NEW YORK – A team of researchers funded by the National Institutes of Health's Rapid Acceleration of Diagnostics (RADx) initiative has developed a genotyping-based approach to monitoring SARS-CoV-2 variants.
Detailed in a study published last month in the Journal of Clinical Microbiology, the method uses PCR to detect single-nucleotide polymorphisms (SNPs) characteristic of specific SARS-CoV-2 strains and, according to its developers, could offer a faster and cheaper way to track known virus variants.
It could also enable more comprehensive surveillance of both known and new variants, said Eric Lai, a RADx team lead and first author on the paper.
The US Centers for Disease Control and Prevention (CDC) has led the country's effort to track SARS-CoV-2 variants, largely by contracting with labs to perform next-generation sequencing on a sample of positive patient specimens. This sequencing-based approach has a relatively slow turnaround time, however, meaning that reports generated through the effort typically reflect not the current variant mix but that of a few weeks prior. Additionally, because sequencing is expensive and capacity is limited, the CDC-led effort samples only around 5 percent of cases, Lai said.
He suggested that a more efficient route would be to use a genotyping approach like that he and his colleagues published in the JCM paper to identify known variants, while passing variants not identified by genotyping on for sequencing.
"We are not saying to replace sequencing," he said. "If you want to find and identify new variants, you have to do sequencing. What we are saying is that the genotyping approach is faster and cheaper for identifying the variants that are known, and then any samples that fall out of the known variants should be sent for sequencing."
This approach could also avoid wasting money and sequencing capacity during times when known variants are dominant, said Richard Creager, senior author on the paper.
"For example, when Delta was dominant last November, December, it was like 95 percent or 99 percent of all cases in the US, and yet we kept sequencing," he said. "It didn't make a whole lot of financial sense."
Creager was a RADx team lead through March 2022. In February 2022, he became CEO of Walnut Creek, California-based Seegene Technologies, a US subsidiary of Korean molecular diagnostics firm Seegene that sells PCR tests for infectious diseases, including COVID-19, and sells research-use only PCR assays for detecting different SARS-CoV-2 variants.
To develop their surveillance assay, the researchers identified a set of SNPs highly specific for different variants of interest. They built a 48-SNP panel and tested it retrospectively on 1,031 SARS-CoV-2 positive samples, finding that it identified the top 10 most prevalent variants during the time period in which the samples were collected with a positive-percent agreement (PPA) of between 96 percent and 100 percent and a negative-percent agreement (NPA) of between 99 percent and 100 percent.
They then explored whether the panel would retain its effectiveness if they reduced the number of SNPs tested, finding that panels using 24 and 16 markers performed similarly to the full 48-SNP panel. Using a 12-marker panel, they could identify eight of the 10 strains, while using an eight-marker panel they could identify six out of the 10 strains. With the 48-marker panel, roughly 7 percent of samples were called as undetermined. With the eight-marker panel, roughly 27 percent were undetermined.
The researchers also developed a four-marker panel for distinguishing between the Delta and Omicron variants of SARS-CoV-2 and used it to analyze roughly 15,000 positive patient samples collected between November 2021 and January 2022, observing during that time Omicron's prevalence grow from around 40 percent to more than 95 percent.
They have continued to track variant prevalence since then, using genotyping data from University of Washington Medicine, Helix, Aegis Sciences, and Ovation. Since the middle of January they have gathered data on around 81,000 samples, observing the rise in recent months of the Omicron BA.4/BA.5 subvariants.
Bruce Tromberg, director of the National Institute of Biomedical Imaging and Bioengineering (NIBIB) and leader of the RADx Tech program, said the program hopes to continue funding the effort at a similar scale for at least another year.
The genotyping assay is currently done as a reflex test on positive samples, but Tromberg said ideally it would be incorporated into the primary COVID-19 test so that you could get positivity and variant data at once. He noted, though, that labs typically want to run the primary test in a single well, which means ensuring that all the targets required can be read in a single-well experiment.
That shouldn't be a huge technical challenge, he said, but it will require some validation work, and, he noted, test vendors might be leery of investing too much into validation of an assay that could quickly become obsolete as the variant landscape changes.
Thermo Fisher Scientific has been generating primers for the RADx effort, but Creager said he and his coauthors hope to interest other vendors in producing reagents for the assay as well.
Tromberg said RADx has had ongoing discussions with the CDC about potentially implementing the genotyping-based approach, but he didn't think it had "crossed the threshold yet" to where the agency saw it as impactful.
The CDC declined to comment.
Genotyping could also prove useful for tracking SARS-CoV-2 variants globally. As the JCM authors note, roughly 80 percent of SARS-CoV-2 sequences in the GISAID variant database have come from just 10 countries, and the roughly 7.5 million sequences in the database represent less than 2 percent of the estimated COVID-19 cases worldwide.
Lai noted that the UK is one country that has moved to a genotyping approach similar to what he and his colleagues have proposed. The UK announced in March 2021 that it aimed to genotype all positive COVID samples to help it identify known variants while continuing to use sequencing to characterize new variants and mutations.