NEW YORK ─ Researchers at Johns Hopkins University are combining surface-enhanced Raman spectroscopy, nanotechnology, and machine learning to develop a sensing platform to detect and differentiate SARS-CoV-2 and other viruses.
The researchers described the preliminary validation of the test in Nano Letters on Tuesday, saying their tool detected SARS-CoV-2 and differentiated it from other viruses including H1N1, Marburg, and Zika at 92 percent sensitivity in saliva samples spiked with the viruses.
With an eye on commercialization, the group has started evaluating its platform — a single-use chip that plugs into a machine learning-enabled handheld spectrometer — in saliva samples from SARS-CoV-2 patients.
If they can attract the interest of suitable diagnostic industry collaborators, the researchers "expect to obtain [US Food and Drug Administration] approvals in about six months and hopefully be ready for commercial launch in the following three or four months," said Ishan Barman, an associate professor at JHU, who is one of the study authors.
The group expects to seek either FDA 510(k) clearance or Emergency Use Authorization for the platform and a single-use SARS-CoV-2 test, depending on the agency's policy for clearances and authorizations at the time, Barman said.
Into the third year of the pandemic, nucleic acid amplification and lateral flow antigen technologies continue to dominate the landscape for SARS-CoV-2 detection, but each has its own drawbacks.
"RT-PCR is the gold standard for SARS-CoV-2 detection, but it can take too long to provide results and requires a high level of operator expertise," Barman noted, adding that "lateral flow antigen tests, on the other hand, are much faster but don't have the sensitivity and specificity of PCR."
By comparison, the JHU tool "combines the salient features of RT-PCR and rapid antigen testing for the detection of viruses including but not limited to the SARS-CoV-2 coronavirus," offering high levels of sensitivity and providing results from saliva in less than 15 minutes, Barman said.
Such a platform could be ideal for point-of-care applications where many people need testing, such as airports, schools, shopping centers, football stadiums, and hospitals, he added.
The JHU platform acquires spectroscopic signals from saliva samples placed on a metal-insulator-metal nanostructure, also known as an antenna array. The antenna array is plugged into a handheld Raman spectrometer and machine learning software on the spectrometer detects spectroscopic patterns that differentiate SARS-CoV-2 from other viruses.
While antigen and PCR tests use probes that leverage reactive chemistry to detect virus components, "our approach is quite distinctive in the sense that we directly observe the virus with a laser, similar to the way temperature can be read using an infrared thermal probe," said David Gracias, a JHU professor and one of the study authors.
"Viruses have very complicated spectra," he said, "but the metal nanostructure antenna array enhances viral signals to increase sensitivity." Specifically, the antenna array amplifies acquired signals by nearly eight orders of magnitude, enabling the detection of trace quantities of viruses in saliva, he said.
Meanwhile, the use of machine learning to differentiate disease patterns contributes to the tool's high specificity. "Every virus has a slightly difference signature, so machine learning enables us to not only detect SARS-CoV-2 but also screen for new variants and other pathogens," Gracias added.
Igor Lednev, a professor of chemistry at the University at Albany, State University of New York, said he believes the JHU technology has the potential for use as a screening test for many viral and other types of infections.
The reported results are "impressive and convincing," said Lednev, who is not affiliated with the JHU group. However, the platform and its tests will require "significant" validation to demonstrate that it can eliminate "false negatives and false positives due to environmental interferences," he said.
Other developers are exploring the launch of biosensor-based tests for SARS-CoV-2 detection. Last April, Greensboro, North Carolina-based Qorvo Biotech received FDA Emergency Use Authorization for a biosensor platform and test that uses bulk acoustic wave technology for SARS-CoV-2 detection. Additionally, University of Illinois at Chicago researchers are leveraging graphene nanomaterial to develop diagnostic tests for SARS-CoV-2 active infection and other indications.
Grip Molecular Technologies is developing a single-use, multiplex biosensor testing platform for upper respiratory infections including SARS-CoV-2. And researchers at the QIMR Berghofer Medical Research Institute in Brisbane, Australia, have developed a proof-of-concept instrument that leverages spectroscopy to screen for SARS-CoV-2 and potentially identify people who could become sicker from the infection.
The JHU group has plans to develop its platform to simultaneously detect and differentiate other viruses and noted that the nanosensor arrays can be adhered to both flexible and rigid surfaces, meaning they could be integrated with wearable devices, or placed on doorknobs, masks, or walls to detect viruses on surfaces.
The researchers are in discussions with potential collaborators interested in commercializing and manufacturing the tests or licensing the technology, and they are talking to private investors who may be interested in supporting a startup that would launch the platform and tests, Barman said.
"Our ongoing efforts have allowed us to further improve this to the point where we are approaching the sensitivity of RT-PCR," he added.
In the Nano Letters study, the researchers reported detecting "slightly higher copies per [milliliter] than RT-PCR," which would enable the detection of SARS-CoV-2 in symptomatic patients, but recent development of the platform "allows us to obtain higher sensitivity and provide early detection as well as detect the virus in asymptomatic patients," Barman said.
Additionally, based on limited production in the laboratory, the cost per test is estimated at $30, he said, adding, "We believe this can be reduced to half or one-third that cost as we scale up to mass production."
Costs are reduced by applying high-volume nanoimprint lithography and transfer printing for the manufacturing of the disposable chip and by eliminating reagents and antibodies that drive up the cost of testing in other diagnostic testing platforms, Barman said.
"Because our method requires no sample preparation, it reduces the need for operator expertise, and because it does not require reagents, probes, and antibodies, it eliminates the potential for supply shortages," Barman added.