The Best may not have been Better: PCR versus Antigen Covid Testing
If you have been among the millions who had your brain tickled by a very long swab, your swab subsequently had the gold standard of testing, PCR, aka Polymerase Chain Reaction Covid testing. We’ve assumed that since PCR was the best performing test, that it was the best test to contain the Covid epidemic, maybe wrongly. I want to share with you why the “second best,” antigen testing may have been the better choice.
Earlier this week, the New York Department of Health kicked off the discussion with their answer to, “What is the difference between a SARS-CoV-2 antigen test and a SARS-CoV-2 PCR test?”
The main difference is that the antigen test does not have PCR’s amplification system which increases the signal to background noise ratio. Without this complex and demanding laboratory step, Covid antigen tests get results back faster (typically 15 minutes actual testing time), but by definition are less sensitive than PCR tests.
Covid antigen test instruments are about the size of tissue boxes and are called Point of Testing instruments, because they can render results before the patient leaves, almost crude compared with highly-complex PCR lab testing platforms. But how does the slightly decreased testing ability to pick up Covid stack up against the PCR gold standard in real life? The answer might surprise you and reveals an oversight by national testing gurus early in the outbreak.
Let me share the New York Department of Health’s patently correct answer:
“SARS-CoV-2 antigen tests detect a part of the virus called viral proteins, which make up the virus’s structure. SARS-CoV-2 PCR tests detect a different part of the virus called viral RNA (nucleic acid), which is the virus’s genetic material.”
Translated, the tests test different parts of the virus. But there as we already started discussing, there are other aspects. Before I opine on their answer, let me speak as a pathologist, a specialist in the nitty-gritty aspect of diseases. We often think of infections as binary, (you either have it or you don’t) and the viruses as perfect little gremlins, (they attack and cause disease upon contact). Neither of these statements is true.
Viral infections traditionally demonstrate a phenomenon called “infectious dosage.” Below a certain number of viruses, whether they are HIV or Covid, an infection cannot take hold. That threshold has several known factors (immunocompromised status, existing illnesses, fitness, etc.) that lower that susceptibility threshold. To that, we add somewhat unknowable and unquantifiable factors.
Among the latter include clearance from the nasal cavity. The nose and sinus cavities are the unsung heroes of our immune defense systems. Unlike antibodies which only work after being honed to defend the body, the “drip, drip” runny nose is not just an annoyance, but a misjudged first-line defender. Within that runny snot are a toxic cocktail of enzymes, (lysozymes which breakdown things like a disinfecting cleaner), antibodies (IgA type which latch onto most everything), mucus (the sludge-like mud which captures and whisks the offending stuff out the front door. No wonder why our nose runs more when the temperature drops or when pollen spews forth from trees.
The eddies and slipstreams of air current making their way through the sinus passageways with each inhalation deposit more than 80% of their particles into this rapacious embankment of a defense system. Where the virus able to penetrate into the bedrock of this stream, to the cells lining the sinus passageways, it has to deal with waiting T-cells. Specialist T-cells attack viruses days before over-heralded antibodies make their appearance.
So if the number of Covid viruses is sufficiently few, it is entirely possible to capture and contain an infection within the nasal sinuses, which brings us back to the antigen versus PCR test question.
By now, you have heard of the conundrum of antigen-positive and PCR-negative test results. In theory, the PCR test should become positive before the antigen test turns positive. In scientific parlance, this is called the “limit of detection.” This window is on the order of less than a day, measured in hours. In other words, an antigen test should turn positive a few hours later than a PCR test. I wish to come back to this in a few paragraphs when I assume the role of Monday Morning Quarterback’ing, the national testing strategy.
We should not be seeing antigen-positive and PCR-negative test results, if this is our framework of understanding. However, many patients, healthcare professionals, and even doctors hold incorrect views of viruses.
The Covid virus is not the perfect magic broom of Disney’s “The Sorcerer’s Apprentice.” If the assaulting party is not overwhelming in number, they will get trapped within the nose, not establish a beachhead and get swept out to the waiting Kleenex. Therefore, if you received a relatively small virus dose from a passing positive Covid contact, then those first-line nose-sinus defenders will make short work of the invaders. It is entirely possible that you will be antigen positive from a nasal swab, but negative on PCR.
Take a look at the diagram above. The antigen test loses much of the sensitivity on the backend of the infection, where the tail of viral shedding is prolonged. The PCR excels here. Or does it?
Toward the end of a Covid infection, we have been seeing patients whom we cannot grow virus, even out of a luxuriant viral culture medium, even though the PCR test is positive. What I think is happening at this point in time is the body is interrupting the ability of the virus to properly assemble viruses.
That’s ‘Take Home Point C,’ from our same diagram, which now shows vertical interval bars. The PCR test gives us test positive results, even when patients are past their infectious period.
Instead of the nicely formed Covid viruses you see adorning the pages of scary articles, if we were to see them on special microscopes would be misshaped viruses. To use Micky Mouse’s “The Sorcerer’s Apprentice’s “broom example, you would be looking at a dowel of wood, some straw bundles haphazardly bundled by twisted wire.
Which brings me to playing Monday’s (Morning? Afternoon?) Quarterback. When I learned lab testing principles, we were taught about test specificity, sensitivity, positive and negative predictive values as measures of test performance. Covid has taught us that more important than traditional medical school test performance measures are two factors. The first is
Test Positive: meaning retaining the ability to be infectious.
PCR while more sensitive, is overly positive. People who are in the non-infectious late phase are positive with the PCR test. The antigen test loses points on paper, but maybe not performance when it comes to what we really want to know, “Who is infectious right now?”. A PCR test positive overestimates who is infectious positive.
The reason is related to the second point which was not apparent to me during residency and only later this year I have learned:
Test Turnaround Time can be more important than test sensitivity or even test volume.
What is the utility of a test which you cannot act upon for several days? What is the use of not being able to identify and inform people who are spreading the virus if it takes several days to get in touch with them after they get swabbed? The mainstream media’s message of more testing fell short because we had virtually non-actionable results when the PCR’s Test Turnaround Time exceeded 48 hours.
The antigen test trades off sensitivity with shorter Test Turnaround Time and cost. That’s the touted difference in the bars shown in ‘Interval A,’ A few hours, perhaps morning versus afternoon testing on the same day when you really desired higher test sensitivity — a few hours of sensitivity is what an antigen test gives up.
We could test more with the same resources because antigen tests are cheaper and can be rendered as patients wait. As a nation, we can afford more of the cheaper antigen tests, thereby identifying more infectious patients, if we can get results with more frequent, faster, actionable Covid results. Not much of a loss, especially if you can catch that superspreader. Most infections are spread by just a few individuals. Most Covid people do not spread their viruses around.
That superspreader might be more willing to come in for testing if they didn’t have to get an eye-watering nasopharyngeal, long-swab poke. The antigen test uses a shorter swab which goes up the front of the nostril, which you can do yourself.
Finally, go up one last time and look at ‘Interval B.’ Notice that the antigen test does a better job at defining when someone is infectious compared to the PCR test. There is no ‘best test.’ There only is the best test for the answer you desire.
The wisdom which our leaders and all healthcare professionals needed to have learned much earlier than I have is the adage: “Don’t let the perfect be the enemy of the good.” PCR tests are the most accurate, but the antigen test might have been better.
References:
SARS-CoV-2 Point of Care (POC) Antigen Tests, Frequently Asked Questions for Health Care Providers, October 16, 2020, NYSDOH Bulletin
Test sensitivity is secondary to frequency and turnaround time for COVID-19 surveillance, https://doi.org/10.1101/2020.06.22.20136309.t
