The Host Response to SARS-CoV-2 Infection Differs by Age

Background. Infection with SARS-CoV-2 and the resulting host immune response has been primarily characterized in middle and older aged populations due to a higher incidence of symptoms in these age groups. Due to reduced severity of disease, children were poorly studied and assumed to be less frequently infected compared to older age groups. We measured the viral load and adaptive immune response across the age-spectrum to define the age-dependent viral and host responses. Methods. From March 2020-March 2022, we enrolled individuals across the age spectrum who presented to U.S. military medical treatment facilities with COVID-19-like symptoms. In this longitudinal cohort study, demographic and clinical data were collected in addition to nasopharyngeal swabs and peripheral blood. Magnitude of viral RNA was measured by quantitative PCR (qPCR) from nasopharyngeal samples and SARS-CoV-2-specific IgG antibodies were measured from blood with multiplex microsphere immunoassays. Results. 4,768 SARS-CoV-2 positive participants were enrolled, among whom 42, 64, 89, 380, 948 and 245 individuals were in age brackets 0-4y, 5-11y, 12-17y, 18-44, 45-64y, and >65y, respectively. Viral load as measured by qPCR was determined to be similar across age groups within the first week post symptom onset. The magnitude of the IgG antibody response against the spike protein was also compared across age groups at early and convalescent time points and was higher in those over the age of 65 years. Conclusion. Early viral load during acute infection did not correlate with age in individuals who experienced COVID-19. These findings diverge from other respiratory viruses, such as respiratory syncytial virus and influenza where children tend to have higher viral loads. In contrast, the magnitude of the antibody response against the spike protein correlated with older age at acute and convalescent time points. Together our data suggest that the host response against SAR-CoV-2 differs with age and is not associated with the acute viral load. Defining age-dependent immunity against SARS-CoV-2 has the potential to identify key immunologic responses that can be used to optimize treatment and vaccine strategies.

Precision phenotyping of 'long COVID' through machine learning

Background. Characterizing, diagnosing, and caring for 'long COVID' patients has proven to be challenging due to heterogenous symptoms and broad definitions of these post-acute sequelae. Here, we take a machine learning approach to identify discrete clusters of long COVID symptoms which may define specific long COVID phenotypes. Figure 1: (A) Principal component analysis followed by K-means clustering identified three groups of participants. (B) Heatmap depicting three distinct clusters (high values are in red and low value are in blue);Cluster 1 exhibits sensory symptoms (e.g., loss of smell and/or taste), Cluster 2 exhibits fatigue and difficulty thinking (e.g., changes in ability to think) symptoms, and Cluster 3 exhibits difficulty breathing and exercise intolerance symptoms. (C) Clinical and demographic characteristics of 97 military health system beneficiaries by identified clusters Methods. The Epidemiology, Immunology, and Clinical Characteristics of Emerging Infectious Diseases with Pandemic Potential (EPICC) study is a longitudinal COVID-19 cohort study with data and biospecimens collected from 10 military treatment facilities and online recruitment. Demographic and clinical characteristics were collected using case report forms and surveys completed at enrollment and at 1, 3, 6, 9, and 12 months. For this analysis, we identified those who reported any moderate to severe persistent symptoms on surveys collected 6-months post-COVID-19 symptom onset. Using the survey responses, we applied principal component analysis (PCA) followed by unsupervised machine learning clustering algorithm K-means to identify groups with distinct clusters of symptoms. Results. Of 1299 subjects with 6-month survey responses, 97 (7.47%) reported moderate to severe persistent symptoms. Among these subjects, three clusters were identified using PCA (Figure 1A). Cluster 1 is characterized by sensory symptoms (loss of taste and/or smell), Cluster 2 by fatigue and difficulty thinking, and Cluster 3 by difficulty breathing and exercise intolerance (Figure 1B). More than half of these subjects (57%) were female, 64% were 18-44 years old, and 64% had no comorbidities at enrollment (Figure 1C). Those in the sensory symptom cluster were all outpatients at the time of initial COVID-19 presentation (p < 0.01). The difficulty breathing and exercise intolerance symptom-clusters had a higher proportion of older participants (Age group >= 45-64) with more comorbidities (CCI >= 1-2). Conclusion. We identified three distinct 'long COVID' phenotypes among those with moderate to severe COVID-19 symptoms at 6-months post-symptom onset. With further validation and characterization, this framework may allow more precise classification of long COVID cases, and potentially improve the diagnosis, prognosis, and treatment of post- infectious sequelae.

