Investigations Into The Molecular Pathogenesis Of Cervical Glandular Neoplasias - ABSTRACT
More than 12,000 women are diagnosed with cervical cancer, an HPV-associated malignancy, annually in
the U.S. There are two major histologic forms of cervix cancer: squamous cell carcinoma (SCC) and
adenocarcinoma (ADENO). Importantly, SCC can be prevented through cervical cytology screening
programs, whereas ADENO cannot. While there is a prophylactic vaccine against HPV, uptake of HPV
vaccination in the U.S. has been poor and there are several generations of women who are already exposed
to HPV infection and will not be helped by HPV vaccination. Therefore, cervical screening at a cost of over $6
billion per year will be needed for several decades at least. Understanding the molecular pathogenesis of
ADENO has important public health value in order to facilitate the prevention of ADENO, for which current
screening strategies are inadequate. In the US and Europe, HPV16, 18, and 45 infections account for >95%
of AIS/ADENO. We examined viral methylation of HPV16/18/45 AIS (n=27) and cervical intraepithelial
neoplasia grade 3 (CIN3) (n=63) cases compared to 90 HPV16/18/45 control infections with <CIN2, and found
AUC's of 0.99 and 0.82 for HPV16, respectively. In addition, we have preliminary data showing that
methylation levels of a host gene panel is significantly higher in AIS lesions (p<0.001) compared to controls
(<CIN2). Furthermore, we have employed an HPV-DNA capture NGS assay to prospectively evaluate viral
integration and found a specific HPV18 viral integration that is maintained for the entirety of the screening. We
further found that this integration site has a progressive increase in HPV-integration reads leading up to, and
with a dramatic spike in coverage directly preceding, the histological identification of an incident CIN3 lesion.
The NGS detected integration likely represents an expanding malignant cell clone and shows strong potential
for clinical application.
We propose to investigate the natural history of HPV and ADENO precursors (i.e., samples prior to the
diagnosis of ADENO and from AIS) with the goal of defining molecular precursors of ADENO, which could be
amenable to immediate treatment and thus prevent ADENO. We will leverage cohorts with large numbers of
HPV16, 18, and 45-positive women (n > 20,000) in whom the vast majority of ADENO occur, to measure host
and HPV viral methylation changes and HPV viral integration to molecularly define the natural history of
ADENO. We will compare and contrast these findings with the better known natural history of SCC. Ultimately,
we will develop a risk-prediction model for AIS/ADENO precursors using HPV genotypes, HPV viral and host
gene epigenetics, and HPV viral integration.