Persistent symptoms after SARS-CoV-2 infection may reflect ongoing viral presence, immune dysregulation, or both. Current evidence suggests that certain measurable signals — viral markers, immune signatures, and reactivation of other viruses — are the most promising predictors of whether a patient with long COVID will respond to antiviral therapy. Clinical studies remain incomplete, so these signals are candidate predictors rather than proven guides for treatment.
Candidate viral markers
Detectable SARS-CoV-2 RNA or antigen in blood or tissues is the most direct rationale for giving antivirals. Work by Michel C. Nussenzweig at Rockefeller University documented persistent viral antigen in tissues after acute infection, supporting the hypothesis of tissue reservoirs as a source of prolonged symptoms. Plasma antigenemia or RNAemia has been associated with symptomatic persistence in several cohort reports; when present, these markers plausibly identify patients whose symptoms are driven by residual replication or protein expression and who might therefore benefit from antivirals. Measurement methods and thresholds vary, and positive detection is not universal in long COVID, so findings are context-dependent.
Immune and host biomarkers
Immune profiles that point to ongoing antigen exposure or dysregulated responses are also informative. Akiko Iwasaki at Yale University and colleagues have described distinctive immune activation patterns in people with long COVID, including altered T cell phenotypes and persistent inflammatory signals. Elevated inflammatory cytokines and specific B- and T-cell signatures may indicate a host response sustained by residual viral antigen and correlate with clinical response when antiviral pressure reduces antigen load. Conversely, detection of autoantibodies or strong evidence of immune-mediated damage suggests that antivirals alone may have limited benefit because self-directed immune processes, not active virus, drive symptoms.
Reactivation of latent viruses, especially Epstein–Barr virus, has been reported in multiple studies and may both mimic and amplify long COVID. When EBV reactivation is demonstrated, antiviral strategies targeting SARS-CoV-2 may be insufficient, and antiviral response prediction becomes more complex.
Clinical relevance and consequences hinge on validation. Using these biomarkers to select patients for antivirals could increase benefit and avoid unnecessary treatment, but requires standardized assays, prospective validation, and equitable access. Environmental and territorial factors — such as laboratory capacity and drug availability in underserved regions — will affect implementation. Until randomized trials incorporate biomarker-based enrollment, these markers remain informed, biologically plausible tools to prioritize research and individualized care rather than definitive clinical algorithms. *