more covid 19: vitamin D, helper T cells, testing

I’m continuing with my gleanings from the Medcram Covid-19 updates presented by Dr Roger Seheult, though I’m not up to date with them, because they’re quite comprehensive and nuanced, and I want that detail more than anything. I’m also reading the book Outbreaks and epidemics: battling infection from measles to coronavirus, by Meera Senthilingam, which among other things, highlights the importance of preparedness, co-ordination and resourcing to deal with new and unexpected pathogens but also upsurges and cross-border spread of diseases we haven’t sufficiently dealt with in the past. As we hurtle at an unprecedented rate towards a number of vaccines against SARS-CoV2, for example, we may have to deal firmly, on a governmental level, with the anti-vaccination movement and its disinformation campaigns, but we also have to deal with grossly uneven levels of healthcare within and across nations. This current pandemic has been revelatory, for all but those on the front lines, of the variable impact such outbreaks have on the different levels of empowerment within societies. To take a stark example, Boris Johnson, the British Prime Minister, very likely owes his life to the fact that he is the British Prime Minister. Had he been a fifty-something person of colour living in Dagenham (or most anywhere outside of a UK city), his Covid-19 case would surely have turned out quite differently.
Update 74 is quite brief and mainly touches on vitamin D, the ‘sunlight’ vitamin, also obtained from foods such as fish, especially salmon and tuna, and egg yolks, and mushrooms raised using UV light – but mostly from the sun’s UV. Vitamin D enhances bone and muscle strength and function. A Lancet article is discussed, which correlates ‘vitamin D status’, presumably meaning bodily levels, with Covid-19 mortality. Some surprises in the data – vitamin D deficiency was common in ‘sunny’ Italy and Spain, but less of a problem in Nordic countries, perhaps due to a high vitamin D diet. Deficiencies were greater in poorer regions and in black communities, as of course were higher Covid-19 mortalities. in fact, ‘black people in England and Wales are 4 times more likely to die from Covid-19 than white people’ according to the UK’s Office for National Statistics.
The Lancet article referred to points out two aspects of vitamin D’s possible protection against Covid-19. First, it ‘supports production of antimicrobial peptides in the respiratory epithelium’, which sounds positive, and second it may help to reduce the inflammatory response to the virus because it’s known to interact with and promote the ACE-2 protein, which the virus suppresses. Other articles emphasise the benefits, with no attendant harm, of vitamin D supplements, particularly for the elderly. There have been no systemised, detailed trials as yet relating vitamin D levels to Covid-19 outcomes, but it seems like a no-brainer.
Update 75 continues the argument about SARS-Cov2 attacking the lining of the blood vessels, i.e. the endothelium, with the resultant effect on von Willibrand factor. This happens in the lungs as well as the vascular system, creating clots as well as the growth of new blood vessels as a type of immune response. This essentially marks it out from any kind of influenza. The New England Journal of Medicine has an article, published late May, looking exactly at these differences in the autopsies of Covid-19 victims – endothelialitis (inflammation of the endothelium) and angiogenesis (the formation of new blood vessels). They compared Covid-19 lungs with the lungs of ARDS (acute respiratory distress syndrome) victims, associated with influenza A (H1N1), and with uninfected lungs. They found ‘alveolar capillary microthrombi’ – often difficult to detect with scans – in the Covid-19 lungs at nine times the level of the influenza lungs, and new vessel growth at almost three times that of the influenza lungs. Clearly the new vessel growth is caused by the blockages, and the need to circulate around them. Microscopic analysis shows lymphocytes infiltrating the lungs, adding to inflammation, stiffness and tissue damage. The clotting prevents oxygen being picked up from the alveolar space, leading to low oxygen saturation of the blood. Scanning electron micrographs of the lung endothelium revealed viral particles in the extracellular space, suggesting strongly that the virus itself, and not simply the immune response to it (perivascular inflammation) is causing damage. Dr Seheult brings up NAC again here, as a possible disruptor of the cascade of events, especially in the suppression of superoxide and in the cleaving of disulphide bonds in VWF.
