Long Term Side Effects of mRNA Vaccines Unlikely, Part 3

Now we are moving on to consider possible long-term side effects from the genetic payload itself, the mRNA inside the lipid nanoparticle.

Purity of the genetic payload

The first question to ask about the safety of the genetic payload is, how sure can we be sure that the payload is what is intended and not defective or mixed with dangerous impurities?  Some concerns people have had in the past about vaccines include mercury impurities or worries about protein remnants from chicken eggs in which some vaccines are grown.  So it's worth looking at the manufacturing process for mRNA.  (A lot of the information from this section I got from these two videos: here, and here.  They are long watches but pretty informative.)

The manufacturing process

All living beings create messenger RNA from DNA as part of the regular life of the cell.  In the nucleus of the cell, you have DNA, free floating genetic "building block" material (the four basic molecules from which all RNA is built, the "nucleotides"), and an enzyme (called "RNA polymerase") that reads (transcribes) the DNA and builds mRNA from the nucleotides based on what it reads.

The manufacturing process mimics that setup.  First, you create a custom strand of DNA (your "template") that contains the genetic instructions for the protein you want.  Then you put that in a tub along with a soup of nucleotides and add a bunch of those enzymes (RNA polymerase) and some salts.  The enzymes will do their thing, read your DNA and convert all the nucleotides around them into the RNA you want.

The product of this is a tub full of some template DNA, some enzymes, some leftover nucleotides, and a bunch of RNA, some of it complete and some of it failed or partial.  The major challenge is then to strain and/or chemically separate the complete RNA from all the other stuff in the tub.  You then quality-control the result to ensure you have achieved sufficient purity of the final product.  Most of the manufacturing progress in this area has to do with clever ways of getting higher yields from the straining / purification part of the process.

Commoditization

A very important character of this process is that it relies entirely on commercial, off-the-shelf equipment and ingredients.  Creating small custom pieces of DNA is routinely done nowadays via various processes and any big genetics lab can do it.  You can buy both nucleotides and RNA polymerase online if you want to: here's a site where you can buy nucleotides and  here's a site where you can buy RNA polymerase.  Equipment for doing RNA transcription, including the filtering, are commercially available.  Hardware and software for doing the purity checks are industry standard.

This has very important safety implications, because it means that the manufacturing process for mRNA is not new, even though the specific mRNA for the vaccines is new.  Ways to guarantee sufficient purities of the end product have already been worked out by the industry.  Safety problems that can come about because of surprises in a new manufacturing process are therefore eliminated. 

This was a major time saver for the mRNA vaccine approval process as well.  Moderna, for example, when asked by the FDA about their manufacturing process, was able to say "it's mRNA in lipid nanoparticles: the same equipment and process as they did for Patirisan, which you already fully reviewed and approved".  This got them a quick "thumbs up" from the FDA on the manufacturing front and saved a lot of time.

Chemical vs. biological

Another important part of this manufacturing process is that it is entirely chemical.  It relies on no living organisms, nor on any cells from living organisms.  This means that you can theoretically crank out product much faster than vaccines that rely on growth cycles of living cells in "bioreactors".  It also eliminates a whole class of potential health problems with proteins from host organisms getting into the vaccine.

In conclusion, we can give the manufacturing process a "pass" in terms of innate safety.  It is tried-and-true in a way in which the specifics of the genetic payload itself is not, and it is not any more likely to have weird or new contaminants in it than any previous genetic therapeutic has been.  Furthermore, its chemical nature precludes a whole class of potential contamination to which older vaccines might have been susceptible.  

So now let's move on to considering the mRNA itself, and what possible long-term safety consequences it might have.  Theoretically, we can consider the mRNA in three stages: before it enters the cell, after it enters the cell, and the protein that is generated from it by the cell.

mRNA in the intercellular region

While the mRNA is packages in a lipid nanosphere to keep it from interacting with the body before it gets into a cell, what if some of the mRNA is unintentionally released into the intercellular region?  You can assume that at least a certain amount of this will happen, because nothing is perfect and some nanoparticles *will* break down before being absorbed by cells as designed.

I originally had a much larger argument planned for this section with multiple ways of showing that random mRNA fragments in the body aren't a problem, bringing in things like how the body handles natural cell replication failures and cellular disintegration being a normal occurrence.  Those arguments are all quite valid, but I think a simpler, common-sense argument is good enough: we can know that mRNA fragments aren't a problem in the intercellular region because we've been trying to get mRNA to do useful things for us for decades and we've found that they're so fragile that we *need* to use lipid nanospheres to get them to do anything.  The challenge of getting mRNA to survive the intercellular region has been described in this article: 

The complexity of the problem is enormous. Naked RNA or DNA molecules are rapidly degraded in biological fluids, do not accumulate in target tissues following systemic administration, and cannot penetrate into target cells even if they get to the target tissue. Further, the immune system is exquisitely designed to recognize and destroy vectors containing genetic information

Even using lipid nanoparticles, the mRNA vaccines have to be treated very carefully or the mRNA can be degraded just by the temperature.  It might not actually need the deep freezing that the vaccine as a whole needs (I think that's more due to the lipid nanoparticles than the mRNA), but nonetheless, it's not super-sturdy stuff.

There is only danger to the mRNA strands from the intercellular region, not the other way round, and this is true even on a minutes-to-hours time scale, let alone in the long-term.

mRNA in the cell

Some people have been very nervous about the vaccines because they hear it contains mRNA, which the think will modify the DNA in their cells, which they associate I think with comic-book-style mutations?  I'm not going to talk about this concern much, since other people are debunking it ably.  Just the fact that mRNA vaccines are injecting genetic material into your cells is not a rational cause for fear, any more than the fact that the common cold does the same sort of thing is a rational cause for fear.  I talk about this sort of irrational fear a bit in an earlier post, so I'm not going to say anything more about it now.

