More than 400 million people worldwide live with Long COVID. Five years after SARS-CoV-2 emerged, there is still no FDA-approved treatment. But a new generation of clinical trials — better designed, more targeted, and informed by hard-won failures — is now underway. Here is where the science stands in March 2026.
The Graveyard of Good Ideas
Before we look at what's coming, we need to understand what hasn't worked — because these failures are teaching us how to do better.
Paxlovid (nirmatrelvir-ritonavir) was the first great hope. The antiviral that shortened acute COVID seemed like a natural fit for Long COVID, especially if viral persistence was driving symptoms. But when tested in the PAX LC trial, 15 days of Paxlovid did not significantly improve outcomes compared to placebo. The failure didn't disprove the viral persistence hypothesis — the dose may have been too low, the duration too short, or the reservoirs too inaccessible. Longer courses (25 days) are still being tested in the PROLIFIC and RECOVER-VITAL trials.
BC007 (rovunaptabin) targeted a different mechanism: functional autoantibodies against G-protein coupled receptors. The German startup Berlin Cures ran a Phase II trial, and in November 2024 announced it had failed to demonstrate superiority over placebo. The trial was suspended. However, an independent investigator-led study (reCOVer) found some positive signals for fatigue specifically — hinting that the drug might work for an autoimmune subgroup of patients, even if it doesn't help everyone.
Temelimab, a monoclonal antibody targeting the HERV-W envelope protein (originally developed for multiple sclerosis), was tested in 203 patients with neuropsychiatric Long COVID. Despite a compelling biological rationale, researchers did not observe clinically significant improvement compared to placebo.
Ensitrelvir, an alternative 3CL protease inhibitor already marketed in Japan for acute COVID, showed a trend toward improvement but did not reach statistical significance in reducing Long COVID symptoms at three months. A dedicated Long COVID trial (PREVAIL-LC) at UCSF is still underway.
And the RECOVER-NEURO trial — the first results from any RECOVER clinical trial — tested cognitive training, cognitive-behavioral rehab, and brain stimulation across 328 adults. All groups improved modestly. None outperformed the others. As lead author David Knopman of the Mayo Clinic put it: “None of our rehabilitation approaches proved effective.”
Five trials. Five failures. These aren't just disappointments — they're data. And they're reshaping how the next generation of trials is being designed.
The Lesson: Long COVID Is Not One Disease
The emerging consensus from these failures is stark: you cannot treat Long COVID as a single condition with a single drug. A commentary in The Lancet Infectious Diseases warned that trials which don't account for Long COVID's heterogeneity risk false-negative results — real responder subgroups can be invisible in the overall statistics.
UCSF researchers are now drawing explicit parallels to HIV treatment: single-drug therapies failed for years before triple-drug combinations transformed the disease from a death sentence to a manageable condition. Long COVID may need the same approach — not one magic bullet, but targeted combinations matched to biological subtypes.
The Big Four: RECOVER-TLC Trials
The RECOVER Treatment to Lessen the effects of long COVID (TLC) program represents the next wave: four large, well-funded trials designed with the lessons of failure in mind. The Foundation for the National Institutes of Health (FNIH) is coordinating, with $1.8 billion in total RECOVER funding behind them.
1. Baricitinib (REVERSE-LC) — Now Enrolling
The JAK inhibitor baricitinib, already FDA-approved for rheumatoid arthritis and severe COVID, is being tested in 550 participants across 17 sites, led by Dr. Wesley Ely at Vanderbilt. The trial is targeting both neurocognitive and cardiopulmonary symptoms over a 6-month treatment course — significantly longer than most failed trials. Neurocognition results are expected November 2026, with complete data in July 2027.
The rationale is strong: baricitinib blocks JAK-mediated inflammatory signaling, and mounting evidence implicates persistent inflammation as a central driver of Long COVID. This is the furthest along of the RECOVER-TLC trials.
2. Low-Dose Naltrexone (LDN) — Enrollment Expected Summer 2026
LDN has been one of the most widely used off-label treatments in the Long COVID community, and the observational data is encouraging. A meta-analysis found moderate effects on fatigue (Hedges' g = -0.74), while a mechanistic study showed LDN restores TRPM3 ion channel function in natural killer cells — a specific, measurable biological effect.
Critically, the RECOVER-TLC LDN trial is designed as a “pivotal” trial — large enough that positive results could directly initiate the FDA approval process. It targets youth ages 6-25, an underserved population in Long COVID research. At doses below 5mg, naltrexone acts as a glial modulator rather than a full opioid antagonist, triggering endogenous opioid upregulation through a brief 4-6 hour receptor blockade.
