From the start of the SARS-CoV-2 outbreak in early 2020 and included in World Health Organization guidelines urging governments to “test, test, test” as part of their contact tracing and containment strategy, the PCR test was promoted as the “gold standard” of COVID-19 detection.
COVID-19 cases confirmed by diagnostic test alone
The first step in any disease outbreak investigation is to create a case definition. Epidemiologists define the uniform criteria that classify a confirmed case of a disease for public health surveillance. A clear, concise definition is essential when deciding what to measure in terms of frequency and severity of illness. Case definitions for communicable disease should contain both clinical symptoms and confirmatory laboratory testing . An example of a case definition for influenza is: an individual who meets influenza-like illness (ILI) criteria of fever (≥100.4℉) and cough and/or sore throat and has a positive culture or RT-PCR test for influenza from a respiratory specimen.
In April 2020 the Council for State and Territorial Epidemiologists (CSTE) issued an interim position statement for the COVID-19 case definition . COVID-19 became the first communicable disease for which no clinical signs or symptoms were required for classification: a case was confirmed if it “meets confirmatory laboratory evidence.” The confirmatory laboratory evidence was “detection of SARS-CoV-2 RNA in a clinical specimen using a molecular amplification detection test.” The approved diagnostic test for COVID-19 was real-time or quantitative polymerase chain reaction (RT-PCR), a technology used for estimating the number of copies of a specific DNA sequence in a sample in real-time.
What is polymerase chain reaction (PCR)?
Viruses are simply genetic material, either DNA or RNA, surrounded by protein. There are several hundred viruses that can cause respiratory illness in humans. In most instances, doctors do not test for respiratory viruses. Prior to the 1980s, methods for testing were culture and serology (blood serum) based. A typical culture could not grow over seventy of the respiratory viruses which infect humans, and it was always a lengthy, subjective and insensitive method for those viruses that could grow. After the 1980s, a technique called polymerase chain reaction (PCR) became the gold standard for testing for viral infection.
PCR was invented by Dr. Kary Mullis in 1984 during a moonlit drive through the mountains of California . PCR is a technique that harnesses the power of DNA replication – a basic biological process in cells. It is a powerful technique, very sensitive and highly specific, which is used in forensics and research because very few copies of a gene sequence in the starting material are required for detection. This technique uses DNA polymerase and synthetic bits of DNA to amplify a specific region of a target DNA sequence in vitro, which then becomes a part of further reactions . Each round of replication increases the region of interest exponentially, doubling the product with each cycle until it can be detected.
Three steps of PCR─denaturation, annealing, and extension─as shown in the first cycle, and the exponential amplification of target DNA with repeated cycling. Image source
In standard PCR, the amplified material is detected by separating the mixture by gel electrophoresis – a laboratory method used to separate mixtures of DNA, RNA or proteins according to molecular size. However, with quantitative PCR, the reaction products can be monitored in real-time using fluorescent labelling. This enables the collection of data as each cycle progresses. The fluorescence signal increases proportionally to the amount of replicated DNA. Hence, the DNA is quantified in “real time.”
The threshold line is the level of detection or the point at which a reaction reaches a fluorescent intensity above background. The cycle at which the sample reaches this level is called the cycle threshold (Ct). The Ct value is inversely proportional to the amount of starting material. This means that the higher the Ct value, the lower the number of copies in the starting material; conversely, a lower Ct value would indicate a higher number of copies in the starting material. When using real-time PCR for detection of virus, this would mean that less virus is present when the Ct value is higher because it takes more cycles to detect the given target. Therefore, Ct values can give an indication of where a person may be in the infectious cycle. However, just because the test detects the viral genome, there is no way to know if what is being detected is from an infectious, actively replicating virus.
The plot resembles a sigmoidal amplification pattern, as indicated by the green line. This green line demonstrates the change in fluorescence over the number of cycles. Typically, in a qPCR experiment, there are 40 cycles. That is 40 rounds of amplification. Image source
Many of the PCR platforms being used for detection of SARS-CoV-2 had established a Ct value of forty (also recommended by the CDC) as the cut-off value for a positive test, out of a total of forty-five cycles. At high Ct values (more than thirty), there is usually non-specific binding due to a host of things (primer dimers, environmental contaminants, etc.) . Also, depending on the type of detection system, fluorescent dyes will bind to any double stranded DNA in the reaction tube . These types of reactions would be considered false positives: with a positive result in people who are neither infectious or contagious.
