The Pathogenesis of Long COVID

We are well into my blog series on the health consequences of COVID-19 to survivors, including long COVID or PASC. In prior parts of this series, we have discussed what long COVID or PASC is, the fact that not all health consequences from infection with SARS-CoV-2 fit within this category, the potential magnitude of the problem long COVID or PASC, and in my last blog piece we dived into the pathogenesis and pathophysiology behind COVID-19 as preparation for this and my next blog piece that will delve into what we know about the possible pathogenesis of long COVID and the other health consequences that we see in some people who have recovered from COVID-19, many of whom had “mild” infections.

Before we do, I already have an update. A pre-print article (not yet peer-reviewed or published in a scientific journal) was released just two days ago from the day I am writing this. https://doi.org/10.1101/2022.05.26.22275532. This article is titled: A global systematic analysis of the occurrence, severity and recovery pattern of long COVID in 2020 and 2021. So, as we jump in, let’s put our science hats on and remember the information I presented to you earlier in this blog series about interpreting clinical studies. The first point to note is that this article examines cases of long COVID across the world. This should already raise two concerns for us. First, there is not yet a clear, consistent, or globally agreed upon case definition for long COVID or a diagnostic test or criterion (in other words, this remains a diagnosis of exclusion):

World Health Organization (WHO) clinical case definition: (It refers to long COVID as post-COVID-19 condition). Post-COVID-19 condition occurs in individuals with a history of probable or confirmed SARS-CoV-2 infection, usually 3 months from the onset of COVID-19 with symptoms and that last for at least 2 months and cannot be explained by an alternative diagnosis. Common symptoms include fatigue, shortness of breath, and cognitive dysfunction but also others and generally have an impact on everyday functioning. Symptoms may be new onset following initial recovery from an acute COVID-19 episode or persist from the initial illness. Symptoms may also fluctuate or relapse over time.

US Centers for Disease Control and Prevention (CDC): Post-COVID conditions are a wide range of new, returning or ongoing health problems people can experience four or more weeks after first being infected with the virus that causes COVID-19. Even people who did not have COVID-19 symptoms in the days or weeks after they were infected can have post-COVID conditions. These conditions can present as different types and combinations of health problems for different lengths of time.

UK National Institute for Health and Care Excellence: (1) Ongoing symptomatic COVID-19 for people who still have symptoms between 4 and 12 weeks after the start of acute symptoms; and (2) post-COVID-19 syndrome for people who still have symptoms for more than 12 weeks after the start of acute symptoms.

Allow me to point out some of the fine points of differences:

1. WHO: requires a history of probable or confirmed COVID-19.

CDC: acknowledges that long COVID can occur in people who had asymptomatic COVID-19 or may have had mild symptoms, but were not diagnosed as having COVID-19.

UK: Implies that the person must have had symptomatic COVID-19 with references to “ongoing symptomatic” and “still have symptoms.”

• WHO: long COVID symptoms occurring 3 months from onset of infection and lasting at least 2 months.

CDC: long COVID symptoms four or more months after infection, but without specifying duration.

UK: long COVID symptoms are those persisting from initial infection for more than 4 weeks or the development of new symptoms characteristic of post-COVID-19 syndrome lasting more than 3 months following infection.

Thus, looking at the authors’ criteria for what they consider to be long COVID will be critical and depending which case definition is used, it may artificially limit the number of people who are included as cases- in other words, there is a risk that this will undercount cases. Second, we must remember that health systems vary greatly from country-to-country, with some having nationalized health care systems that have robust health records of their entire population (e.g., U.K., Israel) and others that have almost no health care infrastructure and are unlikely to have complete information on their populations. Thus, we run the risk of undercounting cases in developing countries, simply because they don’t have the public health infrastructure to test and identify cases, but in those countries, we also may face the risk of disproportionately high cases if they are able to be identified due to the fact that they were likely unvaccinated, resulting in more wide-spread infections, which will mean more of the population was at risk for long COVID. Finally, the time frame of 2020 and 2021 means that this will not include large numbers of infections from the various omicron surges and thus, if there are a large number of resulting cases of long COVID from omicron infection, those will not be included in this study.

This study is a cohort study (see my earlier blog post on understanding clinical trials) conducted in ten countries based upon the occurrence of three major symptom clusters of long COVID among representative COVID cases, but they use a meta-analysis methodology, which means that the authors are gathering their data in large part from published studies, and while this is an important tool and can provide very important insights, we have to remember that differences in definitions or methodologies in each of the studies can also introduce error. We should also be cautious about generalizing the occurrence of long COVID in ten countries to the rest of the world, in identifying long COVID cases by three major symptom clusters (they use the three major groups of symptoms identified by WHO, but we will need to keep in mind that long COVID patients often will not fit nicely into one of these three symptom clusters because many will have overlapping symptoms and others will have symptoms that don’t fit neatly in any of the symptom clusters) and their reference to COVID cases (which raises the concern that they are only looking at symptomatic cases that were identified, diagnosed and reported, which will exclude cases that occurred, but were not diagnosed or reported as well as asymptomatic cases unless we see that they identified cases based upon serologic testing).

