Skip to main content
Premium Trial:

Request an Annual Quote

Stanford Team Developing Low-Cost, At-Home, Molecular COVID-19 Saliva Test


NEW YORK – Researchers at Stanford University have been developing a low-cost, molecular COVID-19 saliva test that they plan to commercialize for use at home as well as in low-resource settings.

The test, called SnapDx, could help enable the kind of mass screening advocated by a number of organizations to curb the spread of the SARS-CoV-2 virus, allowing individuals with your without symptoms to test themselves at home. However, some experts say that the path to authorization for at-home tests is thorny, and that making it sufficiently robust and easy to use is not a trivial task.

SnapDx uses reverse-transcription loop-mediated isothermal amplification (RT-LAMP) with colorimetric readout and requires no lab instrumentation. It comes in two versions that use the same cartridge: one that only needs a cup of boiling water and another that relies on a $10 heating device. The inventors are hoping to develop the technology not only for SARS-CoV-2 testing but also for other infectious disease tests, such as tuberculosis or sexually transmitted diseases.

Stanford's Catalyst for Collaborative Solutions program has been funding the pilot phase of the development project, which is headed by Manu Prakash, an associate professor of bioengineering at Stanford.

Prakash said his goal was to develop a molecular COVID-19 at-home test that costs $5 or less. His lab has been focusing on so-called "frugal science" for about a decade. "The idea for us is to really make technology affordable," he said. The challenge, he added, is to design a molecular test that can amplify specific genetic targets but at the same time "is as easy as making a cup of coffee."

The researchers are in the midst of running usability and clinical validation studies for both versions of their test that will involve a total of about 1,000 individuals. Those studies are expected to wrap up by the end of February, after which they plan to publish the results and assay design as a preprint.

"It's an iterative cycle, so we've been able to take feedback from the first batch of users, improve the test, and implement this in a manner that is allowing us to build a real-world test," Prakash said. His team is also planning a parallel study for TB testing that will commence soon.

The studies are taking place at Stanford's COVID-19 testing sites, both on campus and in the surrounding area, such as the Central Valley, and include "a very broad range of society," said Euan Ashley, a professor of medicine and genetics at Stanford and an advisor to the SnapDx project.

The developers are already in discussions with the US Food and Drug Administration and plan to apply for Emergency Use Authorization for both versions of the COVID-19 test. Within the coming months, they also want to form a company to commercialize the platform and are in talks with potential investors. Moreover, they have detailed plans for scaling up production with an industrial partner, Bic, said Ashley, adding that even if demand for COVID-19 testing eventually subsides, there will be other applications. "We imagine COVID testing is here to stay but the vision for the technology is much broader," he said.

If authorized, SnapDx would likely compete with several other COVID-19 tests that have recently obtained EUA for at-home use, including an RT-LAMP test from Lucira Health and antigen tests from Ellume and Abbott. Ellume recently won almost $232 million from the US Department of Defense to scale up production of its test.

"I think this is great to see, since testing at home is the safest way to catch viruses and prevent others from getting sick," said Chris Mason, an associate professor at Weill Cornell Medicine who has helped set up several COVID-19 testing labs and is intimately familiar with LAMP-based testing. "I can personally attest to running daily at-home saliva testing with RT-LAMP, using just a few pipettes, a kettle, and the [New England Biolabs] kit, and we have already been able to catch a family member the very first day she was shedding the virus, and we kept everyone else safe," he said.

Commenting on printed instructions for SnapDx posted on the researchers' website last week, which were replaced by shorter instructions this week, he said it seems to be similar in principle to the Lucira Health test but appears to involve more steps. "The FDA will surely look to ensure that usability is optimal for home use, including across a range of ages and educational backgrounds," he said.

Peter Yager, a professor of bioengineering at the University of Washington who said he greatly respects Prakash's work on technologies for low-resource settings, had similar reservations based on the printed instructions posted as of last week. "[The test] seems to have sacrificed user friendliness on the altar of low cost," he said. Specifically, it involves too many time-critical steps that need to be carried out with great precision. "This might work in some low-resource laboratories in the developing world, but I cannot imagine it being used in homes in the US, for example," he added. Yager also noted that the colorimetric readout may not work for colorblind individuals, and that the recommended time window for reading out the result is short.

Prakash said that based on feedback from the ongoing usability study, a new version of the assay instructions was recently released that has reduced the number of main steps to seven, though the overall assay essentially remains the same. Also, a smartphone app is now guiding users through each step of the process. This app could also be used to report test results to public health authorities.