The spike protein

So, now for the end-product of the mRNA once it is produced by the cells: the spike protein.  Could it cause some long-term issues?  Let's run it through the same mechanisms we laid out in the previous post:

• Material accumulation: no.  We know from PCR tests of Covid-19 infected individuals (which are extremely sensitive and can detect even fragments of individual viral particles) that the body eventually eliminates all the spike particles from the system.  And this is for people who have system-wide infections, some of them lasting months, with the virus replicating all throughout the body and pumping out spike-covered viruses the whole time.  Once the infection is beaten, it only takes a matter of days or at most weeks before all of the particles are cleared.  A two-time shot in the arm will also be completely cleared.
• Persistence by self-replicating colonies: impossible.
• Damage to non-regenerative tissue: no.  All the same arguments that applied to lipid nanoparticles apply to the protein as well.
• Cumulative effect: no.  Same arguments apply.
• Interaction with adaptive or "learning" body mechanisms: no.  Again, the same arguments apply.

That again leaves the immunological response, which *is* a valid manner in which the spike protein could cause a long-term effect.  So now we are going to deal with that question.

Immunological response

Lipid nanoparticles

I said I would return to the immunological response to lipid nanoparticles in the previous post.  Here I will point out that the immunological response to foreign objects in the body relies on detection that those objects are foreign, and depends solely on the surface of the object in question.  The body is not able to probe into an object for foreign DNA, it must touch the surface of the thing.

For lipid nanoparticles, the surface of the particle is the PEG coating.  This is why the nanoparticles are coated in PEG in the first place, because it reduces immunogenic response.  Because of this, the surface of mRNA vaccines should look identical to all of the older liposome injections which were also coated in PEG, from the standpoint of the body's immune response.

This is not to say that there won't be a reaction--there will be some reaction.  However, what the immune system will be reacting to is exactly the same thing it reacts to in much older medicines: a tiny round particle coated in PEG.  We should therefore expect that the long-term possible side effects of the immune response to the lipid nanoparticles to be no different than the long-term side effects of the immune response to those older liposomes--which is to say, nothing.

The spike protein

The spike protein, on the other hand, is a new protein and we do not have long-term data on the immune reaction to it.  One concern that has been raised is, suppose the spike protein is similar in some ways to good, human proteins elsewhere in the body.  When the immune system develops a memory of how to neutralize the spike protein, couldn't it sometimes accidentally generate an antibody that attacks these human proteins as well?  In other words, couldn't the vaccine cause some sort of auto-immune condition to develop?

The answer to this is, possibly.  Auto-immune disorders of various kinds are certainly possible rare side-effects of all vaccines.  Now, I do not personally know (and I'm not sure if scientists do either) whether the mechanism which triggers an auto-immune disorder is what I have just described: bad antibody generation because of *similarity* between the antigen and some human protein.  It's possible that what happens is more of a simple mistake: the immune system gets excited by the presence of an invader, and then during the course of the immune system arousal some over-excited antigen presenting cell grabs a human protein of some sort and presents it to a T-cell.  But whatever the mechanism, we do know that this sort of thing does happen some times.

The long term?

However, knowing that this sort of thing could possibly happen isn't enough.  We know that this doesn't happen at any meaningful frequency in the short-term because of the safety data and adverse side effect tracking.  So for there to be a long-term risk of this, it has to be possible for an auto-immune disorder to arise after a vaccine is given, but not in the first year.

This does not match with how the immune system works.  In order for the immune system to build a long-term memory of some antigen which will trigger the multiplication and release of specific antibodies, it needs those antigen particles to be physically present.  Special cells grab bits of the offending material and present them to other cells, and this is the origin point of the immune "learning" process.  It doesn't proceed if there is no more offending material to present.  If the body's immune system is going to learn some incorrect antibody recipe to an antigen, it has to be while that antigen is present.  And we already said that the spike protein is completely eliminated from the body within days or weeks of the vaccine injection, so there is a short time-span in which this could happen.

Moreover, the nature of the human immune system response is to ramp up antibody production over the course of a few days to a week, keep up the antibodies while it sees that there is still an infection and for a while after that, and then spin down antibody production again to some background level.  We have studies on the timing of this surge and spin down with the new vaccines; here's an important study on this for Moderna.  You can see the antibodies spike shortly after the booster shot of the vaccine (about 5 weeks after the first shot) and begin steadily declining after that.

Since this is the case, the time of maximum danger from antibodies produced by a vaccine is 4-6 weeks after the injection.  If any auto-immune problems are going to be caused by vaccine-generated antibodies, they are for sure going to show up no later than that period.

Vaccine safety timeline

Because issues of auto-immune disorders or other problems stemming from the immunological response are common concerns for all vaccines, this question of timing of possible problems has been considered for a long time.  The history of all failed vaccines has been examined, and on that basis it is accepted among vaccine experts that if there are going to be any long-term side effects from a vaccine, they will manifest within six weeks of administration.  There are no known exceptions to that timeline yet in the entire history of vaccines.  Based on that, the FDA required two full months of safety data from the new vaccines before considering them for emergency use authorization, and this *wasn't* considered an accelerated timeline for approval.  The real reason the approval is considered "emergency use" rather than a full approval is not because of lack of long-term safety data--in fact, the reason is that for full approval, the FDA normally asks for a vaccine to show efficacy for a longer period of time than what we know so far for the Covid vaccines (great interview that talks about this among other things here).  They are more concerned with the effects of the vaccine possibly waning over time than they are about it possibly gaining more negative effects.

Conclusion

The process by which mRNA is mass produced is not new, but rather a known and well controlled procedure which we can trust.  The mRNA itself is a safe material which should not be expected to do anything unless it gets into the cytoplasm of the cell.  The spike proteins generated will cause an immune reaction, and the effects of this immune reaction will fully manifest by at least six weeks after injection.  We therefore have very good knowledge of what this reaction entails.  Rare long-term side effects are possible, but at this point we know that whatever auto-immune reactions do happen to some people are going to be extremely rare, at about the same frequency as they happen with ordinary vaccines that have been with us for decades at the most.

Long Term Side Effects of mRNA Vaccines Unlikely, Part 2

What are the possible long-term side effects of lipid nanoparticles? 

Lipid nanoparticles are not new technology invented for the mRNA vaccines; there is history and safety data for them.  Therefore I will start with what is already known about long-term effects of lipid nanoparticles, and then move on to consider whether the constituent materials of lipid nanoparticles even allow for any sort of long-term side effect.

