NEW YORK – Most diagnostic COVID-19 tests today – based on real-time PCR, CRISPR, or antigens – simply tell whether or not a patient is infected with SARS-CoV-2. Meanwhile, numerous research studies are underway that are sequencing the viral genome in patient samples in order to study how the virus spreads and evolves over time.
Several academic and commercial laboratories, though, are betting on diagnostic viral genome sequencing assays to complement existing testing options. Such tests, they argue, provide additional information to help trace the origin of an infection. At least one hospital has already employed a research-use-only SARS-CoV-2 genome assay to determine whether an infection of its healthcare workers came from within or outside.
Viral genome diagnostic tests could also avoid some of the supply issues plaguing quantitative PCR assays and potentially offer increased sensitivity. Because these tests analyze all or most of the viral genome instead of just focusing on a few specific regions, they are distinct from high-throughput diagnostic COVID-19 sequencing assays currently in development by a number of other groups.
As of this week, no sequencing-based diagnostic COVID-19 assay had obtained Emergency Use Authorization from the US Food and Drug Administration yet. However, several laboratories, including Fulgent Genetics and Clear Labs, have submitted EUA applications for their tests.
Fulgent's NGS assay
Genetic testing company Fulgent Genetics said in March that it was developing two diagnostic coronavirus tests, an RT-PCR-based one and a sequencing-based assay.
The company now offers both, running thousands of tests per day, though the majority are RT-PCR tests, mainly because of their shorter turnaround time. Emergency use authorizations for both tests are pending.
One reason the firm also offers an NGS test is that RT-PCR tests may face reagent shortages. "With the NGS test, we don't have that issue, we don't need to fight with anyone for supplies," said Hanlin (Harry) Gao, the company's CSO and lab director.
The first COVID-19 assay the company launched was actually the NGS assay because it was afraid of potential RT-PCR reagent supply issues and had a lot of sequencing capacity available. "We did run some clinical samples [by NGS] and continue to do so," said Brandon Perthuis, Fulgent's chief commercial officer. But to control the spread of SARS-CoV-2, speed is more of the essence, so the RT-PCR assay is currently the preferred option with a turnaround time of 24 hours versus three days for the NGS assay. The NGS test is also more expensive, he said, "but not in a prohibitive way."
So far, the company has not run into any reagent supply problems, "but should there ever be any issues with supply chain, our NGS test can serve as a backup in our laboratory because it is up and running and validated as a clinical test," Perthuis said.
In addition, the company may use the NGS test to analyze samples for clinical trials where investigators are interested in viral mutation and not just infection status.
Also, the NGS assay, which is amplification-based and primarily runs on Fulgent's Illumina NextSeq sequencer, would be less prone to errors if the viral genome were to mutate than the RT-PCR assay, which might be affected by mutations in the amplicons it analyzes.
The test could also easily be expanded in the future to include other pathogens, Gao said. Overall, he said, it provides more information, which "can help you to understand the virus better" instead of just providing a yes-or-no answer.
Fulgent currently tests swab, not saliva samples, and has found that mid-turbinate swabs, which were initially not authorized, quickly gained in popularity and now make up the majority of its samples. "What we can tell from our clients is, the mid-turbinate [swab] is actually easier and more readily accepted by patients," Perthuis said.
The company's clients for the COVID-19 tests include private physician practices but also large hospitals and academic institutions, as well as companies that want to screen their employees, he said.
The bottleneck remains patients' access to testing. "We currently have more capacity than we have samples, so the lab is not the bottleneck, it's the accessibility and the sample collection," he said. "These drive-through facilities can only get to so many vehicles per day, per minute, per hour. Only so many people can get into a clinic. So, really, the bottleneck to doing more testing is the actual collection of the sample." Seeking to improve this issue, Fulgent recently submitted an EUA application for an at-home sample collection kit to the FDA. Once approved, that swab kit will be delivered through Fulgent's consumer genomics brand, Picture.
Fulgent plans to remain in the COVID-19 testing business "so long as there is a demand," Perthuis said, and is currently building additional capacity, with the goal to be able to provide 10,000 tests per day in a few weeks.
Besides Fulgent, other companies have also started to explore NGS-based COVID-19 diagnostics. Food testing firm Clear Labs, for example, said this week that it plans to sell an assay system for COVID-19 testing that includes library preparation and nanopore sequencing.
Also, cancer liquid biopsy firm Guardant Health said last week that it is developing an NGS-based COVID-19 assay that analyzes saliva samples, though it provided few details about its approach.
Consumer and population genomics company Helix, too, said that it is working on an NGS-based coronavirus diagnostic and is considering both a high-throughput assay and a viral genome approach. "Our focus is on highly sensitive diagnostic testing to enable 'back to work' applications (which will require massive scale) with surveillance as a secondary priority," Justin Kao, the company's cofounder and chief business officer, said in an email.
Baylor's POLAR
Academic laboratories are also working on diagnostic SARS-CoV-2 genome sequencing assays. Last month, a team led by Aviva Presser Aiden and Erez Lieberman Aiden at Baylor College of Medicine put out a preprint on BioRxiv, updated earlier this week, about a whole-genome SARS-CoV-2 diagnostic assay.
The test, called Pathogen-Oriented Low-Cost Assembly & Resequencing (POLAR), starts with RNA extraction and combines an enrichment method for the virus that was originally developed for nanopore sequencing by the ARCTIC Network with sequencing on an Illumina instrument.
