BioVie identifies Long COVID biomarker subgroups in Phase 2 trial

BioVie has reported two biomarker-defined Long COVID subgroups in its Phase 2 bezisterim trial, alongside a treatment-associated reduction in neuropeptide Y.

The findings from the ADDRESS-LC study suggest that patients with different biological characteristics may experience different responses to bezisterim, an investigational oral treatment targeting inflammation and insulin resistance.

Presented at the 4th Long COVID International Conference in Nice, France, on 13 November, the analyses identified an inflammatory subgroup associated with improvements in objective cognition and a low white blood cell subgroup showing larger effects on fatigue and post-exertional malaise (PEM).

The company also reported a reduction in circulating neuropeptide Y (NPY), a signalling molecule involved in nervous system, endocrine and immune activity.

However, the findings are exploratory. The overall Phase 2 trial did not achieve statistical significance on any endpoint in its intention-to-treat population.

Biomarker analysis identifies two patient subgroups

The ADDRESS-LC trial enrolled 203 participants, randomised to receive either bezisterim at 20mg twice daily or placebo for 12 weeks.

BioVie analysed 380 plasma biomarkers to investigate whether biological differences could help explain variations in treatment response.

The first subgroup, comprising 67 participants, was characterised by higher levels of serum amyloid A1 (SAA1), greater interleukin-6 (IL-6) and myeloid activity, and higher high-sensitivity C-reactive protein (hs-CRP) and body mass index.

In this high-inflammatory group, bezisterim was associated with an effect size of approximately 0.83 for objective global cognition at week 12, with a p-value of 0.001.

The second subgroup included 44 participants with lower white blood cell and myeloid cell levels, including lower neutrophil and monocyte counts.

In this group, treatment was associated with effect sizes of approximately 0.80–0.90 for fatigue and PEM measures. However, the p-value was approximately 0.11, meaning these findings did not reach conventional statistical significance.

The results suggest that different clinical measures may be more informative in different biological subgroups. Objective cognitive assessments could be particularly relevant in patients with higher inflammatory activity, while fatigue and PEM measures may be more useful in those with lower white blood cell and myeloid activity.

Neuropeptide Y reduction provides further biomarker evidence

A second finding involved NPY, which has roles in regulating stress responses, autonomic activity and communication between the nervous and immune systems.

Across the ADDRESS-LC study population, bezisterim treatment was associated with a reduction in circulating NPY at week 12.

The treatment contrast was -0.58 log2 normalised protein quantification units, corresponding to approximately 33% lower relative abundance. The finding had a false discovery rate-adjusted q-value of approximately 0.01.

BioVie said NPY is being added to the National Institutes of Health’s RECOVER biomarker prioritisation list, which identifies potential biological markers for further investigation in Long COVID.

Previous research has reported elevated circulating NPY in people with myalgic encephalomyelitis/chronic fatigue syndrome (ME/CFS) and Gulf War Syndrome. However, whether the reduction observed with bezisterim represents normalisation of an underlying disease process remains a hypothesis requiring further investigation.

NPY is not currently a validated diagnostic or predictive biomarker for Long COVID, and its reduction does not establish that patients experienced clinical improvement.

Overall Phase 2 trial did not meet statistical significance

The ADDRESS-LC study was designed as a signal-finding, proof-of-concept trial rather than a definitive assessment of clinical efficacy.

It evaluated 22 clinical outcome measures, with no single endpoint designated as the primary determinant of study success. In the overall intention-to-treat population, none of the endpoints reached statistical significance.

Earlier exploratory subgroup analyses had suggested greater treatment effects among participants with more severe baseline fatigue, PEM or cognitive impairment.

The newly presented biomarker analyses provide additional biological context for those observations, but subgroup findings remain hypothesis-generating and will need to be confirmed in further studies.

Bezisterim demonstrated a safety and tolerability profile comparable to placebo. Treatment-emergent adverse events were reported in 41.6% of participants receiving bezisterim, compared with 55.9% in the placebo group. No deaths occurred, and no serious treatment-emergent adverse events were reported in the bezisterim group.

Penelope Markham, who leads BioVie’s Long COVID clinical development programme, said: “Long COVID is unlikely to represent a single biological disease state, and our findings suggest that differences matter and may play an important role in treatment response.”

She added that the combination of biomarker-defined subgroups and NPY findings provides a rationale for investigating more targeted approaches in future trials.

The study was funded through a $13.13m award from the US Department of War. BioVie is investigating bezisterim in other neurological conditions, including Parkinson’s disease and Alzheimer’s disease.

Mail Icon

news via inbox

Sign up for our newsletter and get the latest news right in your inbox