influencing the course of cervical lesions in tion of antibodies to capsid L1 protein
these women by continuously reinfecting them indicates that productive infection does takes
with HPV. Castellsagué et al (060), in a pooled place (Feltkamp et al, 102). Persistent infec-
analysis of data from the seven IARC multicen- tions are obviously activated by sunlight expo-
tre case control studies of CIN3 and invasive sure, immunosuppression, or hyperprolifera-
cancer, showed that the prevalence of penile tion of the epithelium such as in psoriasis. An
HPV was generally higher in husbands of cases interesting report on the transient generation of
than in husbands of controls. However, penile anti-HPV-5 antibodies after epidermal repair
HPV was not associated with the risk of cervi- in burns supports the view of transient
cal neoplasia in high or intermediate risk coun- activation of persisting viruses as a result of
tries, but was associated with an eightfold epidermal proliferation (Favre et al, 676). A
increase in risk in Spain. similar scenario has been proposed for the
cervical-vaginal mucosa by Jin et al (197), who
HPV AND SKIN CANCER found a strikingly high prevalence of HPV
During the past few years, the association of types 61, 72, and 11 related, novel sequences in
HPV with skin cancer in the general population lavages of HIV infected women. These previ-
apart from the epidermodysplasia verruci- ously rare members of the phylogenetic sub-
formis (EV) syndrome has been a subject of group A3 seem to be activated during AIDS
intense research. Non-melanoma skin cancer is associated immunodeficiency, but must cer-
the most frequent cancer among white indi- tainly have a reservoir in the general popula-
viduals worldwide and is particularly wide- tion. The growing awareness of this type of
spread among the increasing number of immu- ubiquitous infection will have a strong impact
nosuppressed patients. Using newly developed, on future plans for disease prevention and
highly sensitive PCRs, a large and still growing treatment.
number of new partial sequences could be
identified in non-melanoma skin cancers,
which all probably represent new PV types. HPV AND CANCERS OF THE ORAL CAVITY AND THE
Most of them are related to EV associated ANAL AND PERIANAL REGIONS
viruses but, depending on the PCR used, In relation to epidemiological presentations on
HPV-4 related types have also been identified HPV and cancer sites other than the cervix, the
(Antonsson et al, 116). These days, genital study by Björge et al (148) provided the first
HPV types are less frequently detected. The prospective evidence that previous infection by
HPV prevalence is highest in squamous cell HPV-16 (as indicated by the presence of IgG
carcinomas of immunosuppressed patients (up antibodies in serum samples collected on aver-
to 90%) and lowest in basal cell carcinomas of age 10 years before cancer diagnosis) is associ-
immunocompetent patients (about 40%, ated with an increased risk of anal or perianal
which is close to the prevalence in normal skin cancers.
samples) (Rust et al, 497; Wieland et al, 115). There were several presentations on various
Viral load determinations revealed only one aspects of HPV and oropharyngeal cancers. An
HPV DNA copy/20–1000 cells in the tumour early report by Herrero et al (166) from the
(Wieland et al, 115; Stockfleth et al, 125), indi- IARC international case control study of HPV
cating that probably not every tumour cell har- and cancer of the oral cavity and pharynx sug-
bours an HPV genome. This suggests that the gested that HPV is not a major aetiological fac-
contributions of cutaneous HPV types to tor, based on the comparison of cases and con-
carcinogenesis are quite diVerent from genital trols with respect to HPV prevalence in
PVs in cervical cancer. It underlines the need exfoliated cells. However, the HPV detection
for studies of the molecular biology of these rate was three times higher in tumour biopsies
viruses so that their activity can be better than in exfoliated cells. In contrast, Mork et al
understood. Preliminary experiments to this (065), in a case control study nested within a
end were presented that dealt with the Nordic cohort of almost 900 000 subjects who
biochemical functions of cutaneous HPV had donated serum samples, showed that
proteins, with the ultraviolet light eVect on viral HPV-16 positivity in serum collected on
promoter activity, and with the eVects of cuta- average 9.4 years before diagnosis was associ-
neous HPV oncoproteins in organotypic cul- ated with a 14-fold increase in the risk of sub-
tures (Caldeira et al, 604; Ruhland and de Vil- sequent oropharyngeal cancer (OR, 14.4; 95%
liers, 661; Mulder et al, 629, respectively). CI, 3.6 to 58.1). Seropositivity to HPV types
18, 33, and 73 was not associated with altered
NATURAL HISTORY OF HPV INFECTION IN THE risk. The reason for the discrepancy between
SKIN the findings of these two studies is unclear—
The frequent detection of viral DNA identical possible explanations include diVerences in
with or related to EV associated HPVs in study design or specimen types, diVerences in
healthy skin samples or plucked hairs of the methods of assessing HPV status, or lack of
individuals in all age groups supports the good information on confounders in the
notion that probably everybody is infected over Nordic study. The preliminary results of the
long periods, if not throughout life, with these IARC study are consistent with previous stud-
viruses. The general need for highly sensitive ies indicating that HPV might have aetiological
www.jclinpath.com

5. Highlights of HPV 2000 167
relevance for tonsillar carcinomas. This is sup- protein sequences were identified (Baines et al,
ported to some extent by the results of HPV 126). Specific primers were designed which, in
analysis of tumours in the case series reported contrast to the MY 09/11 and the GP5+/6+
by Weissenborn et al (114) and Summersgill et system, allowed the detection of PVs in all
al (172), who found HPV DNA in 58% and groups of the HPV Sequence Database. To
63%, respectively, of tonsillar tumours. improve sensitivity, a PCR system was de-
signed by Kornegay et al (282), where a
CERVICAL CANCER IN DEVELOPING COUNTRIES non-degenerate pool of five upstream and
The 18th international conference included for seven downstream primers was designed in
the first time a special workshop entitled “HPV conserved regions of the L1 gene, resulting in a
in developing countries: screening, diagnosis, 170 bp product. In clinical specimens, the
therapy and prevention”. The workshop was short amplification detected up to 10% more
chaired by T Wright (University of Columbia) HPV positives compared with the MY 09/11
and R Sankaranarayanan (IARC) and included system.
138 abstracts.
The presentations in this workshop dealt LOCALISATION OF HPV DNA WITHIN TISSUE
mainly with the prevalence and type distribu- SPECIMENS AND VIRUS ASSOCIATED PROGRESSION
tion of genital HPV infections in central MARKERS
America and in Africa, and in the quality of A combination of the SPF10 system with in
cytological screening for cervical neoplasia in situ hybridisation (Vermeulen et al, 236)
developing countries. The former did not allow allowed the detection and localisation of multi-
firm conclusions to be drawn concerning ple HPV infections in up to 20% of HPV posi-
diVerences in HPV type distribution from tive cervical cancers. In situ hybridisation
those reported elsewhere, and the latter results indicated the presence of multiple HPV
highlighted the need for improvements in the subclones.
quality of screening, which is handicapped by Several studies looked at the potential value
the lack of resources in developing countries. of virus associated progression markers, such
Clearly, these are important topics given the as the presence of certain viral RNA species or
relatively high rates of cervical cancer in these the E4 protein (Sotlar et al, 195), viral load
areas, and most presentations highlighted the (Swan et al, 280), and integration (Klaes et al,
need for further investigations in these areas. 194; Counio Coste et al, 223; Blennow et al,
Visual inspection of the cervix was discussed 231). Lamarcq et al (187) showed that the
extensively at the meeting. Ferreccio et al (200) expression of E6/E7 transcripts and vascular
described a cervical cancer prevention pro- endothelial growth factor (VEGF) mRNA was
gramme in Peru that uses visual inspection of higher in women with dysplastic cervical
the cervix with acetic acid and immediate smears than in those with normal smears, sug-
treatment—it is anticipated that this will lead to gesting that these markers might be a useful
a far higher detection and treatment rate for prognostic and diagnostic tool for cervical dis-
high grade lesions than in neighbouring control ease. Because for most of these markers only
areas in which Papanicolaou (Pap) based data from cross sectional studies are available
programmes operate. It is expected that further at this point, their clinical usefulness is still
developments in this important area of field unknown.
research and implementation will be stimu-
lated by the growing presence of international NEW BIOMARKERS AS A QUALITY CONTROL OF
cooperative networks such as the Alliance for THE HISTOLOGICAL OR CYTOLOGICAL DIAGNOSIS
Cervical Cancer Prevention (PAHO, PATH, A non-viral marker might be useful for improv-
IARC, AVSC Int. JHPIEGO (www.alliance- ing the quality of the histological or cytological
cxca.org)), the Challenge/ESO Group (www. evaluation of cervical specimens. Enhanced
cancerworld.org), and the International Net- expression of a cyclin dependent kinase inhibi-
work for Control of Gynaecological Cancers tor p16ink4a can be used to increase the
(www.matweb.hcuge.ch/matweb). specificity and interobserver agreement of the
histological assessment of cervical lesions
Diagnosis and detection of HPV DNA (Klaes et al, 188). In addition, it was shown that
Several papers concerned new PCR based immunocytological staining for p16ink4a can
techniques for HPV detection. The accurate identify dyskaryotic exfoliated cells in cervical
detection of various HPV types was described smears (Friedrich et al, 129).
by van Doorn et al (254 and 255), where a Few studies compared the quality of HPV
broad spectrum PCR primer set (SPF10) is detection systems, and there were contradic-
used to amplify a fragment of 65 bp from the tory results with respect to sensitivity and type
L1 region. Genotyping is done by reverse specificity of the various systems (de Sanjosé et
hybridisation on a line probe assay discriminat- al, 220; Perrons et al, 302; Molijn et al, 254).
ing between 25 HPV genotypes. Seth et al These discrepancies between diVerent HPV
(253) described the application of the same detection systems, which are in wide clinical
technique to 12 year old archived smears; 31% use, show that there is an urgent need for
of cases were HPV positive, with higher rates in standardisation and quality control—an indis-
women born after 1940 compared with those pensable prerequisite if HPV testing is to be
born before 1940. The SPF10 system was integrated into clinical practice. This problem
more sensitive than the GP5+/6+ system. To has been recognised and led to the establish-
facilitate the detection of a wide range of vari- ment of the “first human papillomavirus inter-
ous HPV types, 11 blocks of conserved L1 national quality control proficiency panel” at
www.jclinpath.com

6. 168 Bosch, Rohan, Schneider, et al
this meeting, an activity which should be high grade CIN. HPV screening in addition to
supported by the scientific community and by cytology increased eYcacy significantly.
industry. In the Dutch screening programme (182),
the prevalence of HPV DNA in cytological
negatives was 3.4%. Cytology negativity and
HPV TESTING AS A TRIAGE METHOD FOR
high risk HPV positivity was associated with
CYTOLOGY RESULTS OF UNDETERMINED NATURE
HPV persistence in 60% and led to abnormal
Several studies concerned the value of HPV
cytology in 26% of HPV positives. Cytology
testing for triaging women with a diagnosis of
positivity and high risk HPV positivity was
ASCUS or low grade CIN (Solomon et al, 181;
associated with HPV persistence in 70% and
Ronco et al, 224; McGoogan et al, 256; Ponce
led to high grade CIN in 24%.
et al, 294). Ponce et al (294) reported an equal
In the Seattle cohort (183), Thin-Prep
value of HPV testing (hybrid capture II (HC
cytology showed a sensitivity of 56% for the
II)) and expert cytology in predicting HSIL in
detection of CIN2/3, compared with a sensitiv-
women with ASCUS/LSIL. Ronco et al (244)
ity of 67% by PCR based HPV DNA testing
showed that the use of HPV detection to triage
and 68% for HC II. Specificity for cytology was
women with cytological evidence of ASCUS
93% and for PCR 91%, compared with 79%
and LSIL could reduce substantially the
for HC II. With cytology, 12.5% of women had
number of colposcopies, with little loss in
to be referred for colposcopy. Comparable fig-
histological detection of CIN2/3. Solomon et al
ures were 15% using PCR and 23% with HC
(181), reporting from the ALTS trial (a
II. Thus, PCR had a higher sensitivity and
randomised, multicentre trial comparing three
comparable specificity when compared with
management strategies for women with cyto-
liquid based cytology.
