High-risk HPV is not associated with epithelial ovarian cancer in a Caucasian population

The carcinogenic potential of high-risk HPV is well described and includes several
mechanisms. Viral oncogenes E6 and E7 are pivotal elements, since expression of these
genes impairs the function of host-cell tumor suppressors p53 and retinoblastoma protein,
thus favoring malignant transformation 23], 24]. Furthermore, molecular studies have demonstrated the capability of E6 and E7 to
target cellular factors like paxillin, tuberin, E6-AP, E6-BP, E6TP1 and TNFR-1 as
well as cell cycle regulators such as cyclins, cyclin-dependent kinases and CDKs-inhibitors
24]–26]. The resulting suppression of apoptosis and the prolongation of the cell life span
maximize HPV DNA replication in the infected cell. High-risk HPV is an established
etiological factor in anogenital and oropharyngeal cancers 9], 10]. The role of HPV in the carcinogenesis of EOC is controversial. Our study indicates
that HPV is unlikely to be involved in EOC carcinogenesis in the Danish population
as only one of 191 DNA sample was HPV 18 positive and none were positive for HPV 16.
Furthermore, additional testing for several other HPV subtypes confirmed that these
HPV subtypes were also found to be negative. This is in line with several previous
studies, although results are conflicting (Table 2) 11], 12], 15], 16], 18], 26]–35].

Table 2. Studies published during the last 20 years, reporting on HPV prevalence in EOC tissue

The strength of the present study was the large number of samples compared to previous
studies. In addition, the detailed description of the characteristics of the patients,
including information on tumor histology and FIGO stage, should be noted. Some limitations
should also be addressed. Indeed, HPV DNA detection may be less optimal in FFPE tumor
material compared to fresh frozen tissue due to nucleic acid degradation 35], 36]. Recently we have, however, successfully demonstrated HPV DNA in 90 % of anal cancers
using a comparable FFPE material in the same laboratory 23], 36]. Moreover, the previous studies with the highest reported HPV prevalences used FFPE
material 11], 13]32], 33] (Table 2). Our study did not include a control group. However, since we report a very low
prevalence of HPV DNA, it is our opinion that a control group would not alter the
conclusions of the study. Another limitation is that signs of HPV infection may be
lost in the interval between the primary infection and the diagnosis of cancer. For
comparison, the time between primary HPV infection and the development of cancer of
the cervix uteri has been estimated to be 15–20 years 37]. However, in other HPV-related cancers, such as cervical and anal cancer, viral DNA
is present in the malignant tumor tissue and not exclusively in the precancerous lesions
38], 39]. Moreover, continuous expression of viral oncogenes is considered necessary for the
sustenance of the malignant phenotype in cancers associated with HPV 40]. In the event that HPV is involved in the pathogenesis of EOC, we would therefore
expect that HPV DNA would still be present. A number of aspects could account for
the conflicting results of the available studies on EOC and HPV. Firstly, different
analyzing methods were used. For instance, Wu et al. used both In situ hybridization
and immunohistochemistry on the same 50 FFPE EOC samples and detected HPV 16 DNA in
52 % and 36 % of samples, respectively 13]. In addition, differences in race or country of origin may play a role. Studies from
Asia generally report higher prevalence of HPV in EOC tumor tissues than studies from
the Western countries 35], 36]. This is in support of our findings. Several factors could account for this difference.
Firstly, HPV may be more prevalent in some regions of the World and most notably in
the developing countries 41]. Another factor could be the distribution of more virulent high-risk HPV strains
in Asia 42], or that the Asian populations have a higher genetic susceptibility to HPV-induced
carcinogenesis, e.g. through expression of certain variants of polymorphisms like
the TNFA-308G/A (rs1800629) and -238G/A (rs361525) or the p53 codon 72 polymorphism 43]. Despite the high Asian prevalence of HPV in EOC tissue, the overall prevalence of
ovarian cancer is lower in Asian countries compared to Western countries 44]. This speaks against a connection between EOC and HPV even though the etiology of
EOC is most likely multi-factorial.

Our results do not support the theory that HPV is associated with EOC. However, other
microorganisms may play a role in the development of EOC. Thus, pelvic inflammatory
disease (PID) has been associated with an increased risk of EOC in some epidemiological
studies that also report a dose–response effect, with more episodes of PID associated
with a stronger risk of EOC 45], 46]. Other studies, however, have not confirmed this association 47]–49]. Inflammatory cells can promote neoplastic transformation by the induction of angiogenesis,
invasion and metastasis and through the release of mutagenic reactive oxygen species
50]. If an association between EOC and PID exists, it is therefore still unclear whether
the microorganisms involved are directly carcinogenic or indirectly promotes a carcinogenic
microenvironment by inducing tissue inflammation. Indeed, inflammation is also a key
point in the incessant ovulation theory that is characterized by repetitive damage,
resulting in inflammation in the ovarian surface epithelium 33]. Pelvic inflammation is also a characteristic of endometriosis, and the condition
has been associated with increased risk of especially clear-cell and endometrioid
carcinomas 51]. However, a direct or combined mechanism is also possible since in vitro models have
demonstrated a carcinogenic potential of common bacterial and viral pathogens involved
in PID 52]–54]. Serous EOC is suspected to originate in the precursor lesions in the distal fallopian
tubes 55], and the tubes are often affected by PID 56]. Therefore, neoplastic or precancerous lesions from the fallopian tubes may be more
suitable candidates for future studies on EOC and PID. Conclusively, the role of PID
in EOC is still controversial, and more studies including a broader range of microorganisms
are warranted.