SARS-CoV-2 Associated Multisystem Inflammatory Syndrome in Adults (MIS-A): a Multicenter Case Series

Background. MIS-A frequently presents with symptoms including fevers, cardiovascular, gastrointestinal mucocutaneous, and neurologic involvement in both children and adults. Predominantly occurs in previously healthy children and young adults, disproportionately affecting individuals of African, Latino and Asian descent. Represents one of the most significant and yet under-recognized clinical consequences of SARS-CoV-2 infection. Methods. A retrospective review was performed of the identified MIS-A cases admitted to two institutions between January 2020-December 2021. Included cases required evidence of present or previous SARS-CoV-2 infection by either PCR or antigen testing;and meeting the CDC definition for MIS-A. After case identification, we collected clinical and laboratory data to analyze demographics, clinical symptoms, laboratory result, imaging, length of hospitalization and therapeutic interventions. Results. Ten cases were identified in the established period of time. The average interval to hospital admission was 31days. Clinical symptoms included fever, malaise, gastrointestinal complaints, dyspnea, diffuse rash. Males outnumbered females in a ratio of 7:3, with African American and Latino patients comprising 80% of the cases. All patients had elevated inflammatory markers and the most commonly manifesting organ dysfunction was cardiovascular system. Conclusion. This case series helps reinforce the characterization of an as yet incompletely understood and challenging clinical syndrome, providing insight to guide diagnosis and treatment of MIS-A. A greater understanding of MIS-A may facilitate earlier syndromic recognition in order to improve outcomes, reduce unnecessary antimicrobial use, and reduce costs.

SARS-CoV-2 infection is associated with decreased reported physical fitness in a US military longitudinal cohort

Background. COVID-19 may have deleterious effects on the fitness of active duty US military service members. We seek to understand the long-term functional consequences of SARS-CoV-2 infection in this critical population, and in other military healthcare beneficiaries. Methods. The Epidemiology, Immunology, and Clinical Characteristics of Emerging Infectious Diseases with Pandemic Potential (EPICC) study is a longitudinal cohort study to describe the outcomes of SARS-CoV-2 infection in US Military Health System beneficiaries. Subjects provided information about difficulties experienced with daily activities, exercise, and physical fitness performance via electronic surveys. Subjects completed surveys at enrollment and at 1, 3, 6, 9, and 12 months. Results. 5,910 subjects completed survey fitness questions, 3,244 (55%) of whom tested SARS-CoV-2 positive at least once during the period of observation. Over 75% of subjects were young adults and over half were male (Table 1). 1,093 (34.3%) of SARS-CoV-2-positive subjects reported new or increased difficulty exercising compared to 393 (14.8%) SARS-CoV-2 negative subjects (p < 0.01) (Table 2). The most commonly reported symptoms related to problems with exercise and activities were dyspnea and fatigue.Among the active-duty members who answered the question about their service-mandated physical fitness test scores, 43.2% of SARS-CoV-2-positive participants reported that their scores had worsened in the study period, compared with 24.3%of SARS-CoV-2 negative participants.Among SARS-CoV-2-positive subjects, reports of difficulty exercising and performing daily activities were highest within one month of the first positive test, decreasing in prevalence among the cohort only slightly to 24% and 18%, respectively, at 12 months (Figure 1). Conclusion. A substantial proportion of military service-members in this cohort have reported impairment of their service-mandated physical fitness scores after COVID-19;this proportion is significantly higher than those who are SARS-CoV-2 negative and persists to 12 months in many;similar complaints were reported among non-active duty. Further objective evaluation of post-COVID fitness impairment in this population is warranted. (Figure Presented).

