The Pandemic Set Off a Boom in Diagnostics

A decade ago Willy Ssengooba began crisscrossing Uganda, training health-care workers on how to use a new machine to detect tuberculosis. The deadly lung disease infects around 90,000 people in the East African nation annually, but it can sometimes take months to diagnose using traditional methods such as culturing samples of coughed-up sputum. These new machines used rapid molecular testing to yield results within a couple of hours, meaning patients who tested positive could immediately be referred for lifesaving treatment. Ssengooba, who is scientific director of the mycobacteriology research unit at the Makerere University College of Health Sciences in Kampala, helped to set up 265 of the devices in clinics around the country. By increasing the number of early diagnoses—especially among vulnerable groups such as children and people living with HIV—deaths associated with tuberculosis dropped, too. Ssengooba saw this as a major success and wanted to deploy more machines. But it was hard to make politicians aware of the technology’s power.

Then COVID arrived. Not long after the first case was reported in Uganda on March 21, 2020, Ssengooba received a message from a commissioner under the Ministry of Health. It had quickly become apparent that most of the new cases were coming through the border crossings, so screening people there would be a priority. Could Ssengooba make it possible to test everyone who wanted to enter Uganda? His entire country was counting on him.

Ssengooba and his team began facilitating the collection of nasal swabs taken from truckers at popular entry points, where imported goods are brought into the landlocked country. Those samples—sometimes more than 1,000 a day—then needed to be shuttled 150 miles to Kampala. The capital city was the nearest place with laboratory technology set up to run a process known as polymerase chain reaction—or PCR. Using fluorescent probes that latch onto portions of the coronavirus genetic sequence, these massive machines could determine whether a sample was positive for the genetic material of SARS-CoV-2.

Ssengooba’s team had to shuttle the samples themselves. A crew of about 50 workers collected the samples in pickup trucks and delivered them to the lab, then turned around to go back for the next batch, spending long, exhausting nights on the road. As the pandemic intensified, they couldn’t keep up. Truckers awaiting their test results were stalled at the border for days, in part because the sample analysis in Kampala would sometimes take up to 72 hours to return a result. A queue of trucks formed, stretching for kilometers, holding up the import of everything from home appliances to construction materials to replacement parts for cars. Making matters worse, authorities had closed the airport.

The government was desperate to alleviate the backlog. Ssengooba considered the 265 machines he had set up throughout Uganda over the years to test for tuberculosis. He realized he could repurpose some of those small PCR machines to test for the coronavirus by using a different sample-processing cartridge. He relocated that equipment directly to the border entry points and engineered some basic infrastructure (electrical power; benchtop safety spaces) to support their use. Unlike the lab setup in Kampala, which requires multiple machines spread across different rooms and experienced technicians to prepare and process the samples, these so-called GeneXpert modules were automated and about the size of a printer. They still used PCR technology but could return results on the spot in around half an hour.

By May the first COVID testing systems were working at the crossing point in the Kenya-Uganda border town of Malaba, reducing the waiting time for truck drivers from days to around half an hour. Within a week the equipment was set up at two other major border points. Many countries, even wealthy ones, struggled to get COVID screenings off the ground in the early months of the pandemic. But Ssengooba understood how to balance the needs of public health and the economy. By cobbling together testing infrastructure where it was most needed, he was preventing disease while keeping critical goods flowing into the country.

Ssengooba’s creative repurposing of Uganda’s limited testing resources was “tremendous,” says Wilber Sabiiti, an expert at the University of St. Andrews in Scotland in the development of diagnostics tests. After Ssengooba had worked tirelessly for years to try to make PCR testing more accessible for tuberculosis, his efforts finally seemed to be getting validated. The pandemic, he says, has clarified the urgency of deploying PCR technology more widely for all kinds of infectious diseases, especially to politicians, who are now allocating more money to the technology. “The SARS-CoV-2 outbreak has been like a blessing in disguise for the scale-up of molecular assays,” he says.

Ssengooba is hardly alone in his thinking. Alex Greninger, assistant director of the clinical virology labs at the University of Washington Medical Center, says that in the past, the facility where he works typically ran 50,000 PCR tests a year for diseases such as influenza and HIV. Between early 2020 and the end of 2021, it had done four million PCR tests, mainly for COVID. Unlike in the past, however, the results have been vital to informing immediate behaviors: those who test positive self-isolate or are kept in a special ward of a hospital. This has created a massive new reliance on testing to guide such decisions. “We [did] 81 years’ worth of molecular testing in the virology lab in the first 22 months” of the pandemic, Greninger says. He expects the demand for PCR testing to stick around even after COVID ebbs. The general public is much more aware of virology now, he says, and as rapid antigen tests become a more routine part of living amid COVID and its potential future surges, people will seek out PCR tests to confirm positive results.