logical diagnoses of ASCUS or LSIL—
Two German studies compared the perform-
immediate colposcopy, conservative manage-
ance of routine cytology and HPV testing by
ment, or colposcopic triage using HPV DNA
GP5+/6+ PCR (Schneider et al, 184) or HC II
testing added to repeat thin layer cytology),
(Iftner et al, 185). The sensitivity of cytology
showed that the negative predictive value of a
for the detection of HSIL was 20% (184) and
negative HPV DNA test was high in all groups
34% (185), respectively. HPV testing showed a
of study subjects, but that the positive predic-
sensitivity of 89% and a specificity of 94% in
tive value of finding HPV was greatest in older
both studies. The positive and negative predic-
women (> 30) with ASCUS as a referral diag-
tive values for HPV testing using PCR were
nosis on cytology. Thus, HPV testing was
35.8% and 99.6%, respectively.
clearly superior to cytology at the expense of a
referral rate of 56% for colposcopy. In another
SELF SAMPLING FOR HPV TESTING: A
study including 270 women with a borderline
POTENTIALLY IMPORTANT DEVELOPMENT
or mildly dyskaryotic smear, no cases of HSIL
Several studies reported the value of self
would have been missed by HPV testing
collected samples for HPV testing by PCR sys-
(McGoogan et al, 256). The authors conclude
tems (Harper et al, 186; Nobbenhuis et al, 265;
that HPV testing is likely to be useful in triag-
Mire et al, 275) or HC II (Salmerón et al, 270;
ing women with low grade cervical abnormali-
Hillemanns et al, 303) compared with physi-
ties.
cian directed sampling. These studies con-
The value of HPV testing in the triaging of
cluded that HPV detection rates in self
women with ASCUS seemed to be consistently
collected vaginal samples were similar to those
superior to repeated cytology for the detection
in physician collected cervical samples, and
of HSIL. HPV testing in the triaging of LSIL
that the use of self collected samples might be
cases is less consistent because of the high
useful in cervical cancer screening in under-
prevalence rates of HPV DNA found in some
served women and women in developing coun-
studies, such as ALTS. However, many authors
tries (303 and 265). In a similar vein,
noted the diYculties of comparing studies
Brinkman et al (274) showed that HPV detec-
because of the additional variability introduced
tion rates using a PCR based urine test agreed
by the cytology methods used (standard v
reasonably with those derived from matching
expert v liquid based) and the use of diVerent
cervical swabs, raising the possibility that such
axes of classification of the neoplastic lesions
samples might be useful.
(CIN, the Bethesda system (TBS), Pap, and
others).
ACCEPTANCE OF HPV TESTING BY THE
POPULATION AND THE MEDICAL COMMUNITY
HPV TESTING AS A PRIMARY SCREENING METHOD An important prerequisite for the acceptance
The validity of HPV testing in screening of HPV testing by individuals undergoing
programmes was investigated in various co- screening is the evaluation of socioeconomic
horts (Franco et al, 180; Meijer et al, 182; and psychological factors, counselling of af-
Kulasingam et al, 183; Schneider et al, 184; Ift- fected individuals, and information to the pub-
ner et al, 185). In the Newfoundland study lic, issues that were considered in various stud-
(180), the adjusted sensitivity for cytology for ies (Johnson et al, 228; Phillips et al, 229;
detection of HSIL with cut point LSIL was Whynes et al, 230; Flores et al, 296). The clini-
28%, and for HPV testing 68%, at the expense cal workshop included special presentations on
of specificity, which was 90.6% for HPV testing the topics of “Counseling persons with HPV
versus 99.1% for cytology. A combination of infections” (Steben, 008) and “Medical, social
HPV testing and cytology yielded a negative and ethical issues and challenges of HPV
predictive value of 100% for the detection of infections: what can we do?” (Alexander, 007).
www.jclinpath.com

7. Highlights of HPV 2000 169
HPV IN IMMUNOCOMPROMISED INDIVIDUALS immunodeficient (SCID) mouse model (Bon-
The natural history of HPV might have to be nez et al, 339; 340; 341). It was shown that the
redefined based upon the information gathered antisense oligonucleotides ORI-1001 comple-
in immunosuppressed individuals (Hoesley et mentary to the 5' end of HPV-6 E1 mRNA has
al, 154; Jin et al, 197): 76% of HIV positive an inhibitory eVect on the growth of HPV-6
women were found to be HPV positive by induced human condylomata (339).
nested PCR, with 65% harbouring multiple In vivo studies in humans: experiences with
infections and with one third of isolates indole-3-carbinol (I3C), a compound found in
belonging to the phylogenetic subgroup A3. It cruciferous vegetables that detoxifies carcino-
is concluded that HPV infection might be gens and alters oestrogen metabolism, were
acquired early in life, and that it might be part reported by Bell et al (311) and Da-Zhi Chen et
of the commensal viral microflora of the skin al (312). It was shown that I3C and its
and the anogenital mucosa, with sporadic sub- metabolite DIM caused apoptotic changes in
clinical shedding caused by transient reactiva- cervical cancer cells in vitro. In vivo, HPV
tion in specific settings such as immunosup- transgenic mice exposed to oestrogen and fed
pression and overt disease being unusual. I3C showed evidence of apoptotic cells in their
cervical epithelium (312).
