After weeks of inexplicably low PCR testing capacity in Wuhan, new reports out of China are showing that the country’s PCR testing effort has been yielding false negative rates as high as 70%, both in Wuhan and more developed parts of China whose health systems are under less stress.
It’s hard to know what exactly is happening, but it’s clear that something is seriously wrong with the Chinese PCR testing campaign for 2019-nCoV. This really hurts the effort to contain and treat 2019-nCoV, and if any other country or group can step in to help, they should.
How PCR works, and why it should be high capacity and reliable
PCR stands for Polymerase Chain Reaction. Invented in 1985 and the topic of the 1993 Nobel Prize in Chemistry, it has become a ubiquitous method for detecting, quantifying, and amplifying nucleic acids in a sequence-specific manner.
It works by exploiting DNA polymerases and the sequence complementarity between a template and primer. A DNA sample containing (or not containing) a small amount of the template sequence is put into a reaction vessel containing DNA polymerases, DNA nucleotides, and short DNA primers which are uniquely complementary to sequences at the opposite ends of the template sequence.
The reaction vessel, most often a tiny tube with only a few microliters of fluid, is thermocycled between three temperatures, one which allows the primers to anneal to the template, a second which allows the DNA polymerase to extend each primer to make another copy of the template, and a third which allows the new copy to detach (“de-anneal”) from its partner. With each cycle, the number of copies of the template doubles (approximately), and after many cycles, even a single molecule of the template can be expanded into enough molecules to be used for many purposes, including detection.
The type of PCR most often used for detecting RNA viruses in the clinic is called rt-RTqPCR. This stands for “real time, reverse transcriptase quantitative PCR.”
In this method, PCR is performed using a reverse transcriptase, which allows copying of RNA in the viral genome onto DNA, in addition to the DNA polymerase. The concentration of amplified template is measured in real time using photospectroscopic methods inside the reaction chamber, and the time at which the detection is made allows an approximate quantification of the number of template sequences which were present in the initial sample.
For example, if the template is 1kb long, the efficiency of amplification is about 90%, the detection threshold is 1gb of templates, and detection is achieved after 15 thermocycles, this would suggest that the original sample contained about 66 copies of the template. If no detection is obtained after 25 samples, this would show that the original sample did not contain any copies of the template.
The bottom line: Modern PCR is a highly developed technology. It is used as the positive control by which other biological methods are evaluated, because it’s among the most reliable methods in biology. Even when performed by hand on benchtop machines by inexperienced personnel, PCR is relatively reliable.
Modern medical PCR laboratories with machines from firms like Roche perform both the extraction of DNA from biosamples and the PCR amplification and detection themselves with almost complete automation. Single machines fitting in one room can process thousands of samples a day under the supervision of only a few technicians.
If, for example, one such machine and the supporting infrastructure were present in each of Wuhan’s ten major hospitals, along with the appropriate technicians and supplies, then Hubei province would have all the testing capacity it needed during the current outbreak (so far).
China’s 2019-nCoV PCR test campaign has been a disaster
Earlier, we covered the perplexing fact of China’s medical PCR capacity in Wuhan being stubbornly low, and lower than projections.
The PCR capacity in Wuhan was reported in the last week of January at two thousand samples per day with the expectation of growth, a low number but in the correct order of magnitude. However, the Caijing report last week showed that this number was not being achieved in practice, with Wuhan’s ten major hospitals being offered as few as ten tests per day, rather than the hundreds (and growing) that they’d been promised.
More recent reports from Wuhan have not shed light on how (if at all) this situation has changed. We have no confirmation of whether or not PCR capacity in Wuhan has undergone the rapid large increase that modern technology should allow.
The shortfall in capacity has had significant negative impacts on both reporting (and the research and social trust based on reporting) and the care of patients in Hubei.
With plenty of capacity and reliable tests, patients could be diagnosed in hours and sent home if they tested negative, clearing hospital resources for real cases. Instead, cases have backed up in hospitals, resulting in 2019-nCoV cases going without care for days (or at all), resources being used on patients who turn out to be flu patients, and, worse, cold and flu patients being quarantined along with 2019-nCoV patients, and likely contracting the disease themselves.