COVID has not just increased the scale of disease testing and created demand from pretty much everyone; it has also spurred the adoption of more advanced versions of the tests themselves. Hospital systems have begun purchasing PCR machines that are small enough to install at a doctor’s office so that samples do not have to be sent offsite to a huge, centralized lab. That means patients can get a diagnosis on the spot and isolate immediately, if needed—and take whatever antiviral or antibiotic is most appropriate. Companies and university researchers around the globe who are working on PCR technologies say there is escalating interest in their innovations, such as handheld versions that would make testing anywhere—from supermarket parking lots to remote villages—more feasible.

All of this may benefit more than just individual patients. As the emergence of the quick-spreading Omicron variant demonstrated in late 2021, ramping up testing at the population level is crucial to keep tabs on how COVID might morph and push hospital systems (among other sectors of society) to a breaking point. The sweeping uptake of PCR for COVID could also pave the way for stronger public health surveillance systems that can spot future pandemics by scanning for dozens of pathogens at once. According to Jeffrey Townsend, a biostatistician at the Yale School of Public Health, PCR is a powerful tool to use for disease surveillance, and “there’s a lot of people who say we need to be doing it even more.”

A Strange Trip

In 1983 Kary Mullis was driving up to his cabin on the northern coast of California with his girlfriend, a chemist at the biotechnology company where they had been hired to synthesize genetic fragments. Mullis had spent earlier years completing a Ph.D. at the University of California, Berkeley, where he would trip on LSD while making new chemicals. His girlfriend had fallen asleep, and as he drove he had a vision of molecules dancing on the mountain road. It was then that the idea for polymerase chain reaction came to him. He pulled over the car and scribbled down his thoughts. It won him a Nobel Prize a decade later.

At its heart, PCR is a method of making copies of genetic sequences. There are now many dozens of different kinds of PCR, but the most basic form that Mullis devised started with a tiny bit of DNA and then used various cycles of heating and cooling to replicate it. First, the process would heat the DNA to break its double helix structure into two strands. Next, it would cycle to a cooler temperature that would allow specially tailored primers to bind to specific target sequences within the strands. The samples would be warmed up again, and enzymes would get to work building off those primers to finish replicating the complementary DNA sequences. The cycle would then repeat. Ultimately it yielded a lot of copies of the target strands. Special fluorescent tags were later added to the process to flag the presence of those amplified short sequences of interest.

It became possible to use this method to detect the presence or absence of pathogens: if a virus was present in a person’s blood sample, for example, the PCR machine would make a lot of copies of its sequence, and the fluorescent tags would shine brightly. If there was no virus, there would be only darkness.

Unassuming machines: GeneXpert modules use PCR technology to test for all kinds of infectious diseases, including COVID. Credit: Esther Ruth Mbabazi

The incorporation of fluorescent tags meant that the PCR machines could also indicate how much virus was in a person’s system. If the fluorescent light shined more strongly and sooner in the replication cycling, it meant more virus was present. A PCR could not only detect DNA, it could also detect genetic material known as RNA. This opened up a whole new world of diagnostics because many viruses, such as HIV, are RNA-based organisms. As the AIDS pandemic tore through the globe, doctors wanted to know how much HIV was circulating in their patients’ bodies and whether the antiviral drugs they prescribed were working to keep the levels low. PCR could finally give them an answer.

The machines that did the analyses, though, required lab technicians with highly specialized expertise to prep samples and took half a day or more to return results. That changed after the U.S. Postal Service launched a competition for technology that could quickly screen mail for deadly anthrax spores, which a bioterrorist sent in letters to the offices of U.S. senators and journalists after 9/11. The winner, announced in 2002, was a GeneXpert prototype from Cepheid, a Silicon Valley diagnostics company founded in the late 1990s. The system automated many of the previously laborious sample preparation steps by using cartridges and valves that pull liquids through tiny channels and mix them together. And it returned results in minutes rather than hours. In the decades since, the GeneXpert platform has received approval to test for pathogens such as norovirus, chlamydia, tuberculosis and SARS-CoV-2.

Cepheid says there are now more than 40,000 GeneXpert machines around the world, up from 23,000 in 2020. (The diagnostics branch of the major biomedical company Roche also has a PCR machine for clinics that is about the size of a coffee machine.) Increasingly, they are found at doctors’ offices and at locations such as the border crossings in Uganda—rather than just at centralized labs. In September 2020 Cepheid received FDA authorization for a GeneXpert test that looks simultaneously for influenza A and B, SARS-CoV-2 and a pathogen that is particularly dangerous in young kids called respiratory syncytial virus. The test results, which can come back in about half an hour, help clinicians know what specific antiviral to give if a patient is sick—Tamiflu for influenza, for example, and Paxlovid for COVID. That is all the more crucial during a pandemic when your infection determines your isolation behavior.