HPV ECONOMICS To evaluate the eVectiveness of I3C in
HPV 2000 introduced the topic of HPV humans, 27 patients with biopsy confirmed
economics to stress the relevance of conduct- CIN2 or 3 were randomised to oral I3C or pla-
ing cost–benefit analyses and studies evaluating cebo (311). Eight of 17 patients treated for 12
quality of life. Most of the work presented did weeks had complete regression of lesions. None
so in relation to screening and treatment of of the 10 patients in the placebo group experi-
preinvasive neoplastic lesions through conven- enced regression. In another study, a fusion
tional cytology screening programmes or protein consisting of the heat shock protein
through primary HPV testing. Hsp65 and Mycobacterium bovis BCG linked to
The addition of HPV testing was favourably the HPV-16 E7 protein was constructed. In a
evaluated as an adjuvant to the follow up of randomised double blind placebo controlled
HIV positive women (Goldie et al, 267), and as trial, three monthly subcutaneous doses were
a single lifetime screening method in low applied to 10 patients with high grade anal
resource settings (Goldie et al, 693; Sherlaw- intraepithelial neoplasia (Goldstone et al,
Johnson et al, 260; Kulasingam et al, 272). HPV 347a). All patients had a clinical response after
testing when combined with a low technology a course of three doses.
treatment, such as cryosurgery for HPV Immunomodulatory substances such as imi-
positive women, resulted in cost saving (Goldie quimod (Ikenberg et al, 310), leukonorm
et al, 693). An ongoing project in Vietnam (Metzner et al, 322), viscum album (Romero
showed that five yearly cytology based screen- Montes et al, 327), and a combination of
ing programmes were both feasible and cost iscador and interferon (IFN) (Basta et al, 318)
beneficial (Chan Hung et al, 694). Quality of were used in various studies for the treatment
life issues about HPV testing (self sampling of CIN and anal condylomata in phase 2 trials
acceptability, cost acceptability, diagnostic re- and showed promising responses with tolerable
lated stress) form a novel research area that side eVects. Medical treatment of LSIL
presents several challenges before primary emerged as an area of growing research interest
HPV testing can be implemented (Flores et al, and a plausible alternative to reduce surgical
296; Johnson et al, 228; Philips et al, 229). overtreatment.
Some exercises based on mathematical models
for predicting outcomes of screening pro- HPV DNA AS A PROGNOSTIC FACTOR IN CIN CASES
grammes under a variety of assumptions were AFTER TREATMENT
also presented (Gallivan et al, 259; Whynes et Studies evaluating the presence of HPV DNA
al, 230) and may become an area of increasing after surgical treatment of CIN (Kjellberg et al,
interest. 196; Tachezy et al, 242; Nobbenhuis et al, 317;
Ramírez Porras et al, 320; Simon et al, 323;
Bettinger et al, 193; Elfgren et al, 329) showed
Treatment of HPV infections and of HPV that the detection of HPV increases the risk of
related cancer recurrent disease. HPV persistence is rare after
In vitro studies: various gene therapeutic conisation with free margins, but not after
modalities were defined such as HSV thymi- conisation with positive margins (323). A high
dine kinase (HSV-TK) (Sethi et al. 326), HSV risk HPV positive test three months after treat-
vectors (Kari et al, 345), HPV-16 E2 derivates ment is more predictive for residual or
(HoVmann et al, 333), and anti-HPV hairpin recurrent CIN disease than abnormal cytology
ribozymes (Enriquez-Aragon et al, 328). In (317). Persistence of HPV infection after treat-
addition, growth factor inhibition (Michael et ment was more common in patients less than
al, 342) and endothelin receptor antagonists 35 years of age and with a history of
(Venuti et al, 313) were used. The in vitro cryotherapy (329). Conisation resulted in a
results of these studies look promising and decline of cervical HPV IgA antibodies
development to clinical application is a realistic whereas cryotherapy had much less eVect
goal. (329). Using HPV-16 E6 and E7 peptides, a
In vivo studies in animals: for preclinical cellular immune response was seen in disease
studies, HPV-6 or HPV-11 induced human free patients but not in those with disease
xenografts were used in the severe combined recurrence (Sarkar et al, 314).
www.jclinpath.com

8. 170 Bosch, Rohan, Schneider, et al
HPV TESTS AS PROGNOSTIC MARKERS IN INVASIVE capsid specific antibody within four years of
CANCER infection, with a median time to seroconversion
In patients with invasive cancer (Lizano et al, of over six months (Carter et al, 057). HPV-18
325), viral DNA sequences were detected in seroconversion is similarly slow, whereas sero-
the plasma of four of five patients with conversion for HPV-6b is apparently faster,
recurrence and all three patients with tumour and occurs at the same time as the acquisition
persistence, but in only one patient of 17 with of viral DNA (057). The definition of a positive
regression. A prognostic eVect of the presence serological reaction to the capsid remains
of HPV antibodies, HPV DNA (Bravo et al, mathematical (Stone et al, 056) because the
338), certain HPV types (Lo Keith et al, 332; identification of a cohort of individuals never
Dreier et al, 335), VEGF serum concentrations infected with a particular HPV genotype is not
(Bachtiary et al, 336), or marker chromosomes straightforward, and there is also some evi-
(Macville et al, 344) was shown in case series, dence of genotype crossreactivity for serologi-
but larger prospective series are needed to sub- cal reactions (Combita et al, 412). Hence, the
stantiate these preliminary observations. usefulness of serology for epidemiological
studies is limited. For virus encoded non-
Immunology and vaccines structural proteins, serological tests are not yet
Recognition of the association between HPV as well standardised as for PV virions. Antibody
infection and cervical cancer in the 1980s to the E7 protein, which is expressed in acute
fuelled interest in HPV immunology, and in the infection (Hamsíková et al, 397) and also in
development of vaccines to prevent infection. many PV associated tumours, is seen in about
However, data were initially slow to emerge, in 50% of subjects with invasive cervical cancer.
part because two problems hindered research, However, it is not seen after acute HPV infec-
namely: (1) PV could not be grown in the tion with several genotypes (Gaenzler et al,
laboratory as a source of antigen for the devel- 398), even in the presence of antibody to the
opment of serological tests and for conven- L1 capsid protein. Low amounts of antibody
tional killed vaccines; and (2) no PV could be reactive with peptides derived from other open
identified that infected an immunologically reading frames (E2 and E4) are seen in some
well characterised laboratory animal. patients with HPV infection and invasive
Improved systems for expressing genes in cancer (Sehr et al, 399), although sensitivity
eukaryotic cells through the 1990s has allowed and specificity for infection appear to be low.
a partial solution to the first problem, which is These results are reminiscent of the results
exemplified by the production of VLPs and the obtained with L1 protein before the use of cor-
recent resolution of a crystal structure for the rectly conformed VLP substrate.