In addition, less reliable reporting hampers every line of research which uses reporting numbers. Finally, unreliable numbers, hospitals being unable to help those who arrive, and cases and deaths happening outside the medical system all undermine trust in medical authorities, which has a number of negative consequences in a volatile situation like a deadly epidemic.
And we just learned it’s even worse than we thought it was
Now, however, we have learned something even worse: A new report in Caixin confirms what we have heard for the prior two days from sources in Beijing and Hong Kong: an unspecified flaw in China’s PCR testing campaign led to 50% to 70% of otherwise-confirmed cases showing negative results on Chinese PCR tests, including many of those who later transmitted the disease to others. This problem has been confirmed to affect tier one cities like Beijing and Shanghai in addition to more affected provinces like Hubei and Zhejiang. This is not what would be expected, since qPCR is supposed to be able to detect even a single molecule of template DNA.
By contrast, such failures have not been reported outside China. PCR tests by the WHO, CDC, and major first world hospitals are still considered reliable, although this theoretically could change.
As a result, Chinese medical authorities are considering changing the diagnostic criteria for 2019-nCoV infection to allow confirmed diagnosis on the basis of only a symptom profile and chest X-ray, bypassing biochemical methods altogether.
Worse, this means that cities outside Hubei have been diagnosing many 2019-nCoV cases as negative and sending them out into the general population, meaning that the quarantine has less effective than it would otherwise have been in ex-Hubei China up to approximately now, and China is in store for at least another week of case numbers rising more rapidly than they should have. This is a disaster.
It remains totally unclear why this has happened. Theoretically, false negatives in a PCR test of this kind could result from problems with the sampling procedures, sampling equipment, RNA extraction protocol, PCR reagents, PCR machine, PCR protocol, or some combination of these. But given China’s role as a significant player in the global biotechnology scene, it’s unclear why China would suffer such a disaster, just as a week ago it was unclear why PCR capacity would remain stubbornly below what was needed.
Western governments and companies could help China fix its problem
It doesn’t have to be this way. There are a number of ways that biomedical experts outside China can help China’s experts fix their PCR testing program.
One would be to share expertise. China’s experts could easily be consulting with international experts on the exact details of their testing procedures, supplies, equipment, reagents, and protocols, comparing them to the ones used elsewhere and for other diseases, and jointly figuring out how to make them more reliable. In addition, they could be comparing notes on how to rapidly expand testing capacity where it’s needed most, by moving equipment from place to place, running current equipment on more aggressive duty cycles, and pressing research equipment into clinical use on an emergency basis.
But more than this, Chinese and international experts should be working together to put international PCR capacity to work for Chinese patients. There’s no fundamental reason that, e.g., thousands of sample tubes could not be shipped by air from China to be tested in PCR labs elsewhere in the world. This would not allow the same rapid turnaround as onsite PCR labs, but it should allow patients to (e.g.) have samples collected, and then be sent home to receive testing results in a day or two. It would also allow more accurate numbers to track the progress of the epidemic.
It might even be possible to rapidly deploy PCR equipment, reagents, and technicians from other parts of the world to China together, on an emergency basis, to near-immediately establish high-capacity PCR labs for use in the epidemic. Midsized PCR labs fitting in the back of a truck have been used in rural areas of Southeast Asia for years, and the fixed locations and preexisting laboratory capacity with utilities and fixtures should make it even easier to move large PCR equipment into hospitals in China.
If rapid and accurate testing could be made available to everyone in China who needs it, it might make a very large difference in the containment of the epidemic, potentially the difference between containing it and allowing 2019-nCoV to go endemic or pandemic. And it may be possible to make this outsized difference in the overall situation with a relatively modest effort. A PCR test is orders of magnitude cheaper, easier, and simpler than weeks of inpatient care.
Accordingly, efforts along these lines to expand and improve the PCR testing effort in China should be considered a priority in international efforts to help China in the 2019-nCoV epidemic.