History of lipid nanoparticles

Liposomes

According to Lipid Nanoparticles: Production, Characterization and Stability by Shah et al., the idea to surround fragile or reactive drugs in a lipid layer in order to target delivery to specific parts of the body was first theorized in the early 20th century (Shah et al., 4).  It was not until 1961 that this concept was first put into practice, with the development of the liposome:  https://en.wikipedia.org/wiki/Liposome.  The liposome was really the first medicinal nanoparticle: it was a lipid sphere containing an aqueous solution in its center in which a desired drug was dissolved.  Its purpose is to either protect a fragile drug from the body until the drug reaches its desired destination OR protect the body from a dangerous drug until the drug reaches some target for destruction (chemotherapy is the biggest use case here).

Since its introduction in the '60s, the liposome has been used for many types of drugs.  It is the most common vehicle used for any sort of targeted drug delivery, and according to this review of the state of liposomes written in 2017, it is used as a drug delivery mechanism in the areas of analgesics, fungal disease medicine, cancer therapy, viral vaccines and something called "photodynamic therapy".

The fundamental ingredients used for lipid-assisted drug delivery have been around for many decades, even if not in exactly the same form as modern lipid nanoparticles used to deliver genetic material. This should put a severe limit on the doubts as to the safety of modern forms. Some people, for example, have raised allergy concerns with the substance called "polyethylene glycol" (PEG) with which the lipid nanoparticles of the mRNA vaccines are coated.  But this coating has been used on liposomes and other injected drugs since the '70s.  This is not to say that there are no allergy concerns with the substance, but what this does mean is that we have ample long-term safety data on PEG and so far no known long-term effects have been found.  It is not a valid cause for concern about long-term side effects.

Lipid Nanoparticles

If that is the case, what is new about lipid nanoparticles, and why were they developed separately from liposomes?  The key thing that needed to be changed was the aqueous center.  As I mentioned before, RNA is a fragile thing.  One of the fragile things about both RNA and DNA is that they are water soluble;  you can't put genetic material in the center of a liposome, because it would just dissolve.

The solid lipid nanoparticle was thus created as a way of delivering fragile material like RNA that you don't want dissolved in water.  Instead of an aqueous center, it bundles up the drug into a solid lipid "matrix".  These are more challenging to manufacture than liposomes, but scientists have been working on them since the mid '80s, the first drugs commercially delivered by them came out in 1993 and they have been "greatly exploited ever since as drug carriers" (Shah et al., 4).

More recently, lipid nanoparticles have become the most popular option for various gene therapy drugs, like the mRNA vaccines.  This has been enabled by evolutions in the design of the lipids used that increase the stability of the particles and reduce the immune reaction to them, while also reducing the particles' "cytotoxicity" (which is the quality of being toxic to cells--an important thing to do if you want to keep the cell to which you deliver the genetic payload to remain intact and functional).  The primary breakthrough here had to do with neutralizing the surface charge of the lipid nanoparticles, which made them neutral with respect to body chemistry and hence non-reactive at the cellular level.

Note that while lipid nanoparticles have tweaked chemistry with respect to older liposomes, the long-term effects of the material should not have changed at all.  Cytotoxicity, which the older lipid formations could exhibit, is a short-term effect of nanoparticles with a particular surface charge.  Over time, all of these lipids are bio-degradable and will eventually be broken down by the body's organic solvents into the same component parts.  Here is a study from 2013 looking at a representative lipid and studying how rapidly it is biodegraded and eliminated from the system (answer: rapidly and safely, with the caveat that the study was on rats, though that shouldn't make a difference here).

Most of the projects using these newer lipid formations are still undergoing trials, but a few have seen full FDA approval.  Patisiran, for example, was FDA approved in 2018, and the first patients dosed with Patisiran and tracked ever since were recruited no later than 2013 (link to clinical trial here).  Here is a link to a list from 2018 of 22 genetic therapies using lipid nanoparticles constructed just like the lipid nanoparticles used for the mRNA vaccines.

Safety conclusion from history

In summary, the medical world has had 50 years of real-world (i.e., patients dosed) experience with lipid drug delivery systems, 20 years of experience more specifically with lipid nanoparticles, and at least 8 years of experience of exactly the type of lipid nanoparticles which are now used in the mRNA vaccines.  No long-term side effects resulting from lipid nanoparticles have ever been described.  Short-term side effects from older lipid formulations included some cytotoxicity, but newer formulations have substantially removed that side-effect, thus enabling a lot of gene therapy options which were not possible in the past.

Possible mechanisms for long-term side effects

Let's now spend some time thinking about how a medicine might have long-term side effects.  It's theoretically possible that lipid nanoparticles have long-term side effects that have been missed by scientists and doctors, so I'd like to go over the different mechanisms by which drugs can have long-term side effects.  Let's look at a number of real-life examples and see which ones might apply to lipid nanoparticles

Accumulation of material

Example: Lead poisoning.  Lead poisoning causes long-term side effects because lead is a material which the body can only eliminate very slowly.  Small amounts therefore accumulate inside the body and over time become enough to interfere with various biological processes.

Accumulation of material in the body which cannot be easily eliminated has been identified as a concern for some types of nanoparticles, namely gold nanoparticles.  However, lipid nanoparticles cannot possibly act in the same way.  They are all biodegradable.  They have different lifespans in the body depending on their exact composition, but none of them can last long enough to be a long-term concern.  After receiving a lipid nanoparticle injection, no more particles will be left in the body in a matter of days or weeks, let alone months or years.

Persistence by self-replicating colonies

Example: Chicken pox.  The chicken pox virus is able to cause Shingles many years after an initial infection because the virus replicates throughout the body and manages to establish dormant colonies in the nervous system.  At some point, through mechanisms which are unclear, these dormant colonies can re-awaken spread through neuronal tissues.

Lipid nanoparticle obviously cannot cause long-term side effects in this way.

Damage to non-regenerative tissue

Examples: Lyme Disease, Covid-19.  Lyme disease is a bacteria that spreads throughout your body.  If not caught quickly enough, it spreads through your nervous system as well and damages nerve tissue.  Nerve tissue doesn't really regenerate, so the side-effects of Lyme disease can be permanent.

