Unique Differentiation and Homing Features of T Cells in Individuals with Long Covid

Background: A significant portion of individuals experience persistent symptoms months after SARS-CoV-2 infection, broadly referred to as Long COVID (LC). Although the frequencies of subsets of SARS-CoV-2-specific T cells have been shown to differ in individuals with LC relative to those with complete recovery, a deep dive into phenotypic and functional features of total and SARSCoV- 2-specific T cells from individuals with LC has yet to be performed. Method(s): Here, we used CyTOF to characterize the phenotypes and effector functions of T cells from LIINC cohort. The median age was 46, the cohort was 55.8% female, and 9/43 had been hospitalized. Participants were reported a median of 7 LC symptoms at 8 months. SARS-CoV-2-specific total antibody levels were also measured in concurrent sera. Manual gating was used to define T cell subsets, SPICE analyses for polyfunctionality, T cell clustering for phenotypic features, and linear regression for correlation. Permutation tests, Student's t tests, and Welch's t test were used for statistical analysis. Result(s): SARS-CoV-2 total antibody responses were elevated in the LC group (p=0.043), and correlated with frequencies of SARS-CoV-2-specific T cells in those without LC (r=0.776, p< 0.001) but not those with LC. While the frequencies of total SARS-CoV-2-specific CD4+ and CD8+ T cells were similar between individuals with and without LC, those from individuals without LC tended to be more polyfunctional (co-expressing IFNgamma, TNFalpha, IL2, and/or MIP1beta). CD4+ T cells from individuals with LC harbored higher frequencies of Tcm (p=0.003), Tfh (p=0.037), and Treg subsets (p=0.0412), and preferentially expressed a variety of tissue homing receptors including CXCR4 and CXCR5 (p=0.037). SARS-CoV-2-specific CD4+ T cells producing IL6, albeit rare, were observed exclusively among those with LC (p=0.016). In addition, participants with LC harbored significantly higher frequencies of SARS-CoV-2-specific CD8+ T cells co-expressing exhaustion markers PD1 and CTLA4 (p=0.018). Conclusion(s): Long COVID is characterized by global phenotypic differences in the CD4+ T cell compartment in ways suggesting preferential migration of these cells to inflamed mucosal tissues. Individuals with LC also harbor higher numbers of exhausted SARS-CoV-2-specific CD8+ T cells, potentially implicating viral persistence. Finally, our data additionally suggest that individuals with LC may uniquely exhibit an uncoordinated T cell and antibody response during COVID-19 convalescence.

PLASMA-BASED ANTIGEN PERSISTENCE IN THE POST-ACUTE PHASE OF SARS-CoV-2 INFECTION

Background: There is mounting evidence regarding the frequency and spectrum of post-acute sequelae of SARS-CoV-2 infection (PASC), but a search for causes has been elusive. Recently, a plasma-based assay for SARS-CoV-2 antigen has been developed, which in initial use revealed that a high fraction of severely affected patients with PASC had circulating antigen. It is unknown whether detectable SARS-CoV-2 antigen is specific for PASC or how the assay performs in a broader clinical spectrum of patients with PASC. Method(s): We evaluated a cohort of patients with RNA-confirmed SARS-CoV-2 infection enrolled >=3 weeks following initial symptoms. Participants, both with and without PASC at enrollment, were identified via facility- and communitybased advertising and examined every 4 months. An interviewer-administered questionnaire ascertained presence of 30 different symptoms (new or worse compared to pre-COVID) in the prior 2 days at each exam. Using the single molecule array (Simoa) assay, we measured spike, S1, and nucleocapsid SARSCoV- 2 antigens in plasma collected at time of symptom assessment. Result(s): We examined 172 participants (50% men, 46% non-white, median age 46 years) who contributed 667 timepoints from 0.7 to 15.4 months following infection, at which 66% featured report of >=1 symptom. Sixty-one of 667 timepoints (9.1%) representing 24% of persons had >=1 detectable SARSCoV- 2 antigen. Among the 437 timepoints at which >=1 symptom was present, 9.8% had >=1 detectable antigen;this compares to 7.8% of timepoints at which symptoms were absent. In comparison to those without symptoms, individuals with several specific symptom complexes (gastrointestinal, musculoskeletal, and central neurologic) more commonly had detectable antigen (Figure). Hospitalization during acute COVID-19 was strongly related to antigen detection. Conclusion(s): Among a diverse group of SARS-CoV-2-infected persons in the post-acute phase of infection, SARS-CoV-2 antigen is detectable in plasma in both those with and without symptoms but more commonly in those with gastrointestinal, musculoskeletal, and central neurologic complaints. The findings indicate that antigen persists in at least some persons and suggest (but do not prove) that antigen is causally related to symptoms. That antigen is found in only a fraction of those with PASC indicates either that not all symptoms are driven by antigen, current plasma antigen detection is insensitive relative to tissue, or nominal PASC symptoms are sometimes unrelated to SARS-CoV-2. (Figure Presented).

