Abstract
Long COVID, also termed post-acute sequelae of SARS-CoV-2 infection (PASC), affects a substantial minority of people after acute COVID-19 and is now recognized as a distinct clinical entity rather than slow convalescence. This review has two aims: to critically appraise the evidence for the principal proposed mechanisms of long COVID and to evaluate whether heat shock protein 90 (Hsp90) represents a candidate molecular node that intersects several of these mechanisms and may provide a therapeutically tractable target, without implying primacy over other pathways involved in disease pathogenesis. We first summarise the clinical presentation and the diagnostic difficulty created by symptom heterogeneity and by overlap with myalgic encephalomyelitis/chronic fatigue syndrome and other post-infectious syndromes. Then the leading candidate mechanisms, including immune dysregulation, viral antigen persistence, reactivation of latent herpesviruses, autoimmunity, endothelial and microvascular dysfunction, and skeletal muscle and mitochondrial abnormality, grading the strength of the supporting human evidence and stating the principal uncertainties for each, are evaluated. Further, we discuss the roles of Hsp90 on SARS-CoV-2 replication, immune evasion, endothelial barrier dysfunction, and chaperone-directed autoimmunity. Evidence on Hsp90 inhibition is deliberately presented in both directions: several preclinical studies report reduced viral load and inflammation, whereas one inhibitor failed to reduce viral shedding and increased mortality in infected mice, indicating that Hsp90 is a potential therapeutic target because of its involvement across multiple biological processes relevant to long COVID. Management is discussed by symptom domain, with clinical evidence distinguished throughout from in vitro, in silico and animal data, and candidate repurposed agents are summarised with the highest level of evidence currently available for each. In summary, no pharmacological therapy is yet supported by adequate randomised evidence in long COVID, and the value of Hsp90 as a target will depend on isoform selectivity, treatment timing and biomarker-defined patient selection.

This publication has 46 references indexed in Scilit: