Notes from the NIH Workshop on Post-Acute Sequelae of COVID-19

• 

Interestingly, although this graphic came up in quite a few presentations, it does not show the two largely synonymous terms that were most frequently mentioned:  brain fog and cognition (or cognitive dysfunction).  Does this suggest that brain fog and cognition are included in some other term like fatigue? From the comments, they are described as distinct characteristics. If nothing else, this small distinction shows how early we are at reaching a common understand of these longcovid sequelae.

Another way to look at the longcovid sequelae is simply by listing the breakout sessions:

1. Neurological/Psychiatric/Neuromuscular
2. Cardiovascular
3. Pulmonary
4. Renal/GI/Metabolic
5. Immunologic/Rheumatologic
6. Pediatric

One might avoid being accused of facetiousness when asking “Of the major organ systems, what’s not on this list?”  And while pediatric does not represent an organ system per se, the unique aspects of pediatric physiology, particularly around immature immune-response, warrant including it as a separate area of focus.

One of the most encouraging aspects of the discussion, using that term cautiously, was how much the COVID research is informed by past research, particularly that around two earlier coronaviruses, SARS-CoV (referred to as “SARS 1”) around 2003 and MERS-CoV (referred to as “Middle East Respiratory Syndrome” or “MERS”) around 2012.  Realizing that we are still well within the first year of identification, albeit with the global explosion of the virus, it is encouraging to see how the experience of these earlier viruses can inform and guide today’s research.

In addition to learning about the relationship of COVID to SARS 1 and MERS, several speakers identified other conditions that, at least on the face, present similar pathologies.  Dr. Ann Parker, from Johns Hopkins, discussed the similarities with Post Intensive Care Syndrome (PICS), and, more broadly, post hospital syndrome, which show similarities in the impact on physical function, fatigue, cognition and subsequent delirium, as well as mental health challenges including grief, fear and anxiety.  Similarly, other speakers talked about syndromes related to cancer treatment, particularly chemo therapy.  These may provide models that can be applied to longcovid or where the lessons of longcovid may be applied to them.  It is too early to know how well aligned these will be with longcovid.

While the focus of this workshop was more on research to understand longcovid, it was notable that when interventions were mentioned, they were primarily pharmacological or biomedical interventions, not behavioral or compensating interventions (e.g. supporting activities of daily living).  In fact one speaker criticized an otherwise valuable study for suggesting that CBT (cognitive behavioral therapy) might be of value in treating fatigue.  He strongly noted that CBT has no value in treating this type of fatigue.  That makes sense intuitively, but leaves open the question To what extent can behavioral and compensating therapies benefit longcovid patients?  Almost as an aside, Dr. Chris Cox, from Duke University, mentioned mobile apps that Duke is rolling out to help longcovid patients.  It will be interesting to learn if these are behavioral aps and to see what impact they have alongside pharmacological interventions for longcovid patients.

Although it wasn’t an express talking point in the workshop, it was very encouraging how many of the speakers identified themselves as leading post-acute COVID clinics, showing that despite the limitations of our knowledge of the pathogenesis and the dearth of research supporting evidence-based practice for the treatment of longcovid patients, healthcare providers are already formalizing the combination of rapid response clinical care with continued research. 

3.What are the apparent knowledge gaps?

If the workshop demonstrated anything, it is that the scope of what we don’t know is almost impossible to overstate.

• Does the virus itself cause damage during the acute phase and is this damage the cause of longcovid? 
• Does the virus itself persist in the systems, either distributed in the organs or, perhaps, in a reservoir, such as the kidneys that gives the immune system less opportunity to fend it off? 
• Is it the body’s own immune response, rather than the virus itself that leads to longcovid, continuing to disrupt function long after the virus is gone? 
• Or, is it the treatment during the acute phase that causes or exacerbates the complications for longcovid? 
• And then there are the questions of risk factors and disparities that may impact frequency and severity.

Although the evidence is still insufficient, some presenters noted that studies of patients who died post-acute COVID had only residual levels of the virus, suggesting that the virus is at least not the primary cause for longcovid.  (This was a rather technical area that was presented with wonderful clarity.  I won’t try to dive any deeper here, but invite you to listen to the archive discussion.)

As an illustration of the challenge regarding the risk of co-morbidities such as diabetes, Dr. Jonathan Himmelfarb (@xpotasn), of the University of Washington, the host of the renal/GI/metabolic breakout session, reported that compared to people without diabetes, people with type-2 diabetes experience higher rates of longcovid while people with type-1 diabetes do not.  It has not been established whether this is a causal relationship or whether it is co-occurring, along with obesity, hypertension and other potential risk factors.  Conversely, the evidence is insufficient to determine whether COVID contributes to the onset or exacerbation of either type-1 or type-2 diabetes. 

Many speakers commented on the limited but still suggestive impact of gender, race and ethnicity.  Dr. Carlos del Rio (@carlosdelrio) of Emory University, noted that women tend to do better than men, wondering if estrogen provides some mechanism of protection.  He also noted that according to a New York Times article from July, 2020 (that summarizes the results of a study published in Nature), Black people are 2.5 times as likely to die as people with white ethnicity.[1]  Most interestingly, Dr. del Rio noted that SDOH alone do not appear to explain the disparity in outcomes.  Along with potential genetic factors, he went on to list common co-morbidities as well as behaviors with respect to crowding, types of jobs, segregation, housing, resources, income levels and access to health services all of which may play a role.  In particular he wondered about the disparity in having the option to work from home thus reducing exposure for people with more economic flexibility as a driver of the disparity in acute COVID as well as longcovid. 