Long-term clinical outcomes following SARS-CoV-2 infection include persistent symptoms and cardiovascular disease beyond 3 months post-infection

Background. The long-term health effects after SARS-CoV-2 infection remain poorly understood. We evaluated health and healthcare usage after SARS-CoV-2 infection via surveys and longitudinal electronic medical record (EMR) review within the Military Health System (MHS). Methods. We studied MHS beneficiaries enrolled in the Epidemiology, Immunology, and Clinical Characteristics of Emerging Infectious Diseases with Pandemic Potential (EPICC) cohort from March to December 2020. COVID-19 illness symptom severity and duration were derived from surveys initiated in late 2020. In addition, multi-year healthcare encounter history before and after onset of COVID-19 symptoms was collected from the MHS EMR. Odds of organ-system clinical diagnoses within the 3 months pre- and post-symptom onset were calculated using generalized linear models, controlling for age, sex, and race, and including participant as a random effect. Results. 1,015 participants were included who were SARS-CoV-2 positive, symptomatic, and had 3-month follow-up data available in the EMR (Table 1). 625 of these participants had survey data collected more than 28 days post-symptom onset, among whom 17% and 6% reported persistent symptoms at 28-84 days, and 85+ days, respectively. 9.6% had not resumed normal activities by one month. The most frequently reported symptoms persisting beyond 28 days were dyspnea, loss of smell and/or taste, fatigue, and exercise intolerance (Figure 1A). When compared with the period 61 to 90 days prior to symptom onset, the first month post-symptom onset period was associated with increases of pulmonary (aOR = 57, 95% CI 28-112), renal (aOR = 29, 95% CI 10-84), cardiovascular (aOR = 7, 95% CI 5-11), and neurological diagnoses (aOR = 3, 95% CI 2-4) (Figures 1B and 1C). Cardiovascular disease diagnoses remained elevated through 3 months (aOR = 2, 95% CI 1-3). Fig1A. Symptoms reported by EPICC participants with illnesses longer than 28 days;1B. Percent of participants with organ system specific diagnoses on each day, 90 days pre- and post-symptom onset;1C. Odds of organ system specific diagnoses within each month, +/- 3 months of symptom onset, were calculated using generalized linear models, controlling for age, sex, and race and included participants as a random effect. Odds shown are relative to the earliest period included in the model, 61-90 days before onset. Conclusion. In this MHS cohort, a significant proportion of participants had persistent symptoms and cardiovascular disease diagnoses 3 months after COVID-19 illness onset. These findings emphasize the long-term morbidity of COVID-19 and the importance of mitigating SARS-CoV-2 infections. Further analyses will evaluate demographic, clinical, and biomarker predictors of medium-to-long term organ-specific post-acute sequelae.

Pragmatic Assessment of Influenza Vaccine Effectiveness in the DoD (PAIVED): Updates from Year 3 of Multi-Site Trial

Background. The SARS-CoV-2 pandemic has spotlighted respiratory infections and the value of effective vaccines. The SARS-CoV-2 vaccine has been remarkably effective;however, influenza vaccine effectiveness has been reported to be lower among active duty military populations than in the general public (18% vs 36%). The Pragmatic Assessment of Influenza Vaccine Effectiveness in the DoD (PAIVED) study compares 3 FDA-licensed influenza vaccine types (egg-based, cell-based, and recombinant) to assess differences in immunogenicity and effectiveness in adults. Methods. Participants in the 3rd year of PAIVED (2020/21 influenza season) were enrolled from October 2020 through January 2021. Participants received weekly surveys about influenza-like-illnesses (ILI) experienced in the past week;if they reported an ILI, they were queried about symptom duration and severity, and asked to self-collect a nasal swab and dried blood sample. Four weeks later, more information about symptom duration and illness burden was obtained via telephone interview, and the participant collected a second blood sample. Results. PAIVED year 3 enrolled 3,269 participants (Table 1). 278 participants reported 1 ILI , while 60 reported 2 ILIs, and 18 reported 3 ILIs. No pathogen was identified for most processed ILI samples (78%);the most common viruses were SARS-CoV-2 (25, 12%), rhinovirus (24, 12%), and seasonal coronaviruses (4, 2%). No influenza has been identified thus far. Among those participants who had convalescent ILI visits (275), the median duration of the reported ILIs was 9 days (IQR 5, 15), with a median of 4 days (IQR 2, 7) of limited activity, and 2 days (IQR 0, 3) with fever. Three individuals were hospitalized. Conclusion. There have been relatively low rates of ILI identified in this study during this season, with only 11% of the participants reporting an ILI so far, consistent with low rates of non-COVID-19 ILI reported elsewhere during the current pandemic. We anticipate some influenza cases may be identified as more samples are processed. Planned analyses include calculating comparative influenza vaccine effectiveness to inform future vaccine purchasing decisions, as well as comparing serological response to the different vaccines.