Some cases of Covid-19 have also shown to cause some neuronal symptoms, but we don't yet know if it can cause permanent damage to nerve tissue in the way Lyme disease can.  However, we *do* know it can cause scarring to heart tissue.  Almost all bodily tissues regenerate to some degree or another, but some organs regenerate at a much slower rate than others do.  Cardiac tissue regenerates so slowly that it essentially doesn't regenerate at all, so scarring of the heart is a problem you can expect to affect you for the rest of your life.

If a drug were to cause tissue damage to one of these slowly-regenerating organs, it could cause problems in the long term when that organ is under more stress but not be obvious right at the time.

Which organs do we worry about here?  Primarily, these would be: heart, lung, brain and nervous system, and eyes (retinal cells do not regenerate).  There are other cell types which take a fairly long time to regenerate (from 8-10 years for some muscles and intestinal cells, for example), but they do regenerate--and we are looking for long-term effects that would not yet have shown up after the 8 or so years these specific lipid nanoparticles have been administered to test subjects, so we can discount any organs that would naturally self-heal in that time period.

So, can the lipid nanoparticles injected for these vaccines substantially damage the heart, lung, brain or nervous system, or the eyes?  I think the answer to this is pretty clearly, "no".  The course of lipid nanoparticles through the body is well understood.  Here is a study that looked at the safety of lipid nanoparticles delivering a particular genetic payload on rats and monkeys.  They dissected all of the major organs, looking for any evidence of toxicity or organ damage.  The only organ they found to be affected by the lipid nanoparticles directly was the liver, where there was some slight oxidative stress.  Oxidative stress is what damages your liver when you drink alcohol; it's reversible because the liver self-regenerates quickly, and you have to have quite a lot of it all at once to give yourself permanent liver damage.

More specifically in the case of the mRNA vaccines, the course of the vaccine through the body is very predictable.  It's injected into the upper arm muscle, where it's initially trapped by the regular matrix of muscle tissue.  It slowly drains into the nearby lymph nodes, and what is not metabolized at a cellular level is disposed of by the normal waste system: i.e., filtered out by the kidney and liver and then excreted.  So we can eliminate damage to vital organs as a possible way of causing long-term side effects for lipid nanoparticles.

Cumulative effect

Example: carcinogens.  Some substances have a deleterious effect if they are used for a very long period of time.  For example, some materials appear to slightly promote the mutation of human cells, thus increasing the chance that a dangerous mutation will come about in the form of cancer.  In every case of material of this sort that I am aware of, however, a long period of time of exposure to the material is necessary to statistically increase the chance of cancer.  Birth control, for example, may well increase the chance of breast cancer, but that is something that is taken day-in and day-out, constantly, for many years.  Likewise, smoking is something that is done habitually and often, usually over the course of many years.

The mRNA vaccines are given in two doses only.  The lipid nanoparticles exist within the body for a small number of days, twice in your lifetime, and then that's it.  Thus they are not susceptible to this form of unexpected long-term side effect, where each use increases the odds of some negative outcome by some tiny amount, which grows over the course of cumulative uses to be a significant risk.

Interaction with adaptive or "learning" body mechanisms

Example: addictive drugs (oxycodone, et al.).  Certain mechanisms in the body adapt to different circumstances and can thus "learn" a behavior at some biological level.  Once they learn a certain behavior, this behavior can become very difficult to unlearn, thus becoming a long-term side effect.  Pleasure receptors in the brain can become accustomed to the euphoria of a drug-induced high, making it difficult to find pleasure in any normal activity again.  Any strong, mind-altering chemical has the capability to have this sort of side-effect, as the mind is the pre-eminent learning organ.

The endocrine system can probably also be counted as a potential source for long-term side effects of this sort.  While the endocrine system isn't a "learning" system per se, it is a complex interrelation of hormones which is designed to be adaptive to various situations.  It's possible an imbalance among the interrelation of hormones here might cause some long-running, self-perpetuating problem--though I have not actually been able to come up with a real-life example of this sort of thing being the *source* of long term issues yet.

Lipid nanoparticles can't really cause issues with either of these systems, however.  With both brain-driven processes and endocrine processes, drugs that can cause serious side-effects through interaction with them are very chemically specific.  Pain and pleasure receptors require very specific chemicals to activate; hormones stimulate bodily functions using very specific chemical signatures.  Microscopic balls of lipid are not going to interact with any of these systems in the specific ways necessary to trigger them.  And even if they did, I'm not aware of any dangerous interaction with these volatile systems that is not immediately obvious.  You can become addicted to a painkiller because of the power of its effects on your brain and nervous system, but likewise because of that power, you can't become addicted without noticing it.  If there is a long-term side-effect to lipid nanoparticles, however, it has to be something that happens without anyone initially noticing it.  So we can eliminate brain and endocrine problems as possibilities.

There remains, however, one "learning" body mechanism which *can* interact with lipid nanoparticles, and that is:

The immune system

Now we have arrived at the only way in which I think it is legitimately possible that lipid nanoparticles can have a long-term side effect on the human body.  While lipid nanoparticles are intentionally designed and manufactured in such a way as to have as little interaction with the human immune system as possible, and it is know that they are relatively effective at hiding from the immune system in order to be able to deliver their genetic payload to cells, nevertheless we also know that they aren't 100% effective at this.  There is some slight immune reaction purely to the lipid nanoparticles when injected into the body.

And this makes complete sense: it is the job of the immune system to find and eliminate foreign microscopic particles, and lipid nanoparticles are certainly that.  Furthermore, anything that the immune system reacts to, carries with it the danger of a long-term side effect because the immune system has a memory system: for certain pathogens, it learns about the pathogen and makes specific antibodies which it releases to counteract that pathogen, and it can retain this memory for life.

The immune system, therefore, is a mechanism by which any vaccine--including one based on lipid nanoparticles--could plausibly cause a long-term side effect.

Is it likely, however, that lipid nanoparticles do sometimes cause such a long-term side effect?  I'm actually going to leave that question till the next part, because the immune response is a factor for both the lipid nanoparticle and the genetic payload, and it's a very important thing to consider.  So for just right now, I'm going to leave this question open.

Conclusion on plausible mechanisms for long-term side effects

The conclusion, after considering all mechanisms that I can think of which could create a long-term side effect in any way at all, is that the only *real* possibility here is some interaction with the immune system.  No other plausible mechanism exists by which microscopic bundles of biodegradable, chemically inert material which are broken down and eliminated from the body quickly could cause any sort of long-term problem, which was not also immediately evident in the short term.