Dr. del Rio was the first of several presenters to expressly question the role of structural discrimination and the need not just for clinical solutions but also for policy solutions that look beyond COVID and longcovid, particularly system level reforms that address access to care.

Another acknowledged gap across all the speakers and breakout sessions was the need for baseline data for both acute and longcovid patients.  Whether it was lung function or cognitive performance, all researchers expressed the challenge of measuring the extent of change or recovery in the absence of baseline data.  While they were not the only ones, Dr. Eva Feldman (@@EvaFeldmanMDPhD) from the University of Michigan and another panelist whose name I missed but who is affiliated with the Cleveland Clinic both described existing and new data sets that help meet this need. 

It was exciting to hear presenters talk about the beneficial impact of the migration from paper-based records to electronic health records over the past 15 years, including the increased use of standardized tools like the Patient Health Questionnaire (PHQ), with tools like the PHQ-2 and PHQ-9 which measure depression-related information.  While the overall feel was of a need still not fully met, it was impossible not to recognize the value of migrating to electronic health records along with the increased adoption of evidence-based pratices, whatever the challenges.

Along with the gaps in the purely clinical dimensions of our understanding, there was a strong focus on the patient experience.  In a profound moment, when the two patient representatives, Hannah Davis and Chimere Smith (@chilnvs1), a teacher and longcovid survivor from Baltimore, were introduced, the host noted that two other expected patient panelists were not able to attend because they are back in the hospital. That, more than any other moment, drove home both the severity and the advancing/retreating nature of longcovid. 

Ms. Davis said one of the biggest challenges both acute COVID and longcovid patients must contend with is “the struggle to be believed.  Doctors simply don’t understand.”  Ms. Davis described how in March, when her primary early symptoms were brain fog, cognitive dysfunction and fatigue, she could not get her clinicians to give her a COVID test.  “This can’t be COVID.  You don’t present with the right symptoms.”  They wanted to pursue other avenues like ADHD.  As an indication of her personal challenge, she noted that an assessment of her own cognitive processing dropped from the 94^th percentile in 2019 to the 14^th percentile in 2020.  Although she did not say it, we all hope that today that dismissal would not happen.

Ms. Smith described her experience being told by clinicians “This can’t be related to COVID. You’re cured of COVID.”  Ultimately, she said, she was able to engage her clinicians as collaborators in her longcovid struggle, but the additional process of engaging her care providers was itself daunting.  There wasn’t anyone in the audience who didn’t cringe in their support.

Ms. Davis sited her work with a group appropriately named Patient Led Research for COVID-19.  She sited one survey published in May of self-identifying longcovid survivors which, among many important findings, showed:

• 90+% of people felt they had not recovered fully from COVID
• 80+% experienced cognitive symptoms
• 73% experienced memory loss
• 73% were not able to return to work.

Then came Dr. Peter Piot, one of the world’s leading researchers in HIV/AIDS and Ebola, and Director of the London School of Hygiene and Tropical Medicine, who told his own story as a part of his larger presentation.  He told of going to the hospital with COVID and, like “my sisters,” experiencing the challenge of having his words undervalued, although, he noted, that they might have done a better job of hiding it because “after all, I’m an old guy.”  We cringed even more.  He went on to describe that he was in a bed next to a homeless man, receiving the same care, at no cost to either.  Whether intended or not (I don’t think that Dr. Piot is inclined to say things where the meaning is not intended), he quietly drove home the value of a healthcare system that provides equal access to all.

4. What are the next high-priority areas of research and treatment?

Let me say here that the enormity of what we don’t know may have hindered my ability to grasp a succinct list of immediate priorities.  I picked up a few things, like the need for common terms and definitions and a need to move beyond “convenience sample” studies, where people are in the study because they self-select or are in the subset who appear at a clinic, to more valid research models of that lead to better data.  While an overarching priority list may not have been an outcome, clearly every researcher came away renewed in their dedication to their own research-as well as collaboration and coordination with other researchers.

Conclusion

This was an excellent workshop.  I thank the NIH and every participant for their wonderful contributions.  I came away with a clearer understanding of just how early we are in this race.  While the possible availability in the near future of a vaccine that has a high rate of success with little or no side effects will certainly mitigate the harm caused by COVID, it is not enough.  We must continue our research, not only as it applies to longcovid but also to similar syndromes like PICS and chemo therapy.  And as to the role that behavioral interventions can play, supported by technology and AI, I think we are on the cusp of a whole new level of understanding and opportunity.

Finally, I can’t imagine a better set up for closing remarks by Dr. Adaora Adimora, from the University of North Carolina, and Dr. Emily Erbelding, from NIAID.  I didn’t capture who made these particular comments, but the message was clear:  Under all of this is structural discrimination.  The most effective way to address not only COVID and longcovid as well as the next pandemic is by addressing the underlying structural causes of disparity and discrimination.  That is the real challenge we now face.

Post script

This summary does not give recognition to all the many wonderful speakers or ideas over these two days.  I want to thank each of them for their enormous contributions, so clearly communicated.

I also want to reiterate what I said at the top.  I approach this with a limited knowledge base and interest.  Others most certainly would come away with a significantly different set of observations, which I’m certain would celebrate as well.  I am really looking forward to reading the full report.  And I can’t wait for the next NIH workshop on this important topic.

---

[1] My use of the terms “Black people” and “people of white ethnicity”, including the capitalization patterns, follows the conventions used in the underlying study report published in Nature.