Long COVID endotheliopathy: hypothesized mechanisms and potential therapeutic approaches

Abstract

SARS-CoV-2-infected individuals may suffer a multi-organ system disorder known as "long COVID" or post-acute sequelae of SARS-CoV-2 infection (PASC). There are no standard treatments, the pathophysiology is unknown, and incidence varies by clinical phenotype. Acute COVID-19 correlates with biomarkers of systemic inflammation, hypercoagulability, and comorbidities that are less prominent in PASC. Macrovessel thrombosis, a hallmark of acute COVID-19, is less frequent in PASC. Female sex at birth is associated with reduced risk for acute COVID-19 progression, but with increased risk of PASC. Persistent microvascular endotheliopathy associated with cryptic SARS-CoV-2 tissue reservoirs has been implicated in PASC pathology. Autoantibodies, localized inflammation, and reactivation of latent pathogens may also be involved, potentially leading to microvascular thrombosis, as documented in multiple PASC tissues. Diagnostic assays illuminating possible therapeutic targets are discussed.

Figure 1. Clinical signs and symptoms distinguish long COVID (PASC) from acute COVID-19.

(A) Common clinical signs and symptoms as well as comorbidities and other cofactors for disease progression distinguish long COVID, also known as PASC, from acute COVID-19. In particular, female sex at birth is linked to a higher incidence of PASC, while male sex at birth is a risk factor for acute COVID-19 progression. Multiple metabolic and cardiovascular risk factors exacerbate acute COVID-19. Overall, age, BMI, and prior respiratory or cardiovascular history do not affect the incidence of PASC but may influence its clinical phenotype. (B) An acute thromboinflammatory process characterizes acute COVID-19. Clinical progression parallels biomarkers of EC injury and a hyperinflammatory state, including enhanced release of proinflammatory cytokines and chemokines, activation of complement and coagulation cascades, platelet activation, NETosis, and, ultimately, hypoxia. IFN-I signals can promote an antiviral response via MxA and exacerbate inflammation. AKI, acute kidney injury; AMI, acute myocardial infarction; ANC, absolute neutrophil count; ATN, acute tubular necrosis; CRP, C-reactive protein; CVD, cardiovascular disease; DM-1, DM-2, diabetes mellitus types 1 and 2; DVT, deep vein thrombosis; eGFR, estimated glomerular filtration rate; ENT, ear, nose, and throat; ESRD, end-stage renal disease; LFTs, liver function tests; NET, neutrophil extracellular trap; PASC, post-acute sequelae of SARS-CoV-2 infection; PE, pulmonary embolism; POTS, postural orthostatic tachycardia syndrome; TF, tissue factor; vWF, von Willebrand factor.

Figure 2. Potential mechanisms underlying long COVID (PASC).

Multiple factors may contribute to the pathophysiology of at least certain PASC phenotypes, either directly or via microvascular injury. This occurs in the absence of the hypercoagulable state and systemic immune activation that characterize acute COVID-19. It is proposed that cryptic reservoirs of SARS-CoV-2 in multiple tissues, acting via direct EC infection, release of soluble viral products, infection of monocytes adjacent to the vasculature, or activation of other viruses, such as EBV, which can transmit virus to microvascular ECs, lead to cytopathic effects in ECs. Specific evidence for this is outlined in Tables 1 and 2. Autoantibodies, particularly against IFN-I, leading to suppression of the antiviral MxA pathway shown in Figure 1B, or against ECs, may also be involved. Development of SARS-CoV-2 cryptic reservoirs may be facilitated by lower antiviral immunity. Vaccinated individuals have a much lower incidence of PASC than the unvaccinated.

Figure 3. Proposed diagnostic assays and potential treatment targets in PASC.

Abnormalities on lung CT and brain MRI have been correlated to functional changes, including dyspnea with air trapping and cognitive deficits, respectively, but they are not sensitive enough to distinguish an inflammatory versus a microthrombotic process. As PASC, like COVID-19, is a systemic process, we hypothesize that a simple 4 mm cutaneous punch biopsy of normal-appearing deltoid skin, a diagnostic method our group has employed for over a decade to investigate thrombotic microangiopathies linked to atypical hemolytic-uremic syndrome and hematopoietic stem cell transplantation and, most recently, acute COVID-19, should permit pathophysiologic explorations of PASC. Direct biopsy of other accessible tissues, including lung and peripheral nerve, could also be pursued. This could enable collection of evidence for vascular damage, microthrombi, and direct SARS-CoV-2 infection. Viral signals in stool and peripheral blood and hematologic/immunologic abnormalities linked to PASC may also be followed longitudinally. It should be recognized that the possible treatments illustrated are based on pathophysiology hypotheses and have not been evaluated in clinical trials. IVIG, intravenous immunoglobulin; sTM, soluble thrombomodulin.