Differential Association of Cytomegalovirus with Acute and Post-Acute Covid-19

Background: Asymptomatic Cytomegalovirus (CMV) infection reshapes systemic immune responses and its replication can be both a consequence and cause of inflammation. As CMV resides in the same tissues affected by SARSCoV- 2, we hypothesized that asymptomatic CMV co-infection might modify the pathogenesis of both acute and post-acute COVID-19. Method(s): Participants had current or prior nucleic acid-confirmed SARS-CoV-2 infection in the COVID-19 Multi-Phenotyping for Effective Therapies (COMET, n=219), Immunophenotyping Assessment in a COVID-19 Cohort (IMPACC, n=244) or the Long-term Impact of Infection with Novel Coronavirus (LIINC, n=327) cohorts. We assessed the relationship between CMV serostatus and odds of hospitalization and plasma SARS-CoV-2 N antigen levels during acute COVID-19 as well as post-acute "Long COVID" symptoms, defined as >=1 of 32 COVID-19-attributed symptoms present at least 60 days following initial symptom onset. Result(s): Among 758 participants, 518 were hospitalized for their acute COVID-19 episode. CMV seropositivity was independently associated with a 1.9-fold increased odds of hospitalization for acute COVID-19, after adjustment for age, sex, race, ethnicity, HIV status, prior autoimmune disease, diabetes, and obesity (p=0.01, A). Among those hospitalized, CMV seropositivity was also associated with higher plasma SARS-CoV-2 N antigen levels (median 936 vs. 323 pg/ml, P=0.03, B), which remained significant after adjustment for potential confounders, but not with ICU admission (n=209), death (n=58), or thrombotic events (n=31). In contrast to its relationship to acute COVID-19 disease severity, CMV seropositivity was independently associated with a 48% decreased odds of having neurocognitive Long COVID symptoms such has headache and brain fog 4 months after initial COVID-19 diagnosis (P=0.036). Conversely, serologic evidence of Epstein-Barr Virus (EBV) reactivation and HIV both increased the odds of these symptoms (C). Conclusion(s): CMV seropositivity is associated with a 1.9-fold higher odds of hospitalization in people with acute COVID-19 and a nearly 3-fold higher SARS-CoV-2 antigen burden in hospitalized patients. In contrast, CMV seropositivity is associated with a decreased odds of neurocognitive Long COVID symptoms, while other chronic viral co-infections like EBV reactivation and HIV are associated with an increased odds of this complication. The biologic mechanisms mediating these relationships are unknown, but warrant further investigation. (Figure Presented).

What We Know Now About Long Covid Syndromes?

It is now widely accepted that SARS-CoV-2 infection can affect long-term health and quality of life. Long COVID, a type of post-acute sequelae of SARS-CoV-2 infection (PASC) characterized by persistent unexplained symptoms, has a major impact on the health of many COVID-19 survivors. Although many individuals (up to 30%) experience some limited symptoms in the weeks and months following COVID-19, the prevalence of severe disabling Long COVID is less common (perhaps <5%). Long COVID syndromes are variable and include general (e.g., fatigue) and organ-system specific symptoms (e.g., shortness of breath, palpitations, neurocognitive symptoms), as well as symptoms resembling other medically unexplained syndromes (e.g., myalgic encephalomyelitis/chronic fatigue syndrome, dysautonomia, post-exertional malaise). For reasons not yet understood, female sex is a strong predictor of Long COVID, as is the presence of certain comorbidities, particularly obesity. Mechanisms that might plausibly contribute to Long COVID include irreversible tissue damage associated with acute infection, persistence of SARS-CoV-2 antigen or possibly a viral reservoir, residual or ongoing immune activation and inflammation, reactivation of other latent human viruses, microvascular dysregulation and thrombotic events, microbial translocation, dysbiosis, and autoimmune phenomena. These mechanisms may act in isolation or in combination to drive Long COVID syndromes. Notably, many if not all of these pathways have been implicated as possible mechanisms for the excess rate of cardiovascular disease and other comorbidities in people living with HIV. Industry engagement in Long COVID research is growing, and NIH funding for clinical trials is emerging through programs such as the RECOVER Initiative. As a result, we are entering an era of experimental medicine, in which potential interventions will be used as tools to probe the biology of the disease. This presentation will provide an overview of the proposed biological mechanisms contributing to Long COVID, with a focus on the current state of evidence, human and animal models, and the emerging therapeutic agenda.

