From the start of the COVID-19 pandemic, diagnostic testing has been highlighted as a key part of global measures to contain the spread of SARS-CoV-2. In a media briefing on March 16, 2020, Dr Tedros Adhanom Ghebreyesus, WHO Director-General, remarked that testing, isolation and contact tracing were “the backbone of the response” and urged all countries to “test, test, test”.
RT-qPCR testing, which detects SARS-CoV-2 genetic material present in a patient sample, quickly became the predominant method used to identify infected individuals. However, several claims and allegations about the capabilities and value of PCR have been circulated throughout the course of the pandemic, leading to questions about its use.
To address these misconceptions and communicate the true strengths and limitations of the technology, a group of PCR experts recently published a commentary in the International Journal of Molecular Sciences. Technology Networks had the pleasure of speaking to Professor Stephen Bustin, professor of molecular medicine at Anglia Ruskin University, and lead author of the commentary, to learn more about some of the facts and fallacies highlighted.
The views and opinions expressed below are those of Stephen Bustin and do not necessarily reflect the official policy or position of Anglia Ruskin University.
Anna MacDonald (AM): RT-qPCR was quickly adopted at the start of the pandemic as the predominant method of detecting SARS-CoV-2. Can you describe some of the key strengths of the technology that made it a suitable choice?
Stephen Bustin (SB): Indeed, the first three PCR tests for SARS-CoV-2 were designed within a day of its genome sequence having been published by Chinese scientists. Such lightning speed is unthinkable with any antigen-based test, hence PCR being the predominant method. Furthermore, a properly designed, optimized and validated RT-qPCR test is the most sensitive, specific, reliable and robust method for detecting a pathogen. Current protocols are not as fast as antigen tests, but they will be so in the future (see below).
Crucially, RT-qPCR tests are also easily modified to accommodate the appearance of mutations and variants and so are characterized by exceptional flexibility, an important advantage when dealing with an RNA virus that, against initial expectation, continually and rapidly changes. Once genetic changes that characterize variants have been identified by sequencing, it is straightforward to design tests and, crucially, use them immediately. Again, this is unlike antigen tests, that require the development and production of new antibodies, a process that takes months and is very expensive. PCR tests can distinguish the original virus from the variant even if they differ by only a single nucleotide change (e.g., variant B.1.1.7 N501Y which has a nucleotide change of A to T within the sequence coding for the spike protein).
Although mainstream PCR instruments take between 30 minutes and 1.5 hours to complete a test, a combination of fast protocols and instruments can reduce that time to less than 15 minutes. Even that does not fully exploit the potential of this technology, with “extreme” PCR shown to complete a test in less than 20 seconds. I have no doubt that in a year or two’s time such instruments will be available commercially. Furthermore, there will be inexpensive hand-held personal and point-of-care devices that will give rapid results when and where needed.
AM: Several claims have been made suggesting that PCR-based testing is not fit for purpose. What has led to these allegations and what effect is this misinformation having?
SB: These sentiments arise from the deliberate spreading of false information, coupled with misleading quotes and incomplete reporting. They are made by people who have little knowledge and no understanding of the technology, but are articulate and rely on the general public’s ignorance and lack of interest in scientific detail. The media also are at fault, because they have generally failed to distinguish between authoritative scientific conclusions based on data and facts versus false and unreliable opinions based on mendacity and sophistry.
The most common falsehood is that the inventor of PCR, Kary Mullis, claimed that PCR should not be used for diagnostics. He did make an unfortunate comment about the reliability of diagnostic testing, but this related specifically to the detection of HIV in AIDS patients, was aimed at conventional gel-based PCR and was made in 1993, before real-time (qPCR) was in use as a diagnostic tool. I knew Kary Mullis personally and I know for certain that this was not his opinion as we discussed, and he was interested, in the use of RT-qPCR as a prognostic tool for colorectal cancer patients.
Without a shadow of a doubt, qPCR is highly suitable for diagnostic testing for pathogens for at least three reasons: (1) it has two levels of specificity that minimize the risk of false positive results, (2) its sensitivity minimizes the risk of false negative results and (3) as a closed-tube procedure amplified DNA is never released into the environment, minimising the chances of contamination. In a way we are seeing the same dishonest campaign resurrected that caused so much unnecessary heart-ache in the late 1990’s concerning the MMR vaccine and autism.
These operations are driven by callous individuals and groups with an agenda, and unfortunately are not open to persuasion of any kind. PCR testing is just another topic that has become entangled in today’s doctrine that feelings and opinions matter as much or, indeed, more than facts, a phenomenon that has become familiar in politics, education as well as in a health-related environment. Whilst it is easy to come up with a simple slogan that denigrates something and reduces a complex issue to a single catch phrase, it takes much longer to explain that matter adequately and sensitively. Ultimately, there is an obnoxious fringe element in our society that is not interested in listening or learning and, sadly, there is neither vaccine nor cure for stupidity. A serious consequence of this trend is that the public is unable to distinguish genuine scholarship and balanced expert advice from snake oil peddled by fraudsters and charlatans.
AM: Can you explain the main causes of false-positive or false-negative results and what can be done to reduce the chances of them occurring?
SB: False positive results in a properly validated and CE-marked PCR test are caused by contamination or by inappropriate interpretation of the results. Contamination can be introduced during the sampling procedure, the extraction process or the dispensing of reagents for the PCR test. This is routinely detected by including appropriate negative controls at every stage, i.e., samples that do not contain the target RNA or DNA. These controls must always be negative; if they are and a PCR result is positive, this means that whatever target was being amplified is present. This is especially important when there is very little target present and the analysis generates results near the test’s limit of detection. The “analytical sensitivity” of a PCR test refers to the smallest amount of target in a sample that can be detected and in theory this is a single molecule. It goes hand in hand with its “analytical specificity”, which describes the ability of the test to detect one specific target, e.g., SARS-CoV-2, as distinct from, say SARS-CoV.