They defined their symptom clusters based upon the WHO clinical case definition (so that is good) and they came up with fatigue, cognitive problems and shortness of breath as their three clusters (so this could leave out many patients who do not have these symptoms, but have symptoms or conditions that we do currently believe are the result of SARS-CoV-2 infection (e.g., new-onset diabetes, postural orthostatic tachycardia syndrome, etc.), unless they also identify as having one of these other three symptoms. The authors also use in their criteria for inclusion, the duration of symptoms for at least three months. I don’t find that overly problematic, other than keep in mind, many other studies use one month or two months, so this will simply create some difficulties in comparing studies if there are a significant number of cases of long COVID or PASC that resolve themselves in this time-frame.

The results are based on “detailed information” on 1906 community infections (a really low number from 10 countries during this almost two year period) plus 37,262 cases published in the literature (total 39,168) and “detailed information” on 10,526 hospitalized patients (so, we can see that the results are going to be skewed towards sicker (and likely older) patients, when we have a lot of information to suggest that many patients with long COVID had “mild” infections and were not hospitalized) plus 9,540 published hospitalized cases (total 20,066). In addition, the authors indicate that they had medical record information data concerning 1.3 million infections. So, ultimately, the authors do get to a large number of cases.

The authors find that in 2020 and 2021, there were 144.7 million cases of long COVID fitting into one of the three symptom clusters, corresponding to 3.69% of all infections, with 51.0% of these people with long COVID suffering from fatigue, 60.4% suffering from respiratory symptoms and 35.4% suffering from cognitive problems (so, we see that the long COVID patients identified in this study do have symptoms that overlap among clusters since the total adds up to > 100%).

Interestingly, the authors find that those with milder acute COVID-19 infections recovered from their long COVID symptoms (median 3.99 months) than those with long COVID following hospitalization (median duration 8.84 months). However, at one year, 15.1% of those with long COVID continued to experience symptoms.

Interestingly, these authors also found what has been previously reported in terms of a female predominance in long COVID – 63.2% (see below why this is interesting).

The risk for long COVID after COVID-19 that did not require hospitalization was:

• 2.73% in children
• 4.76% in adult males
• 9.88% in adult females

Alarmingly, the peak ages for developing long COVID-19 were in those young adults who probably were not concerned about risk for hospitalization or death from COVID-19 – those ages 20 – 29 (this is lower than some earlier studies, which seemed to find the highest risk among those in their thirties and early 40s).

However, do not be deceived by the references to the symptoms, which for those unaffected, might think, these are no big deal. The average disability score reported was 0.231 – equivalent to moderately severe traumatic brain injury!

Despite the limitations of this study, we once again can see that for many people, COVID-19 is not merely a cold or the flu, and that they can suffer significantly for extended periods of time. Unfortunately, this study was not designed to answer the question as to how much protection the COVID-19 vaccines provided against long COVID. It also does not answer the questions as to whether infections with different variants are more or less likely to result in long COVID and whether early treatment (with antivirals or monoclonal antibodies) was effective in reducing the risk for long COVID. Notice also, that this study was not designed or intended to assess health outcomes for this population over time.

So, now let’s proceed with our discussion about the pathogenesis of long COVID. First, let me point out that SARS-CoV-2 is not the first virus to cause a post-acute infection syndrome. There is a nice review of post-acute infection syndromes at https://doi.org/10.1038/s41591-022-01810-6. While we do not fully understand the pathogenesis of these post-acute infection syndromes (PAISs), there are some similarities in many cases caused by very different pathogens, suggesting that perhaps there are common etiologies at play.

While PAISs appear to affect only a minority of patients who suffer acute infection, that can still be a huge number when you consider how many people the pathogen infects, and we also know that many patients go undiagnosed due to the nonspecific nature of their symptoms, inadequate access to expert medical care, imprecise case definitions for the specific PAIS, and the fact that most of these PAISs do not have an identified marker of disease (in other words, a specific test that confirms or rules out the presence of the PAIS in question).

Common shared symptoms among sufferers of PAISs include exertion intolerance, disproportionate levels of fatigue, neurocognitive and sensory impairments, flu-like symptoms, unrefreshing sleep, myalgia (muscle pains)/arthralgias (joint pains) and a plethora of nonspecific symptoms, however, the relative frequencies of these symptoms following different acute infections appears to vary in a manner that may reflect the underlying tropism of the pathogen causing the infection and/or the pathogenesis of the acute infection.