Long Term Side Effects of mRNA Vaccines Unlikely, Part 1

Some people are concerned about the possible side effects of the new Covid-19 vaccines.  With tens of millions of doses of these vaccines administered, we now have definitive evidence that these vaccines are quite safe--in the short term.  But what about possible long term side effects?

While it's not true that the development of these vaccines was "rushed" (a topic for another post), nevertheless it is true that the first people ever injected with these specific mRNA vaccines were Phase 1 trial participants on March 16th, 2020 (https://www.cidrap.umn.edu/news-perspective/2020/07/hopeful-results-phase-1-moderna-covid-vaccine-trial).  Therefore the longest running data we have on the effects of these vaccines is only about 11 months old only; we can't definitively prove on the that basis that there won't be side effects that only surface after a year or more of taking the vaccine.

However, in the case of the mRNA vaccines, the medicine in question is quite straightforward and specific in its action.  It is a very simple compound of two parts: an RNA genetic payload, surrounded by a lipid nanoparticle sphere.  This simplicity makes it reasonable to attempt to think through all the side effects that the two components of this vaccine might possibly produce.  What do we know about the possible long-term effects of the components of this vaccine?  Can we think of a mechanism by which either might produce some long-term effect?  If what we know about the components of the vaccine tends to rule out long-term side effects, and if we cannot come up with reasonable mechanisms by which the vaccine could possibly cause long-term side effects, we might then conclude that the vaccine is very unlikely to produce long-term side effects.

As it turns out, there are good reasons to predict that no long-term side effects will manifest from these mRNA vaccines, which I will begin presenting now. I plan on breaking this up into four parts: 

1. An overview of how the mRNA vaccines work (compared to the virus against which they inoculate).
2. What are the possible long-term side effects of the lipid nanoparticle.
3. What are the possible long-term side effects of the genetic payload.
4. Summary of possible risks and comparison to risks of infection.

How the mRNA Covid Vaccines Work

It's useful to compare the working of the mRNA vaccines side-by-side with the workings of the coronavirus itself, because in many ways the vaccine is a simplified, mechanical version of the virus.  Both are genetic material (specifically RNA) delivery mechanisms.

Envelope and Payload

Both the Covid virus and the mRNA vaccines are composed of small particles: roughly the same size of about 1/10th the diameter of the typical human cell.  The virus is composed of an outer membrane layer that's studded with spiky protrusions (the "spike protein") and contains a large single strand of RNA inside which holds the genetic instructions for making copies of the complete virus.  

The particles inside the mRNA vaccine are just super-tiny spheres of lipid with a coating (a "surfactant") that keeps the sphere of lipid together.  Instead of containing a single continuous strand of RNA with instructions for making the whole virus, the vaccine particle contains multiple much smaller strands of RNA which are each *part* of the virus RNA--specifically, the part that contains the instructions for making the spike protein (about 13% of the total genetic code of the virus, specifically from about the 21563th nucleobase in the chain to the 25384th, see full genetic sequence of virus here).

That the two structures are very similar is due to a simple fact that both structures have to deal with: RNA, being single-stranded, is much more fragile than double-stranded DNA.  Both particles are aiming to deliver their RNA payload to human cells, but the inter-cellular space in the human body is hostile to the survival of free-floating RNA.  Without a more durable delivery particle, the RNA would disintegrate quickly rather than penetrate into any cells.  So both systems implement an envelope-and-payload structure to get the job done.

Cell Infiltration and Replication

The Covid virus infiltrates the human cell using the spike protein.  This binds to a receptor on the cell surface (the "ACE2" receptor) and allows the virus to open up the cell wall and inject its payload RNA.  This RNA is then taken up by the cell's protein manufacturing process and whole new copies of the virus are created and "bud" out of the infected cell.

Since what is produced in the end here is more fully functional copies of the virus, this process repeats over and over again as more cells are infected and replicate the virus.  The infection spreads throughout the body, able to multiply wherever there are cells that have ACE2 receptors.

Lipid nanoparticles do not require a spike protein.  Instead, they merge with the cell naturally because of compatible chemistry between the lipid particle and the lipid membrane of the cell.  Once inside the cell, the lipid structure dissolves in the higher pH setting and releases the mRNA particles into the protein manufacturing structures of the cell.  The cell then proceeds to manufacture only the spike proteins.

Because only the code for the spike protein is included in the nanoparticle, this process continues only until all of the nanoparticles that were injected into the muscle have either merged with a cell and produced spike proteins, or broken down in the intercellular region.

Immune System Response

In the case of both the virus and the nanoparticles, the immune response is roughly the same.  The spike protein is identified as a foreign invader of the body and the standard immune response is invoked against it.  Importantly, the body also learns how to produce a neutralizing antibody: a protein shaped just right so that it latches on to the spike protein.  Once the body's immune system has learned how to do that, whenever it senses that same spike protein again, it will ramp up production of these antibodies.  The viral particles will then pick up these antibodies all over on their spike proteins, which will then no longer able to fit into the cell wall receptacles.  The virus will therefore lose its ability to penetrate cells and reproduce, and thus the infection is halted before it can progress very far.

The New Vaccines, DNA "damage", and Infertility

With the previous discussion on basic vaccine mechanics under our belts, we can now consider what sorts of dangers these vaccines might entail.

Can a vaccine cause a systemic problem in the body because of DNA damage of some sort?

The oldest type of vaccine, which is simply a milder disease, replicates throughout the body and--as viruses do--modifies and "damages" DNA in cells throughout the body as it does so.  So, there is some danger of systemic damage throughout the body.  But DNA damaged cells are disposed of throughout the body all the time, so this isn't usually a big deal.  Cells die and are disposed of naturally; even mild diseases like the cold "damage" DNA and cause more cells to die and be disposed of all the time and no one thinks twice about it.

More modern vaccines are not a milder disease, but rather an attenuated form of the disease: neutered virus particles.  It is *intended*  that the effects of these virus would be limited to a smallish number of cells because the virus replication ability has been (hopefully) mostly destroyed by whatever the neutralizing procedure was (UV lighting or something else).  How well you can trust that only a limited number of cells will be effected depends on how well you trust this attenuation process.