Trajectory of Post-Covid Neurocognitive Recovery in People Living with and without Hiv

Background: Neurocognitive dysfunction is common in long COVID and in people living with HIV (PWH). It is unknown whether PWH experience different disturbances in neurocognitive function following COVID-19 compared to HIVseronegative people. Method(s): The amfAR-Johns Hopkins University COVID Recovery Study is a prospective observational cohort study consisting of four groups: participants who had SARS-CoV-2 infection for the first time within 30 days prior to enrollment with HIV (PWH, arm 1) and without HIV (arm 2);participants with no history of SARS-CoV-2 infection with HIV (arm 3) and without HIV (arm 4). 93.5% of the cohort had received a COVID-19 vaccine prior to enrollment. Cognitive tests were administered at 1-and 4-months post symptom onset (arms 1-2) or post-enrollment (arms 3-4) in seven domains. Age standardized scores (all tests) and age-sex-and education-standardized scores (verbal fluency) were obtained. Standardized scores were compared using the Mann-Whitney U Test and the Kruskal-Wallis test. Result(s): PWH scored lower than HIV-seronegative participants at 1 and 4 months post-COVID on three tests: the Hopkins Verbal Learning Test-Revised (HVLT-R) learning (M1, p=0.011, M4, p=0.015), HVLT-R memory (M1, p=0.029, M4, p=0.007), and category-cued verbal fluency (VF;M1&4, p< 0.001). For the majority of timepoints, PWH who were post-COVID produced equivalent scores as PWH who never had COVID (p-levels > 0.05). Comparing post-COVID HIV-seronegative people to those who never had COVID, post-COVID participants scored lower than never-COVID participants on the Oral Trail Making Test part A (OTMT) test of processing speed at month 1 (p=0.033). Between month 1 and 4, HIV-seronegative people who were post-COVID showed improvements in HVLT-R Recognition (p=0.039), OTMT A (p=0.003), and OTMT B test of executive function (p=0.032). Conclusion(s): Neurocognitive scores in PWH were independent of COVID status, suggesting that higher frequencies of post-COVID neurocognitive dysfunction in PWH compared to HIV-seronegative people are due to HIV-associated factors more so than COVID. HIV-seronegative, post-COVID people demonstrate diminished recognition memory, processing speed, and executive function at 1 month post-COVID that improves by 4 months. Post-COVID neurocognitive dysfunction is present, if temporary, even in a highly vaccinated cohort of people.

