
Topline Phase 2 ADDRESS-LC Data Show Subgroup Benefit for Bezisterim in Long COVID
Key Takeaways
- Intent-to-treat analyses showed no statistically significant separation on any of 22 clinical outcome measures, despite favorable directionality on 21 of 22 endpoints.
- Pre-specified severity strata (~78% of participants) identified significant improvements in fatigue, post-exertional malaise, and objective cognition, with Cohen’s d reportedly more than doubling versus overall.
Subgroup analyses from the phase 2 ADDRESS-LC trial showed bezisterim significantly improved fatigue, malaise, and cognition in long COVID patients with more severe baseline symptoms.
BioVie reported topline results from the phase 2 ADDRESS-LC trial (NCT06847191) evaluating bezisterim (NE3107) for neurological symptoms associated with long COVID, with prespecified subgroup analyses in patients with more severe baseline symptoms showing statistically significant improvements in fatigue, post-exertional malaise, and objective cognitive measures.1
Across the full intent-to-treat population, none of the 22 individual clinical outcome measures reached statistical significance, though 21 of the 22 trended in favor of bezisterim over placebo.
Subgroup findings
BioVie said the subgroup analyses, covering roughly 78% of the approximately 200-patient study population, were pre-specified in a statistical analysis plan submitted to the FDA before unblinding, grouping patients by baseline severity of fatigue, post-exertional malaise, or objective cognitive impairment.1
In the high baseline fatigue subgroup, 5 endpoints reached statistical significance, including 3 fatigue-specific measures, with additional favorable trends in post-exertional malaise and sleep. In the high post-exertional malaise subgroup, the company reported statistically significant improvements across malaise, fatigue, and objective cognitive measures, and in the high objective cognitive impairment subgroup, 4 clinically coherent cognitive endpoints reached significance. Treatment effect sizes, expressed as Cohen's d, more than doubled in these subgroups relative to the intent-to-treat population, though BioVie did not report the specific numerical values.1
“These results are among the most promising we have seen in this field so far,” said Michael Peluso, MD, MHS, of the University of California, San Francisco, who said the findings warrant progression to a confirmatory phase 3 trial.1
Safety
Adverse events (AEs) were reported in 41.6% of patients receiving bezisterim compared with 55.9% of those receiving placebo. The most frequent drug-related AE was headache, occurring in 4% of the bezisterim group versus 4.9% of the placebo group. No serious AEs were reported in the bezisterim arm, compared with 1 in the placebo arm. BioVie characterized the overall safety profile as similar to placebo, a characterization that comes from the sponsor and should be attributed accordingly.
“To our knowledge, this is the first time a drug candidate has shown the ability to help patients improve across these persistent long COVID symptoms," Cuong Do, president and chief executive officer of BioVie, said in a statement.1 Grace McComsey, MD, of Case Western Reserve University, added, “It's great to see a trial showing such potential effectiveness while using a well-tolerated oral drug.”
READ MORE:
Bezisterim is an oral, blood brain barrier-permeable insulin sensitizer with anti-inflammatory activity, designed to bind extracellular signal-regulated kinase (ERK) and selectively modulate nuclear factor kappa B (NF-kB) activation and tumor necrosis factor alpha (TNF-alpha) production without broader immunosuppressive effects.3
ADDRESS-LC enrolled adults with a long COVID diagnosis, cognitive impairment, and self-reported fatigue in a randomized, 1:1, placebo-controlled design, with bezisterim dosed as a 20-mg oral capsule twice daily over approximately 3 months.2-4 The trial began enrolling in May 2025 and completed enrollment of approximately 200 patients in May 2026, fully funded through a $13.13 million award from the U.S. Department of War, administered through the Peer-Reviewed Medical Research Program and originally granted under the agency's prior name, the Department of Defense.2,3
At the full-enrollment stage, BioVie described the primary endpoint as change in performance on the Cogstate Cognitive Battery, with secondary endpoints spanning PROMIS cognitive function, fatigue, and sleep disturbance measures, the SF-12, and the DePaul Symptom Questionnaire; in its topline announcement, the company described the trial as evaluating 22 clinical outcome measures with no single endpoint pre-specified as primary.1,2
No treatments are currently approved specifically for the neurological symptoms of long COVID. BioVie said the topline results support advancing bezisterim into a confirmatory phase 3 trial, though the company did not provide a timeline or design details for that next stage.1

