At the heart of SnapDx lies a syringe-like cartridge that consists of a fat plastic tube with caps at both ends and a plunger filled with reagents that runs through its center. The researchers have filed three patents on different aspects of the device, Prakash said, adding that "there are several things that are happening behind the scenes that have never been tried in a diagnostics context."

He stressed that the cartridge does not require a device to perform the test. "This is the device, and this is the test itself," he said. "Our primary source of energy to run the test is just hot water."

That hot-water-version of SnapDx would be ideal for global health applications in countries where power might be scarce. The other, more automated version uses the same cartridge but puts it into a low-cost heating device, one design of which would fit several cartridges for group testing.

According to the instructions, users first attach a funnel to the tube and collect their saliva sample. They then partially immerse the tube in a cup of boiling water for three minutes to kill any active virus.

Following five minutes of cooling, they push the plunger through the tube up to a certain mark and then bend the plunger until they hear a glass vial snap inside, which releases the reagents into the tube. After pushing the plunger to the end and tapping the tube a few times for mixing, they fully immerse the tube into the cup of now less-than-boiling water and leave it there for 30 minutes for the reaction to proceed.

After that, the results can be read out by looking at the color of the plunger near the cap: if it's purple or orange, no virus is present, but if it's yellow, the test has detected SARS-CoV-2. However, results read more than five minutes after taking the tube out of the water are considered unreliable. The instructions emphasize that SnapDx is currently a screening test, not a diagnostic, and that users with a positive result should follow up with their healthcare provider.

Shannon Clark, CEO of UserWise, a consultancy that helps companies perform human factors and usability testing of medical devices and submit them to the FDA, said that "there is a lot of excitement around home use and molecular solutions [for COVID-19 testing], though the pathway is extensive." She said that FDA has elevated its requirements for home-use tests and now asks developers to have at last 100 untrained individuals use the test, who are not required to look at the instructions. "In that context, having a lot of different complex steps really is a deal breaker and will preclude a company from having success passing the rigorous human factors testing required by the FDA," she said.

It is also important how error-tolerant the different steps are. "If you can get away with making mistakes every step of the way and still get an accurate result, then obviously, it would still be a great product," she said.

Having a user be instructed by an app is a good intermediary solution between full telemedicine oversight and having to follow written instructions, she said, but can cause problems of its own. "In our studies with elderly individuals, we find that there are some pretty critical issues with those individuals being able to use a smartphone," she said. "But it's certainly a good direction to take."

Prakash's lab has so far manufactured 1,500 SnapDx kits and plans to make 10,000 in house eventually. Cartridge production will be easy to scale up to millions, he said, as it relies on common technologies that are used, for example, in ballpoint pen production. "Every aspect of the design of this has really been around industrial manufacturing," he said.

Manufacturing costs are hard to pin down exactly but Prakash said they are around $1 per cartridge at the moment. The target cost for SnapDx, including the LAMP chemistry, is less than $5 per test. The assay uses LAMP reaction reagents from New England Biolabs but the researchers have optimized some of the primers. The heating device used in the automated version would add another $10 or so to costs, he said, but that device would be reusable many times.

While the SnapDx test and the Lucira Health test, which already has EUA, both use RT-LAMP, there are important differences between them. For example, Lucira's test requires a prescription from a healthcare provider, which the Stanford team wants to avoid, and the company has said it will cost around $50. Also, each Lucira test uses a disposable battery-operated device, which not only drives up costs but also potentially creates a lot of electronic waste. "What we need is to be able to have over-the-counter testing, where you can buy 10 of these at once and take them home and do them at home before you go to work," Ashley said.

The two tests also use different sample types – the Lucira test a nasal swab and the SnapDx test a saliva sample. Lucira's test appears to be easy to use, with only a single timed step that is measured by the device. Since no performance data for SnapDx have been published yet, the sensitivity and accuracy of the two tests cannot be compared at the moment.

The two COVID-19 antigen tests authorized for home use – both lateral flow immunoassays – are fast and inexpensive. While the Abbott BinaxNow test requires a prescription and oversight from a telehealth company, the Ellume test was the first over-the-counter COVID test to obtain EUA.

However, rapid antigen tests generally have lower sensitivities than molecular tests, and sometimes deliver false-positive results. "I think the antigen tests have their role, but what struck us about Manu's invention here, and why I was excited by it personally, is, this is clinical grade testing, essentially," Ashley said.

Home-use COVID-19 tests like SnapDx might give the at-home diagnostic testing field in general another nudge, he said, just like the pandemic has pushed telehealth forward. "We've seen clearly with respiratory viruses, and this particular one, that what we don't want is bring people who are sick into the community for testing," he said. "What we want for them is to stay at home until they're no longer infectious, so it is really important for respiratory viruses to have home-based testing."