When the German team (Cormen et al. (aka “Drosten Report”)) published primers capable of detecting synthetic oligonucleotide that matched parts of the SARS-CoV-2 genome sequence published from the first clinical sample from the first patient diagnosis of pneumonia associated with a novel coronavirus, there was hope that the virus might be controlled at least until healthcare facilities could be made ready using the training and preparations they conducted during the Ebola care of 2014.
The use of PCR to detect specific sequences is trivially non-controversial, if it is done in an expert manner. First, the primers are chosen computationally to match only the target species (or quasi-species, in the case of viruses). Second, multiple targets can be selected that help nail down the results in case one of the single. Finally, if the primers are nested or hemi-nested – that is if the targets sequences overlap, a local sequence can reliably be determined on a percentage of the samples as a direct gold-standard check to estimate both the true positive rate (the probability of detecting a virus that is truly present) and the false positive rate (the probability of detecting the target when it is truly not present).
I expected when CDC’s PCR test failed that there would likely be a variety of approaches that would be developed by commercial nucleic acid technology companies. I started advocating for private mass testing, with an emphasis on “private”, expecting that accurate testing would emerge.
I was astonished when I saw that the FDA had only requested that commercial suppliers provide data on the true positive rate, but no data on the false positive rate. Worse, I was astonished to learn, upon reading the documents being submitted to the FDA, that the test kit manufacturers were not including positive control sample material to allow the determination of cycle threshold (Ct) by which one makes the call for a given patient that the virus was present or absent. Cycle thresholds are the number of rounds of amplification necessary to reach a specific point in the exponential growth of the number of copies of target sequences. This must be done for each patient separately – even if the test is the same kit done in the same lab, on the same day, by the same technician – because the amount of starting material in each swab varies.
This was the first time I had ever seen any RT-PCR-based test NOT use a positive control sample. The lack of a positive control sample means that the assay was qualitative, not the robust and rigorous qRT-PCR (“q” stands for “quantitative”. For example, even the RT-PCR test for the Monkeypox virus uses a positive control sample.
Instead of using empirically derived Ct values per patient, generic Ct values were used. These were part of the kits, but they were not published. This was unusual. I asked medical freedom activists to request from their local health department what Ct values were being used to determine a diagnosis for a patient. Multiple people did, and the reply was the same: that information is proprietary. This was ridiculous; the exact Ct being used is not a top-secret part of a test, but is, instead, an essential aspect of checking the reliability of the results of an RT-PCR test.
Go to top link for a video on RT-PCR.
The absence of a positive control target sample was a big deal because, without any data on whether tests were hitting human sequences via off-target amplification, many people would be “diagnosed” with COVID-19 who were not infected. This would not only be disruptive – the resulting huge numbers of false positives could be used to justify police actions as was going on in China at the time.
Further, people who had other respiratory illnesses like influenza, respiratory syncytial virus, bacterial pneumonia, the common cold, or other coronaviruses might not be given appropriate medical care given the isolate-and-do-nothing approach to COVID-19. Many would die from severe pneumonia that could have been prevented via medical intervention, just as antibiotics for bacterial pneumonia.
Also, it was clear that many people who were sick but did not have COVID-19 would believe thereafter they were immune, and they might risk exposures that they otherwise might not risk.
In addition, those that did then become infected after their false positive COVID-19 test might then doubt their actual case of SARS-CoV-2 infection was worth testing for and they might fail to protect others who were at the highest risk of death from COVID-19 infection – the immunocompromised and the elderly.
So I wrote to the FDA with my concerns. Dr. Peter Marks wrote back with a terse “Thank you, we will take your concerns up with my team”.
At the same time, Dr. Sin Hang Lee independently saw the same issues and was motivated to develop a Sanger sequencing-based test. His approach circumvented the risk of false positives altogether by using primer pairs that targeted parts of the SARS-CoV-2 genome in an overlapping mapper. This way, he could tell if he truly had the sequence of interest in a sample, or if one of the primer pairs was failing due to a mutation in the primer site.