Real-Time Warning Systems

It was not until the past decade or so that scientists established global surveillance systems that rapidly tracked outbreaks of viruses. Testing for pathogens fell to individual labs, and molecular diagnostics approaches such as PCR were expensive or unavailable. Furthermore, to do PCR testing for viruses of interest, scientists needed specific probes that would recognize a genetic sequence in the pathogens. But they lacked easy tools to create these probes. The barriers to conducting PCR and the dearth of repositories to upload such data made tracking the ebb and flow of viruses in populations spotty.

In 2012 the California Department of Public Health received several reports of a mysterious poliolike disease striking children. It manifested as a sudden onset of muscle weakness in the limbs, sometimes also leading to slurred speech and difficulty moving the eyes. The sick children did not have poliovirus, and health authorities ruled out other possible culprits, including West Nile virus, stroke and botulism. What the children did have was an obscure virus called enterovirus D68, or EV-D68, which had first been identified decades ago. It had recently been linked with acute flaccid myelitis. Although some children make a full recovery from this condition, it can cause permanent paralysis and even death.

Around the same time that acute flaccid myelitis became associated with EV-D68, BioFire Diagnostics, a Utah-based molecular biology company that is now a subsidiary of the global diagnostics giant bioMérieux, began offering a comprehensive PCR-based respiratory test. It looked for 17 viruses and three bacteria in a single deep nasal swab taken from a patient.

Although the respiratory panel does not test specifically for EV-D68, it tests for the presence of the general family of viruses to which it belongs. BioFire wanted to find a way to catch EV-D68 outbreaks so that doctors and public health officials could know to keep patients from infecting others. Along with its academic partners, the company developed and tested an algorithm that was trained on past data to predict hotspots of EV-D68. The real proof of the approach came in 2018, when the algorithm alerted researchers to the emergence of EV-D68 that summer. Nationwide Children’s Hospital in Columbus, Ohio, was one of the first places the algorithm identified with a possible uptick in cases of the virus; the team there confirmed the algorithm was right. As a result, the hospital implemented EV-D68 testing to catch cases early and prevent it from spreading.

A related surveillance platform that uses BioFire’s PCR test collates data from different sites across the U.S. and other countries around the world on respiratory viruses such as influenza, rhinovirus and now coronavirus, as well as more than a dozen gastrointestinal pathogens. Unlike the cumbersome data-collection protocols of the past, surveillance systems that continuously collect data directly from connected PCR machines have the potential to be used to detect outbreaks, including those of foodborne disease.

In many ways, this approach—combination PCR tests that cast a wider net to look for more possible pathogens in a given sample—signals the future of PCR. “Their instruments are phoning home, which is totally cool,” Greninger says of the BioFire disease-tracking platform, explaining that the broad net it casts could help show where unexpected outbreaks are occurring. The COVID pandemic has made it clear that testing people for viruses even if they are asymptomatic can help identify those who would not otherwise know they are infected, prompting them to isolate before they pass the pathogen unknowingly to others in their community.

Viral evolution can sometimes create a challenge for PCR. Because the primers and probes used in the tests are tailored to look for specific, telltale sequences within a virus, sometimes a new viral variant can evade detection because its sequence has evolved beyond what the test is looking for. Test developers have to constantly ensure the primers and probes are up-to-date. “You need to have a very good understanding of emerging genomes in populations throughout the globe if you’re going to have a globally applicable and accurate diagnostic PCR-based test,” explains Alexandra Valsamakis, head of clinical development and medical affairs at Roche Diagnostics Solutions.

Yet once scientists have identified new viral variants, they can use PCR testing to track the spread of those variants. This is a capability that the antigen-based testing methods—which look for proteins unique to a particular pathogen—cannot do. The emergence of the Omicron variant has shown how vital it is to track variants. The data pouring in from PCR tests revealed that Omicron was spreading like wildfire compared with the Delta variant that preceded it. As a result, some governments began updating their guidelines and pushing for more booster shots, and some people took the data as a cue to reconsider their social interactions and upgrade the efficacy of their masks.

Some experts worry that even if PCR testing capacity expands to make more of this kind of surveillance possible, it will be hampered by insurance companies that might be unwilling to pay for asymptomatic testing or that hesitate to reimburse tests for pathogens for which no drugs or treatment are yet available. In most cases, insurance companies “pay for vaccines and diagnostics based on individual benefit,” says Dan Wattendorf, the Innovative Technology Solutions team director at the Bill & Melinda Gates Foundation. “But we don’t really have payment schemes or reimbursement and coverage to find transmission in the community.” The problem with coverage for PCR testing has already been a sticking point in the coronavirus pandemic. The U.S. government set up requirements for health insurers to cover PCR testing for COVID, but consumers both with and without coverage have still been left with surprise bills in the thousands of dollars. Whereas PCR technology itself is undoubtedly powerful for disease surveillance, the question of who will foot the bill remains largely unanswered.