Adenovirus vector vaccines (AstraZeneca and Johnson & Johnson, among others) certainly do "damage" the DNA of cells at the injection site, by design.  They are limited to a set number of cells which they effect because their replication ability has been genetically removed.  How well you can trust that only a limited number of cells will be effected depends on how well you trust this genetic alteration.

Finally, the mRNA vaccines (Pfizer and Moderna) don't affect DNA at all.  They cause the production of spike proteins without needing virus particles at all; they deliver mRNA to the cell via mini lipid drops rather than via a virus of some sort.  mRNA vaccines are therefore absolutely incapable of reproduction, no matter what the circumstance.

Bottom line: new vaccines are less likely to cause systemic problems than older vaccines, because they are more precisely targeted to specific effects.  None of them, though, have a built-in mechanism that would cause a systemic problem throughout the body, despite some of them working via DNA.

How could a vaccine cause problems with infertility?

DNA "damage" is therefore not an issue, but that doesn't mean there is no mechanic by which a vaccine might cause some body-wide problem.  Almost all plausible mechanics of that sort are related to the immune response to the vaccine, however, not the vaccine itself (which only exists in the body for a short time).    Here's how something like infertility could potentially be caused by a vaccine:

As mentioned, the body's immune systems learns how to produce virus-specific neutralizing antibodies.  As we have also seen, sometimes these antibodies are not actually specific to a single virus--and this is sometimes the point.  The cowpox antibodies happen to match the smallpox viruses as well, which is why cowpox infection ends up in smallpox immunity.

The danger then exists that the body will produce antibodies in response to a virus, or in response to a vaccine, that will attack something good as well as the virus that it is intended to attack.  This is the mechanism by which it has been theorized that a vaccine could interfere with fertility.  There is a protein called "syncytin-1" that is used in the placenta, and it happens to have small stretches of genetic code that are similar to parts of the code for the spike protein of a coronavirus.  What if this similarity is enough that the antibodies produced against the spike protein are also able to attack syncytin-1?  This would, in effect, create an auto-immune response to the placenta and prevent or imperil pregnancy.

While not impossible, this scenario is very unlikely, for several reasons:

1. The amount of similarity between the proteins is actually quite small.  I know that scientists nowadays are actually pretty good at determining how likely it is that you can get cross-reactions between different proteins based on genetic structure.  I'm not aware of an actual study in which the similarities between syncytin-1 and the spike protein were computer-analyzed to estimate *how* similar they would be, but if you hear scientists say that the similarity is too small to matter, they probably have good reason for saying that.
2. Pfizer actually has some data on pregnancy from its phase 3 trial data.  No trial participants were recruited who were pregnant, but some became pregnant in the course of the trial.  Pfizer had 12 people who took the vaccine end up getting pregnant and 11 people who took the placebo end up getting pregnant.  This is a good indication that no effect on fertility is to be expected from the antibodies.
3. If the vaccine caused infertility, it would be because of the immune response and not the vaccine per se.  This means that recovering from Covid-19 itself would also cause infertility, because it would be the immune system's antibodies that would attack syncytin-1.  Worldwide, there have been about 95 million reported cases of Covid-19.  We're pretty sure that this is a large understatement of the total number of cases, and that there are a lot of people who had mild or no symptoms who are not accounted for by that number.  These people would also have generated these neutralizing antibodies.  So there should be something like 200 million people out there at this point who have neutralizing antibodies to the Covid-19 spike protein.
   
   If that many people were out there with antibodies that cause a fertility problem, we should have already started seeing some of these problems materialize by now.  No greatly increased reports of fertility problems since the widespread proliferation of Covid-19 have, however, been forthcoming.
4. Covid-19 isn't the only virus that has a spike protein.  Many common colds / flus are caused by coronaviruses that have very similar spike proteins.  Presumably, these virus spike proteins *also* are similar to syncytin-1, and yet there have been no indications over the years that people who have suffered from a coronavirus of one sort or another have had any sort of fertility problems more than other people.

Therefore, it is extremely unlikely that the new vaccines cause any sort of fertility problem.

How Various Vaccines Work--Simplified

I was asked recently about whether the new vaccines could cause some unexpected trouble because they "modify our DNA", as opposed to previous vaccines.  To "modify our DNA" certain *sounds* kind of risky, I agree.  There was also some concern about these vaccines being able to damage the reproductive system and cause infertility.

The short answer is that this isn't a concern, but as I gave the answer, I realized there might be some important misconceptions behind this concern that could be cleared up by a longer answer.  So I'm going to give a very simplified (and therefore necessarily incorrect) explanation of some basic biology and how it relates to viruses and vaccines.

I'm going to follow this up immediately with an evaluation of the inherent risks of some of these different vaccine types and address the infertility question.

DNA, mRNA, and Proteins

DNA

Everyone knows that DNA is in some way the "master plan" for the whole body: it's the set of instructions that makes each person unique.  And I think it is a partial understanding of this that makes people very nervous when they hear about something that messes with their DNA.  I think they imagine their DNA as some single master plan that controls their entire body, and therefore anything that messes with their DNA has the potential to mess up literally anything about their body.

But there is no single "master" copy of DNA--that's not how it works.  Your DNA is replicated billions of time in cells all throughout your body.  There is nothing that can change your DNA throughout your whole body; it only exists one cell at a time and if you want to change it you have to change it one cell at a time.

Furthermore, aside from the very beginning of life when you exist as a very small set of cells, there's never actually just one version of your DNA.  DNA mutations happen all the time at the time of cell replication.  A mutation is usually just an error in cell replication--an incorrect copy.  And since these happen all the time, it's certain that you have many slightly different versions of your DNA in different cells all throughout your body.

The cellular replication process does have several error correction procedures it goes through.  For serious problems, badly messed up DNA or malformed cells are destroyed.  *Slightly* modified DNA can get through this all the time though.  Cancer exists in the grey area of this process, being mutated sufficiently enough to cause problems but *not* enough to be attacked and disposed of by the body's natural mutation clean-up system.

