NEW YORK – Two months into the COVID-19 pandemic, the number of SARS-CoV-2 molecular diagnostic tests across the US remains woefully inadequate.
According to the COVID Tracking Project, a volunteer effort to track testing across the country headed by The Atlantic, more than 4 million coronavirus tests had been conducted in total as of Monday and about 150,000 per day in the last couple of weeks. Experts say this is not nearly enough to track new infections adequately, which would be needed to gradually lift stay-at-home orders and other restrictions on public life and businesses.
While there are several reasons for the continuing testing shortage, one challenge has been a shortage of reagents, in particular RNA extraction kits. In addition, many SARS-CoV-2 assays are difficult to perform in high throughput to allow for a real boost in test numbers.
Such kit shortages may not affect tests from vendors that sell fully automated systems like Roche, Cepheid, Hologic, Abbott, or GenMark, but they do apply to the dozens of labs around the country that have developed their own tests or run the Centers for Disease Control and Prevention's assay.
To try to relieve this bottleneck, several labs and at least one company have been working on assays that add samples directly to the PCR reaction and don't require RNA extraction, a time-consuming step.
A group at the University of Vermont College of Medicine and their collaborators at the University of Washington, for example, published a BioRxiv preprint this month — an update of an earlier version they posted in March — in which they described an RNA-extraction-free SARS-CoV-2 test. Validating it on 155 known positive nasopharyngeal samples, they found that it correctly called 92 percent of them.
Jason Botten, an associate professor at the Vermont Center for Immunology and Infectious Diseases and co-corresponding author of the study, which is currently under review at a journal, said his virology research lab never expected to be getting into COVID-19 PCR diagnostics but jumped into it in order to help testing labs in Vermont that were short of RNA extraction kits.
His lab learned about the critical shortage of Qiagen viral RNA extraction kits in early March and originally set out to determine whether another type of Qiagen kit that was still available could also be used. "During the discussions, a couple of us said, 'What if we just exclude the extraction step?'," he recalled. Doing this would not only eliminate the kit shortage but also speed up the test. "This extraction step is the most limiting step of the whole assay," he said. "It takes more time to run the extractions than to do the downstream PCR."
Within less than a week, his team designed an RNA extraction-free assay, tested it on two NP samples from positive patients to show that it works, and posted the results as a preprint. After testing it on a few more samples, his group teamed up with Keith Jerome, who runs a clinical testing lab at UW, and validated the assay with 150 additional samples from patients with high, medium, and low viral loads. They found that the only samples the test missed had low levels of virus, a category that about a quarter of all samples that tested positive at UW fell into.
In the meantime, at least two other groups, based in Chile and Denmark, have independently reported similar RNA-extraction-free protocols, and one of them showed it also worked with nasal swabs. "It seems like people are getting this to work everywhere," he said.
Botten said he could envision the RNA-extraction-free assay being used as a screening tool that could pick up the great majority of people who are virus-positive while saving time and labor, and not requiring extraction kits. Individuals who test negative but are suspected of having the disease would need to be retested with an assay that does use RNA extraction, he added.
Testing labs in Vermont ultimately did not adopt his protocol, he said, because they were able to start using large automated systems, for which they could obtain sufficient reagent kits. Likewise, the UW collaborators have not seen the need to adopt the assay yet.
However, in response to the preprint, Botten has received dozens of emails from labs around the world saying they could not start testing because of a lack of RNA extraction kits. "I think the real benefit of this might be in all of the areas in the world that just don't have access to these kits or don’t have access to these really high-throughput automated machines to do the testing for them," Botten said.
"You are going to want as many of your big, automated, robust systems as you can have online, but that's only going to cover a certain amount [of labs]," he said. "For everybody else, which might be the majority, you're going to need something as simple as you can make, so that all these constraints are removed and people can just rapidly test."
The assay could also be further optimized to increase its sensitivity, he said, for example by using less swab medium, which would increase the concentration of viral particles in the sample.
Nathan Ledeboer, medical director of the microbiology laboratory at the Medical College of Wisconsin, agreed that RNA extraction "is a significant challenge for clinical laboratories" doing SARS-CoV-2 testing, at least for those that run their own lab-developed tests or the CDC test, rather than fully automated systems.