SARS-CoV-2 Viral Load Does Not Predict Incident Venous Thromboembolism in COVID-19

Background. The risk factors of venous thromboembolism (VTE) in COVID-19 warrant further study. We leveraged a cohort in the Military Health System (MHS) to identify clinical and virological predictors of incident deep venous thrombosis (DVT), pulmonary embolism (PE), and other VTE within 90-days after COVID-19 onset. Methods. PCR or serologically-confirmed SARS-CoV-2 infected MHS beneficiaries were enrolled via nine military treatment facilities (MTF) through April 2021. Case characteristics were derived from interview and review of the electronic medical record (EMR) through one-year follow-up in outpatients and inpatients. qPCR was performed on upper respiratory swab specimens collected post-enrollment to estimate SARS-CoV-2 viral load. The frequency of incident DVT, PE, or other VTE by 90-days post-COVID-19 onset were ascertained by ICD-10 code. Correlates of 90-day VTE were determined through multivariate logistic regression, including age and sampling-time-adjusted log10-SARS-CoV-2 GE/reaction as a priori predictors in addition to other demographic and clinical covariates which were selected through stepwise regression. Results. 1473 participants with SARS-CoV-2 infection were enrolled through April 2021. 21% of study participants were inpatients;the mean age was 41 years (SD = 17.0 years). The median Charlson Comorbidity Index score was 0 (IQR = 0 -1, range = 0 - 13). 27 (1.8%) had a prior history of VTE. Mean maximum viral load observed was 1.65 x 107 genome equivalents/reaction. 36 (2.4%) of all SARS-CoV-2 cases (including inpatients and outpatients), 29 (9.5%) of COVID-19 inpatients, and 7 (0.6%) of outpatients received an ICD-10 diagnosis of any VTE within 90 days after COVID-19 onset. Logistic regression identified hospitalization (aOR = 11.1, p = 0.003) and prior VTE (aOR = 6.2 , p = 0.009) as independent predictors of VTE within 90 days of symptom onset. Neither age (aOR = 1.0, p = 0.50), other demographic covariates, other comorbidities, nor SARS-CoV-2 viral load (aOR = 1.1, p = 0.60) were associated with 90-day VTE. Conclusion. VTE was relatively frequent in this MHS cohort. SARS-CoV-2 viral load did not increase the odds of 90-day VTE. Rather, being hospitalized for SARS-CoV-2 and prior VTE history remained the strongest predictors of this complication.

A Machine Learning Approach Identifies Distinct Early-Symptom Cluster Phenotypes Which Correlate with Severe SARS-CoV-2 Outcomes