One condition that has now been recognized as a PAIS from certain infections including SARS-CoV-2 is myalgic encephalomyelitis/chronic fatigue syndrome (ME/CFS). I remember being in practice and seeing patients that would ultimately fit this diagnosis, but before this condition was identified or it was understood what was causing it, though we suspected many were experiencing this as a consequence of Epstein Barr virus infection. The patients I saw had their lives upended, both in the normal activities of life that most of us take for granted, but also their ability to work or work as effectively as they previously did. Characteristically, these patients will have systemic exertion intolerance along with chronic fatigue that is unresolved by rest or sleep. Importantly, whereas we often prescribe physical therapy, rehabilitation and conditioning for persons who are recovering from many conditions in a weakened state, these patients often experience a worsening of their symptoms following physical, cognitive and emotional exertion and exercise actually poses the risk of a decline in their condition.

Patients with ME/CFS may have other prominent features of their illness including neurocognitive impairments (such as impaired memory, impaired concentration or what is colloquially referred to as “brain fog”), pain, sensory disturbances and various forms of dysautonomia (we’ll discuss this in a later blog post).

One can understand that given the pandemic with COVID-19 is still unfolding and changing, we might not have good data on prevalence and the prognosis of long COVID, which is certainly the case. But, we don’t really have good data on any of the PAISs, despite having seen these cases for decades. This may be for many reasons – no precise case definitions or diagnostic tests, no required reporting to centralized data bases, and likely, no significant funding devoted to research of this kind. This is unfortunate, because given our suspicion that there are likely common etiologies of the pathogenesis resulting in these PAISs, since we tend to see similar PAISs with differing pathogens, had there been more robust research over the past decades, we might have a better understanding of the pathogenesis (and therefore potential treatments) of long COVID by now, as well as answers to a very important question as to why some people recover from these illnesses apparently just fine, but others develop these PAISs.

We do have some data from adolescents and young adults who developed ME/CFS following infectious mononucleosis (these are most often the result of Epstein Barr virus infection; but in a small number of cases can be due to cytomegalovirus infection). Some studies showed that of approximately 30-40 percent of patients with persisting symptoms following infectious mononucleosis, most would recover over time with a drop to 8 – 14 percent still symptomatic at 6 months and 7 – 9 percent at 1 year. Another study showed that 4 percent were still symptomatic at 2 years, but we have no idea why those in these studies developed ME/CFS, why most improve, and why some remain symptomatic.

West Nile virus can also lead to persistent symptoms. A study that followed persons infected in Texas over 8 years found that the frequency of persistent symptoms seemed to depend upon which form of disease people got – West Nile-related fever vs. West Nile meningitis or encephalitis. Again, there tended to be a number of patients whose symptoms resolved over a 2-year period, but with this infection, there was a much higher level of patients with persistent ongoing symptoms – 40 – 70% of those with persistent symptoms following their initial infection remained symptomatic past the 2-year mark.

On the other hand, we do see PAISs in some infections, such as Q fever and Lyme disease where we see little improvement, if any, over many, many years.

Fortunately, as reported above, as well as in a number of other studies, although we do see patients with thus far persistent long COVID, we do see improvement, and in some cases, resolution of their symptoms in many patients over the course of months to a year or so.

So, what do we know about the pathogenesis of other post-acute infection syndromes? Not much. Mostly, we just have hypotheses (educated guesses). What are those hypotheses:

