Four factors shown to predict long COVID
Post-acute sequelae of COVID-19 (PASC), aka ‘long COVID’, is an emerging health care crisis. Clinical factors have been identified that predict the risk of progression to PASC.
Research just published in the journal Cell based on a deep analysis of biological factors present early in the course of infection resolves four factors that anticipate PASC with distinct clinical expressions.
The authors state:
“Post-acute sequelae of COVID-19 (PASC) represent an emerging global crisis…We executed a deep multi-omic, longitudinal investigation of 309 COVID-19 patients from initial diagnosis to convalescence (2-3 months later), integrated with clinical data, and patient-reported symptoms. We resolved four PASC-anticipating risk factors at the time of initial COVID-19 diagnosis.”
They gathered data on patients at clinical diagnosis (T1), acute disease (acute, T2), and 2-3 months post onset of initial symptoms (convalescent, T3). Their analysis included autoantibodies and SARS-CoV2-specific antibodies, global plasma proteomic and metabolomic profiles, and single-cell multi-omic characterizations of peripheral blood mononuclear cells (PBMCs). This was paired with nasal swab and plasma measurements of SARS-CoV-2 viral load. All of this was integrated with electronic health records (EHR) and self-reported symptoms.
Four risk factors predicting long COVID
Type 2 diabetes
SARS-CoV-2 RNAemia
Epstein-Barr virus viremia
Specific autoantibodies
Type 2 Diabetes
This comes as no surprise given the already observed heightened risk for severe COVID due to the adverse metabolic, immunologic, and vascular harms of pre-existing type 2 diabetes.
SARS-CoV-2 RNAemia
SARS-CoV-2 RNAemia at T1, higher levels of COVID RNA in the blood at the time of clinical diagnosis, proved to be not only a risk for mortality in the acute phase of infection but also for PASC. Interestingly, nasal-swab viral load associated only with anosmia/dysgeusia (loss of smell / taste disorders) and did not correlate well with blood levels.
“T1 SARS-CoV-2 RNAemia also provided a biomarker of mortality (Figure S1H), as reported (Gutmann et al., 2021). SARS-CoV-2 nasal-swab viral load significantly associated only with anosmia/dysgeusia”
Epstein-Barr virus viremia
The observation of latent EBV reactivation is particularly interesting because the expansion of the remnants of earlier EBV infections is associated with both immune system polarization that promotes autoimmune diseases (including multiple sclerosis as documented in a recent study) and immunosenescence—the aging of immune cells resulting diminished effectiveness for their primary ‘jobs’ plus the increase of systemic inflammation associated with a legion of chronic diseases.
“While memory PASC were significantly associated with T1 measures of both EBV viremia and SARS-CoV2 RNAemia, the fatigue and sputum PASC were specific for EBV viremia…In sum, reactivation of latent EBV and SARS-CoV-2 RNAemia at T1 are factors that anticipate, to varying degrees, PASC at T3.”
Autoantibodies associate with PASC and mortality
Autoantibodies (autoAbs) are antibodies that target self-tissue in the process of autoimmunity. Both autoimmune conditions in general and autoimmune diseases triggered by infection have increased dramatically in the past couple of decades and recently skyrocketed. That ‘long COVID’ would have an autoimmune component comes as no surprise.
“We interrogated for such a link by measuring a panel of autoAbs at T1 and T3 and comparing against anti-SARS-CoV-2 Abs of different isotypes. The autoAb panel included anti-IFN-α2, and five anti-nuclear autoAbs (ANAs) (Ro/SSA, La/SS-B, U1-snRNP, Jo-1, and P1) commonly associated with Systemic Lupus Erythematosus (SLE) (Choi et al., 2020; Pisetsky and Lipsky, 2020). SLE is an autoimmune disease that shares certain symptoms with PASC (Raveendran et al., 2021) and has also been reported to manifest following COVID-19 (Zamani et al., 2021).”
Importantly, their data confirm that some subjects already had autoimmune antibodies (predictive or reactive phase antibodies’ if in the early asymptomatic phase) prior to COVID infection.