Ct values should not be used to inform public health policy
There is now enough evidence in terms of published data showing that the Ct cut-off values used to determine a ‘case’ were too high. Meaning, the positive results may not have been meaningful and the person was well past the transmissible stage of infection. Viral culture is the reference against which any diagnostic index test for viruses must be measured and calibrated, to understand the predictive properties of that test. The CDC noted that infectious virus is no longer detected 8 days post infection. The peak of SARS-CoV-2 replication is around day 5 post infection and drops off, to be no longer detectable via culture by day 8 . In terms of surveillance, the high Ct values may indicate a past infection, but certainly are not indicative of a current infection. Specimens with Ct values of >30 have been shown to not contain culturable (actively replicating) virus. A study by La Scola et. al. correlated the Ct values obtained using RT-PCR technique based on amplification of the E gene of SARS-CoV-2 (Figure 1). The results of the culture show, “that at Ct = 25, up to 70% of patients remained positive in culture and that at Ct = 30 this value drops to 20%. At Ct = 35, the value we used to report a positive result for PCR, <3% of cultures are positive .”
Percentage of positive viral cultures of severe acute respiratory syndrome coronavirus 2 polymerase chain reaction–positive nasopharyngeal samples from coronavirus disease 2019 patients, according to Ct value (plain line). The dashed curve indicates the polynomial regression curve. Abbreviations: Ct, cycle threshold; Poly., polynomial. Image source
Another study investigated viral culture in samples from a group of patients and compared the results with PCR testing data and the time of their symptom onset . This study analysed ninety RT-PCR positive samples using cell culture. The SARS-CoV-2 cell infectivity of respiratory samples from SARS-CoV-2 positive individuals was only observed for RT-PCR with Ct values < 24 and symptom onset to test of < 8 days. The probability of obtaining a positive viral culture peaked on day 3 and decreased from that time on. Therefore, infectivity of patients with Ct values >24 and duration of symptoms >8 days was considered to be low. They also determined for every one-unit increase in Ct value, the odds of a positive culture decreased by 36%. This study showed that the probability of SARS-CoV-2 infectious virus is greater when the cycle threshold is lower. When symptoms to test time is shorter and confirmed beyond 8 days, no live virus is detected.
Another study published in August 2020 tested specimens from the first 3 months of the COVID-19 pandemic in the United Kingdom: late January to early April 2020 . Using RT-PCR cycle threshold (Ct) values as a semiquantitative measure of SARS-CoV-2 viral load, they found that the level of SARS-CoV-2 RNA in the upper respiratory tract was greatest around symptom onset, steadily decreased during the first 10 days after illness onset and then plateaued. Virus culture was attempted from 324 samples, extracted from 253 cases that tested positive for SARS-CoV-2 by RT-PCR. They observed a strong relationship between Ct value and the ability to recover infectious virus. The estimated odds ratio (OR) of recovering infectious virus decreased by 0.67 for each unit increase in Ct value (95% CI: 0.58–0.77). Detection of cultivable virus peaked around the time of symptom onset, with a median duration of 4 days of virus shedding as measured by culture. They concluded that using Ct values provides a valuable proxy for infectious virus detection, and may help to inform decision-making on infection control.
Traditionally clinicians do not test for respiratory viruses because testing does not inform patient treatment. In other words, testing for respiratory virus infection would not change how a doctor would treat an individual, since treatment would be symptom-based and mostly supportive. Influenza testing is routinely performed because it is the only respiratory virus for which there is an approved antiviral therapeutic for treatment. As SARS-CoV-2 was not associated with a constellation of specific symptoms, and symptom-free individuals were suspected to be infected, confirmation of a case was based solely on results from a molecular test. This is the first communicable disease for which a person with no signs or symptoms of illness may be classified as a confirmed case. A confirmed case based solely on a molecular test means due diligence must be done to ensure the proper classification of each case. The detection of some bit of genetic material does not indicate whether an individual is infectious or contagious. This means more details are required for the positive result.
The high number of people who tested positive with no symptoms indicated that positive PCR results were not meaningful for diagnosis and the patients could be well past the transmissible stage of infection. The widespread use of PCR testing greatly skewed and corrupted the data on SARS-CoV-2 infection and death rates. The take home message is that results from PCR testing should be evaluated in the context of clinical symptoms and possible exposure. Under no circumstances should the PCR test be used as a broad surveillance tool, or an indicator of infection prevalence, or to evaluate cause of death. Furthermore, PCR test results should never be justification for imposing population-level measures, mitigations or restrictions.