NEW YORK (GenomeWeb) – Chinese genomics firm Genetron Health saw important data published this week from a prospective study of its combined circulating cell-free DNA and protein liquid biopsy test to detect individuals with incipient virus-associated liver cancer.
In the study, published this Monday in the Proceedings of the National Academy of Sciences, investigators from the Chinese Academy of Medical Sciences working with Genetron followed more than 300 individuals with hepatitis B who had normal liver ultrasonography and serum alpha fetoprotein (AFP) levels to see if the company's test could identify those who would go on to be diagnosed with hepatocellular carcinoma (HCC).
According to the authors, at the time of the study endpoint, the test had maintained 100 percent sensitivity and 94 percent specificity with a 17 percent positive predictive value.
If this performance is borne out in future validation, such a test would offer a significant improvement over current screening paradigms for individuals infected with hepatitis viruses and known to be at an elevated risk of developing cancer, including various strategies of imaging surveillance and protein testing.
Although various companies advancing genomic and other molecular cancer early detection tests have said they are collecting prospective data of this type, very little has yet been published actually demonstrating detection of early or incipient cancers in asymptomatic individuals, making Genetron's data notable.
That said, several other firms are targeting HCC and could be competitors to a clinical test. These include Epigenomics, which has collected early evidence that its methylated biomarker approach can offer sensitive detection of liver cancers, as has the Laboratory of Advanced Medicine's IvyGene. Meanwhile, the founders of Chinese firm SeekIn have also said they hold IP around liver cancer early detection.
Others still are advancing techniques for other virus-associated cancers — Grail, for example — which could potentially be translated to the setting of other viral cancers like HCC.
Genetron describes its own approach as combining both cfDNA mutation detection and measurement of plasma proteins.
Founded in 2013, the company has clinical testing labs in several Chinese cities and a research facility in North Carolina. It markets a wide variety of genomic research services, and clinical genomic testing including whole-genome sequencing and targeted panels for tumor tissue genotyping, corresponding blood-based sequencing tests, and germline analyses.
For the HCC assay described in the PNAS study, the group decided to combine sequencing of a predetermined panel of genetic alterations that are known to be highly prevalent in viral HCC with measurement of serum AFP and des-gamma-carboxy prothrombin (DCP).
"In previous cancer genomic studies, most HBV-associated HCCs harbor at least one mutation in … TP53, CTNNB1, AXIN1, or the TERT promoter," the study authors wrote. They also considered the hepatitis B virus integration breakpoint as an additional target. "Since the HBV integration site should be unique in each individual cell, the detection of multiple copies of a specific integration site … could be an indication of clonal expansion," they added.
In the resulting assay, extracted cfDNA is ligated to a customized adapter with a DNA barcode and amplified. Target point mutations and HBV integrations are enriched via a method the authors said is "similar to rapid amplification of cDNA ends (RACE)," using multiple primers covering coding regions of the named genes and the HBV sequences.
Before applying the Genetron test prospectively, the investigators first applied it to a retrospective training cohort, which included 135 individuals who had liver nodules and/or elevated serum AFP and then went on to be either diagnosed or not diagnosed with HCC. In this group, the assay separated cancer cases from non-cancers with 85 percent sensitivity and 93 percent specificity.
They then tested 331 AFP- and ultrasound-negative individuals, among whom 24 were positive based on the algorithm established in the training set. They then tracked these individuals over an eight-month period. Seventeen individuals received dynamic CT scan, four had additional AFP testing and ultrasound, and three were followed with telephone interviews.
Overall, four of the 24 test-positive individuals were diagnosed with HCC during this time period, yielding a positive predictive value of 17 percent.
No cases of cancer were discovered in the test-negative group, including 70 individuals who agreed to also receive dynamic CT examination at six to eight months after testing; another 172 subjects that had follow-up AFP and ultrasound, and 65 who were followed up with telephone interviews.
Based on these results, the researchers calculated a 100 percent specificity and 94 percent sensitivity.
All four tumors were less than 3 cm when diagnosed by CT scan and were found in patients that didn't have liver cirrhosis at baseline.
Blood was drawn a second time from 13 out of the 24 positive cases and re-tested at six to eight months.
One case diagnosed with HCC continued to show a positive result, with a higher score than seen at the first testing timepoint. Another HCC case had already had the tumor surgically resected by the time of the second blood draw and the liquid biopsy yielded a negative result, consistent with this status, authors wrote.
Seven out of the 11 individuals who had positive tests but had not been diagnosed with cancer were negative on their second test, though two were close to the positive threshold.
The remaining four cancer-negative, test-positive cases were still positive at the second time point, and all are continuing to be followed to further refine the assay performance numbers.
Test-negative patients are also still being followed, a Genetron spokesperson said in an email. This will be important to confirm the test specificity that was calculated in the current study. If individuals who tested negative do at some point go on to develop cancer, it would mean that the assay is not 100 percent specific, or that it only offers this high specificity in predicting near-term diagnosis.
According to the firm, investigators are also now working on a larger-scale clinical trial to further improve and validate the method. The firm declined to discuss its commercialization timeline for the HCC early detection test, or its regulatory strategy in either the US or China.