EXERCISE CAPACITY IS REDUCED IN HIV INDEPENDENT OF SARS-CoV-2 INFECTION

Background: Reduced exercise capacity occurs as a post-acute sequela of COVID-19 ("PASC" or "Long COVID"). Cardiopulmonary exercise testing (CPET) is the gold standard for measuring exercise capacity and identifying reasons for exercise limitations. Only one prior study used CPET to examine exercise limitations among people living with HIV (PLWH). Extending our prior findings in PASC, we hypothesized that PLWH would have a greater reduction in exercise capacity after SARS-CoV-2 co-infection due to chronotropic incompetence (inability to increase heart rate). Method(s): We performed CPET within a COVID recovery cohort that included PLWH (NCT04362150). We evaluated associations of HIV and prior SARS-CoV- 2 infection with or without PASC with: (1) exercise capacity (peak oxygen consumption, VO2) and (2) adjusted heart rate reserve (AHRR, marker of chronotropic incompetence) using linear regression with adjustment for age, sex, and body mass index. Result(s): We included 83 participants (median age 54, 35% female, 10% hospitalized, 37 (45%) PLWH) who underwent CPET at 16 months (IQR 14-17) after SARS-CoV-2 infection. Among PLWH (median duration living with diagnosed HIV 21 years (IQR 15-28), all virally suppressed on antiretroviral therapy), 14 (39%) had not had SARS-CoV-2 infection, 12 (32%) had prior SARSCoV- 2 infection without PASC, and 11 (30%) had PASC (Long COVID symptoms at CPET). Median CD4 count was 608 (370-736) and CD4/CD8 ratio 0.92 (0.56-1.27). Peak VO2 was reduced among PLWH compared to individuals without HIV with an achieved exercise capacity only 80% vs 99% (p=0.005, Fig.), a difference in peak VO2 of 5.5 ml/kg/min (95%CI 2.7-8.2, p< 0.001). Exercise capacity did not vary by SARS-CoV-2 infection among PLWH (p=0.48 for uninfected vs infected;p=0.25 for uninfected vs no PASC;p=0.32 no PASC vs PASC). Chronotropic incompetence was present in 38% of PLWH vs 11% without HIV (p=0.002), and AHRR (normal >80%) was significantly reduced among PLWH vs individuals without HIV (60% vs 83%, p< 0.0001, Fig.). Heart rate response varied by SARSCoV- 2 status among those with HIV: namely, 3/14 (21%) without SARS-CoV-2, 4/12 (25%) with SARS-CoV-2 without PASC, and 7/11 (64%) with PASC (p=0.04 PASC vs no PASC). Among PLWH, CD4 count, CD4/CD8 ratio, and hsCRP were not associated with peak VO2 or AHRR. Conclusion(s): Exercise capacity is reduced among PLWH, with no differences by SARS-CoV-2 infection or PASC. Chronotropic incompetence may be a mechanism of reduced exercise capacity among PLWH. (Figure Presented).

IMPACT OF SARS-CoV-2-MEDIATED CD4 T CELL ACTIVATION AND HIV DNA PERSISTANCE IN VIVO

Background: Antigen-driven CD4+ T cell proliferation is a proposed mechanism of HIV-1 reservoir persistence. We previously reported that SARSCoV- 2 infection leads to increased detectable low-level HIV-1 plasm RNA blips months after COVID-19, but the impact of SARS-CoV-2-mediated T cell activation on expansion of HIV-1 reservoirs is not known. We sought to identify if SARSCoV- 2 infection leads to expansion of preferentially HIV-infected CD4+ T cells in people with HIV (PWH) on ART. Method(s): Five PWH with samples collected prior to and approximately two months after SARS-CoV-2 infection were identified. We performed a surface activation induced marker (AIM) assay using a CD4-optimized overlapping SARS-CoV-2 peptide pool to measure OX40/CD137 expression following peptide stimulation and sorted CD4+ T cells based on surface marker expression. ddPCR quantification of genomic HIV-1 DNA was performed on sorted subsets. Result(s): We observed an increase in the frequency of SARS-CoV-2 AIM+ non-naive CD4+ T cells following COVID-19 in samples from 4 of 5 participants (mean AIM+ % 0.13 pre- vs 0.31 post). A large percentage of non-naive AIM+ CD4+ T cells expressed PD1 compared with total non-naive cells before (76% vs 36%) and after (65% vs 19%) COVID-19;PD1 expression was lower following SARS-CoV-2 in both AIM+ and AIM- CD4+ T cell subsets (although very few cells were AIM+ prior to COVID-19). HIV-1 DNA levels in non-naive AIM- CD4+ T cells prior to COVID-19 unexpectedly decreased following infection (mean 3,522 to 766 copies/106 cells). The numbers of AIM+ cells obtained by cell sorting were overall low ( 3,863 mean) and only one participant had detectable DNA in post-COVID AIM+ CD4+ T cells. However, a large majority of this participant's post-COVID AIM+ cells harbored HIV-1 DNA (0.89 copies per cell) whereas HIV DNA in their AIM- cells decreased from 8,387 to not detected following SARSCoV- 2 infection. No HIV-1 DNA was detected in the small number of AIM+ cells obtained prior to COVID-19 in this participant. Conclusion(s): COVID-19 in PWH led to a modest SARS-CoV-2-specific CD4+ cell response approximately two months following acute presentation. One participant may have preferentially expanded HIV-1-infected, SARS-CoV-2- specific CD4+ T cells following COVID-19 but studies involving larger numbers of participants and larger numbers of cells will be needed to fully understand the impact of SARS-CoV-2 on clonal expansion and HIV persistence.

