Opinion The next phase in Long COVID research: addressing the ethical challenges in trials of disease-modifying treatments, 2026, Hendriks

[rvallee]

The next phase in Long COVID research: addressing the ethical challenges in trials of disease-modifying treatments

https://www.thelancet.com/journals/eclinm/article/PIIS2589-5370(26)00166-5/fulltext

Spoiler: Authors

Saskia Hendriksa ∙ Christine Grady ∙ Megan L. Fitzgerald ∙ Rachel S. Gross ∙ Christine Maughan ∙ Michael J. Peluso ∙ Sumeeta Varma ∙ Avindra Nath ∙ Annette Rida

Summary

Almost five years after COVID-19 emerged, multiple scientific uncertainties remain about why some people experience ongoing symptoms long after being infected with SARS-CoV-2 (Long COVID). The pathophysiology underlying Long COVID and its potential to represent several endotypes are still under investigation. These scientific uncertainties around Long COVID have been cited as a reason to delay treatment trials until the disease is better understood. In this paper, a group of bioethicists, clinician-scientists and people with lived experience with Long COVID argue that it is ethically imperative to conduct trials of disease-modifying treatments for Long COVID now. Furthermore, we argue that although conducting such trials can pose ethical challenges, these challenges can be overcome through careful research priority-setting, rigorous trial design, fair participant selection, and ensuring that the risk-benefit profile is favorable.

 

[Utsikt]

There’s nothing special about LC in terms of urgency. All disease is urgent to the people it affects.

And there is nothing special about trialling drugs for LC compared to other diseases, other than that lots of trials are done in vain due to too broad inclusion criteria and/or poorly chosen outcome measurements.

 

[Jonathan Edwards]

Furthermore, we argue that although conducting such trials can pose ethical challenges, these challenges can be overcome through careful research priority-setting, rigorous trial design, fair participant selection, and ensuring that the risk-benefit profile is favorable.

Except that if you have no idea what the problem is you don't have anything you can justify trying. With vast numbers of people having LC there must be a few who have incidentally been put on almost every drug in the book and we have not heard of any dramatic improvements.

The risk-benefit profile for all drugs at present is negative.

 

[Hutan]

The authors include Michael Peluso and Avindra Nath.

Furthermore, we argue that although conducting such trials can pose ethical challenges, these challenges can be overcome through careful research priority-setting, rigorous trial design, fair participant selection, and ensuring that the risk-benefit profile is favorable.

Yes, it's a weird paper. 'Let's do good research!'. Well, yes. (Although it is in the Lancet, so perhaps that will be news to some there.)


They are still at the stage of treating Long COVID as one thing:

Some common symptoms of Long COVID in adults include fatigue, brain fog, dizziness, gastrointestinal symptoms, palpitations and post-exertional malaise.7,8 Currently, it is not possible to clearly exclude any particular symptom from representing Long COVID. Multiple different phenotypes are framed under the umbrella of Long COVID, yet whether these are all caused by SARS-CoV-2 is uncertain.

So, yeah, I think we can expect a whole lot more trials in inadequately characterised and stratified samples.

Taken together, there is a strong ethical case that the time is ripe for more trials of disease-modifying treatments for all PWLC, including groups who might require special protections in clinical trials. For example, in our view, more treatment trials in children are needed now because children experience a high disease burden from Long COVID and some of the identified disease-modifying treatments are likely to have a reasonable risk-benefit profile for them in clinical trials (e.g., low-dose naltrexone).62–64 Moreover, delaying trials in children may risk wide implementation of treatments based on adult data only, whereas pediatric trials are needed to rigorously assess candidate treatments in children given that clinical outcomes in children may differ from adults.

Oh, my goodness. We still don't have any idea what works in adults and they want to put children through trials? Low-dose naltrexone.....

In pediatric trials with a prospect of benefit, U.S. regulations generally require that risks are justified by the anticipated benefits to participants, and that the risk-benefit profile of the treatment be at least as favorable as existing alternatives (45 CFR 46.405a/b). This will often favor treatments that are either already used in pediatric clinical practice or have significant supportive data from adults. However, testing novel treatment candidates in children and adolescents can be justifiable and avoid delays.

 

[Hutan]

There's a bit of nuance about the Long COVID concept later:

This challenge is further increased by the possibility that Long COVID represents several, potentially overlapping, endotypes with distinct pathophysiological mechanisms which require distinct disease-modifying treatments, as any given treatment candidate may benefit an unknown proportion of PWLC.20,32,38,68,84 For example, PWLC with post-Covid chronic kidney disease likely need different treatments than PWLC with anosmia and fatigue. While unknown endotypes exist in other fields (e.g., oncology), the heterogeneous pathophysiology in Long COVID38 creates complex challenges which distinguishes it from most other diseases.

But really - only the "possibility" that Long COVID represents several.. endotypes? A major problem we have seen in Long COVID research is that there is inadequate effort to select and stratify participants. So many papers are useless as a result, and it's even worse in treatment trials.

For example, inclusive, symptom-based definitions of Long COVID like the one proposed by NASEM may have low diagnostic specificity, as NASEM itself recognized.12 If such definitions underpin trial inclusion criteria, persons with conditions other than Long COVID but similar symptoms might be included, making it more challenging to detect treatment effectiveness.38,44 Additionally, to the extent that Long COVID may represent many, potentially overlapping, endotypes with distinct pathophysiological mechanisms,20,32,38,68,84 treatment effects might be masked in trials that involve participants with multiple endotypes.

 

[Hutan]

Define eligibility criteria and design recruitment strategies in ways that enhance generalizability (e.g., include participants from potentially clinically distinct PWLC subgroups, demographically diverse participants,101 and other groups with potentially distinct risk-benefit profiles such as pregnant persons, immunocompromised persons or children and adolescents102),44 provided this does not undermine the trial’s scientific validity, risk-benefit profile, and protections for participants are appropriate.

They say this about Long COVID, seeming to advocate for the inclusion of pregnant people when knowledge of useful treatments is so inadequate. But, ME/CFS is not mentioned at all.

In a similar vein, although people with infection-associated chronic conditions other than Long COVID might reasonably benefit from being included in some trials (e.g., as an additional comparator group where scientifically appropriate36,49,97), sponsors and investigators might limit inclusion to PWLC to exercise stewardship of limited resources. As long as resources are limited and developing disease-modifying treatments for Long COVID remains urgent, sponsors and investigators should carefully consider whether the social value of using broader inclusion criteria justifies any additional time and resources needed to complete a given trial. How to balance these trade-offs may depend on the trial and context; for example, ensuring representativeness may be more important in large RCTs than in small proof-of-concept trials.

There's a slight acknowledgement of 'infection-associated chronic conditions', but there seems to be encouragement to exclude such cohorts in order to 'exercise stewardship of limited resources'.

It seems to me that it is probably very hard to be sure that someone has Long COVID ME/CFS versus post-2020 ME/CFS triggered by other diseases. The authors are seemingly keen to bundle people with various sorts of Long COVID into a trial, but decidedly not keen to recognise ME/CFS, and the utility of trials with strict ME/CFS criteria.

It's a frustrating and disappointing paper, particularly coming from someone as influential as Nath. On the one hand there is repeated promotion of the inclusion of everyone including children and pregnant people and people with all sorts of post-COVID symptoms, so that they don't feel left out. On the other hand, there seems to be a huge reluctance to acknowledge that a portion of Long COVID is the very recognisable symptom pattern of ME/CFS.

 

[EndME]

And they have the PPI part covered with Patient-Led signing this off.