“The SLE-associated ANAs [anti-nuclear antibodies] have already been detected in acutely infected COVID-19 patients (Chang et al., 2021). The use of the SLE-ANA-panel was additionally supported by the observed expansion of atypical memory B cells in both COVID-19 and SLE patients. In SLE, AtMs are generated during chronic inflammation, enriched with autoreactivities, and correlate with disease activities (Jenks et al., 2018).
First, we observed that patients with autoAbs at T3 (44%) already exhibited mature (class-switched) autoAbs as early as at diagnosis (56%), indicating the autoAbs may predate COVID-19, as reported elsewhere (Paul et al., 2021). Analysis of EHR data confirmed that only 6% of autoAb-positive patients had documented autoimmune conditions before COVID-19, suggesting that the autoAbs may reflect subclinical conditions [predictive or reactive phases].”
The authors also confirmed that different antibody patterns associated with different presentations of PASC.
“A third major finding was that anti-SARS-CoV-2 Abs and specific autoAbs associate with different PASC. For example, patients with neurological PASC exhibited slightly higher levels of anti-SARS-CoV-2 nucleocapsid protein IgG, whereas GI-related PASC and sputum production were associated with elevated levels of multiple autoAbs at T3 and even T1. IFN-α2 autoAbs uniquely associated with respiratory viral PASC, even after correcting for age, sex, and disease severity. These observations suggest that T1 autoAb levels may be anticipating biomarkers of certain PASC.”
Antidepressant effect of transcutaneous auricular vagus nerve stimulation.
Bystander activation against CMV
It’s useful to note that, in addition to the expansion of EBV noted above, the authors describe reactions to CMV. Both EBV and CMV are known to be associated with immunosenescence, the process by which immune cells shift to a ‘depleted’ type that is less effective at controlling the remnants of chronic infection while increasing the systemic burden of inflammation. Though a hallmark of aging, this is occurring at younger ages.
“Bystander activation describes the case when T lymphocytes with specificities for unrelated epitopes are activated during an antigen-specific response...Interestingly, CMV-specific CD8 + T cells from COVID-19 patients displayed distinct transcriptome characteristics relative to unexposed healthy controls, with more cytotoxic and less naïvelike signatures even at T3. Notably, although the absolute numbers of cytotoxic CMV-specific CD8 + T cells decrease from T1 to T3, those cells that do persist at T3 are positively associated with GI PASC. These observations, coupled with the absence of detectable CMV viremia, suggests an association of bystander activation of CMV-specific CD8 + T cells with GI PASC.
Bystander activation also occurs with epitope spreading in autoimmune conditions as new tissue ‘targets’ are added over time. It may also be factor in the phenomenon of polyreactive antibodies where antibodies appear reactive to most foods tested due to an overzealous ‘stickiness’; this has been associated with previous infection exposure and remnant burden.
PASC prevention best begun before infection
This study offers indications for preventing and treating PASC that in agreement with the functional immunology phased approach to COVID case management. This entails, on an individual basis, supporting innate and Th1 immunity to both fight infection and better control remnants of previous infections such as EBV and CMV, optimizing glucose and insulin regulation, diminishing other risk factors for loss of immune tolerance that can lead to autoimmunity, and targeted support to reduce the chemistry of hyper-inflammatory reactions and the persistence of chronic, non-purposeful inflammation.
The authors conclude:
“We…reveal that various PASC-anticipating biological factors (PASC factors) can be measured as early as at initial COVID-19 diagnosis, including pre-existing type 2 diabetes, assessments of SARS-CoV-2 RNAemia, EBV viremia, and autoantibodies from patient blood. Since symptoms can often arise from multiple sources, the identification of discrete and quantifiable PASC factors should be of fundamental importance for understanding PASC and developing treatments…Our analyses provided a framework to understand the heterogeneity of “long COVID” and a rich resource for interrogating the biological factors that contribute to PASC, which can potentially be utilized to monitor and guide interventional trials to treat and prevent post-acute COVID-19 symptoms.”