Symptoms and Biomarkers of Long Covid in People Living with and without Hiv

Background: HIV is a risk factor for severe acute COVID-19, but it is unknown whether HIV is a risk factor for long COVID. Method(s): We conducted a prospective observational cohort study of US adults with HIV (PWH) and HIV-seronegative adults with first SARS-CoV-2 infection within 4 weeks together with people who never had COVID-19. At enrollment, participants recalled the presence and severity of 49 long COVID-associated symptoms in the month prior to COVID-19. The same symptom survey was administered at 1, 2, 4, and 6 months post-COVID or post-enrollment for never- COVID participants. Post-COVID participants donated blood 1 and 4 months post-COVID, and never-COVID participants donated blood 0-1 times. Antibody titers to 18 coronavirus antigens and levels of 30 cytokines and hormones were quantified (Meso Scale Discovery). The Mann Whitney U test was used to compare continuous variables between groups, and Pearson's chi-squared test for categorical variables. Spearman correlation analyses were used to build networks of associations between cytokines and symptoms. Result(s): 341 participants enrolled between June 2021 and September 2022. Of these, 73 were PWH post-COVID, 121 were HIV-seronegative post-COVID, 78 were PWH never-COVID, and 69 were HIV-seronegative never-COVID. Over 85% of participants were vaccinated prior to COVID-19. Most participants with HIV were male sex at birth (83% post-COVID, 59% never-COVID), on ART ( >95%), with median CD4 counts >500. Over 60% of participants reported 1+ new or worsened symptoms 2-6 months post-COVID, with higher percentages in PWH at 2 months post-COVID (p< 0.05). PWH were more likely to report body ache, pain, confusion, memory problems, and thirst and had higher levels of creatine phosphokinase post-COVID than HIV-seronegative people. SARS-CoV-2 and non-SARS human coronavirus antibody titers did not differ between PWH and HIV-seronegative post-COVID participants. Cytokine associations with each other (network density) were significantly enriched at 1 month post-COVID in both PWH and HIV-seronegative people, with significantly less enrichment at 4 months post-COVID and in never- COVID participants. Levels of four analytes (cortisol, C5a, TGF-beta1, and TIM-3) associated with specific symptoms of long COVID. Conclusion(s): PWH may experience more symptoms post-COVID with a slightly different symptom profile than people without HIV. Inflammatory networks were active in PWH and people without HIV at 1 month post-COVID.

Detection of Infectious SARS-CoV-2 in Specimens with High CT Values Is More Common for Omicron than for Delta Variants

Background. Although not validated, cycle threshold (Ct) values from real-time (r)RT-PCR are sometimes used as a proxy for infectiousness to inform public health decision-making. A better understanding of variant-specific viral dynamics, including RNA and infectious virus relationships, is needed to clarify implications for diagnostics and transmission. Methods. Non-hospitalized SARS-CoV-2-infected individuals were recruited <= 5 days post-onset and self-collected nasal swabs daily for two weeks. Sequencing was used to determine variant, an in-house quantitative rRT-PCR targeting N gene was used to produce Ct values and determine RNA load, and cytopathic effect was used to assess the presence or absence of infectious virus (binary outcome). We used a Ct threshold of 30 to define high-Ct (Ct > 30) or low-Ct (Ct <= 30) specimens and assessed the percentage of RNA-positive specimens that had infectious virus;variantspecific percentages were compared by chi2 test. Results. We included 113 and 200 RNA-positive specimens from 18 and 28 Omicron- and Delta-infected participants, respectively;timing of RNA-positive specimen collection was similar in both groups (median = 8d post-onset). Maximum observed RNA levels occurred at median of 5 days post-onset for both variants but were lower for participants with Omicron vs Delta [mean RNA copies/mL = 105.2 vs 107.9]. Despite lower RNA levels, infectious virus was frequently detected for both variants [Omicron: median duration = 4.5d;Delta: median = 6d;p = 0.13]. Omicron specimens with infectious virus had higher Cts vs Delta specimens [mean Ct = 29.9 vs 23.2, p < 0.001]. In high-Ct specimens (Ct > 30;Table), the percentage of specimens with infectious virus was typically higher for Omicron vs Delta, and was significantly higher in adults [27.3% vs 9.5%]. In low-Ct specimens (Ct <= 30), the percentage with infectious virus was similar or higher for Omicron vs Delta, and was significantly higher in children [87.5% vs 53.8%] and in those unvaccinated [94.1% vs 47.4%]. Conclusion. CDC does not recommend the use of Ct values as a proxy for infectiousness. These data further highlight that Ct values may not provide a reliable or consistent proxy for infectiousness across variants.