capsid of HPV-16 (Chen et al, 359), and has led
to a considerable worldwide interest in the
development of vaccines to prevent and to treat CELL MEDIATED IMMUNE RESPONSES
PV infection. In addition, epithelial raft cul- Several presentations describing the conse-
tures (Thomas et al, 535) have recently allowed quences of HIV infection on the natural history
the production of authentic virions from of HPV infection confirmed that HPV infec-
epithelial cells infected in vitro. The lack of an tion is more likely to persist and less likely to
animal model for the study of PV immunology respond to treatment in individuals with
has been circumvented by the development of systemic immune suppression (Palefsky, 349;
more robust tests for cell mediated immunity Ebrahim et al, 350). These data strongly
to defined antigens in humans, including suggest a role for cell mediated immune
tetramer and ELISPOT technology, and key responses in the control of HPV infection, as
experiments on the immunological conse- would be expected from data for other viruses.
quences of PV infection can now be under- The newer assays (such as ELISPOT and
taken in the natural host. Such research has tetramer technologies) allow more precise
demonstrated the ability of PV to elicit both measurement of virus specific cytotoxic T cells
humoral and cellular immunity and, possibly as (CTLs) and to some extent of virus specific T
a consequence of preferential uptake by helper responses (Steele et al, 384). Assay
dendritic cells (DCs) into the major histompat- reproducibility and sensitivity remain impor-
ibility complex (MHC) class I presentation tant issues, and there is only a limited number
pathway, PV VLPs can also induce host of defined viral protein epitopes restricted by
protective cytotoxic T cell responses. Interest defined MHC class I and II haplotypes. T cell
in PV immunology has now extended outside assays, to be suitable for immunoepidemiology,
the narrow confines of viral epidemiology, and or for modelling mechanisms of disease
abstracts on PV immunology and immuno- control, need to: (1) produce positive results
serology, and on vaccine development, ac- that are reproducible from day to day in the
counted for nearly a quarter of the material same patient, (2) correlate with the genotype of
presented at the workshop. clinical HPV infection, and (3) demonstrate
MHC restriction and antigen specificity. Few
SEROLOGICAL RESPONSES TO HPV INFECTION of the presentations gave results from human
Data presented from prospective studies of studies that met these stringent criteria.
infection with genital genotypes of HPV Positive control reagents, including immortal-
suggest that serological responses to the viral ised antigen specific T cell lines, using
capsid are slow to appear, when compared with telomerase reverse transcriptase (TERT)
most viruses. Only 50% of individuals acquir- transfer (Hooijberg et al, 427), promise to help
ing HPV DNA of genotype HPV-16 develop with assay standardisation in the future.
www.jclinpath.com

9. Highlights of HPV 2000 171
Cytotoxic T cell mediated immune re- immune responses including CTLs. This para-
sponses are not easily demonstrable after natu- dox can be explained by noting that PV infec-
ral HPV infection. Low frequencies of CTLs tion is non-lytic and superficial, and because
specific for the E6 protein of HPV-16 were most virus genes are expressed at low levels
associated in one study with the resolution of they might simply be ignored by the immune
cervical infection, whereas E7 specific CTLs system. This has been demonstrated in a skin
were found both in patients with regression and transplant model using HPV-16 E7 transgenic
in patients with persisting infection (Nakagawa skin, where rejection only occurs after the
et al, 354). Of potential importance for vaccine administration of powerful proinflammatory
development, some E7 specific CTL epitopes signals, and is mediated via MHC class II
were shown to be crossreactive between dependent mechanisms (Frazer et al, 378).
HPV-16 and HPV-52 (Youde et al, 357). More Alternatively, if PV antigen expression is
studies in this area will be needed to determine increased, presentation in skin without inflam-
whether host protective crossreactive responses mation may be tolerogenic, because skin asso-
exist after natural HPV infection or vaccina- ciated E7 antigen can be cross presented by
tion. Virus specific T helper cells are, predict- bone marrow derived DCs and anergise
ably, harder to demonstrate than specific CTL specific T cells, at least in the mouse (Doan et
responses, given the much lower frequency of al, 356). In addition, and in common with most
T helper cells after viral infection. The E7 pro- viruses, PV genes encode mechanisms for
tein contains sequences recognised as MHC modulating both the induction of the host
class II restricted epitopes by most normal immune response and virus specific eVector
subjects (van der Burg et al, 358), as does E2 mechanisms. Initiation of the innate and adap-
(de Jong et al, 383; Dillon et al, 408). Patients tive immune responses to PV antigens may be
with cervical cancer but not patients with impaired by E7 mediated downregulation of
squamous intraepithelial lesions (SIL) respond the host response to IFN (Barnard and
to E6 peptides in vitro (Timmins et al, 353), McMillan, 579; Nees et al, 630). E5 and E7
and patients with progressive cervical disease also interfere with the TAP (transporter associ-
respond less to E7 peptides (Luxton et al, 400), ated with antigen presentation) antigen trans-
although multiple in vitro restimulations are porter mechanism in infected cells (Vambutas
required to elicit positive proliferative re- et al, 360), and with MHC class I presentation
sponses. Responses to E6 and E7 peptides are of antigen. Local proinflammatory chemokines
better in patients who are disease free after (Fife et al, 388) and cytokine B (Bermúdez
treatment (Sarkar et al, 314). Cytokine secre- Morales et al, 379; Jacobs et al, 416) production
tion by T cells in response to E6 and E7 is altered in PV infected as opposed to
proteins is altered in recurrent respiratory pap- uninfected keratinocytes. Additional mutations
illomatosis (RRP) (DeVoti et al, 406), suggest- in the antigen processing and presentation
ing that failure of resolution after infection pathways are seen in cervical cancers (Garrido,
might reflect immune deviation to a more T 370): SCID/human (HU) tumour grafts, facili-
helper type 2-like immune response. In vivo tated with steroid, might prove an interesting
testing of delayed type hypersensitivity (DTH) model to examine defects of the antigen
to antigen is a cheap, simple, and robust tech- presenting machinery in tumours, free of
nique for measuring CD4+ T cell responses to stroma (Baines et al, 625).
