Put to the test

A squinting eye and an urge to gag have been two unpleasant experiences shared by many during the pandemic, as the testing process for Covid-19 PCR test requires a vigorous swabbing of the tonsils, followed by lodging the same swab so far up your nose it feels like the brain itself is getting a prodding.

Most people accept that it’s necessary, of course – a little short-term discomfort endured to prevent yourself from becoming a spreader. In some countries, like the UK, there’s likely to be a lateral flow test prior to the confirmatory PCR, but it’s the latter that is considered highly accurate by virologists. Although the breadth of studies on PCR suggest its exact sensitivity is hard to pin down, a recent study entitled ‘Screening for SARS-CoV-2 by RT-PCR: Saliva or nasopharyngeal swab? Rapid review and meta-analysis’ looked at 48 studies and found an overall sensitivity of 92% for nasopharyngeal samples and 86% for saliva samples. But despite the advantage of high sensitivity, which translates into a low rate of false negatives, PCR tests could never power the mass testing strategies many governments have used to try and control the spread of Covid-19.

An RT-PCR test is the most common form of PCR used to diagnose Covid-19. It only takes between two and three hours of thermocycling to get a result, but that’s not the whole story. There’s also the time it takes to obtain a nasopharyngeal sample, transport it to the nearest available laboratory and prepare the primers and reagents necessary to conduct the test. Add that all together and it’s not unusual to wait two days or more for a Covid test result.

Of course, vaccines supplanted mass testing as the new ‘plan A’ at the start of this year, but with several mutations already increasing transmissibility, there’s still a convincing case for accurate point-of-care tests that can provide a result in hours, not days.

PCR in a box

Dr Chris Toumazou was an early innovator looking to cut turnaround times and the reliance on laboratories – but his solution was more, not less, PCR. The difference is his PCR test doesn’t take place in a lab, but in a small box about the size of a toaster, in which microfluidic cartridges automatically perform all of the necessary preparation and amplification steps. The device returns results in 90 minutes.

As suitable as it might seem, the process wasn’t developed to test for viruses. Instead, it was designed for Toumazou’s consumer business, DnaNudge, which uses PCR to test for genetic variables that contribute to the risk of certain diseases, like obesity and diabetes.

“We had a solution before the coronavirus came along,” he says. “The box was designed for PCR thermocycling. It could be a lot smaller [...] but if you want to take what’s done in a laboratory out of the lab, you have to do exactly what’s done in a lab, to the same standard and with the same accuracy.”

The UK government was convinced, buying 5,000 NudgeBox units and the additional equipment needed to run 5.8 million tests back in August 2020 for a total of £161m. But DnaNudge had to enlist the capabilities of contract manufacturer Jabil Healthcare to meet the increased demand for its product, and with a manufacturing rate of only 1.2 million cartridges per month, it was clear DnaNudge couldn’t be the moonshot solution either. Instead, the boxes and cartridges, which were sold under the product name CovidNudge, found their homes on UK hospital wards, where the morbidity and mortality risk of Covid spreading was much higher than in the wider world. Toumazou says these wards, which typically provide maternity, elective surgery or mental health care, are still using CovidNudge, but the number of orders from the UK government has decreased with the positive impact the vaccines are having.

Mass testing

For mass testing, the UK government opted for lateral flow tests – a cheap to produce assay that uses the same technique found in pregnancy tests to detect SARS-CoV-2. The chorionic gonadotropin (hCG) hormone detected via pregnancy tests tends to be easy to pick up due to the high volume of it present in the urine of pregnant women. However, the viral load present in a Covid swab can vary significantly, and without an amplification method, such tests are always going to be less sensitive to lower viral loads than RT-PCR.

“Departments like the MHRA [Medicines and Healthcare products Regulatory Agency] would never usually consider antigen tests like lateral flow or even LAMP [loop mediated isothermal amplification] tests for Covid because they were so inaccurate,” says Toumazou. “But then we got to a point where beggars couldn’t be choosers, so we had to use these technologies.”

The UK government bought into LAMP technology at the same time it purchased CovidNudge. Specifically, it bought the equipment needed to perform 450,000 LamPORE tests from University of Oxford spin-off company Oxford Nanopore Technologies (ONT). Certain figures suggest this assay, which combines reverse transcription loop mediated isothermal amplification (RT-LAMP) with a DNA sequencing method (nanopore sequencing) developed by ONT, has a sensitivity and specificity equal to that of PCR – but Toumazou says the technology also has disadvantages that limit its use in the current pandemic. “The sequencing takes place in a lab, and there’s no multiplex that can sequence out of the lab and in the community,” he explains. “LAMP works at a fixed temperature, so unless your assays or primers amplify at that fixed temperature, then you can’t put many of them on the chip, whereas thermocycling can amplify many different assays.”