Biological Determinants of Post-Acute Sequelae of SARS-CoV-2

Background. The biological determinants of post-acute sequelae of SARS-CoV-2 infection (PASC), defined as the persistence or recurrence of symptoms not explained by an alternative medical diagnosis, are poorly understood. We assessed viral and immunological determinants during acute SARS-CoV-2 infection for an association with PASC at 4 to 8 months. Methods. From September 2020 to February 2022, symptomatic nonhospitalized individuals with laboratory-confirmed SARS-CoV-2 infection were identified within 5 days of symptom onset. We used anterior nasal biospecimens to measure the magnitude and duration of RNA and infectious viral shedding as well as blood samples to measure soluble markers of inflammation during the acute phase (first 28 days post-enrollment). PASC was defined as self-report of 1 or more COVID-19 attributed symptoms between 4 and 8 months after initial illness. We compared virologic and inflammatory markers, GFAP (a marker of neuronal damage) and neutralizing antibody levels from the acute phase between those with and without PASC using Mann-Whitney U tests or repeated measures mixed effects linear models. Results. Among 71 SARS-CoV-2-positive participants with a completed follow-up visit between 4 to 8 months, we included 69 with virologic data and 61 with inflammatory marker data. Median age was 37 (IQR: 29 to 48) Overall, 16/72 (23%) reported at least one qualifying PASC symptom. Report of PASC was associated with >9 days of RNA shedding (p=0.04);all participants stopped RNA shedding by day 20. During acute illness, those with subsequent PASC had increased levels of INF-alpha, INF-gamma, IP-10, IL-10, and MCP-1;these differences were greatest in the early period and normalized over 2 to 3 weeks post-illness onset. Compared to those without PASC, during the acute illness those with PASC had increased levels of GFAP and decreased levels of neutralizing antibodies but these differences were not statistically significant. Conclusion. We found indications that viral and immunological factors during acute illness may be associated with PASC, suggesting acute immunologic response to SARS-CoV-2 may have longer term effects and play a role in PASC. Further understanding of the clinically significance of these observations is needed.

NEUROLOGIC and IMMUNOLOGIC BIOMARKERS ASSOCIATED with POST-COVID NEUROLOGIC SYMPTOMS

Background: The biologic mechanisms underlying neurologic post-acute-sequelae of SARS-CoV-2 infection (PASC) are incompletely understood. We measured plasma markers of neuronal injury (glial fibrillary acidic protein [GFAP], neurofilament light chain [NfL]) and inflammation among a cohort of people with prior confirmed SARS-CoV-2 infection at early and late recovery following the initial illness (defined as < and > 90 days since COVID-19 onset, respectively). We hypothesized that those experiencing persistent neurologic symptoms would have elevations in these markers. Methods: The primary clinical outcome was the presence of self-reported central nervous system (CNS) PASC symptoms during the late recovery timepoint. We compared fold-changes in marker values between those with and without CNS PASC symptoms using linear mixed effects models and examined relationships between neurologic and immunologic markers using rank linear correlations. Results: Of 121 individuals, 52 reported CNS PASC symptoms. During early recovery, those who went on to report CNS PASC symptoms had elevations in GFAP (1.3-fold higher mean ratio, 95% CI 1.04-1.63, p=0.02), but not NfL (1.06-fold higher mean ratio, 95% CI 0.89-1.26, p=0.54). During late recovery, neither GFAP nor NfL levels were elevated among those with CNS PASC symptoms. Although absolute levels of NfL did not differ, those who reported CNS PASC symptoms demonstrated a stronger downward trend over time in comparison to those who did not report CNS PASC symptoms (p=0.041). Those who went on to report CNS PASC also exhibited elevations in IL-6 (48% higher during early recovery and 38% higher during late recovery), MCP-1 (19% higher during early recovery), and TNF-alpha (19% higher during early recovery and 13% higher during late recovery). GFAP and NfL correlated with levels of several immune markers during early recovery (MCP-1, IL-6, TNF-a, IFN-g);these correlations were attenuated during late recovery. Conclusion: Self-reported neurologic symptoms present approximately four months following SARS-CoV-2 infection are associated with elevations in markers of neurologic injury and inflammation at early recovery timepoints, suggesting that early injury can result in long-term disease. The correlation of GFAP and NfL with markers of systemic immune activation suggests one possible mechanism that might contribute to these symptoms. Additional work will be needed to better characterize these processes and to identify interventions to prevent or treat this condition.

