The testing that can help us put the coronavirus genie back in the bottle

The genie is out of the coronavirus bottle and has officially arrived in Minnesota.

Although this particular genie is not easily seen based on clinical symptoms, fortunately, science has provided us with genie glasses. If you believe in that kind of thing … you know … science.

From a testing standpoint, viruses have always been hard to detect. They're difficult to grow in a culture, like we often do with bacteria, and checking for antibodies against a specific virus is often imprecise, particularly early in an infection, when the body hasn't had enough time to produce antibodies.

Then came reverse transcription-polymerase chain reaction (RT-PCR) testing, a lab technique that allows us to identify fragments of genetic material (RNA in the case of coronavirus) in a body fluid sample. Like giving a search-and-rescue dog a sniff of the victim's clothing, a specific RT-PCR test has to be developed for each specific viral or bacteria: It has to be told what it's supposed to be looking for. (Visit tinyurl.com/molecular-assays for more information.)

RT-PCR testing isn't exactly new, but like so many other things in modern life, technological advances have made it much more accessible and affordable. In scientific terms, it's a damn good test. It rarely misses a virus that's there, and it rarely mistakes another virus (or something else) for the virus it was seeking. It won't mistake influenza or strep throat for coronavirus. RT-PCR testing isn't perfect, but it's the best we've ever had.

Ideally, we would have made a lot of RT-PCR genie glasses before the genie actually arrived on our shores. The delay wasn't because we couldn't figure out the gene sequence of this new coronavirus. In a tribute to the stunning sophistication of modern genetics, scientists in China released the virus's mug shot — its entire genome sequence — on Jan. 10, a month after they became aware of the infection (or so we are told, China being China).

An aside here for freaky cool genome science: Because RNA viruses mutate at a somewhat predictable rate, scientists can use mutations to estimate the age of a virus, like counting growth rings on a tree stump. A young virus will have few if any extra mutations. Scientists believe the COVID-19 virus was "born" no earlier than Oct. 30, 2019, and no later than Nov. 29th (tinyurl.com/coronvirus-genome).

This genome map provides the template for developing both vaccines and RT-PCR testing. When Germany flew 126 of its citizens home from Hubei province on Feb. 1, they PCR-tested each of them. Two of the 126 tested positive, and they were not among the 11 people with symptoms.

In medicine, we are trained to avoid testing in situations where the likelihood of finding anything is low. Routine testing for rare or low-likelihood diseases, or in healthy people, just wastes a lot of money and generates more testing for those inevitable cases where the test returns falsely positive. It rarely saves lives. Men do get breast cancer, but not at a rate that justifies routine mammograms.

All that is different with the COVID-19 virus.

At the beginning of the outbreak, when the virus was thought to be isolated to a single province in China, travel history screening was an incredibly cheap and powerful "test." But as the virus continues on its worldwide tour, travel history will become an increasingly nondiscriminating test (example: "Have you or a close contact traveled to any one of the following 30-and-counting states?"). And thanks to widespread RT-PCR testing in other countries, we now know that our other test — signs and symptoms of an upper respiratory infection — has been a crude one from the beginning. It appears that a majority of those infected have either no symptoms or minor symptoms — and yet remain capable of passing the virus to others. Never mind that we are still in the midst of our winter cold and flu season, where COVID-19 impostors abound.

So now that we know travel history and signs/symptoms are no more than beer-goggle-quality tests, it's time to break out the genie glasses: RT-PCR. If the airline industry alone stands to lose $100 billion, it could offer the Centers for Disease Control and state and private labs $20 billion to rev up RT-PCR production, and still save itself $80 billion.

I asked Joanne Bartkus, director of the Minnesota Department of Health's Public Health Laboratory, exactly how does a RT-PCR lab "rev up" capacity? Is it more machines, more reagents (substances used in chemical reactions), more staff?

"Yes to all of those things," she replied as she went on to explain the COVID-19 virus testing process.

Somewhat surprisingly, the RT-PCR test itself is the simplest part of the process: The samples are placed in a machine called a thermal cycler, and in 2.5 fully automated hours, the data is ready to be interpreted. The rate-limiting, most-tedious part of the process is getting the specimens ready for the cycler.

As test kits arrive from around the state, each sample must be "accessioned": Information such as patient identifiers, location and ordering physician must be logged into the computer, and the sample must be tagged with a bar code. It might sound mundane, but imagine the implications of attributing a particular test result to the wrong patient. This is serious stuff.

Then comes the process of extracting the viral RNA out of the immunological fog of war found in any snot sample. In the beginning, the RNA extraction process certified by the CDC was labor-intensive.

"It really limited the number of samples we could turn around in a day, and that's one of the reasons that initial testing was limited to individuals with known risk factors," Barkus told me. "But as we move toward community surveillance, we'll need higher throughput extraction methods," and who could ever argue with "higher-throughput extraction methods?" Good news: On the day I spoke with Barkus, the CDC had just greenlighted a technical change that will allow 96 tests to be done in a platform that previously handled 24.

So things are heating up at MDH labs, and elsewhere, too, as the Mayo Clinic and private labs like Quest Diagnostics and LabCorp have all recently rolled out their own RT-PCR testing. Under normal circumstances, every new RT-PCR would need to be tested and approved by the FDA, but under emergency conditions like COVID-19, the FDA can loosen the reins and allow labs to internally validate their new test and get the FDA "seal of approval" later.

All of that, and the prospect of broader testing, is welcome news. Then we'll have some real data to guide our containment strategy, rather than hunches and case reports, or the difficult, labor-intensive, gumshoe detective work by public health workers.

And then we'll have a much better chance of putting this genie back in the bottle.

Craig Bowron is a physician and writer in the Twin Cities. On Twitter: @billcarlosbills.