The bottom line is, something altering your DNA isn't the equivalent of mutating you into a different creature.  It's something that happens on the level of individual cells and the impact depends on how successfully the new cells replicate and how successful the body's immune system is at disposing of these mutations.

DNA, mRNA, and the production of proteins

Let's do a quick review on how the microscopic structures of the body are built from DNA.

Imagine DNA as a large book written in Braille.  It contains instructions for all the different things that the body builds at a microscopic level: the proteins.  It sits in the center of the cell.  When a new structure needs to be built, a piece of Play-Doh comes along and presses against the paragraph that has the instructions for this structure.  It then goes over to the cellular protein factory, which reads the impression and creates the protein by a reverse impression process.  This bit of "Play-Doh" is the "messenger RNA" or "mRNA", so called because it takes the production order "message" from the DNA to the protein factory.

How viruses hijack this process

Viruses survive by invading and hijacking this process.  To be successful, a virus needs to do the following:

1. Spread: survive outside the body for a bit of time
2. Penetrate: get into the body and survive there among the cells but not in them for a while
3. Infiltrate: gain access to the inside of reproducing cells
4. Replicate: Use the cell's own replication ability to replicate themselves, using one of various strategies:
1. Some viruses have their own DNA replication abilities; they just need to get into the protein factory in order to hijack the facilities to duplicate themselves using their own DNA.
2. Some viruses modify the DNA in the cell.  They stick their own instructions into the DNA "book" in a place where the cell will read with the mRNA.  The mRNA then takes the virus instructions to the protein factory and the factory produces more virus instead of the normal things it would instead.  NOTE: in this case, the modified DNA is not replicated.  It's just been modified so that the *virus* will be replicated.
3. Some viruses just produce their own mRNA and arrange so that their mRNA gets sent to the protein factory instead of the normal mRNA.  The end effect is basically exactly the same as modifying the original DNA; it just skips a step.
5. After replication, each copy starts over again at step 2.

How the immune system defeats viruses

This is really complicated, but to overly-simplify (and skip a bunch of steps), the body determines that there is a foreign invader and starts working on developing new proteins that are custom-shaped to latch on to the foreign particles and somehow neutralize them ("neutralizing antibodies").

In the case of Covid (and other coronaviruses, actually), it seems that the primary thing that the body's immune system attacks is called the "spike protein".  This is the spiky protrusion around the outside of the virus, and it is used by the virus for step 3 ("Infiltrate")--it works kind of like a skeleton key to break into the cell.  The neutralizing antibodies that the body produces bind to the spike protein, so that it no longer fits into the key hole.  The virus particles are thus not able to enter cells and not able to reproduce.

How vaccines also cause the immune system to produce neutralizing antibodies

The goal of a vaccine is to trigger the same immune response that the real virus does, but without the virus actually replicating throughout the body, causing the disease.  There are several strategies that have worked for this:

1. Infect the person with a disease that is really similar to the original disease, but much less severe.  The smallpox vaccine was like this; cowpox is a disease that works much like smallpox, so that the antibodies that the immune system produces to counteract it happen to also work against smallpox.
2. Take a virus and somehow neuter it so that its ability to replicate is destroyed.  I believe the MMR vaccines work in this way; I think they use UV radiation to damage the virus.  Then they inject you with a quantity of the virus.  The virus particles retain the protein structures necessary to invade cells (I think) and the immune system generates neutralizing antibodies to those structures.  There is a risk in this method that the culture is not completely neutralized and enough viable virus is left to replicate and cause the disease you are trying to vaccinate against.  (It's a low risk; about 1 in 1 million).
3. There are new kinds of vaccines that try to be much more targeted in how they operate.  Rather than using a full virus (either a weak relation or a weakened form) for the body's immune system to react to, these vaccines get cells to produce just small parts of a virus.  There are two main new technologies being employed to do this, which I will now describe.

The newer vaccines

Adenovirus vector vaccines

Adenovirus vector vaccines are essentially genetically modified cold viruses.  They still operate like normal viruses up until they invade your cells.  After this, when it comes time for them to inject their genetic replication DNA into the cell, this genetic replication code has been replaced with a specified set of instruction instead.  So the messages that get sent to the protein factory aren't "build another copy of this virus", but rather just "built this specific structure" instead.  In the case of a coronavirus vaccine, the specific protein that they cause to be produced is the spike protein.  The infected cell therefore generates that spike protein, which the body's immune system recognizes as foreign material.  The immune system therefore generates antibodies to the spike protein of a virus without ever having had exposure to the whole virus.

Since an adenovirus vaccine does not have replicating code, it will not replicate throughout the body.  The modified virus will infiltrate those cells with which it comes into contact at the site of the injection, those cells will operate differently for a time, and those cells will then be destroyed and disposed of by the body's immune and waste disposal systems.

mRNA vaccines

mRNA vaccines aim for the same effect as the adenovirus vector vaccines, but with an even simpler delivery method.  With these vaccines, we just straight up manufacture the messenger RNA fragments that will tell the protein factory to produce the spike protein.  Our chemical gene manufacturing processes have gotten good enough where we can mass produce small segments of RNA--so that's what we do.  We then encapsulate bundles of these mRNA fragments into lipid nanoparticles so that they can survive for a while inside the body.  ("Lipid nanoparticles" sounds exotic, but it is less so than you might think.  It's actually not dissimilar to the process of making mayonnaise, it's just a much finer emulsion.) When these are injected into someone, a lot of them bump into cells and get absorbed into them.  The mRNA fragments then get into the cell protein factory, causing those cells to start producing some spike proteins, thus triggering the immune response.

As with the adenovirus vector vaccines, there is no replication code injected, so the vaccine can only cause a set number of cells to produce spike proteins.  Again, these cells will be disposed of by normal waste disposal processes.

Risk Analysis of Moderna and Pfizer Vaccines: Update

Since I first published my vaccine risk analysis, one person (the first) has died shortly after receiving the vaccine and possibly because of the vaccine: click here for an article on Dr. Gregory Michael.

I think it's worthwhile to take a look at this and see if this affects the bottom-line risk analysis.

Did Dr. Michael die because of the vaccine?

Bottom line: not certain, but quite possibly.

