2nd Annual RECOVER-TLC Research Workshop : A summary

Part I: Workshop Summary

BACKGROUND: RECOVER-TLC is an offshoot of RECOVER, the National Institutes of Health’s (NIH) $1.8 billion Long COVID research program. It launched after advocates demanded that more money go toward clinical trials so that patients get safe and effective treatments as rapidly as possible. TLC was allocated ~$300 million for FY2025-2029. The 2025 workshop included masked in-person and virtual participation. A full recording is now available to stream.

AT THE WORKSHOP: NIH staff announced that RECOVER-TLC will fund four pharmaceutical drug treatment trials. It also gave updates on RECOVER’s older clinical trial platform, observational cohorts, and research projects related to identifying biomarkers, assays, pathogenesis, and future treatment pathways. Patients shared their experiences, knowledge, and expertise on both days, through independent presentations and panel participation. Part I of this memo is a quick run-down of all that was covered. Part II is the community perspective on what needs to happen next. 

THE RECOVER-TLC CLINICAL TRIALS: RECOVER-TLC will begin clinical trials on the interventions low-dose naltrexone (LDN); glucagon-like peptide-1 receptor agonist (GLP-1) and stellate ganglion blocks (SGBs). It also expanded funding for an existing trial, called REVERSE-LC, on the drug baricitinib. NIH staff said their decision to trial these drugs was based on feedback from last year’s workshop, and treatment suggestions solicited from the public. Patients showed up: of 572 total submissions, 79% were from patients or caregivers. People with lived experience also make up 26% of TLC’s working groups. 

Low Dose Naltrexone (LDN) will be trialed in ~1,300 people aged 6-25 across ~100 sites and primarily measure fatigue severity. LDN is already being explored off-label for Long COVID and ME/CFS in multiple trials outside the NIH. The trial aims to develop a “regulatory-grade LDN study product” that may be officially authorized as a Long COVID treatment. This may make it more rapidly accessible to people with Long COVID. 

Glucagon-like Peptide-1 Receptor Agonist (GLP-1)’s trial is in the design stage. GLP-1 may have anti-inflammatory effects in the brain and modulate the immune system. Hundreds of clinicians who treat patients with complex chronic illness are finding that over half their patients have positive effects from GLP-1. 

Stellate Ganglion Block (SGBs) trial is in the design stage. To administer an SGB, a provider injects an anesthetic medicine into the stellate ganglion nerves to “block” overactive sympathetic nervous system activity. It is a minimally invasive procedure. Case studies show that some patients improve with stellate ganglion block therapies, while others have not. 

Baricitinib: REVERSE-LC is the existing clinical trial being expanded from 4 to 15 study sites. Baricitinib may modulate inflammation and immune responses. The trial primarily explores whether it can reverse Long COVID-related cognitive impairment. 

UPDATES ON THE RECOVER-CT CLINICAL TRIALS: Launched in 2023, eight different Phase II trials for adults are testing 13 interventions (pharmaceutical and non-pharmaceutical) across ~100 sites. These trials are organized into five different categories of study they call “platforms.” Patient advocates protested against the “ENERGIZE” platform last year, because it included a CPET exercise study. Since then, RECOVER has added a pacing study arm to ENERGIZE, but they are still carrying out the CPET study, and it is unclear how they are screening for or managing PEM. All these studies should complete enrollment by the end of this year, with results disseminated by summer 2027. 

UPDATES ON OBSERVATIONAL COHORTS - PEDIATRIC, PREGNANCY, AND ADULT: These cohorts are following large groups of people with Long COVID over time and collecting a lot of data and biospecimens. The pediatric cohort (infants to 25 year olds) has over 22,000 participants across 100 sites, and is racially diverse. So far, the study found that some symptom clusters are unique to specific age ranges. The cohort of pregnant individuals and their offspring now has complete enrollment of 1,822 maternal-child dyads. Among those infected with SARS-CoV-2 during pregnancy, the prevalence of Long COVID was found to be about 9%. Future analyses will examine differences in child neural development as well as changes in menstrual cycle and female sexual function. The adult cohort has over 14,000 adults who have been followed for multiple years. The hope is that findings from all cohorts can inform clinical trials by providing information on patient symptoms, natural history, symptom clusters, and self-reported medication use. 

ASSAYS, BIOMARKERS, PATHOGENESIS, TRIAL DESIGN: The second day of the workshop covered important areas of focus in Long COVID research, outside of TLC. Long COVID still has no validated, specific diagnostic tests or singular biomarkers, and while there are leading hypotheses about how Long COVID develops, pathogenesis is also not known. Because Long COVID is complex, it’s possible that a combination of tests or assays will identify Long COVID, or that there is more than one type of Long COVID and each will need to be identified differently. At the workshop, scientists discussed specific projects outside the NIH that are investigating these areas. Scientists also discussed adaptive platform trials, which test multiple treatments in the same trial, as a potentially relevant study design for a complex disease like Long COVID. Some progress has been made in all these areas, with much more needed, and with the need for collaboration between academia, industry, and regulatory bodies to advance biomarker-based drug development. Continuing study into these areas will inform mechanistic trials and therapeutic interventions. 

Part II: Community Perspective

The workshop included patient perspective talks throughout, including a teenager who attends high school at home, caretakers, and a world-class cellist. NIH staff reported to the community on the processes used and progress made in the past year, including how TLC has been shaped by patient input. NIH staff said that they plan to continue advancing meaningful patient engagement.

RECOVER-TLC specifically is also, so far, funding pharmaceutical treatment trials. Patients strongly advocated for this during the first round of RECOVER-CT trials, which included multiple non-pharmaceutical interventions, like exercise. 

WHAT STILL NEEDS TO IMPROVE: The community needs to see continued evidence of improved patient engagement at every stage of the research. RECOVER also needs to be more clear about how it incorporates feedback, and what feedback is and is not incorporated (and why). 

RECOVER also needs to systematically include people with post-exertional malaise (PEM) and the housebound population when designing trials. Of adults who currently have Long COVID, 80% experience activity limitation and 24% experience significant activity limitation, similar to the percentage of those housebound due to ME. If researchers exclude a specific set of patients with a common disease presentation– like being housebound or bedbound, or experiencing PEM– drugs aren’t being tested on those who are representative of the Long COVID population, which could affect results. Instead of excluding them, trials must be designed for people with PEM. Patient and policy researcher Dr. Victoria Copeland advocated for time allotments for testing that allow for pacing, at-home testing when possible, and PPE and proper ventilation at testing sites. Systematically including people with PEM also means that every clinical trial and cohort needs to use an instrument that measures PEM. 

Long COVID disproportionately impacts people who face structural inequities, including QT-BIPOC people, and yet studies do not appropriately represent them. Of the 11 registered, completed Long COVID clinical trials, two didn’t report race demographics, and six had 80% or more white participants. Dr. Leticia Soares, a Brazilian scientist and disability activist, advocated for intentional engagement with QT-BIPOC scientists, diversifying enrollment strategies, thinking intersectionally, and making treatments that are low-cost and globally accessible, for instance by trialing repurposed drugs. People who are pregnant and immunocompromised, with multiple diagnoses, and children also need to be studied in trials. 

Patients also emphasized that studies are not advancing with enough urgency, or being translated to results usable to patients rapidly enough. Millions have been suffering for years, and still there is no FDA-approved treatment. There are challenges specific to the field that make speed difficult. For instance, Long COVID does not have biomarkers or validated instruments to measure outcomes. To understand Long COVID, mechanistic and exploratory trials must take place in addition to treatment trials, with limited resources at hand. The patient community knows that these realities are due to the way science and medicine have dismissed infection-associated chronic diseases like ME for decades. If the US had funded research into ME/CFS and other infection-associated chronic illnesses since outbreaks were reported in the 1980s, the field would have advanced, and answers for those with Long COVID would be much closer to hand. People with Long COVID are just the most recent inheritors of this lack of funding and focus.

Finally, ME organizations have asked that RECOVER use subgroup tracking and analysis for ME/CFS, POTS/Dysautonomia, and MCAS in all trials. Among those with Long COVID, ME/CFS may affect up to 50% and POTS may affect around 70% of those with Long COVID at six months. When studied, MCAS symptoms are significantly increased in Long COVID patients compared to controls. It’s vital to employ clinical phenotyping to determine which interventions are successful for which people with Long COVID. Given their high prevalence and significant contribution to disease presentation, we cannot treat Long COVID without accounting for these diseases in the Long COVID population.