CRISPR

Multiplex analysis refers to the process of mixing several different primers and reagents in the same reaction chamber in order to test for a range of viruses within a single sample. The CovidNudge device provides this functionality with 72 microwells, all of which can be spotted with different samples and reagents. The benefits of doing this include differential diagnosis of symptoms that could be caused by SARS-CoV-2, influenza or respiratory syncytial virus (RSV). Importantly for the current situation, they can also screen for variants of the same virus.

Similarly, the minimally instrumented specific high-sensitivity enzymatic reporter unlocking (miSHERLOCK) test, developed by Harvard Wyss Institute and Massachusetts Institute of Technology scientists, allows users to carry out three assays at once. The product, which is based on previous work from one of the latter institution’s professors, James J Collins, is referred to by its creators as a “one-pot reaction”. This is because the recombinase polymerase amplification (RPA) step that takes place in lieu of PCR, along with the detection step performed using the CRISPR Cas12a enzyme, happen at the same temperature. A colour change prompted by an embedded fluorescence chemical indicates a positive or negative result an hour after sample collection. The researchers say the test had a sensitivity of 96% and a specificity of 95% in testing, making it highly comparable with PCR.

The original test devised by Collins and now sold through his company Sherlock Biosciences, uses the traditional brain-prodding approach to obtain a nasopharyngeal swab. However, the team behind miSHERLOCK added a step to disable enzymes called salivary nucleases, which destroy nucleic acids, allowing it to detect the virus in spit alone. “We wanted as few user steps as possible to make sure there wouldn’t be any sources of contamination,” says Devora Najjar, a research assistant in the Collins Lab at MIT and one of the lead authors on the paper behind miSHERLOCK. “Even the addition of somebody opening it up and dropping a dipstick in is something we wanted to avoid.”

This and other aspects of the test were designed with resource-limited locations in mind, with all components coming to just $15. But, as fellow lead author Xiao Tan, clinical fellow in the Collins lab and instructor in gastroenterology at Massachusetts General Hospital, points out, the pandemic has proven how easily the simple things we rely on for testing can become scarce when demand rises quickly. “Our aim was to make a wholly self-contained device that can be used in low-resource settings, but it’s very apparent that resource-rich settings can also become resource-limited settings very quickly,” he says. “This was very clearly illustrated by the fact that we couldn’t get simple swabs or viral transport media, things that are so basic they would never have been thought to be bottlenecks for testing.”

Decentralised accuracy

The miSHERLOCK and CovidNudge testing platforms are united by a common goal – to decentralise testing. In hospitals, CovidNudge does this by obtaining a result directly from a sample and storing the data in the cloud. From here, the information can be accessed by the hospital and sent to relevant government agencies to help establish prevalence statistics. On the other hand, miSHERLOCK comes paired with a smartphone app that uses the device’s camera to indicate a positive or negative test from the colour of the liquid in the test tube. In this case, the application itself transmits the data to relevant external bodies.

One of the motivations behind miSHERLOCK is that, by decentralising the data-gathering process, it can help governments and health services direct vaccines and boosters to the areas with the highest rates of transmission. The device is currently being pitched as a cheaper and more accurate home testing kit than the alternatives on the market in the US right now, though it is yet to receive FDA approval.

Rose Lee, visiting scholar at the Collins Lab and paediatric infectious disease specialist at Boston Children’s Hospital, as well as a lead author on the study, believes miSHERLOCK could also play a valuable role in hospitals due to its ability to detect Covid variants. “We know some therapies, like the monoclonal antibodies we give to patients, can be affected by the variant,” she explains. In fact, the UK and US paused the use of two monoclonal antibodies – bamlanivimab and etesevimab – in June this year after it was discovered that the P.1/Gamma variant (first identified in Brazil) and the B.1.351/Beta variant (first identified in South Africa) rendered the treatment ineffective. As the virus keeps mutating, Lee notes that “clinicians could use [variant] data not just for broader epidemiology and surveillance purposes… but also to help them manage each patient”.

Despite the high sensitivity of its own device, the team behind miSHERLOCK are quick to emphasise that less sensitive assays like lateral flow tests still have a role to play in ending the pandemic. “In a perfect world, our testing would be very quick, sensitive and specific – and able to be shared everywhere,” says Najjar. “But there’s a realisation, especially by people in the research and clinical fields, that we don’t live in that world yet. Instead, we have to make testing work with the tools that we have.”

Toumazou looks at it from the other direction. Antigen-based diagnostics are important for frequent at-home testing, but he believes there’s no replacement for the accuracy given by RT-PCR. “Democratising PCR is what was missing from the moonshot,” he says. “There’s a category that CovidNudge and other technologies fall into that cover that middle slot of being out of the lab, but not in the home. We can bring that to the mass with current technologies for the next wave or even the next pandemic.”