How Covid Is Shaping the Future of Diagnostics

As COVID created demand for more PCR testing everywhere, it also exposed how most of the technology relies on costly enzymes and single-use plastic parts for sample processing. After successfully setting up the fast-turnaround GeneXpert machines on Uganda’s border in the spring of 2020, Ssengooba soon ran out of the cartridges and reagents the machines rely on. In those early months of the pandemic, Uganda requested 500,000 such cartridges from Cepheid, but Ssengooba says the company sent only 30,000. The test maker, he recalls, said that it was barred from sending more cartridges out of the U.S. “We spent the rest of 2020 without access to additional cartridges,” Ssengooba says.

Modern PCR machines use plastic trays that traditionally have each contained 96 or 384 small wells to hold samples. To circumvent the need for expensive plastic “consumables” such as tubes and caps, U.K.-based company LGC replaces the tray with a long, flexible polymer tape. Only 0.3 millimeter thick, it can stretch up to 40 meters and has room for 106,368 reaction wells. “That allows you to do 100,000 to 150,000 tests per machine per day, which is 10 times more than any machine in the world at 10 times less cost,” Wattendorf says, adding that the Gates foundation has partnered with LGC and Northwell Health, the largest health system in New York State, to try the tape-based method for COVID testing.

Another bottleneck with PCR is that “you have to get the sample very, very purified” before running the test, says biomedical engineer Nicholas Adams. PCR machines are calibrated to run reactions at specific temperatures, and impurities such as salts and proteins from patient samples and added preservatives can throw that off. Removing impurities is tough. To avoid that step, Adams and Frederick Haselton, both at Vanderbilt University, had the idea of adding DNA that is a mirror version of the target genetic sequence the PCR test is trying to detect. These mirror sequences are “left-handed”—meaning that they twist in the opposite direction of naturally occurring DNA, which is right-handed—so they do not interfere with the detection process. By adding a specific amount of the left-handed DNA and tracking how much of it is copied, Adams can use it as a benchmark to calibrate and confirm that the PCR machine is running without worrying about many impurities. Adams says that by reducing the need for purification with the left-handed DNA—which costs about 11 cents per test—labs could save significant labor and material costs.

Now that COVID has shown how important it is for testing to be accessible, there is more enthusiasm for portable PCR devices. Avleo Technologies has designed a handheld molecular testing machine that gives results in 30 minutes. Another device, from Visby Medical, was initially developed to look for sexually transmitted infections such as chlamydia and gonorrhea (and received FDA clearance for those applications) and has since added testing for SARS-CoV-2. Anavasi Diagnostics’ AscencioDx platform, originally developed to detect HIV and flu before the pandemic hit, is being used in trials as a rapid molecular COVID test. In November 2021 the National Institutes of Health awarded $14.9 million to Anavasi to support that initiative.

Diagnostics developers are continuing to tinker with PCR. German engineering company Solarkiosk Solutions is developing a version that runs on solar power, which it is piloting for COVID testing in a remote part of Sumatra where many residents lack access to electricity and diagnostics. Academic labs and start-ups such as Mammoth Biosciences in San Francisco are combining traditional PCR methods with CRISPR gene-editing technology to make the tests more efficient at detecting specific pathogen genes.

At Uganda’s border crossings, Ssengooba says that testing, at this moment, anyway, is “very smooth.” But nearly 40 years after the idea of PCR was born, the technology is evolving rapidly as a result of the pandemic, and Ssengooba is dreaming big. He is eager to try the handheld disease diagnostics because traditional PCR—including the printer-size machines at the border—still require hookups to the electricity grid and various sample-processing rooms. A portable version, akin to the one in development by Indian company Molbio, could bypass some of these requirements and open up fast-testing access to remote areas for the first time. “This is something that is incredible,” he says.

Public health has always been stymied by the hours or days between collecting a sample and delivering the results to the patient; in the meantime, an infected person has left the clinic and gone back to the routines of their life, unwittingly exposing others and delaying treatments that are often more effective if started earlier. COVID, and the stunning transmissibility of Omicron in particular, has laid bare the consequences of that gap—for individual health, community transmission, overburdened hospitals, labor shortages, and so much more. Ssengooba is hopeful that the urgency for closing that gap will persist. When imagining a future where portable PCR tests with on-site results are commonplace, “all of these challenges,” he says, “are going to be left behind.”