NEW YORK – Genomic techniques that could allow early detection and screening for virus-associated cancers are making strides, according to project leaders who described their efforts last week at a conference on liquid biopsy technologies sponsored by the American Association for Cancer Research.
Chinese University of Hong Kong investigator Dennis Lo, a pioneer in the field, updated attendees in his keynote speech at the meeting on his work to develop an early detection test for nasopharyngeal cancers precipitated by the Epstein-Barr virus (EBV).
Other investigators also shared new data from work focused on cancers linked to the human papilloma virus (HPV).
Because viral DNA offers a very specific non-human target, blood-based early detection of these cancers has been seen by some as a potentially easier goal than early detection of other cancer types, which requires catching minute amounts of mutated sequence fragments or gleaning genome-wide signals in DNA methylation or nucleosome patterns.
Lo has been focused on EBV-associated nasopharyngeal cancers for several years, initially founding a company to commercialize a test, which was acquired by liquid biopsy early detection firm Grail, and then licensing its IP back through a newly founded company, Take2 Health, last year.
Last week, Lo shared how he and his colleagues have worked to improve the sensitivity, specificity, and simplicity of their method, as well as to explore the biology behind putative false positives identified in the first population-based study of the test, a 20,000-patient study published in the New England Journal of Medicine in 2017.
Because EBV positivity isn't a perfect correlate for the presence of cancer, the initial version of the test required individuals who were positive for the virus at a single time point to return for a second screen four weeks later.
Doing a second screen helped to weed out individuals for whom viral DNA is only transiently detectable, likely due to a reactivation of the virus in the body, and in the initial study, this reduced false positives around fourfold.
But for clinical implementation, the team believes a test that needs to be administered only once will have a better chance at adoption, Lo said at the meeting, so in subsequent work, some of which was published in 2018, his team has worked to incorporate DNA fragment size and concentration into an algorithm that improved the test's positive predictive value to about 19.6 percent from 11 percent with the initial approach.
Since then, the team has also analyzed how adding epigenetic information to EBV viral DNA detection might improve the test even further. At the AACR meeting, he said that he and his colleagues have identified methylation patterns that are unique to different types of EBV-associated disease — distinguishing the EBV DNA shed from a nasopharyngeal tumor from DNA associated with active infectious conditions like mononucleosis and from other EBV-associated cancers like lymphomas.
Adding this to DNA detection, fragment size, and concentration, Lo and colleagues have now upped the PPV of their test to 36 percent.
Interestingly, the team is also doing important work to try to understand the biology behind putative false-positive calls — individuals who have detectable EBV DNA of the size and concentration that might indicate cancer but who show no signs of a tumor with MRI or endoscopy.
To do this, the researchers have begun a second study, for which recruitment is about half complete at 11,000 individuals, Lo said. In this cohort, the group is retesting patients after four years and looking whether patients with early false positives have an increased chance of a later cancer diagnosis.
According to Lo, the data so far suggest that if someone is negative at their first test, their chance of developing cancer four years later is about 0.1 percent.
For those who have an initial positive result and then a negative result after retesting — what the group calls transient positives — the likelihood of cancer after four years rises to about 0.5 percent. For persistent positives — those who test positive and are still positive several weeks later, despite showing no signs of cancer — the chance of diagnosis at four years rises to 1.4 percent.
Lo said this implies that what the group is seeing is "not just false positives." One possibility at play could be that some of the individuals in question do have very early cancer, but "so early that you can't see it on endoscopy and MRI," he suggested. Another could be that persistent viral reactivation leads to a higher risk of developing cancer.
"Regardless, [this is evidence that] these people are at higher risk and [might] need to be screened more frequently," he said.
Apart from nasopharyngeal cancer and EBV, HPV-associated cancers have also become an area of increasing focus. While there is an existing screening paradigm for some HPV-linked tumors — namely pap smears for cancers of the cervix — other presentations, like oropharyngeal and anal tumors, aren’t current screened for.
Last year, investigators from Memorial Sloan Kettering Cancer Center published a study in JCO Precision Oncology describing a blood-based approach using droplet digital PCR to detect early-stage HPV-associated cancers based on sequential HPV16 and HPV33 assays that account for HPV subtype distribution and subtype sequence variants. The authors said they believe the low-cost approach could serve as an initial screening tool.
During the meeting last week, University of Chicago surgical oncologist Nishant Agrawal made the case for a combined viral and non-viral approach for the head and neck cancer space, highlighting results from a number of HPV-focused studies, as well as efforts by himself and colleagues using targeted sequencing of frequently mutated genes. This type of combined approach could be applied across the spectrum of head and neck cancers — not only HPV-associated cancers but also other squamous cell tumors of the mouth and oral cavity, in which viral DNA plays no role.
Agrawal highlighted that liquid biopsy offers the potential not only to detect cancer, but also to monitor patients after treatment and surveilling them over time. Various other groups have taken up HPV-focused methods in this vein, as well.
Investigators from the University of North Carolina, for example, have completed early work on a method to track patients being treated for HPV-associated oral cancer for signs of remission or recurrence.
In a study of 89 patients with HPV-associated oral and throat cancers who received definitive chemoradiation therapy, investigators found that ddPCR-based detection of circulating HPV DNA could distinguish patients who would later recur from those who remained disease-free.
The Memorial Sloan Kettering team also explored patient monitoring in their ddPCR study, examining the ability of their assay to measure patient responses to chemoradiation and surgery in a subset of 68 patients.
For that, they tested samples obtained every week and observed rapid declines in HPV across their cohort, with signs of viral DNA generally clearing by about seven weeks after the start of chemoradiation in all but three patients.
Also, in a poster at the AACR meeting, investigators from Toronto's Princess Margaret Cancer Center shared new data from a study they conducted using HPV DNA to monitor cervical cancer patients for signs of minimal residual disease, or MRD.
Eschewing PCR for an NGS technique they call HPV-seq, the group retrospectively analyzed samples from a prospective clinical trial cohort of advanced cervical cancer patients treated with radical chemoradiotherapy, comparing results at baseline, immediately post-treatment, and at a three-month timepoint.
According to the authors, detectable HPV ctDNA at the end-of-treatment timepoint was associated with inferior progression-free survival, with 100 percent sensitivity and 67 percent specificity for recurrence.
As highlighted by Agrawal, virus-based testing approaches are inherently limited in the cancer types that they can address. That said, applications aren't limited to EBV and HPV. Liver cancers associated with viral hepatitis could also be a future use case, with at least one company, Chinese genomics firm Genetron Health, now working on a liquid biopsy approach.