The Complexities of Science and Individuals
Limitations and Concluding Remarks from yesterday's post.
This is a follow-up to yesterday’s already long post going over some of the limitations of the study and some concluding remarks with a bit of my perspective.
The article from yesterday was something I’ve been yearning for us to have ever since COVID became a thing.
Being provided a timeline and window into the viral and immunological responses of each individual helps elucidate many of the details that continue to go missing in the literature.
Here, the researchers provided a very limited profile, but it’s a profile that we otherwise would not have, especially after the vaccines rolled out and confounded all of the data.
That being said, it’s important to note some key issues with this study. I’ve noted a few throughout yesterday’s post, but here I’ll make note of some serious issues.
Table of Contents
Study Limitations
Study Limitations
Limitations in PCR
Of course, I’ve mentioned several times that the researchers didn’t actually measure viral load, but rather used PRC values as a proxy.
PCR can be very misleading since it doesn’t provide the scope of actual virions or if they are still infectious. PCR can only detect genetic material, and so rather than viral load it’s more an indication of genetic load so to speak.
This can create a high degree of variability. The researchers noted that when they sent out some of their samples to CLIA-accredited1 labs some of the participants (4 out of the 7 participants' nasal swabs) had negative results come back.
This contradictory evidence can come down to lab practices including how samples are stored (either dry or in VTM) and what probes are used.
The researchers provide a lengthy discussion about this discrepancy. Initially I was going to point out which sentences to emphasize, but I think the entire paragraph is important here:
Some of the study participants obtained complementary, nasopharyngeal RT-qPCR SARS-CoV-2 tests through CLIA laboratories (arrows in Figs. 2A, 3A, B, D, E). It is noteworthy that in four out of seven tests (57%), a negative result was obtained with the CLIA test, while a positive test result was obtained in our study on the same sampling day. Differences in the initial stages of sample preparation in the two tests (i.e., the current study and CLIA laboratory) present potential consequences for assay sensitivity. In our study, swab samples were either collected dry (i.e., no preservative added, processed within ca. 40 min of collection) or preserved with RNA Shield (300 µL, Zymo Research, Tustin, CA) and typically processed within 24 h of collection. Extraction of viral RNA from these samples was carried out by first adding lysis buffer (400 µL) followed by column purification of the entire sample volume (i.e., either 400 or 700 µL). In CLIA laboratory SARS-CoV-2 RNA RT-qPCR tests, the swab is preserved in larger volumes of transport media (typically 3–10 mL), and in the subsequent RNA extraction step, an aliquot of the sample is used. It is therefore expected that, for the same sample, lower quantities of viral RNA are analyzed in the CLIA laboratory test than in our study, likely leading to lower sensitivities for the former. These considerations need to be taken into account when comparing cycle threshold (Ct) values, or viral RNA copies per unit volume, across studies. Consequently, we have reported viral loads as SARS-CoV-2 copies per swab above.
This gets into the weeds as to how different testing procedures can influence PCR results.
In the COVID lab I worked for we would receive samples in viral transport media (VTM) which, ideally, should come in tubes with the same volume of media. One may assume that calculations would address VTM load (i.e. if 2 mL of VTM is standard, calculations on concentration would take into account that 2 mL was used to transport swabs and how much volume was taking out for testing), but if VTM volumes are inconsistent then viral loads could become incorrectly calculated. In several cases some samples may come in hardly any VTM or in a volume far greater than it normally should.
Here, it could be that the different storage and processing methods may influence the genetic load of each sample, as well as the sensitivity of each assay.
Could we argue that the researchers may be overestimating the viral load, or could these CLIA-accredited labs be underestimating?
This becomes an issue when such high variability can lead to inconsistent results.
There’s also the concern in what genes are being probed for these PCR assays.
For the COVID lab I worked in we used probes targeting 3 genes: the S gene, the N gene, and the Orf1 gene.
Here, the researchers used two probes, both of which only targeted the N gene of the virus:
The samples (nasal, stool, saliva) were analyzed for SARS-CoV-2 RNA copy numbers by reverse transcription (RT) and quantitative PCR (qPCR) using primer sequences targeting the SARS-CoV-2 nucleocapsid protein (N) gene transcript fragments (N1 and N2) and one human RNase P (RP) gene transcript fragment (RP).
RNase P is being used as an internal control here. But as you can see such a limited PCR method means that the researchers may not be properly addressing the critical exclusionary/inclusionary principles of using PCR as a testing method.
PCR positivity is dictated by the probes used and which genes are amplified, which means once again the results here may vary greatly from other PCR methods.
Overall, the fact that PCR has been used extensively raises a lot of limitations in how these results can be interpreted, such as the Cmax from these calculations or how to interpret the actual time until viral clearance.
However, this issue can be overcome given the fact that many of the samples show similar trends in PCR, starting with a sudden increase followed by an immediate decrease with a plateau over several weeks.
The researchers provide this explanation for why they used PCR as a proxy:
There are some limitations associated with the results from our study that followed a small cohort of nine participants who developed mild COVID-19 symptoms. Viral RNA quantified in clinical specimens by RT-qPCR was used as a proxy for SARS-CoV-2 shedding, but viral genome copies would not necessarily reflect titers of infectious viruses. Evaluating viral load by plaque assay would have been challenging given the difficulties in isolating the virus from clinical swab samples, as described above.
Additional Limitations
One thing that really hurts this study is the number of unclear data. I’ve mentioned that some data is missing for specific participants, but this also extends to lack of demographic data as well as the type of variant participants were infected with.