However, both are different from “diagnostic sensitivity and specificity”, the former referring to the ability of the test to identify individuals who are infectious and the latter to its ability to identify correctly those without the disease. Hence a true PCR positive may well not be detecting infectious virus, for example if the individual has been asymptomatic for a couple of weeks after coming down with a disease and, whilst a few virus particles or nucleic acid fragments are present, they are not clinically relevant. Hence the interpretation of the test was at fault, not the test itself. This is a real problem for two reasons (1) the infectious dose of SARS-CoV-2 is still unknown and (2) the qPCR result output is the quantification cycle (Cq), which is not an objective value but differs with instruments, reagents and operators. Monitoring a combination of parameters that include symptoms, probability of infection and PCR results is the best approach to minimize false clinical positives.
False negatives, on the other hand, can be caused by sample degradation during handling or storage, which can be especially problematic with RNA, the presence of inhibitors, poor testing protocols, inexperienced staff not following proper procedures and, least likely, reagent failure. It may also be that the sample taken from an individual did not contain any virus or so little that it got lost or excessively diluted during the extraction procedure. This is of course why a negative test result is really a presumptive negative and a second test a day or so later could well result in a positive result, if the individual is in the early stages of infection.
AM: In 2009, you published the MIQE guidelines, with the aim of encouraging better experimental practice and more reliable interpretation of qPCR results. Have COVID-19 tests been developed using these guidelines? How could the guidelines be applied to improve the use of qPCR testing?
SB: The MIQE guidelines were aimed at the research community, where new tests targeting a multitude of targets are being constantly developed and where it is difficult to compare the results generated in one laboratory with those obtained in another one. Furthermore, they deal with quantitative testing, which is somewhat different from testing for SARS-CoV-2.
However, the guidelines are relevant as they call for transparency with regards to assay design, reporting of test performance and interpretation of results. Researchers and companies have done a sterling job in developing a range of tests that are sensitive and specific and continue to monitor the emergence of new strains to ensure that their tests continue to generate reliable results. There should be more openness with regards to sharing the sequence details of the various tests in use, but the information provided is generally sound and informative.
However, the main issue that has caused some problems with the reliability of testing is the sampling, transport and extraction workflow, which is far from optimal. There are numerous different reagents and protocols, variable skill levels of those taking the samples, different time lines within the transport practice and inconsistent storage conditions as well as different RNA extraction and concentration procedures that all combine to increase the variability of test results, especially when comparing different test centres within or between different countries. There is an urgent need to develop a clear set of guidelines for optimal sample collection, which should be from saliva for SARS-CoV-2 as well as for RNA extraction, which would help standardize and so make the whole process more reliable.
AM: The use of qPCR to quantitate viral load is a particular area of concern. What are the main challenges of using the technique for this purpose? Are there any steps that can be taken to reduce the ambiguity of results?
SB: First of all it is important to distinguish between infectious dose and viral load. It is still unknown what the infectious dose is for SARS-CoV-2, i.e., the amount of virus required to make a person sick. Its ease of transmission suggests that it may be quite low, but this might be 100, 1,000 or 10,000 viral particles. A higher infectious dose could result in a higher viral load, which describes the amount of virus replicating within the cells of an infected individual. There is evidence that higher viral loads result in more severe symptoms and may be associated with a worse outcome as well as lead to more shedding of whole virus by the infected individual.
A standard qPCR test cannot record exact viral copy numbers, as the test result for a positive sample only records a PCR cycle number, e.g., 25.5. This means that the instrument first detected the presence of a target after 25.5 cycles, and whilst a lower Cq value suggests that there is more target present, without additional information it is impossible to tell how much more. This requires knowledge of the amplification efficiency, since it is obvious that a more efficient test will record a lower Cq earlier than a less efficient test.
Quantification also requires the addition of standards, ideally certified reference RNA of known copy number, that amplify at specific PCR cycle numbers. The cycle number of the virus can then be related directly to that standard and give a measure of quantity. Unfortunately, such standard still do not exist. There is a related technology, called digital PCR, which can provide absolute counts of viral copy number. However, this technology is not suited to mass testing and it is expensive. The best solution, for the time being is to record a test result as positive or negative and interpret that result within the individual clinical context and, if in doubt, repeat a day or so later.
AM: In what ways do you think the pandemic has shaped the future of PCR?
SB: Undoubtedly there will be a huge demand for continued, rapid testing at hospitals, care homes, nurseries, airports, cruise liners etc, initially for SARS-CoV-2 but extending to other respiratory viruses as well as fungal and bacterial pathogens. This will drive the development of easy-to-use sampling, extraction and testing devices that can rapidly test multiple samples with little or no operator intervention.
At the same time there will be an increasing demand for more personal devices, such as hand-held microfluidics-based systems where you can add a small amount of saliva, blood, urine or faecal matter and get a rapid test result in the privacy of your own home, at the GP’s surgery or even at the chemist’s. The speed potential of the PCR and its ability to detect numerous targets all at the same time are still largely unexplored and the focus on improving current protocols and introducing new instrumentation and improved reagents will only serve to make PCR even more ubiquitous than it is now.
Reference: Bustin S, Mueller R, Shipley G, Nolan T. COVID-19 and diagnostic testing for SARS-CoV-2 by RT-qPCR—facts and fallacies. Int J Mol Sci. 2021;22(5):2459. doi:10.3390/ijms22052459
Professor Stephen Bustin was speaking to Anna MacDonald, Science Writer for Technology Networks.