Genome Sequencing Hints at New Targets for Cervical Cancer

Cervical cancer is a prime example of how early detection of disease can significantly impact patient mortality. Over the past several decades, the rates of cervical cancer have declined dramatically in developed countries with access to effective screening programs. But for patients who slip through the cracks and present with advanced stages of disease, the picture is considerably bleaker and cervical cancer continues to represent a significant burden worldwide.

Notorious for its high metastatic activity and ability to rapidly develop resistance to chemotherapy and radiation therapy, cervical cancer has also proved unresponsive to many of the molecularly targeted therapies that have emerged in recent decades, the exception being the antiangiogenic agent bevacizumab, which the FDA approved in 2014 in combination with paclitaxel and either cisplatin or topotecan for the treatment of patients with persistent, recurrent, or metastatic cervical cancer.

Slow Therapeutic Progress

The lack of success has in large part resulted from a poor understanding of the molecular background of cervical cancer. With technological advancements in genome sequencing, researchers are now gaining a more detailed picture of the genetic drivers of cervical cancer and the important role that the human papillomavirus (HPV), responsible for the vast majority of cervical cancer cases, plays in molding the genetic profile of this disease.At one time, cervical cancer was the most common cause of cancer-related mortality for women in the United States, but major advancements in screening and prevention during the past half-century have significantly improved this picture. The development of the Pap test, allowing physicians to detect precancerous changes in the cervix, has been heralded as one of the most successful cancer screening programs ever. Meanwhile, the Nobel Prize-winning identification of a causative link between the common viral infection HPV and cervical cancer served as a driving force behind the development of HPV vaccines in an effort to prevent infection before it can lead to cancer.

But that’s where the good news ends for cervical cancer. In parts of the United States where access to screenings and uptake of vaccinations are poor, cervical cancer rates remain high, and in developing countries it is still the second- leading cause of cancer-related mortality.

Sequencing Provides New Targets

Mutational Load Studied

Although the overall survival rate for cervical cancer in the United States is 68%, this rate declines dramatically as the disease progresses. Patients who have already developed advanced cervical cancer represent a substantial unmet medical need, and improvements in treatment options have been slow to materialize.The vast majority of clinical trials of targeted therapies in cervical cancer have been performed in unselected patient populations. Researchers have learned from past experience that this can often mask the effects of these drugs, which may only be effective against certain molecular backgrounds.

In an effort to develop more effective therapies, the focus has shifted to improving our understanding of the genetic underpinnings of cervical cancer. The International Cancer Genome Consortium (ICGC) is undertaking efforts to collect genomic, transcriptomic, and epigenomic data from 50 different tumor types, including squamous cell carcinoma (SCC) of the cervix, by coordinating large-scale cancer genome studies.

Significant Mutations in Cervical Cancer

Researchers developed this portrait of significant mutations in cervical cancer using whole-exome, -genome, and —transcriptome sequencing of tumor samples. Mutations in TP53 and ERBB2 in squamous cell carcinomas and in PIK3CA and KRAS in adenocarcinomas were identified from separate analyses focusing on genes previously reported as mutated in the Catalog of Somatic Mutations in Cancer (COSMIC) database.

Ojesina AI, Lichtenstein L, Freeman SS, et al. Landscape of genomic alterations in cervical carcinomas. Nature. 2014;506(7488):371-375.

The cervical cancer study is being carried out in the United States by The Cancer Genome Atlas (TCGA). Among 308 cervical cancer donors, researchers have identified 194 simple somatic mutations thus far. The rate of somatic mutations, a phenomenon known as mutational load, varies substantially between different tumor types, ranging from 0.1 mutations/ megabase (Mb) in tumors like acute myeloid leukemia to >60 mutations/Mb in lung cancer and melanoma. By comparison, the median mutational load in cervical tumors is 4.3 mutations/ Mb.

13 Significant Mutations Identified

PI3K Pathway Genes

Not all of these mutations will be driver mutations that can be causally linked to the development of cervical cancer. A number of other studies have suggested that cervical cancers have an average of 2 to 4 driver mutations, similar to glioblastoma, ovarian, and breast cancers. Having a lower mutational load can present a challenge since it may mean fewer therapeutically targetable alterations.The most comprehensive genomic analysis of cervical cancer to date was performed by a team of international researchers from institutions in the United States, Mexico, Norway, and China and recently published in Nature. Whole-exome sequencing data from 115 tumors was compared with those from healthy tissue from the same individual among a diverse pool of Norwegian and Mexican patients with cervical cancer. Thirteen genes were found to be significantly mutated (Figure), 5 of which in SCCs had not previously been linked to cervical cancer, and 1 (MAPK1) that had never been connected to human cancer at all. Additionally, the study found that the mutational picture differed depending on tumor histology.Among the most notable mutations identified were those in the PIK3CA gene, which encodes phosphatidylinositol-4,5-bisphosphate 3-kinase catalytic subunit alpha, the central kinase in a pathway that is frequently deregulated in cancer. PIK3CA has been linked to cervical cancer in other smaller sequencing studies. A 2013 study that compared the genomic profiles of the 2 most common types of cervical cancer, squamous cell carcinoma and adenocarcinoma, found that PIK3CA mutations were common in both, but slightly more prevalent in patients with SCC (37.5% and 25%, respectively).

Researchers who performed a genomic analysis of cervical cancer patients from Guatemala and Venezuela found a similar trend, with PIK- 3CA mutations in 31% SCC and 24% adenocarcinoma tumors, respectively. Targeted sequencing studies among patients from China, India, France, and the United States also found similar rates of PI3K mutation, several of which noted higher incidence among patients with SCC, and a tendency toward increased PIK3CA mutations in older patients.

Most importantly, a number of these studies highlighted the fact that the most common PIK- 3CA mutations found in patients with cervical cancer are quite different from those observed in other tumor types. Typically, we see mutations in the kinase domain (H1074R, for example), but in cervical cancers this mutation—and mutations in the kinase domain in general—are rare.

Much more common are mutations in the helical domain, predominantly at 2 sites, E542K and E545K.

These mutations are also frequently observed in colorectal, breast, intestinal, ovarian, and endometrial cancers; they do not appear to be associated with an elevation in AKT phosphorylation and may instead involve activation of the AKT-related kinase SGK3. Researchers have suggested that cervical cancers may, therefore, not respond well to AKT/mTOR-directed therapies.

Drugs targeting the PI3K pathway could represent a promising therapeutic strategy. Indeed, the phase II SIGNATURE study of the PI3K inhibitor buparlisib (BKM120) is ongoing, but not recruiting participants, in patients with PI3K-activated tumors including cervical carcinoma (NCT01833169).

HER2 Family

Furthermore, a recent report of phase I experience with PI3K-matched therapy in patients with cervical SCC found that it provided clinically meaningful benefits. Several trials of AKT inhibitors include patients with cervical cancer, including a phase I trial of AZD5363 (NCT01226316) and a phase Ib trial of ARQ 092 in combination with carboplatin and paclitaxel (NCT02476955).Most notable among the newly identified cervical cancer mutations in the Nature analysis were those in the ERBB2 gene, found in 5% of patients. What is potentially so exciting about this small, but significant pool of patients, is that FDA-approved therapies targeting the ERBB2 protein, also known as HER2, already exist and could be repurposed for the treatment of cervical cancer.

Epigenetic Targets

HER2-targeted therapy is currently being evaluated in the preclinical setting for cervical cancer. A recent study identified HER2 protein overexpression among 3.6% of the 412 human cervical cancer tissues analyzed. The researchers used patient-derived xenograft models for human cervical cancer and found that dual administration of a HER2-targeting small-molecule inhibitor (lapatinib) and monoclonal antibody (trastuzumab) significantly inhibited the growth of a HER2-amplified tumor.Recent research has also examined the role of epigenetic regulation of gene expression during cervical cancer development. While the genome can be thought of as the cellular genetic manual, the epigenome essentially tells the cell how to read the manual, by guiding which genes are expressed at a given time. Two of the most commonly observed epigenetic modifications are methylation of the DNA and acetylation of the histone proteins around which the DNA is wound in the nucleus. A delicate balance of these modifications is required and is often disrupted in cancer cells.

The most frequently mutated gene identified in the Nature analysis, and the second most commonly mutated gene identified by the TCGA, is EP300. This encodes a histone acetyltransferase, responsible for acetylating histone proteins, suggesting an important role of epigenetic alterations in the development of cervical cancer.

Immune System Targets

This theory is further reinforced by the fact that a number of other genes among the top 20 simple somatic mutations identified in cervical cancer by the TCGA are epigenetic modifiers, including KMT2C, a lysine-specific methyltransferase. Drugs targeting epigenetic regulatory proteins are under development in a number of other tumor types and also might make a promising therapeutic avenue for cervical cancer.In the Nature study, the researchers also looked at significantly mutated gene sets in 79 cervical SCCs. A key finding from this analysis was the prevalence of mutations in genes regulating the immune system. In addition to finding the HLA-B gene, encoding the histocompatibility leukocyte antigen B protein, which is among the most frequently mutated genes in cervical tumors, the most significantly mutated gene set in the SCC group incorporated immune response genes in the interferon-gamma signaling pathway.

Evolving Understanding of HPV

This suggests that cervical cancers may have a highly dysregulated antitumor immune response and that immunotherapy may also prove to be a good therapeutic option. Indeed, immunotherapies targeting HPV-associated oncoproteins such as the therapeutic vaccines ADXS11-001 and VGX-3100 are showing significant promise in clinical trials.Further adding to the complex molecular background of cervical cancer, genome sequencing has begun to shed more light on the central role of HPV in cervical carcinogenesis.