An article in Science, which refers to the adaptive immune system, is next discussed. The adaptive immune system, as opposed to the innate immune system, is a system that creates a memory of a pathogen in order to develop an enhanced response, a system exploited by vaccines. This system includes T cells, of which there are three types, memory, cytotoxic and helper. These cells are apparently involved in lifelong immunity. Vaccine researchers are concerned to create antibodies as protection against the virus, but T cells are also important in this regard, and researchers have found that many infected patients, and non-infected people, do have T cells that attack the virus, probably because they have been infected with other coronaviruses that share proteins, such as the spike protein, with SARS-CoV2. Researchers in fact found that Covid-19 patients all harboured helper T cells that recognised the SARS-CoV2 spike protein, and other SARS-CoV2 proteins, again suggesting the possibility/probability of lifelong immunity. Many others harboured the same helper T cells, which may be protecting them against the worst Covid-19 symptoms, before the fact. This is possibly a very important, and highly explanatory finding. Or maybe not. T cells are long-lasting, so these findings are certainly positive.
Update 76 starts with antibodies, and it’s a bit difficult to follow. It looks at the CDC’s ‘interim guidelines for Covid-19 antibody testing’, and a CNN health article summerizes it thus:
The CDC explains why testing can be wrong so often. A lot has to do with how common the virus is in the population being tested. For example, in a population where the prevalence is 5%, a test with 90% sensitivity and 95% specificity will yield a positive predictive value of 49%. In other words, less than half of those testing positive will truly have antibodies’, the CDC said.
This is hard to follow, but 5% prevalence is fairly standard for this virus, at least at the outset. And so false positives are a problem. To be clear about testing – a person either has the disease or not. If you have it and you test positive, fine, that’s a true positive. If you have it and test negative, that’s a false-negative. If you don’t have it and you test positive, that’s a false-positive. If you don’t have it and test negative, that’s a true negative.
So we can look at percentages and maths, and I’m following Seheult strictly here. So imagine we’ve tested 2100 people in a particular region – that’s everyone in the region. At this stage the disease has a prevalence of 5%, so about 100 out of 2100 have the disease (strictly speaking that’s 4.76%). The test has a sensitivity of 90% and specificity of 95% as above. 90% sensitivity means that the number of true positives from the test will be 90% of the number of those who actually have been infected by the virus. That means 90 people. 95% specificity is about those not infected. So you divide the true negatives by those uninfected to arrive at the 95%. The true negatives will amount to 1900. So 10 people will be false positive and 100 false negatives. When specificity rises, false positives decrease. When sensitivity increases, false negatives decrease. So with high sensitivity a negative result is more conclusive, and with high specificity, a positive result is more conclusive.
Imagine then that the prevalence of the infection has risen to 52% in the same population of 2100. That gives us 1094 with the disease, 1006 without. With the same values for sensitivity and specificity of testing, you’ll have 985 true positives and 50 false positives, and 956 true negatives and 109 false negatives. What you need to know with these results is how things stand for patient x, the person you’re dealing with. This means you need to know the predictive values, positive (PPV) or negative (NPV). This requires some simple maths. Given a positive test result, what chance is there of x having the disease? Or vice versa for a negative result. This means that for the PPV you divide the true positives by the total number of positives, and the same process applies for NPV. Going back to the situation where the prevalence was 5% we get a PPV of 47% and a NPV of 99%. What this means is that when the prevalence is low, the negative predictive value is much higher than the positive predictive value. The implication is important. It’s just not clear at this stage whether you have antibodies against the virus. So you need to raise the specificity of the test, especially if the virus or pathogen has a low prevalence. But looking at the 52% prevalence case, and using the same simple maths we find that the PPV is up at 95% and the NPV goes down to 90%. Prevalence, then, is the main determinant of predictive values.
For testing, this means, just as the disease is becoming prevalent, that’s to say, as it’s just being detected, you need a test with a very high specificity (admittedly a big ask) and/or you need to test those with a high probability, based on current knowledge, of being infected, and those in contact with them.
References
Coronavirus Pandemic Update 74: Vitamin D & COVID 19; Academic Censorship
Coronavirus Pandemic Update 75: COVID-19 Lung Autopsies – New Data
Coronavirus Pandemic Update 76: Antibody Testing False Positives in COVID-19
Leave a Reply