Using this method, the researchers were able to detect SARS-CoV-2 at a concentration of 84 genome equivalents per milliliter, which they said is better sensitivity than almost all tests with an FDA EUA. In their paper, they provided a list of tests with EUAs, ranked by their limit of detection, which placed their own testfourth, after assays from PerkinElmer, Becton Dickinson, and Viracor. At higher concentrations, they could also assemble the entire viral genome, often with zero gaps.
More recently, in collaboration with another group at Baylor, they have validated their method on 10 clinical nasopharyngeal swab samples, comparing the results to RT-PCR tests, and found the two methods to be completely concordant.
Along with the preprint, the researchers put out their code for automated data analysis and report generation.
Presser Aiden, an associate professor of pediatrics and genetics at Baylor, said her group has been pursuing a more general whole-genome sequencing diagnostic approach to detect pathogens in clinical samples for a while, and this was an extension of this work.
"Beyond a positive or negative result, we actually have information about how these viruses gained or lost certain mutations over time, which can lead to a phylogenetic analysis or some sort of survey of how the virus has changed over time," said Brian Glenn St. Hilaire, a graduate student in the department of genetics and genomics at Baylor and the lead author of the study.
He acknowledged that the test, which has a turnaround time of about 24 hours including data analysis, is slower than other diagnostic COVID-19 tests. "It does take longer," he said, "but we generate [data for] hundreds of patient samples" in a single run.
According to the researchers' estimate, one technician can process 192 samples in 8 hours and the assay consumables cost is about $30 per sample. "Moving forward, we're driving down the cost to less than $20, if not $10," St. Hilaire said.
Another goal is to increase the throughput, so a single technician can prepare 384 or more samples per day, which will involve some degree of automation. Additionally, the researchers aim to further lower the limit of detection to 10 or fewer genome equivalents per milliliter.
Presser Aiden said the researchers are finalizing an EUA application for the assay and are debating whether they want to offer it as a lab-developed test in a CLIA lab, or whether they should develop it into a test kit that individual labs could run on their own. "We're exploring multiple avenues to figure out what's going to be the best way to disseminate it," she said, adding that the team is currently in discussions with several groups.
Researchers at the HudsonAlpha Institute for Biotechnology are also looking into full viral genome sequencing assays for diagnostics. Shawn Levy, CSO of HudsonAlpha Discovery, said his group, like the Baylor researchers, is considering using the ARCTIC primers to PCR-amplify the viral genome followed by Illumina sequencing.
In addition, they are testing a single-tube multiplex PCR assay from Swift Biosciences. Earlier this month, the firm launched its Swift Amplicon SARS-CoV-2 research panel, which has a 2-hour workflow, and said it plans to release an improved version of the assay soon to accommodate low viral titers found in clinical samples. "At the end of the PCR amplification, you get a normalized library ready for sequencing out in one step. So, it's a little bit more streamlined for Illumina sequencing compared to the ARCTIC method, but it's not exactly the same coverage levels," Levy said.
Levy, whose group is also considering a high-throughput COVID-19 sequencing assay in addition to the whole-genome approach, said his team is currently optimizing the assays and is in no hurry to submit an EUA application, though that is the goal.
Illumina's respiratory virus panel
While Fulgent and Baylor are pursuing SARS-CoV-2 whole-genome sequencing approaches, others are going one step further to include the genomes of other respiratory pathogens.
Illumina, for example, has developed a research-use-only Respiratory Virus Oligos Panel, which contains hybridization capture probes to detect SARS-CoV-2 as well as about 35 other viral pathogens, including recent flu strains and other human coronaviruses.
The company has teamed up with metagenomics data analysis firm IDbyDNA, which has developed a rapid data analysis solution that customers can access through Illumina's BaseSpace cloud-based portal.
Clinical laboratories can validate the assay, which can be performed in about 24 hours, internally for diagnostic use and need to apply for their own EUA. "We are in discussion with a number of laboratories that are working on that or considering that," said Robert Schlaberg, IDbyDNA's cofounder and chief medical officer, adding that he expects the first lab to launch the assay as a diagnostic within the next few weeks.
He said that some laboratories, especially those that already use next-gen sequencing for other types of clinical tests, might find it easier to implement a sequencing assay than to establish a qPCR assay.
In addition, the IDbyDNA test provides more information about SARS-CoV-2 than a qPCR test and can help track the transmission of the virus. Moreover, it can identify co-infections. The panel could also be expanded further to include bacterial and fungal pathogens, which are treatable infections.
Another advantage is that the assay is easier to scale than a qPCR assay, said Guochun Liao, IDbyDNA's cofounder, president, and CEO. "PCR is scalable in a lower-throughput context. Most of those instruments cap out at 1,000 samples or so a day," he said. "Above that, you need more of those machines and they are expensive and need to be installed. With sequencing, there are lower-throughput and very high-throughput instruments."
He did not provide the reagent cost for the test but said that using the targeted enrichment approach, "you can drive the cost of goods down to $100 or less," which would be comparable to a multiplex PCR test.
Liao said that several labs he is aware of are considering an NGS diagnostic test for the fall flu season and are in the process of validating the assay.
Megan Lai, a clinical genomics specialist at clinical genetic testing firm Personalized Genomics in Pittsburgh, said her lab is currently in the planning stages to implement the Illumina NGS workflow for SARS-CoV-2 for research use, though it is exploring potential options for clinical use and diagnostics. "We anticipate the main application is to look for genomic diversity of the SARS-CoV-2 virus among different patients," she said in an email.