Background. The novel coronavirus disease 2019 (COVID-19) pandemic remains a global challenge. Accurate COVID-19 prognosis remains an important aspect of clinical management. While many prognostic systems have been proposed, most are derived from analyses of individual symptoms or biomarkers. Here, we take a machine learning approach to first identify discrete clusters of early stage-symptoms which may delineate groups with distinct symptom phenotypes. We then sought to identify whether these groups correlate with subsequent disease severity. Methods. The Epidemiology, Immunology, and Clinical Characteristics of Emerging Infectious Diseases with Pandemic Potential (EPICC) study is a longitudinal cohort study with data and biospecimens collected from nine military treatment facilities over 1 year of follow-up. Demographic and clinical characteristics were measured with interviews and electronic medical record review. Early symptoms by organ-domain were measured by FLU-PRO-plus surveys collected for 14 days post-enrollment, with surveys completed a median 14.5 (Interquartile Range, IQR = 13) days post-symptom onset. Using these FLU-PRO-plus responses, we applied principal component analysis followed by unsupervised machine learning algorithm k-means to identify groups with distinct clusters of symptoms. We then fit multivariate logistic regression models to determine how these early-symptom clusters correlated with hospitalization risk after controlling for age, sex, race, and obesity. Results. Using SARS-CoV-2 positive participants (n = 1137) from the EPICC cohort (Figure 1), we transformed reported symptoms into domains and identified three groups of participants with distinct clusters of symptoms. Logistic regression demonstrated that cluster-2 was associated with an approximately three-fold increased odds [3.01 (95% CI: 2-4.52);P < 0.001] of hospitalization which remained significant after controlling for other factors [2.97 (95% CI: 1.88-4.69);P < 0.001]. (A) Baseline characteristics of SARS-CoV-2 positive participants. (B) Heatmap comparing FLU-PRO response in each participant. (C) Principal component analysis followed by k-means clustering identified three groups of participants. (D) Crude and adjusted association of identified cluster with hospitalization. Conclusion. Our findings have identified three distinct groups with early-symptom phenotypes. With further validation of the clusters' significance, this tool could be used to improve COVID-19 prognosis in a precision medicine framework and may assist in patient triaging and clinical decision-making.

Before and After: The Impact of the COVID-19 Pandemic on Influenza-Like Illness Trends in PAIVED

Background. The Pragmatic Assessment of Influenza Vaccine Effectiveness in the DoD (PAIVED) is a multicenter study assessing influenza vaccine effectiveness in active duty service members, retirees, and dependents. PAIVED recently completed its third year and offers a unique opportunity to examine influenza-like illness (ILI) trends prior to and during the COVID-19 pandemic in a prospective, well-defined cohort. Methods. During the 2018-19, 2019-20, and 2020-21 influenza seasons, PAIVED enrolled DoD beneficiaries presenting for annual influenza vaccination. After collecting baseline demographic data, participants were randomized to receive egg-based, cell-based, or recombinant-derived influenza vaccine. Weekly throughout the influenza season of enrollment, participants were surveyed electronically for ILI, defined as (1) having cough or sore throat, plus (2) feeling feverish/having chills or having body aches/fatigue. Participants with ILI completed a daily symptom diary for seven days and submitted a nasal swab for pathogen detection. Results. Over the three seasons, there were 10,656 PAIVED participants: 1514 (14.2%) in 2018-19, 5876 (55.1%) in 2019-20, and 3266 (30.6%) in 2020-21. The majority were male (68-73% per year) with a mean age of 34±14.8 years at enrollment. 2266 participants reported a total of 2673 unique ILIs. The highest percentage of participants with ILI was in 2019-20 (28.2%), versus 19.6% in 2018-19 and 9.6% in 2020-21. Figure 1 depicts the percent of individuals reporting ILI by week of the season for each of the PAIVED seasons. Notably, after March 21, 2020, the weekly incidence of participants reporting ILI never exceeded 1%. Figure 1. Percent of PAIVED participants reporting ILI by week of season. Conclusion. The low incidence of reported ILI in PAIVED participants during the COVID-19 pandemic is consistent with national influenza surveillance reports of influenza and outpatient ILI activity, suggesting that mitigation measures taken to reduce transmission of SARS-CoV-2 reduced the spread of other respiratory viruses.