viral antigens worthy of further investigation,
and in one study DTH to E7 peptides was
found to be more common in regressing than PROPHYLACTIC VACCINES AGAINST PV: PHASE I
progressing CIN (Höpfl et al, 351), in general AND II HUMAN TRIALS
support of the hypothesis that cell mediated All existing eVective prophylactic vaccines
immunity is important for the regression of against viral infection induce neutralising anti-
HPV infection. In keeping with this, regressing body, and none is thought to produce sterile
as opposed to persistent CIN appears to be immunity. Some epidemiological evidence was
associated with increased numbers of activated presented suggesting that the development of
T cells locally within the lesion (Petry et al, an antibody response in humans protects
352). Local immune responses to PV antigen against repeat PV infection (Ho et al, 058),
in the female genital tract most probably reflect which is reassuring for the development of pro-
systemic as well as mucosal immunity and, phylactic vaccines, because to date the data
perhaps in consequence, IgA anticapsid anti- supporting such vaccines have been based on
body in the genital secretions, although com- studies in animal models. Most work on
mon in HPV infection, is not a marker for per- prophylactic vaccine development has focused
sisting local infection (Pretet et al, 396). on vaccines that are based on VLPs comprising
the PV L1 capsid protein, produced in yeast or
PV EVASION IN HOST IMMUNITY insect cells (Schiller, 348). Were such vaccines
The lack of demonstrable immunity to non- to be available, one study indicated that they
structural PV proteins after acute HPV infec- would be generally acceptable to high risk
tion, together with the slow and variable communities (Lazcano et al, 446). Early phase
response to the capsid proteins, suggests that clinical trials of vaccines for HPV-11 (Fife et al,
there is a problem with the generation of PV 364) and HPV-16 (Harro et al, 362; Poland et
specific immunity. In contrast, PV VLPs are al, 363; Balsley et al, 366) were presented by
highly immunogenic when delivered as a several groups. HPV-11 VLPs produced neu-
vaccine without adjuvant, stimulate DCs (Ru- tralising antibody for up to 18 months. HPV-16
dolf et al, 355; Lenz et al, 395), and induce VLPs are similarly immunogenic even without
www.jclinpath.com

10. 172 Bosch, Rohan, Schneider, et al
adjuvant. In general, antibodies are PV geno- THERAPEUTIC VACCINES
type specific, and immunogenic at relatively Currently, there are no licensed therapeutic
low antigen doses, although some vaccines vaccines for humans, and no clear evidence of
might have stability problems. However, in the which PV viral protein(s) might produce a
beagle dog, VLP delivered intranasally, al- therapeutic immune response to PV infection.
though inducing a good mucosal IgA response, Cottontail rabbits undergo regression of cot-
failed to protect against virus challenge (Yuan tontail rabbit PV (CRPV) induced warty
et al, 369), suggesting that systemic vaccine lesions after the administration of a polynucleo-
delivery may have advantages even for mucosal tide vaccine encoding E1, E2, E6, and E7
PV genotypes. Although no polynucleotide (374), and the same proteins fused to ubiquitin
are equally successful (Leachman et al, 455).
vaccines have been tried in humans, DNA
Although data on eYcacy in natural PV infec-
encoding canine oral papillomavirus (COPV)
tion are limited, a wide array of potential
capsid proteins protects dogs against viral vaccine delivery systems have been tested for
challenge (Stanley et al, 375) and may clear immunogenicity in transplantable tumour
virus from existing infections (Moore et al, models. Immunotherapy with polynucleotide
452): modification of codon usage of the PV vaccines, either unmodified (Velders et al, 382;
capsid genes to match that of more conven- Doan et al, 441; Smahel et al, 445; De Marco et
tional mammalian genes improves the immu- al, 449), codon modified (Liu et al, 451), shuf-
nogenicity of PV DNA vaccines in animals (Liu fled to avoid potentially harmful consequences
et al, 451), and may therefore facilitate their use of gene expression (Osen et al, 372), or fused
in humans. with ubiquitin to improve antigen presentation
Minor HPV variants share epitopes (Pas- (Brandsma et al, 373), has been eVective in the
trana et al, 368), but crossreactivity of neutral- prevention or elimination of PV protein
ising antibody between major types is limited, expressing transplantable tumours. Similarly,
although the requirement for multicomponent vaccines based on cell derived (Navabi et al,
vaccines can be overcome with chimaeric VLPs 447) or recombinant (Giannouli et al, 429;
displaying two or more types of HPV, at least in O’Neill et al, 438) proteins, derivative peptides
animals (Wilson et al, 462). PV neutralisation (Melief et al, 437), recombinant bacteria
assays, which until recently have relied on a (Benyacoub et al, 450; Revaz-Fellay et al, 453;
labourious in vivo xenograft model, are now Maclean et al, 457), recombinant viruses
available for in vitro use, based on reporter (Jifeng et al, 430; Chen et al, 436; Herd et al,
440), and chimaeric virus-like particles (Liu et
genes (Yeager et al, 371) or detection of viral E2
al, 386; Street et al, 439; Liu et al, 442; Shi et al,
mRNA by reverse transcription PCR (Hall et
448; Revaz-Fellay, 453; Kaufmann et al, 460)
al, 426). Such assays should allow the valida- are all protective in at least one animal tumour
tion of vaccines for the less common genotypes model system. Responses to such vaccines can
based on surrogate markers of eYcacy, without be improved by cytokines (Hung et al, 434), by
the need for extensive clinical trials. targeted delivery systems (Wang et al, 435;
Several studies (Chackerian et al, 389; Liu et Michel et al, 458), and by new adjuvants such
al, 390) examined the use of PV VLPs as deliv- as Hsp (Kadish et al, 365; Randall Chu et al,
ery systems for other antigens. VLPs are potent 377; Cheng et al, 433), whereas “mix and
immunogens that can induce DTH in humans match” prime boost strategies (Chen et al, 381)
(Lian et al, 431), and T cell proliferation and produced improved responses in animal mod-
occasional CTL responses in chimps (Pretet et els when compared with single delivery sys-
al, 396). Chimaeric VLPs can activate DCs tems. The human trials of these therapeutic
(Balmelli et al, 391; Lenz et al, 395) and induce vaccines to date have been more limited. They
epitope specific CTL responses with ex vivo have shown safety and some immunogenicity
human cells, and in vivo in mice. Adjuvants of E7 based vaccines. A fusion of Hsp with E7
such as CpG DNA enhance VLP induced induced transient proliferative responses to E7
mucosal immunity (Gerber et al, 465), al- peptides in T cells in four of the eight subjects
though as mentioned above a role for mucosal immunised. E7 peptides induce some immune
immunity in the control of PV infection responses but CIN regression was no more
remains to be demonstrated. Although most than might be expected without vaccine (Kast
et al, 367), and no regression of cervical cancer
VLPs for vaccine use have been produced in
was reported with any vaccine (Khleif et al,
yeast or insect cells, alternative expression sys-
380), although many trials are ongoing. Ex vivo
tems that might better suit developing coun- DCs pulsed with E7 induce E7 specific CTLs
tries are being explored, including bacterial (Evans et al, 410; Nonn et al, 415; Hermonat et
systems (Escherichia coli, salmonella, BCG), al, 464), but eYcacy data in humans are not
which produce immunogenic and host protec- available.
