NEW YORK – Although their highly sensitive platform for antigen detection was developed before the COVID-19 pandemic, researchers in Ping Wang's laboratory at the University of Pennsylvania saw a clear application of the microbubble-based technology once the SARS-CoV-2 virus began sweeping the globe.
Using microbubbling, a digital ELISA method that quantitates protein biomarkers, Wang — a professor of pathology and laboratory medicine at the Hospital of the University of Pennsylvania and the director of the clinical chemistry section and core laboratory at the school's Perelman School of Medicine — said her team "thought this would be a great choice to use to detect SARS-CoV-2 antigens, knowing that detection of antigens for this virus is going to be very critical for … detection of acute infections."
While published research has focused on a laboratory version of the test, there are also plans for a point-of-care version.
According to a study published earlier this month in Clinical Chemistry, the researchers were right. In validating the clinical performance of the technology with 400 patient samples, the sensitivity and specificity of the SARS-CoV-2 test reached 97 percent, with a 95 percent confidence interval.
The listed sensitivity is for symptomatic patients, while the specificity applies to both symptomatic and non-symptomatic patients. For asymptomatic patients, the sensitivity of the test dropped to 45 percent.
The limit of detection was 0.5 picogram per milliliter of the N antigen, or 4,000 copies per milliliter of inactivated SARS-CoV-2 virus.
Wang noted that the sensitivity dropped in asymptomatic people because they "may present to testing at different stages of the infections, sometimes after the antigen has been largely cleared from the body."
The microbubbling test was validated on samples from people at different stages of infection and from special populations, including immunocompromised patients and patients in the intensive care unit. The team has shown "our technology works very well … and is especially very useful in individuals who may have chronic ongoing infections for a long time," Wang said.
One benefit of the test compared to gold standard RT-PCR is that it can differentiate between active and inactive infections, Wang said. PCR is so sensitive that it can pick up nucleic acid fragments that are present but not indicative of an active infection.
"You'll still be able to see positive nucleic acid PCR tests for a long time because you still may have fragments within the body," Wang said.
In contrast, her team's antigen test is "as sensitive as PCR" but won't pick up those fragments, Wang said. "It actually is better in terms of telling you whether somebody is still truly infectious versus [if] somebody is just having some fragments of the virus left behind."
The method has produced high sensitivity for other use cases as well.
In a proof-of-principle study published in Angewandte Chemie two years ago, the team used the technique for post-prostatectomy measurements of prostate-specific antigen, reaching a detection limit of 0.06 picogram per milliliter — a sensitivity level 170 times higher than its comparator test, the Roche Elecsys Cobas Total PSA test.
During the 2019 study, the researchers also used the platform for early pregnancy detection with beta human chorionic gonadotropin as a biomarker. The test had a detection limit of 2.84 picograms per milliliter.
How it works
The SARS-CoV-2 test was validated on nasopharyngeal swab samples mixed with a protease inhibitor to make sure the nucleocapsid antigen hadn't deteriorated in the sample. The sample was then centrifuged and inputted into the team's instrument.
Within the instrument, the protein binds to a pair of antibodies that are attached to magnetic microbeads and platinum nanoparticles, forming a sandwich complex when the N protein is present.
Those sandwich complexes are then pulled into a microwell array on a microbubbling chip via an external magnetic field, where H2O2 is degraded to generate oxygen microbubbles.
Pictures of the plate with the microbubbles are taken with a mobile device — in the recent study's case, an iPhone 11 or iPad — and processed in seconds with an algorithm the team developed to measure a variety of features, including bubble size number and total bubble volume.
A separate machine learning algorithm determines whether a sample is positive or negative for the virus. The entire test takes around two hours to return results.
The team has applied for Emergency Use Authorization to the US Food and Drug Administration for the SARS-CoV-2 test, said Wang, whose lab is integrating the components of the test into a compact and portable platform that automates each of the steps and returns results within 20 minutes that could be used at the point of care.
Wang said the lab has a prototype "pretty much ready" and is in the final stages of developing it, tweaking the parameters to improve performance.
Instead of a plate, the POC instrument would have a readout window for users to take a picture, and that picture could either be locally processed in the mobile device or uploaded to a cloud-based server, using the algorithm to interpret the results.
Wang said the team uses the algorithm so everything is "automated and objective" but added that "in many cases the result is very clear … just by looking at the picture you can see that the positive samples give you lots of microbubbles."
She added that the platform has multiplexing capabilities, where different samples could be run together and combined in the cartridge to return results.
The method is internationally patented and the team is looking for partners to commercialize and codevelop or license the POC version of the platform, Wang said. She added that the team "thinks there's a wide variety of disease areas it could be helpful."
Wang's lab also created a startup, Instanosis, in 2019 that is now pitching the POC test to investors and seeking funding to commercialize it. "We believe there can be a lot of commercial potential beyond COVID-19 detection," she said.
The lab received a $250,000 contract from the National Institutes of Health's Rapid Acceleration of Diagnostics program to develop the platform and has pending grant applications with different federal agencies, Wang noted.
She added that there is funding pending for use of the tech for early cancer detection and said that it could be used "anywhere that sensitive biomarker detection is needed," including with cardiovascular and neurology applications, as well as more expansive infectious disease detection.
Paul Eder, a senior scientific officer with the Concept Acceleration Program for Diagnostics at NIH, was at a US government seminar where Wang's lab presented the method and test. He said that the technical approach of the test is different from other technologies in the space.
While he said Wang's technology has promise, he also recommended caution. "Is it robust enough to work away from the research bench?" he asked. "Uniqueness raises risk that there might be some downstream unknown challenges for which they are not yet prepared."
He noted that the analytical sensitivity of the test is "tremendously exciting" for the in vitro diagnostics arena. The sensitivity "propel[s] this technology ahead of most other signal output chemistries for protein detection," he added.
Beyond SARS-CoV-2 detection, Eder mentioned that the test could be used for bacterial infections manifested from released protein toxins. "The ability to detect low levels of protein make this test valuable" for identifying "low but clinically relevant levels" of these toxins, he said.
"If this technology can truly detect protein levels down to a single picogram per milliliter of blood, it can have a strong positive clinical impact on treating toxic infections," he continued.