Long COVID: Clinical characteristics, proposed pathogenesis and potential therapeutic targets

The paper addresses the emerging clinical syndrome of long COVID, defined as signs, symptoms, or conditions persisting for at least four weeks after SARS-CoV-2 infection. It situates long COVID within the broader context of post-infectious sequelae seen after other viral illnesses (e.g., SARS, MERS, EBV, Q fever, Ross River virus), noting the heterogeneity of symptoms—commonly fatigue, cognitive dysfunction, and respiratory complaints—and the lack of a unifying pathological mechanism. The authors highlight variability in definitions across studies (4–12 weeks), the high prevalence (up to one-fifth of those infected reporting symptoms), and substantial functional impact (e.g., high rates of work incapacity). They outline the purpose: to review clinical phenotypes, organ-specific sequelae, proposed pathophysiologic mechanisms, therapeutic approaches, and research gaps to guide future work.

The review synthesizes evidence on: (1) clinical phenotypes identified through both manual subgrouping and unsupervised clustering, consistently revealing cardiorespiratory, high-symptom musculoskeletal/pain, low-symptom/anosmia-predominant, and neuropsychiatric clusters; (2) organ-specific sequelae, including pulmonary fibrosis and diffusion impairment after severe disease, pulmonary vascular defects and potential chronic thromboembolic pulmonary hypertension, persistent dyspnea despite normal tests in mild cases, cardiovascular complications (ischemic heart disease, pericarditis, myocarditis, arrhythmias, heart failure), musculoskeletal weakness and myalgia, gastrointestinal symptoms with transient liver enzyme elevations, dermatologic conditions including pernio and telogen effluvium, increased risk of incident diabetes, and neuropsychiatric outcomes (anosmia/ageusia, stroke, anxiety/depression, and cognitive impairment); (3) proposed mechanisms: immune dysregulation (persistent T cell activation/exhaustion, altered B cells, innate activation, cytokine elevations including IL-6, TNFα, IP-10, markers of BBB disruption), viral persistence and latent virus reactivation (SARS-CoV-2 antigens/RNA in tissues and fluids; EBV serologic evidence), autoimmunity (diverse autoantibodies including anti-interferon, anti-receptor antibodies), microbiome dysbiosis with microbial/fungal translocation, dysautonomia with potential impaired peripheral oxygen extraction, and endotheliopathy/coagulopathy including sustained D-dimer and endothelial activation; (4) therapeutic landscape: no unifying treatment; exploration of antivirals (nirmatrelvir-ritonavir, remdesivir), anti-inflammatories and antihistamines, anticoagulation for select high-risk patients post-discharge, and symptomatic management (pacing, targeted agents like guanfacine plus NAC for brain fog). The review underscores heterogeneity in cohorts, symptom measurement, and analytic methods, and notes potential influences of variant and vaccination on phenotypes.

This is a narrative review synthesizing published studies and reports on long COVID. The authors collate findings from observational cohorts, electronic health record analyses, unsupervised clustering studies, organ-specific investigations (imaging, pulmonary function tests, CPET), immunologic profiling (cellular and cytokine studies), microbiome analyses, and early therapeutic trials or case series. No formal systematic review methodology is described; rather, representative and influential studies are summarized to delineate clinical phenotypes, mechanisms, and therapeutic implications.

