A Boost for Cervical Cancer Screening

Despite a declining incidence rate in the US, cervical cancer remains the fourth most common cancer in women worldwide, causing 270,000 deaths in 2015, according to the World Health Organization. High-risk, persistent HPV infections are often a precursor to the disease, with about 0.8 percent of patients developing cancer over 10 to 30 years. The current HPV test with Pap samples, although very sensitive, can be relatively nonspecific and requires extensive follow-up testing for high-risk, HPV-positive patients. To tackle that problem, we and our colleagues at the National Cancer Institute (NCI) decided to create a more specific alternative by developing a blood-based HPV circulating cell-free DNA (ccfDNA) assay (1) – a test we believe would be a potentially valuable resource for the early detection of cervical cancer.

Our involvement in cervical cancers came about thanks to two main driving factors. First of all, there is a significant need for clinical biomarkers for the development of novel immunotherapies, including checkpoint antibody-based and T-cell based therapies (both investigational therapies originally invented at NCI). Specifically, there is an immediate need to select patients based on their HPV genotype for experimental T-cell therapies. Second, we recognized the real potential of using circulating cell-free HPV DNA for patient monitoring; the high copy numbers of virus genome per cancer cell and the lack thereof in normal cells led us to consider the use of such an assay for treatment assessment and recurrence evaluation.

In our initial blinded tests in 2017, most patients had already undergone local surgical or radiation therapies, which meant that the tumor cells were no longer at the original site. We suggested that an HPV ccfDNA assay would complement CT scans in metastatic cervical cancer patients as a routine follow-up monitoring method. Why? Because such an approach is low-cost and non-radioactive. Droplet digital PCR (ddPCR) was the obvious choice for single DNA molecule counting – it’s the most sensitive method for single molecule detection, and it provides accurate quantification without the need for test calibration. We observed the long-term clearance of HPV ccfDNA only in cervical cancer patients who had complete responses to an experimental T cell therapy.

Although our aim was to create a tumor marker, we were still surprised by the rapid induction of HPV ccfDNA immediately after the infusion of cervical cancer tumor-infiltrating lymphocytes (TILs) isolated for patients. Our results suggest that peak tumor killing occurs two to three days after TIL infusion, which means that ccfDNA analysis may provide useful proof-of-concept information for an early investigational therapy.

Our data further show that ddPCR-based HPV genotyping can also be very accurate in determining the viral genotype: we were able to correctly identify the HPV tumor genotypes in 87 out of 87 blood samples from HPV-positive cervical cancer patients. This would allow the selection of cervical cancer patients for HPV-targeted T-cell therapies. Quantitative PCR for ccfDNA has already been approved by the FDA for detecting EGFR mutations in lung cancer patients and is used for Epstein-Barr virus in nasopharyngeal carcinoma, so the precedent has already been set for its use in similar scenarios. Additionally, our ddPCR test is currently under clinical evaluation for patient selection in T-cell immunotherapies to replace the need for invasive biopsies. Over the next few years, the successes of these clinical trials will determine the test’s potential as a companion diagnostic.

If our trials are successful, lab implementation should be relatively straightforward and could even translate to a number of applications beyond cancer diagnosis. For example, after liquid biopsy reveals genotyping, ccfDNA analysis can let you know how compatible the intended therapy is with the individual patient, thus bypassing an invasive tissue biopsy. Such an advance would mean that only patients with specific viral genotypes would be enrolled for the corresponding treatment. HPV ccfDNA could also be used to monitor a patient’s response during therapy and provide information on the likelihood of disease recurrence. Our approach isn’t intended as a replacement for CT scans, of course, but rather as a low-cost, non-radioactive alternative that could be used as a marker for routine cervical cancer testing.

Our assay is traveling beyond our own research. Colleagues within NCI are currently working on the commercialization of T-cell therapies by using the novel approach of T-cell receptor (TCR) transfer, using our assays to assist with their work.

Molecular diagnostics – either via tissue or liquid biopsy – is playing an increasingly significant role in determining cancer treatment, either with targeted or immunotherapies. Many cancers in the near future will be treated based on molecular or mutational analysis, rather than histological classification, so assays like ours are useful in facilitating that transition.