tive L1 pentamers, and plant derived VLPs are It will not be feasible to conduct human eY-
also under development (Accardi et al, 456). cacy studies of all the available PV therapeutic
Although PV capsid based vaccines are cur- delivery systems and antigens, and to date no
rently favoured for prophylaxis against PV single immunological assay has shown a corre-
infection, vaccines based on multiple non- lation with eYcacy of immunotherapy for PV
structural proteins were found in one study to induced disease in humans or in natural animal
protect against infection in the rabbit (Han et PV infection. How can we progress from here?
al, 374), an observation worth following up for The usefulness of further studies of vaccine
combined prophylactic/therapeutic vaccines. eYcacy using transplantable E7 expressing
www.jclinpath.com

11. Highlights of HPV 2000 173
tumours in the mouse is limited, and although expressed in HPV transformed cell lines and in
the cottontail rabbit and the beagle dog provide high grade squamous intraepithelial lesions,
attractive models for testing immunotherapy, indicating that loss of this complex and
the natural history of PV infection in these ani- increased phosphorylation of E7 might con-
mals does not closely resemble that of HPV tribute to malignant transformation (Tugizov et
infection in humans. Tumour xenografts to al, 589).
SCID/HU and transgenic skin/tumour models Regarding genetic instability of persistently
might more closely resemble natural human infected keratinocytes, it is of interest that E7
PV infection, but are not ideal for establishing uncouples centrosome duplication from the
vaccine eYcacy because the immune system of cell division cycle and that E6 allows the accu-
these animal models is compromised. It will mulation of centrosome abnormalities (Duens-
probably be necessary to validate surrogate ing and Münger, 471). This will increase the
markers of vaccine therapeutic eYcacy in likelihood of chromosomal mis-segregation
phase I clinical trials, and discard non-valid followed by genomic changes.
markers. Phase I trials of immunotherapy for In addition to its role in the epidermal
warts/anal intraepithelial neoplasia (AIN)/ growth factor receptor signal transduction
CIN/vulvar intraepithelial neoplasia (VIN)/ pathway, HPV-16 E5 has been shown to down-
RRP will be preferable for evaluating surrogate regulate the cyclin dependent kinase inhibitor
eYcacy markers against outcome because p27 (Pedroza-Saavedra et al, 583).
problems with mutated antigen presentation All oncoproteins are involved in the protec-
machinery, and with the eVects of previous tion of the infected keratinocyte from the host
treatments, complicate interpretation of the immune response. HPV-16 E6 leads to re-
data in patients with cancer. True validation of duced expression of E-cadherin, which has
the surrogate eYcacy markers will not be pos- been postulated to play a role in the retention
sible from such studies unless studies are and localisation of Langerhans cells (Matthews
blinded, placebo controlled, and randomised et al, 665). The E7 protein blocks the antiviral
because of the subjective nature of the and antigrowth activities of IFN- , possibly by
outcomes from such studies. binding to cellular p48 (Barnard and McMil-
lan, 579). However, it does not alter the activ-
Biology of PV and virus and host ity of IFN- . In cells transformed by E5 or the
interactions homologous E8 protein of bovine PV-4, the
Contributions to biology improved the basic export of MHC class I molecules to the cell
knowledge of this virus family and paved the surface is downregulated, impairing antigen
way for epidemiology, diagnosis, and immunol- presentation, and thereby helping the virus to
ogy by identifying and cloning HPV types, escape immune recognition (Ashrafi and
demonstrating their association with benign Campo, 592).
and malignant tumours in humans, and reveal-
ing molecular mechanisms of immortalisation REPLICATION AND TRANSCRIPTION
and tumour progression. This provided rea- The viral proteins E1 and E2 are the major
gents and tools for epidemiology and HPV players in the control of replication and
specific, diagnostic screening programmes. It transcription. E7 seems to promote DNA syn-
also defined targets for prophylactic and thera- thesis by directly activating the cyclin depend-
peutic vaccination. ent kinase CDK2 and enhancing the phospho-
rylation of histones and possibly E1 and E2
VIRAL ONCOPROTEINS (He and Fisher, 586). The interaction of E1
Relevant to oncogenicity are the viral proteins and E2 with CBP/p300 has been shown to
E6, E7, and to a lesser degree E5. They are stimulate transcription and replication in
mitogenic and modulate apoptosis; in addition, HPV-18 and HPV-8 (Lee and Choe, 526;
E6 and E7 induce genetic instability. A prime Müller et al, 540). Furthermore, E2 interacts
cause for chromosomal instability is E6 medi- with the transcription factor NF-IL6 to syner-
ated degradation of the cellular protein p53 by gise in transcriptional activation (Hadaschik et
the ubiquitin–proteasome pathway. De- al, 541). A new mechanism for the repression
genkolbe et al (639) showed that E6 targets p53 of transcription by E2 has been proposed. The
in yet another way by binding to the transcrip- E8–E2C fusion protein of HPV-31 not only
tional coactivator CBP/p300 required by p53, represses by competition for DNA binding
thereby inactivating this factor. It is interesting with transactivation competent full length E2,
to note that E6 of bovine PV-1, which does not but also possesses an active repression domain
induce the degradation of p53, can also down- in the E8 part, which mediates repression over
regulate CBP/p300, and that this interaction long distances (Zobel et al, 531).