1. Persistence of the pathogen that either is in such low levels as to escape our currently available tests or the residence of these pathogens is in places of the body where they can escape our detection. (This is one of the potential etiologies for long COVID and there is some evidence that we will discuss in the next blog post). Persistence of pathogens, or even remnants of the pathogen (in the case of SARS-CoV-2, RNA from the virus, which is no longer infectious – this is why a PCR test may remain + for weeks or months after your acute infection has resolved and you are no longer contagious) may provide antigenic stimulation for the body’s immune system. The residual virus or remnants of the virus that are still recognized as antigens generate pathogen-associated molecular patterns (PAMPs), which can continue to stimulate the innate immune system, which in turn leads to ongoing inflammation, as well as chronic stimulation of lymphocytes as part of the cellular immune system, that can lead to T-cell exhaustion and a diminishment of the immune modulation role that T-cells play in preventing the immune response from becoming overly exuberant and causing more harm than good.
2. Autoimmune reactions. When a pathogen enters our body, our immune system tries its best to specifically target the pathogen with the antibodies it produces and the T-cells that are activated. However, infections involving certain tissues may induce local innate immune responses that can trigger T-cells to be directed at so-called “self-antigens,” antigens that belong to our own cells and tissues and not the pathogen. Other pathogens exhibit “molecular mimicry,” that is to say that the antigens belonging to the pathogen are so similar in their molecular structure to naturally occurring antigens in our own bodies that our immune system is tricked into attacking our own cells believing that they are those of the pathogen (this is believed to be one possible mechanism by which insulin-producing cells of the pancreas are destroyed resulting in diabetes). There is evidence that autoimmunity plays a role in the development of severe COVID-19 and thus, reason to believe it could be playing a factor in the development of long COVID. Interestingly, for reasons unknown to me and I think others, women seem to be at higher risk than men for the development of autoimmune disorders (conditions like lupus, scleroderma and rheumatoid arthritis), and it is then interesting to consider that most studies do show a female predominance among those that develop long COVID. My suspicion is that there are genetic predispositions accounting for some or all of this, though the influence of hormones has not been ruled out. We will discuss the evidence for autoimmune antibodies caused by COVID-19 in my next blog post.
3. Another hypothesis is that the pathogen and/or our treatments of it disrupt the normal balance of bacteria and viruses in our body, causing disruption of various physiologic processes in our bodies. We do see cases where infection with one pathogen disrupts the normal immune processes that keep other latent viruses in check such that they are reactivated. Generally, these latent viruses are DNA viruses such as Epstein Barr virus, cytomegalovirus, and herpes simplex virus (remember that SARS-CoV-2 is a RNA virus). Many people who have been plagued by recurrent sores and blisters on their lips, tongue, roof of their mouths and gums have herpes simplex virus type 1. They will tell you that it erupts, then comes under control and then can erupt again weeks, months or years later. Often, they will note that reactivation occurs at a time of another illness (leading many to call these cold sores, or at a time of significant stress, both situations that can temporarily weaken our immune systems). Another example, in those of us who were infected with chickenpox as children, is that we never really get rid of that virus (varicella zoster virus), rather it becomes what we call latent (basically hanging out in our peripheral nervous system, but not actively reproducing and causing the chickenpox lesions on the skin) and our immune system helps keep it in check. However, as we age and our immune system becomes “senescent” or if we develop an illness that causes compromise to our immune system or we are treated with medication that can suppress our immune system, the varicella zoster virus can reactivate and when it does, it is manifested as shingles. The same thing can happen with certain infections. They may cause our immune system to be so focused on this new infection, that it lets up its guard against these latent viruses and they can reemerge. This is another potential etiology in long COVID, as we have been able to demonstrate that in some cases of COVID-19, the Epstein Barr virus (EBV) that most of us were infected with when we were young, and our immune systems brought under control, but never fully cleared from our bodies, with the virus becoming latent or dormant, has become reactivated in some long COVID patients. We will review that evidence in the next blog post, but keep in mind, chronic fatigue resembling ME/CFS is common in long COVID and is also something we saw as a PAIS with EBV infections.
4. Another proposed etiology is the tissue damage itself caused by the infection. Our bodies are amazing in their ability to heal, but not all injury can be healed. For example, while some cells and tissues can be repaired or replaced with new cells, others become scarred and non-functional. Certain of the variants of SARS-CoV-2 have had the ability to inflict such severe damage on lung tissue that some people have developed a disabling condition called pulmonary fibrosis, basically lung scarring. When this happens, the lung tissue cannot do its primary function sufficiently of transporting oxygen from the air we breathe into the capillaries of the lung beds that will then transport oxygen to our other tissues. In many cases, these people will require supplemental oxygen and you have probably seen some of these patients who require a portable tank of oxygen with tubing that fits around their ears going to their nose. In other even more severe cases, we have had patients have such bad pulmonary fibrosis from their COVID-19 that they required double lung transplantation.

Let’s look at the list of putative etiologies for long COVID:

• Persistence of the virus in certain sanctuaries of the body (failure of the body to clear all of the virus produced during the acute infection resulting in an overly exuberant immune response).
• Immunopathology (damage to the immune system itself) – either resulting from the initial infection or due to the persistence of the virus in the body and the resulting persistent antigenic stimulation.
• Autoantibodies (damage caused to tissues by an aberrant immune response)
• Micro-clotting and the resultant damage to tissues by reducing oxygen delivery to those tissues.
• Reactivation of latent viruses, such as EBV.
• Damage to the nervous system and other tissues resulting from direct infection and the resulting immune response.

What remains unclear is whether all of these play a role in all persons or whether different etiologies or combinations of etiologies occur in different people that in turn account for the myriad presenting symptoms we see in patients with long COVID. Further, we don’t yet know whether the risks for long COVID are additive with each reinfection, or potentially orders of magnitude increased by reinfection.

In my next blog piece, we will examine the evidence that we have for these etiologies and discuss some ways that these abnormal reactions have impacted some patients with long COVID.