When HPV infects a host cell, it hijacks the cellular machinery to generate 8 viral proteins, encoded by 6 early genes (E1, E2, E4, E5, E6, and E7) and 2 late genes (L1 and L2). It has long been understood that HPV drives tumorigenesis through the actions of E6 and E7, which act as oncoproteins by targeting numerous cellular proteins, including the tumor suppressor proteins p53 and pRb, and essentially reprogramming the basic functions of the cell.

In addition to being found free in the cytoplasm (episomal HPV), HPV inserts itself into the host genome (integrated HPV), and it is increasingly appreciated that this latter form is critical to cervical carcinogenesis. HPV integration is observed in almost all invasive tumors and it was initially thought that it occurred at a late stage of carcinogenesis, but more recently mixed forms of both episomal and integrated HPV have been found in precancerous lesions.

The HPV genome is circular and, in order to insert itself into the host genome, must be broken apart. Early studies suggested that HPV commonly split apart around the E1 and E2 genes, disrupting their activity. Since these genes were responsible for inhibiting the activity of E6 and E7, this led to increased expression of these oncoproteins and conferred a growth advantage to the host cell, contributing to marked genomic instability and the accumulation of secondary mutations that promoted cancer progression.

More recent studies have revealed that, while the viral DNA does commonly break apart within the E2 gene, breakpoints are also widely distributed throughout the viral genome. These studies have also examined where HPV integrates into the host genome and revealed that this process may be less random than originally thought and may be equally important to the development of cancer. Though previous studies suggested that the virus showed some preference for fragile sites in the host genome, it was basically thought to integrate randomly.

The results of several genome sequencing studies reveal a potential method behind the madness. Using transcriptome analyses, researchers have been able to search for potential patterns of integration across an array of cervical tumor specimens. The integration of viral DNA into the host genome appears to commonly occur next to sites where the host genome is disrupted in some way. The authors of these studies argue that HPV integration may contribute to cervical carcinogenesis not only by disrupting the HPV genome, but also by disrupting the host genome.

HPV integration could confer a selective growth advantage to the host cell and drive carcinogenesis in a number of ways: by integrating at a site that leads to loss of expression of a tumor suppressor; by integrating upstream of an oncogene and promoting its expression; or by integrating in a way that leads to chromosomal rearrangements.

It has been proposed that identification of integrated forms of HPV could be a useful marker of progressive disease. Though technically challenging, clinical studies are ongoing (NCT02576158, NCT02576262, and NCT02554565).

Jane de Lartigue, PhD, is a freelance medical writer and editor based in New Haven, Connecticut

Key Research

• Akagi K, Li J, Broutian TR, et al. Genome-wide analysis of HPV integration in human cancers reveals recurrent, focal genomic instability. Genome Res. 2014;24(2):185-189.
• Bumrungthai S, Munjal K, Nandekar S, et al. Epidermal growth factor receptor pathway mutation and expression profiles in cervical squamous cell carcinoma: therapeutic implications. J Transl Med. 2015;13:244. doi:10.1186/s12967-015-0611-0.
• Hou MM, Liu X, Wheler J, et al. Targeted PI3K/AKT/mTOR therapy for metastatic carcinomas of the cervix: a phase I clinical experience. Oncotarget 2014;5(22):11168-11179.
• Hu Z, Zhu D, Wang W, et al. Genome-wide profiling of HPV integration in cervical cancer identifies clustered genomic hot spots and a potential microhomology-mediated integration mechanism. Nat Genet. 2015;47(2):158-163.
• International Cancer Genome Consortium. ICGC Data Portal: Cancer Projects. https://goo.gl/LtLEUJ. Accessed December 21, 2015.
• Lou H, Villagran G, Boland JF, et al. Genome analysis of Latin American cervical cancer: frequent activation of the PIK3CA pathway. Clin Cancer Res. 2015;21(23):5360-5370.
• Oh DY, Kim S, Choi YL, et al. HER2 as a novel therapeutic target for cervical cancer. Oncotarget. 2015;6(34):36219-36230.
• Ojesina AI, Lichtenstein L, Freeman SS, et al. Landscape of genomic alterations in cervical carcinomas. Nature. 2014;506(7488):371-375.
• Rusan M, Li YY, Hammerman PS. Genomic landscape of human papillomavirus- associated cancers. Clin Cancer Res. 2015;21(9)2009-2019.
• Xiang L, Li J, Jiang W, et al. Comprehensive analysis of targetable oncogenic mutations in Chinese cervical cancers. Oncotarget. 2015;6(7):4968-4975.
• Wright AA, Howitt BE, Myers AP, et al. Oncogenic mutations in cervical cancer: genomic differences between adenocarcinomas and squamous cell carcinomas of the cervix. Cancer. 2013;119(21):3776-3783.