In more detail:

Dr. Michael developed a condition 3 days after having been vaccinated called "acute immune thrombocytopenia" (ITP), which happens when your immune system begins attacking your own blood platelets.  In about 5% of adult cases, this can lead to death by hemorrhaging, which is what happened to Dr. Michael.

Since vaccines work by triggering the body's immune system, I had already identified immune system overreactions as the most likely cause of adverse side effects from the new vaccine.  It is therefore quite plausible that the vaccine was the cause of, or at least the trigger for, the condition that lead to Dr. Michael's death.

However, we also cannot rule out tragic coincidence at this time.  As of today's date (January 14th), about 11 million doses of the new vaccines have been administered in the U.S. alone.  ITP is rare, but it occurs in about 2 adults per 100,000 each year.  This means that if you picked 11 million adults at random in the U.S. and did nothing to them for one month, you would expect about 18 of them to develop ITP just by chance, and given a 5% fatality rate, one of those 18 or so would be likely to die.  One death from ITP out of 11 million doses in one month therefore is not above the level where it could be an unfortunate coincidence.

However, I feel that it is somewhat more likely than not that the vaccine triggered the condition in Dr. Michael in this case.  ITP is normally triggered by something that activates the immune system: either a pathogen or a medicine such as a vaccine which triggers the immune system.  In Dr. Michael's case, the only such cause we know of is the vaccine; he wasn't otherwise sick, nor had he taken another medicine that could have triggered the condition (that I could tell by the report, anyway).  So I think it is reasonable to suspect that the vaccine had a part in causing his death, even if we can't rule out coincidence and an unknown other cause.

Was Dr. Michael susceptible to ITP even aside from the vaccine?

Probably, yes.  In my risk analysis, I emphasized that many of the mechanisms by which the vaccine might cause serious side effects in its recipient are also possible as side effects of the virus itself.  It stands to reason that if the side effect (ITP in this case) is caused by an immune reaction to the vaccine and if the whole point is for the vaccine to trigger the same immune reaction that the virus would, then that side effect of the vaccine would also be a side effect of the virus.

This does turn out to be the case with ITP.  Although not listed in my short-list of known serious side effects of Covid-19, it turns out that ITP has already been identified as one of the myriad possible symptoms and syndromes that can be triggered or caused by Covid: see this article here.  We don't know the specific reasons why some people can be susceptible to ITP and other people aren't, but genetics are vaguely suspected.  It is therefore more likely than not that Dr. Michael was in some danger of developing ITP from catching Covid, as well as from taking the vaccine.  This is, however, impossible to know for sure at this point.

Should I update the risk analysis bottom line because of this occurrence?

No, that isn't necessary.  Previously, I had set the risk of death from getting both doses of the vaccine at a maximum of 1 in 1 million, which was based on 2 million doses administered at the time with no deaths.  We now have 11 million doses administered and 1 known death.  If we attribute this death completely to the vaccine, that means I should adjust the chances of dying from the vaccine down to 1 in 6.5 million.

However, as I purposefully decided to under-rate the risks of the vaccine and over-rate the risks of the virus and data is still relatively sparse, I will keep the risk at 1 in 1 million for now.

Asymmetrical Processing of Ideas: Part 1

In connection with theories that the recent presidential election was fraudulent (with which I strongly disagree), I've had some discussions recently about how people can become very mistaken about reality, holding theories as certainties when the actual evidence for those theories are very thin.  Someone commented that they wondered whether people who do this were just unaware of the concepts of burden of proof or Ockham's Razor (aka the "principle of parsimony").

I disagreed, for the reason that I believe you can usually see a perfectly good understanding of those principles in operation when the person is critiquing someone else's theories.

The problem is therefore not a lack of understanding of correct intellectual procedures, but a radical asymmetry in the application of those procedures. Theories that fit into a certain a priori favored worldview are subjected to token scrutiny only, while anything that challenges that worldview is subjected to skepticism of the most stringent kind. I think there are several common manifestations of this:

1. A too-complete rejection of arguments coming from entire classes of people because they are of the opposition. "Wait, you're seriously quoting something from the Washington Post? LOL! Don't you know you can't trust anything they say?" Or, "I only need to see somebody quote from Matt Walsh to know I can safely ignore anything they ever say again."
2. The complement to this: an extreme generosity in forgiving the failures of someone or some set of ideas because they pushed some idea or goal that is very important to you. "Fr. So-and-so is accused of sexual molestation?? But he's a conservative priest who loves Latin . . . this must be a setup." Or, "OK, so yeah, some people have taken this ideology to an extreme and it resulted in socialist hell-holes of a country. But this is *so* much more generous to the poor, there must be some good in it we can keep!"
3. Excessive focusing on one aspect of a theory you like that you know to be true, while ignoring crucial parts that are way less certain. "The media is absolutely biased against conservatives, and they hate Trump with a passion. This would need to be the case for a widespread conspiracy to defraud Trump of the election to be successful. Therefore, it's reasonable to believe that this is absolutely happening!" Or, "The actions of this policeman against this black man is an obvious, complete outrage. It's reasonable to believe that the police in general are a complete hotbed of racism and hatred!" Emotionally, you can take certainty in one thing and spread it out to a lot of related things, without a rational justification.
4. The complement to the above, finding one aspect of your opponent's argument that you know with certainty is false and focusing excessively on it, but without questioning how badly this error damages the overall argument. Not all flaws in an argument are fatal, but to someone determined to reject an argument, oftentimes any flaw is sufficient reason to throw away the whole thing in disgust.
5. Radical disengagement from people or activities that could give you sources of information that support your opponent's argument over yours. The time someone is willing to spend listening to his opponent, or looking for evidence that will support his opponent, is a fantastic indication of how objective he is willing to be overall. For the social media conspiracist, the amount of time he is willing to spend reading lengthy, often repetitive treatises from his favorite internet talking heads is usually vastly greater than the time he is willing to spend reading opposing viewpoints.

I would like to look at examples of all of these types of thinking in the current social media debate over the Capitol Building riot.  Since I am in the camp of "Trump deserves to be impeached for what he instigated" and "the violence of this riot was primarily coming from extremists who have been part of the Trump movement, not primarily from leftist plants", I am going to try to make more of an effort to find examples of this type of asymmetric thinking from proponents of these viewpoints.  However, I will be looking at this from both sides as well.