appears necessary but not suYcient for cell The very low concentrations of E1 in PV
transformation. Regarding the interaction of infected cells always suggested that this replica-
E7 with the retinoblastoma protein pRb, the tion initiator is tightly regulated. It has now
importance of enhanced pRb degradation in been shown that E1 is degraded by the
comparison with the molar E7–pRb interaction ubiquitin–proteasome pathway unless pro-
has been stressed (Gonzalez and Münger, tected in a complex with cyclin E–CDK2
587). Phosphorylation of the E7 protein has before the start of DNA synthesis (Malcles et
been shown to be crucial for promoting S phase al, 622). E1 independent replication has been
entry (Chien et al, 588). Phosphorylation is described for HPV types 16, 18, and 31, start-
inhibited by a macrophage inhibitory factor ing from origins outside the long control
related protein complex, which is no longer region, and is possibly functional during the
www.jclinpath.com

12. 174 Bosch, Rohan, Schneider, et al
non-productive state of the viral life cycle or prevented infection (Giroglou et al, 619). The
during latency (Kim and Lambert, 528). This N-terminal region of L2 has been shown to
is important news regarding antiviral therapeu- bind to cells and to enter the cytoplasm
tic approaches targeting the E1 protein. (Kawana et al, 621). This suggests that the
The E1–E4 fusion protein has so far been minor capsid protein plays an important role in
regarded mainly as a factor in virus maturation virus–host cell interaction.
but now also seems to play a role in replication. Organotypic raft cultures of keratinocytes
It interacts with the human replication initia- were introduced long ago into PV research but
tion protein Mcm7 and thereby inhibits cellu- it is only now possible to produce infectious
lar DNA synthesis (Roberts et al, 631). This viral stocks with titres of 1000 and more in
should allow the recruitment of host cell vitro (Meyers et al, 632). This will be useful in
factors to sites of viral DNA synthesis. future studies of the viral life cycle.
New ways to control transcription and repli-
cation have been revealed by the analysis of RELEVANCE OF CELLULAR GENES
nucleosomes specifically positioned on the viral The importance of cellular genes in tumour
enhancer and the E6 promoter, and flanking progression has been undisputed. Using the
nuclear matrix attachment regions, which bind array technique, several groups have started to
the diVerentiation regulated CCAAT displace- collect data on diVerential gene expression, at
ment protein (CDD) factor (Bernard et al, the moment mainly using archival material.
532). In transient transfection assays, these act Other approaches are already more focused.
as silencers of transcription by deacetylating The E6 mediated induction of transcription of
the nucleosomes. However, after integration of human TERT (hTERT), the catalytic subunit
the viral DNA, the matrix attachment regions of the telomerase, is known to be an important
change into potent enhancers, which may par- step in immortalisation. This is followed by the
tially explain the deregulation of viral gene loss of cellular telomerase inhibitors. Using
expression seen after the integration of viral microcell mediated chromosomal transfer,
genomes into cellular DNA in the course of Snijders et al (485) and Backsch et al (486)
tumour progression. demonstrated telomerase suppression by chro-
mosomes 3, 4, and 6, thereby mapping repres-
INTRATYPE VARIANTS sors, the loss of which is involved in HPV
A newly identified variant of the CRPV diVers mediated immortalisation. Another putative
in biological behaviour from the prototype. senescence gene locus has been mapped to the
CRPVb induced papillomas regress in the short arm of chromosome 10 (Poignée et al,
vicinity of still proliferating CRPVa induced 476).
papillomas (Nasseri et al, 674; Nonnenmacher A great success has been achieved in charac-
et al, 479). DiVerences in risk for persistence terising the genetic abnormality of patients
and malignant conversion were mapped to the with EV, who are susceptible to infection with
control and E6–E7 regions of the variant the subgroup of HPV associated with this dis-
genomes (diVerences in wart regression espe- ease. Genetic linkage analyses on EV consan-
cially to E6) and were strongly dependent on guineous families has recently identified a first
the DR–DQ haplotype of the host. susceptibility locus on chromosome 17q25 and
Similar important diVerences in oncogenic a second locus has now been mapped to chro-
potential seem to exist among variants of mosome 2 p21–24 (Ramoz et al, 628). This is
HPVs. The oncogene promoter of so called the first identification of genomic regions
Asian–American and North American variants involved in the control of HPV infections. It is
of HPV-16 turned out to be about three times of particular interest that the EV susceptibility
more active than the European prototype locus on chromosome 17 overlaps with a
(Kämmer et al, 548), and African 1 isolates psoriasis susceptibility locus, in view of recently
from penile cancers were more active than the published papers on the high prevalence of EV
respective prototype (Tornesello et al, 518). associated PVs in skin lesions of patients with
Certain E7 variants of HPV-45 have in- psoriasis.
creased transforming potential (Geisbill et al,
515). The association of HPV-16 E6 variants Summary and prospects
with cervical cancer appear to be linked to CLINICAL
human MHC (HLA) class II distribution In terms of prospects for the near future, it
(Voglino et al, 650), which may indicate that seems that the most exciting developments will
the risk of specific E6 polymorphism is linked relate to the addition of HPV testing to cervical
to host immune surveillance, analogous to the cancer screening programmes. In this regard,
above data on CRPV. the Dutch trial, which ultimately will involve
44 000 women (Meijer et al, 182), will yield
LIFE CYCLE important data in the next few years. Also, sev-
Adsorption of the virus to its host cell is the eral cost–benefit analyses of HPV detection
first step in the life cycle. So far, 6/ 4 integrin (Sherlaw-Johnson et al, 260; Goldie et al, 267;
has been put forward as a potential receptor Kulasingam et al, 183) for screening for cervi-
but turned out to be not obligatory for bovine cal neoplasia were presented at the meeting,
PV-4. Studies with HPV-11 VLPs pointed to and analyses of this type will be important in
glycosaminoglycans as receptors. A convincing evaluating the feasibility of this approach to
study with HPV-33 pseudovirions showed that screening. The other important development
heparan sulphate proteoglycans are crucial for in relation to screening is the potential applica-
HPV binding because heparinase completely bility of HPV testing, performed perhaps on
www.jclinpath.com