- Prevalence and impact: Up to ~20% of people with prior COVID-19 report long COVID; substantial functional limitations with 44% of respondents in one survey unable to return to work. - Phenotypes: Consistent data-driven clusters include (1) cardiorespiratory (predominant breathlessness with cough/chest pain/palpitations); (2) high-symptom musculoskeletal/pain (joint pain, myalgia); (3) low-symptom or anosmia-predominant; (4) neuropsychiatric (brain fog, depression, insomnia). Patients may transition between phenotypes over time. - Organ-specific sequelae: - Pulmonary: Post-ARDS fibrosis and diffusion impairment after severe disease; perfusion defects/angiopathy; potential chronic thromboembolic pulmonary hypertension; dyspnea common even with normal imaging and PFTs in mild cases. - Cardiovascular: Elevated risks of ischemic heart disease, pericarditis, myocarditis, arrhythmias, and heart failure up to 12 months; chest pain/palpitations with prolonged relative tachycardia (median ~79 days post-onset in one study). - Musculoskeletal: Myalgia, arthralgia, muscle weakness (including post-ICU syndromes); weakness also reported after mild illness. - Gastrointestinal/Hepatic: New GI symptoms up to months; ~25% with elevated liver enzymes at 1 month post-discharge, usually transient without underlying pathology. - Endocrine/Metabolic: Increased risk of incident diabetes (type 1 and type 2) within 12 months; risk correlates with acute severity and exceeds that after other URTIs. - Dermatologic: Rashes of variable duration; pernio often >60 days; hair loss reported in ~22% of hospitalized at follow-up (commonly telogen effluvium, but also alopecia areata). - Neurologic/Psychiatric: Persistent anosmia/ageusia; increased cerebrovascular events post-infection; objective cognitive impairment documented even without subjective complaints; increased anxiety/depression up to 12 months; increased prescribing of psychoactive agents. - Mechanisms: - Immune dysregulation: Persistent activated/exhausted T cells (PD-L1, TIGIT), reduced naive B cells, innate activation (HLA-DR+CD38+ myeloid cells), elevated interferons and cytokines (IL-6, TNFα, IP-10); in brain fog, circulating markers of BBB disruption (TGFβ, IL-8). Organ-level signatures (e.g., interferon and neutrophil signaling in lungs). - Viral persistence/reactivation: SARS-CoV-2 nucleocapsid/spike detected months later in gut, blood, monocytes, lung, urine; fecal RNA despite negative nasopharyngeal PCR; EBV reactivation markers associated with long COVID symptoms in some studies. - Autoimmunity: Autoantibodies (e.g., anti-interferon, anti-angiotensin/ACE/muscarinic receptors); afucosylated anti-spike IgG linked to inflammation; associations with specific long COVID symptoms remain inconsistent. - Microbiome/translocation: Persistent gut dysbiosis with reduced immunomodulatory species; increased intestinal permeability; elevated zonulin and beta-D-glucan correlating with IL-6, TNFα, IP-10 and reduced quality of life. - Dysautonomia and impaired oxygen extraction: Evidence of orthostatic intolerance in some; invasive CPET implicates impaired peripheral oxygen extraction as a driver of exertional intolerance in select patients despite normal imaging; possible small fiber neuropathy-like mechanism. - Endotheliopathy/coagulopathy: Elevated D-dimer up to 4 months; reduced fibrinolytic potential and increased thrombin generation post-hospitalization; sustained endothelial activation (factor VIII, soluble thrombomodulin, elevated von Willebrand factor). BBB disruption and cerebral blood flow changes may underlie neurocognitive symptoms. - Therapeutics: - No unifying treatment; supportive strategies include physical/cognitive pacing. - Trials underway for antivirals (e.g., nirmatrelvir-ritonavir) and anti-inflammatories/antihistamines (e.g., colchicine, H2RA famotidine shows benefit in acute disease but no RCTs yet in long COVID). - Anticoagulation: Rivaroxaban reduced VTE at 1 month post-discharge in high-risk patients; broader/longer use remains unproven; guidelines currently cautious pending more RCT data. - Targeted neurocognitive therapy: Small case series suggests guanfacine plus N-acetylcysteine may improve brain fog in some patients. - Prevention: Vaccination and healthy lifestyle adherence associated with lower risk; early antivirals and metformin during acute infection may reduce long COVID risk (emerging evidence).

The synthesis supports the concept that long COVID encompasses multiple clinical phenotypes likely driven by distinct, overlapping mechanisms rather than a single pathophysiologic pathway. Persistent immune activation, potential antigen persistence or latent virus reactivation, autoantibody formation, dysbiosis with microbial translocation, autonomic dysfunction, and sustained endotheliopathy provide biologically plausible links to the diverse symptom clusters and organ-specific sequelae observed. However, most associations are correlative, and causation is unproven in many domains. Phenotyping approaches (including data-driven clustering) can help align patient subgroups with hypothesized mechanisms and guide targeted diagnostic workups and intervention trials. The review emphasizes that heterogeneity in definitions, cohorts, and outcome measures hampers cross-study comparison and therapeutic development. It argues for integrated mechanistic and clinical research, including organ-specific sampling (e.g., CSF, respiratory, gut), to clarify causal pathways and enable precision treatment strategies.

This review consolidates current knowledge on the clinical heterogeneity, organ-specific sequelae, and leading mechanistic hypotheses of long COVID and outlines the nascent therapeutic landscape. Key contributions include identifying recurring phenotypic clusters, summarizing evidence for immune dysregulation, viral persistence/reactivation, autoimmunity, microbiome disruption, dysautonomia, and endotheliopathy, and highlighting early therapeutic directions and prevention strategies. The authors call for coordinated, international, interdisciplinary research programs combining epidemiology, basic science, translational studies, and clinical trials—conducted in parallel—to establish causal mechanisms, validate biomarkers, refine phenotypes, and evaluate targeted therapies. Future research should include longitudinal designs, diverse populations, standardization of definitions and outcomes, and mechanistic studies at organ level to improve generalizability and clinical impact.

The literature is limited by heterogeneity in definitions and timing of long COVID, variability in symptom ascertainment, and reliance on self-reported outcomes. Many studies over-represent previously hospitalized and older individuals, complicating separation of post-critical illness effects from SARS-CoV-2–specific sequelae. Minority groups are under-represented. There is a paucity of longitudinal and mechanistic studies, with most findings being associative rather than causal. Organ-level pathologies may occur without systemic inflammatory signatures, and invasive sampling (e.g., CSF, pulmonary) is limited. Publication and selection biases, multiple comparisons in immune correlates, and changing variants/vaccination status further constrain generalizability.