3. GLP-1 Receptor Agonist (Semaglutide) — Enrollment Expected Late Summer 2026
The "dark horse" of the treatment race. GLP-1 receptor agonists — the same drugs behind the weight-loss revolution — have accumulated remarkable anecdotal evidence for Long COVID. Dr. David Kaufman presented data from 350 Long COVID patients at a RECOVER-TLC workshop, reporting 60-90% improvement rates.
The evidence goes beyond anecdote. Scripps Research has launched the LoCITT trial testing tirzepatide (a dual GIP/GLP-1 agonist), co-led by Eric Topol. Solve M.E. is funding a semaglutide study at Berlin's Charité hospital under Prof. Scheibenbogen. And a 47-patient case series found 89% of patients with mast cell activation syndrome reported overall clinical benefit from GLP-1 drugs.
The proposed mechanisms are multifaceted: anti-inflammatory, neuroprotective, improved microvascular blood flow through reduced capillary inflammation, and anti-thrombotic effects. Whether the hype matches the biology will depend on rigorous trial data.
4. Stellate Ganglion Block (SGB) — Protocol Design Phase
The most unconventional of the four: a local anesthetic injection to the cervical sympathetic chain, essentially “resetting” the autonomic nervous system. A 41-patient retrospective study reported 86% symptom reduction, and a prospective pilot showed significant improvements in post-exertional malaise, sleep, cognition, and orthostatic intolerance.
But the data is very preliminary, and there's a practical challenge: Horner's syndrome (temporary drooping eyelid) as a side effect makes blinding difficult. About 25% of participants reported adverse events including headaches and worsened post-exertional malaise. The protocol is still being designed.
Beyond the Big Four
Several other trials are generating data that could reshape the landscape:
RECOVER-AUTONOMIC — perhaps the most anticipated near-term result. This trial tested ivabradine for moderate POTS (postural orthostatic tachycardia syndrome) in approximately 380 patients. It is the largest randomized controlled trial ever conducted for any POTS treatment in Long COVID. Results will be presented as a late-breaking clinical trial at the American College of Cardiology conference on March 28, 2026 — just weeks from now.
Anktiva (nogapendekin alfa) — ImmunityBio's IL-15 agonist takes a different approach entirely: rather than targeting inflammation or viral persistence directly, it aims to boost the immune system's ability to clear lingering virus. Already approved for bladder cancer, the Phase 2 COVID-4.019-Long trial is enrolling up to 40 participants. Patrick Soon-Shiong, ImmunityBio's founder, has argued that “antiviral strategy alone looks insufficient” — the immune system needs reinforcement.
Bezisterim has received FDA IND authorization for a 200-patient Phase 2 trial. And maraviroc + atorvastatin — a combination of an HIV drug and a statin — has had its RCT proposal accepted by the FDA, representing one of the first formal combination therapy trials.
The Breakthroughs That Aren't Drugs
While we wait for pharmaceutical answers, diagnostic science is making remarkable progress. A team at Yokohama City University developed a PET imaging technique that can identify Long COVID brain fog with 100% sensitivity and 91% specificity by measuring AMPA receptor density in the brain. This isn't just a diagnostic tool — it points toward a potential treatment. AMPA receptor antagonists like perampanel are already FDA-approved for epilepsy. Repurposing them for Long COVID brain fog is now a testable hypothesis.
Meanwhile, researchers at Karolinska Institutet and Cardiff University identified soluble biomarker signatures that can detect lung damage in Long COVID patients — even when cellular immunity appears intact. And an Ohio State study using objective sensory testing found that 65.5% of Long COVID patients had measurable smell deficits, 53.4% had hearing loss, and 31.6% had balance impairment — revealing a pattern of shared inner ear and central nervous system involvement that self-report alone would miss.
These discoveries matter because they transform Long COVID from a syndrome defined by symptoms into one defined by biology. You can't design targeted treatments without targets.
The Road Ahead
Where does this leave 400 million patients?
The honest answer: in limbo, but with more reason for cautious optimism than at any point in the past five years. The failed trials weren't wasted — they've taught us that Long COVID requires the same precision medicine approach that revolutionized cancer treatment and HIV management. One size does not fit all.
The timeline is sobering. Even if the RECOVER-TLC trials succeed, the earliest realistic window for an FDA-approved Long COVID treatment is late 2027. The baricitinib trial won't have complete data until July 2027. The LDN and semaglutide trials haven't started enrolling yet.
But the RECOVER-AUTONOMIC results on March 28 could change the conversation. If ivabradine shows efficacy for POTS — a well-defined, objectively measurable symptom — it would be the first positive signal from a large, rigorous Long COVID treatment trial. That matters enormously, even if POTS is only one piece of the puzzle.
The treatment race isn't a sprint. It's a marathon with obstacles that keep reshaping the course. But for the first time, the runners have a map.
This post synthesizes publicly available research findings and trial data. It is not medical advice. Treatment decisions should be made with qualified healthcare providers. Clinical trial information is current as of March 9, 2026.