His own lab started out with the CDC assay and was initially unable to get RNA extraction reagents for its two extraction systems, from BioMérieux and Thermo Fisher Scientific. Other labs across the country, he said, had the same experience.
"The fact that in this particular study, they were able to demonstrate a very high degree of sensitivity without having to do extraction is a really a promising alternative to having to extract, particularly as those reagents are very, very limited," he said.
While the extraction kit shortage has improved somewhat over time, he said, there is still a need to test more people and "as we start to expand that testing capability, we will see additional stress being put on the system."
Besides the CDC test, his lab has meanwhile also adopted the Roche and Cepheid platforms for SARS-CoV-2 testing, which has helped to alleviate some of the supply constraints. While more and more labs appear to be shifting away from lab-developed tests and towards commercial platforms, Ledeboer said, and manufacturers are ramping up production, "there is still a tremendous need and the vendors can't entirely supply the number of tests that are needed."
Labs that want to adopt the RNA-extraction-free protocol, he said, would need to validate it internally and submit it to the US Food and Drug Administration as a modification to their current Emergency Use Authorization.
In the meantime, at least one commercial player, BioSkryb, has also developed an RNA-extraction-free SARS-CoV-2 test. In late March, the company tweeted that it had created "a cheaper, faster, safer COVID-19 test, removing the extraction step," adding that it would cost about $15 per sample and could be scaled up to 190 samples per run.
BioSkryb CEO Jay West explained that his company, which is based in North Carolina and has been developing whole-genome amplification technology, has been working on single-cell analysis methods that don't require nucleic acid extraction for several years, "so I knew we could get a PCR test to work on this [viral] material."
The goal was not only to develop a test that doesn't require RNA extraction but also one that is fast and can be multiplexed. The test his team came up with — using reagents and input from New England Biolabs, Twist Bioscience, and Integrated DNA Technologies — inserts nasopharyngeal swabs into a stabilization buffer that inactivates the virus but preserves the viral RNA.
After a brief heating step, the sample is added directly to the PCR reaction. The company has been able to run up 184 samples in parallel, along with a positive control, and the test has a turnaround time of about 2.5 hours, which West said could be reduced to further. The assay was developed on the Thermo Fisher Scientific QuantStudio 7, which West said is used by many clinical labs, but could in principle also run on other qPCR machines. "We developed the base of the test within two and a half days," he said.
The firm was able to detect down to 10 copies of viral RNA, which translates to a limit of detection of 200 copies in the sample. Following the initial validation, Bioskryb shipped the test kit to collaborators at the University of Tennessee who work with live virus to assess its specificity. That group found the test detected SARS-CoV-2 but not the coronavirus that causes MERS. With these data in hand, BioSkryb now plans to apply for FDA Emergency Use Authorization.
It has also provided the kit free of charge to two hospital laboratories for additional testing and may send it to others. "We're taking requests as they come in," West said. "The reality is that we are not marketing this really hard."
These labs are currently verifying the assay and might decide to scale it up, West said, adding that BioSkryb would be capable of producing thousands of tests per week "without any real burden" if needed.
While the company never planned to get into the COVID-19 testing business, "if it's a value-add for the customer, we certainly will make it available," he said. "It's not our overarching objective to become cash-positive on such a test. But in terms of expanding our capability around our core technology, it's a reasonable application to experiment with, and to provide as a public service."
Going forward, he said, BioSkryb is interested in integrating the assay with its core technology for sequencing single viral genomes, with the goal of determining the clonal diversity of SARS-CoV-2.
Meanwhile, other labs have been exploring other ways to relieve coronavirus testing bottlenecks. One lab at Rutgers University, for example, recently started using saliva samples as input material because they are much easier to collect than NP swabs and don't require swabs and viral transport medium, which have also become scarce. However, a lack of time to validate modifications like these in their own labs might be holding others back from implementing some of these newer workflows.
"As the first wave of the outbreak begins to slow down, many of us that would otherwise be doing those studies will get the time to breathe and be able to start thinking about doing that," Ledeboer said. "One of the challenges up until now has really been, clinical laboratories have been running full throttle, if you will, focused on being able to provide as many tests as they can, and as broadly as they can. That has largely been driven by the very rapid expansion of the number of cases that we're detecting. As we start to see this first phase begin to slow down and wane, we really have the opportunity to begin those studies, doing them in well-designed fashions so we can get useful viable data."