INFLAMMATION, EXERCISE CAPACITY, CHRONOTROPY, and SYMPTOMS in POST-ACUTE COVID-19

Background: Cardiopulmonary symptoms and reduced exercise capacity can persist after SARS-CoV-2 infection. Mechanisms of post-acute sequelae of COVID-19 ("PASC" or "Long COVID") remain poorly understood. We hypothesized that systemic inflammation would be associated with reduced exercise capacity and pericardial/myocardial inflammation. Methods: As part of a COVID recovery cohort (NCT04362150) we assessed symptoms, biomarkers, and echocardiograms in adults >2 months after PCR-confirmed SARS-CoV-2 infection. In a subset, we performed cardiac magnetic resonance imaging (CMR), ambulatory rhythm monitoring (RM), and cardiopulmonary exercise testing (CPET) >12 months after acute infection. Associations between symptoms and oxygen consumption (VO2), cardiopulmonary parameters and biomarkers were evaluated using linear and logistic regression with adjustment for age, sex, BMI, and time since infection. Results: We studied 120 participants (median age 51, 42% female, and 47% had cardiopulmonary symptoms at median 7 months after acute infection). Elevated hsCRP was associated with symptoms (OR 1.32 per doubling, 95%CI 1.01-1.73, p=0.04). No differences in echocardiographic indices were found except for presence of pericardial effusions among those with symptoms (p=0.04). Of the subset (n=33) who underwent CMR at a median 17 months, all had normal cardiac function (LVEF 53-76%), 9 (27%) had pericardial effusions and none had findings suggestive of prior myocarditis. There were no differences on RM by symptoms. On CPET, 33% had reduced exercise capacity (peak VO2 <85% predicted). Individuals with symptoms had lower peak VO2 compared to those reporting recovery (28.4 vs 21.4 ml/kg/min, p=0.04, Figure). Elevated hsCRP was independently associated with lower peak VO2 after adjustment (-9.8 ml/kg/min per doubling, 95%CI-17.0 to-2.5;p=0.01, Figure). The predominant mechanism of reduced peak VO2 was chronotropic incompetence (HR 19% lower than predicted, 95%CI 11-26%;p<0.0001, Figure). Chronotropic incompetence on CPET correlated with lower peak HR during ambulatory RM (p<0.001). Conclusion: Persistent systemic inflammation (hsCRP) is associated with pericardial effusions and reduced exercise capacity > 1 year after acute SARS-CoV-2 infection. This finding appears to be driven mainly by chronotropic incompetence rather than respiratory compromise, cardiac pump dysfunction, or deconditioning. Evaluation of therapeutic strategies to target inflammation and/or chronotropy to alleviate PASC is urgently needed.