Typical of these studies, the researchers never indicate which variant participants were infected with. We can make some assumptions given some of the figures. For instance, the T-cell blood samples were taken on April 1, 2021 and included participants who were infected up to 4 months prior. These participants are likely to have had the D614G variant, and possibly the Alpha variant. However, such ambiguity may influence results as different immunological responses may correspond to different variants.
Both Subjects 38 and 48 were infected at the end of 2020 which may infer the D614G variant, but context is missing for the other participants.
This includes the lack of demographic data, as we only know the gender and ethnic makeup of the 9 participants. We know that a great deal of genetic diversity exists across ethnic lines, which includes HLA alleles that influence T-cell responses. Having such limited demographic data doesn’t help elucidate the effects genetics plays on SARS-COV2 susceptibility as well as the varied immunological responses.
Once again, we are left in the dark here, and it’s frustrating given how close these researchers are to constructing specific profiles.2
Concluding Remarks
Even with the limitations, this is one of the first studies I have seen to capture so much information from individuals.
These are the type of studies that we need if we are to get close to figuring out how people respond to a SARS-COV2 infection.
This includes the following remarks, which contradicts some of the established information on time until symptomatic:
Our clinical study produced a number of unexpected results. One participant (subject 48), believed to have been exposed to SARS-CoV-2 on December 25, 2020, started to experience symptoms consistent with COVID-19 on December 27 (Sunday) and tested positive for viral RNA by RT-qPCR in a nasal swab sample the following day (Monday) at a scheduled test. The low Ct values measured on December 28 (N1, 15.36; N2, 16.31) corresponding to a viral load of 9.0 × 107 copies per swab (Fig. 3B) suggest this individual likely would have been infectious the previous day when symptoms first presented, only two days following exposure. Another participant (subject 38) tested negative for SARS-CoV-2 RNA on December 28, 2020, but experienced symptoms consistent with COVID-19 the following day and tested positive a day later, on December 30, with low Ct values (N1, 16.56; N2, 18.17) and viral loads of 3.2 × 107 copies per swab (Fig. 3A). As with subject 48, the viral proliferation trajectory suggests that this participant would have been infectious the previous day (December 29) coinciding with the onset of symptoms, and only 1 day after a negative test result.
Based on these interesting results the researchers made this comment:
These results contradict the dogma that COVID-19 symptoms manifest following an infectious, asymptomatic phase spanning multiple days. Furthermore, viral replication following exposure is so rapid (Td ca. 3 h) that a negative test result may only provide a one-day safe window prior to becoming infectious.
Remember that 2-week quarantine period that was hammered into our heads, which only recently came down to 5 days?
The evidence here, albeit limited, suggest that the time between exposure and symptoms could be as quick as one day. More studies would help expand on this finding, but these results were only possible due to the constant surveillance done in this study.
This raises some serious questions as to how accurate those established ideas of COVID infection are, especially since many people have rightfully pointed out the absurdity of a 2-week quarantine period, and we may even argue that a 5-day quarantine is excessive.
I should also point out that the researchers included T-cell data on participants who never tested positive up to the collection of blood on April 1, 2021 which showed an adaptive cellular response suggesting that these participants may have been infected prior to March 23, 2020 and may indicate an earlier spread of the virus than was assumed (Figure 9).
Embrace Ambiguity; Embrace Complexity
If we were to summarize this study, we could just conclude with the following statement:
Everybody is Different
and even the commonly adopted phrase popularized by Bret Weinstein and Heather Heying:
Welcome to Complex Systems
Or put in the ways the researchers did:
Our results underscore the highly heterogeneous nature of these processes across individuals.
It doesn’t quite roll off the tongue in as nice of a manner, but it explains the results of this study rather well.3
I’m continuously reminded of Joomi Kim’s post in which she suggested that we may not find answers to the effects of these COVID vaccines for many years, with her paying particularly close attention to issues in methodology describing spike in the nucleus of cells.
In the same ways that we may argue that results may be ambiguous, we also have to remember the complexities of being an individual.
We each carry our own medical history, comprised of the pathogens we came into contact with while growing up, the genetic profile that was passed down to us through many generations, and the nutrition and habits that constitute our daily leaving all influence how we respond to infections.
In the same ways that we can argue against “one-size-fits-all” approaches to COVID, we should argue against trying to find conclusive explanations for why some people do worse with COVID while some are hardly affected.
Everyone is unique, and it’s each individual’s uniqueness that makes it hard to provide concrete explanations.
We can only hope that studies can provide some form of consilience where the evidence converges onto a theory, but we must always remember that the takeaways from most studies may just be, “it depends”.
So let’s try to embrace ambiguity, and embrace complexity as well.
We are all complicated beings, so let’s try not to assume that this wouldn’t be the case for science as well.
If you enjoyed this post and other works please consider supporting me through a paid Substack subscription or through my Ko-fi. Any bit helps, and it encourages independent creators and journalists outside the mainstream.
CLIA is an acronym for Clinical Laboratory Improvement Amendments, which provides regulatory guidelines and information on how clinical labs should run.
I will argue that lack of in-depth demographic data could be a consequence of having participants remain anonymous. Given that these participants came from a small non-profit, a report that states that Sub X was a Hispanic woman in her 40s who got COVID in July 2020 may point to “Lisa” from accounting (it’s always someone from accounting!). So this could be a consequence of recruitment wanting to keep participants anonymous.
I didn’t include it here, but the researchers titled most of their sections with the word “heterogenous”.
Thank you, this seems like an important start. 👍🏽😊