Pragmatic assessment of influenza vaccine effectiveness in the dod (PAIVED): Updates from year 2 of multi-site trial

Background. Despite nearly universal influenza vaccination for active duty military personnel, breakthrough influenza infections occur. We are reporting on the second year of the Pragmatic Assessment of Influenza Vaccine Effectiveness in the DoD (PAIVED), comparing three FDA-licensed influenza vaccine types (eggbased, cell-based, and recombinant) to assess differences in immunogenicity and effectiveness. Methods. Participants in the second year of PAIVED were enrolled from Oct 2019 through Jan 2020 at 9 military facilities. Participants received weekly inquiries about influenza-like-illnesses (ILI) experienced in the past week, and if the participant reported having a cough or sore throat and a) muscle/body aches or fatigue and/or b) being feverish or having chills, they were scheduled for a clinic visit. During this visit, a blood sample and a nasal swab were collected, as well as information about symptom duration and severity. A second (convalescent) visit was conducted approximately 4 weeks later, which involved collecting additional information about the duration of symptoms and illness burden, as well as a second blood draw. Due to the COVID-19 pandemic, acute and convalescent visits were disrupted at most sites in March and April due to COVID-19 precautions. Results. PAIVED year 2 enrolled 5,892 participants who completed demographic forms (Table 1). Among those who reported any ILIs, most reported one ILI (1,345), while 264 reported two ILIs, and 42 reported three ILIs. Nasal swabs were processed from 273 ILIs at four sites (Fig 1), and 14 cases of influenza were identified thus far. The median duration of ILIs was ten days, with a median of three days of limited activity, and two days with fever. Nine individuals were hospitalized. Table 1. Demographic characteristics of individuals enrolled in PAIVED 2019/20 Figure 1. Lab results as of 5/15 (N=273 samples) Conclusion: Over 25% of participants reported an ILI, and 5% of the nasal swabs that have been tested thus far have been positive for influenza. While most samples have not yet been analyzed, we have identified some breakthrough cases of influenza among vaccinated participants. Planned analyses include comparative vaccine effectiveness in order to inform future vaccine purchasing decisions.

Biomarker elevation during COVID-19: Differences between ambulatory and hospitalized individuals

Background: While the majority of illness due to COVID-19 does not require hospitalization, little has been described about the host inflammatory response in the ambulatory setting. Differences in the levels of inflammatory signaling proteins between outpatient and hospitalized populations could identify key maladaptive immune responses during COVID-19. Methods: Samples were collected from 76 participants (41% female, mean 46.8 years of age) enrolled at five military treatment facilities between March 20, 2020 and June 17, 2020 in an ongoing prospective COVID-19 cohort. This analysis was restricted to those with positive SARS-CoV-2 (severe acute respiratory syndrome- coronavirus 2) RT-PCR testing and included hospitalized (N=29;10 requiring an ICU stay) and non-hospitalized (N=43) participants. Severity markers (IL6, D-dimer, procalcitonin, ferritin, ICAM-1, IL5, lipocalin, RAGE, TNFR, VEGFA, IFNγ, IL1β) were measured in plasma (mg/dL) using the Ella immunoassay and natural log transformed. Univariate negative binomial regression was performed to determine relative risk of hospitalization. Using the full marker panel, we performed a Principal Component Analysis (PCA) to determine directions of maximal variance in the data. Pearson's correlation coefficient was determined between analytes and each axis. Results: Participants requiring ambulatory-, hospital-, and ICU-level care had samples collected at 44.0 (IQR: 35.0-51.0), 40.0 (13.0-51.0), and 47.5 (21.0-54.0) days, respectively. Higher unadjusted levels of IL6, D-dimer, procalcitonin, or ferritin were each associated with hospitalization (Table 1). The PCA showed a separation along axes between level of care and duration of symptoms (Fig 1). While significant correlations were noted with a number of biomarkers, PC1 most correlated with TNFR1 (r=0.88) and PC2 most correlated with IL6Ra (r=0.95). PC1 axis variation accounted for 36.5% of variance and the PC2 axis accounted for 20.0% of variance. Conclusion: TNFR1 and IL6Ra levels correlated with differences in the proinflammatory states between hospitalized and non-hospitalized individuals including time points late in the course of illness. Further analysis of these preliminary findings is needed to evaluate for differences by stages of illness.