PREDICTING COVID-19 CLINICAL PROTECTIVE IMMUNITY from SARS-CoV-2 T-CELL RESPONSES

Background: SARS-CoV-2 produces variable immune responses leading to different levels of immune protection. The relationship between neutralizing antibody level (NAb) and protective immunity has been well characterized after infection and vaccination. While comparatively specific T cell responses tend to be more variable, the impacts of these responses have broad implications on long-term immunity and their role in protective immunity has not been as clearly defined. Using data from our prospective cohort study and studies of clinical protective immunity/efficacy (from vaccines), we predicted protective immunity over time in relation to SARS-CoV-2-specific T cell dynamics. Methods: With linear mixed-effects models from our published immune data from people recovering from COVID-19, we simulated the Spike (S)-specific interferon-γ (IFNγ)+ CD4+, S-specific IFNγ+ CD8+, and nucleocapsid (N)-specific IFNγ+ CD8+ T cells over time (n=500 individuals). We then predicted NAbs from linear regression models developed from the same cohort. Finally, protective immunity from NAb titers was simulated from a published model. We similarly simulated 25, 50, and 75% lower T cell responses than those observed post-COVID-19 to understand how immune response variation may impact protective immunity. Results: Virus-specific T cell responses resulted in similar protective immunity across T cell subsets, but with differences in variability over time. Protective immunity for IFNγ+ S CD8 T cells spanned from 86-95%, while for IFNγ+ S CD4 T cells and IFNγ+ N CD4 T cells it ranged from 81-96% and 84-95% respectively. Further, based on simulated dampened T cell responses, protective immunity overall did not drop below 81% less than nine months after infection even with a 75% reduction in specific T cell immunity. Conclusion: NAbs are often the singular focus to predict protective immunity and the role of virus-specific T cell immunity has often been discussed as a secondary immune response. Our analysis demonstrates that for SARS-CoV-2, certain T cells responses can reliably predict protective immunity and may be intrinsically linked. Simulating dampened T cell response to mimic a more virulent strain or inadequate immune response, demonstrated that dampened T cell response may not be responsible for inadequate protective immunity in these scenarios. In the absence of prospective clinical data, similar models may be utilized to explore the impact of potential therapeutics on immune responses and resulting protective immunity.

Lower SARS-CoV-2 IgG and Pseudovirus Neutralization Titers Post-mRNA Vaccination among People Living with HIV

Background. Limited data are available on whether there are differences in the immune response to SARS-CoV-2 vaccination by HIV status or by mRNA vaccine type. Methods. We saved residual outpatient laboratory samples of all previously mRNA-vaccinated individuals in the adult medicine clinics of a public hospital with a large outpatient HIV clinic during May 2021, and then excluded individuals with prior SARS-CoV-2 infection. We next 1:1 matched 100 PLWH to 100 outpatient HIVnegative adult medicine patients receiving care for chronic medical conditions on days since completion of second vaccination (minimum 10), sex, age +/-5 years, and the type of mRNA vaccine received. We defined a non-response as reciprocal pseudovirus neutralizing titer< 10 and anti-RBD IgG< 10 relative fluorescent units, and compared non-response by HIV status using mixed models. Results. In each matched group there were 13 women;25 received the mRNA-1273 vaccine and 75 received the BNT162b2 vaccine;the median age was 59. The median time from second vaccination was 35 days (IQR: 20-63). Among PLWH, the median CD4+ T-cell count was 511 (IQR: 351-796) and 5 individuals had HIV RNA > 200. We found 2.4-fold greater odds of pseudovirus neutralizing antibody non-response among PLWH compared to people without HIV (95% CI=1.1-5.4). Although few individuals in each group did not mount an IgG response (12 among PLWH vs. 5;p=0.08), continuous anti-RBD IgG concentrations were 43% lower among PLWH (95% CI=0.36-0.88). Among PLWH, when adjusting for age, sex, and days post-vaccination, each 100-cell increase in CD4+T-cell count was associated with 22% higher neutralizing antibody titers (GMR 1.22;95% CI=1.09-1.37). Unsuppressed HIV RNA >200 was associated with 89% lower neutralizing antibody titers (GMR 0.11;95% CI=0.01-0.84). Receipt of the BNT162b2 vs. mRNA-1273 vaccine was associated with 77% lower neutralizing titers (GMR 0.23;95% CI=0.08-0.65) among PLWH. Post-mRNA Vaccination SARS-CoV-2 IgG Concentrations and Pseudovirus Neutralizing Titers by HIV Status and Vaccine Conclusion. PLWH had lower than expected response to mRNA SARS-CoV-2 vaccines, with the highest non-response among those with low CD4+ counts, unsuppressed HIV RNA, and those who received the BNT162b2 vaccine. Immunization strategies to improve immune responses among PLWH should be studied, and may include booster vaccination or preference of the mRNA-1273 vaccine in this group.