Skip to main content

Our take —

Variations in immune response are suspected to underlie some COVID-19 disease outcomes. In this important study, about 10% of patients with severe COVID-19 pneumonia had neutralizing IgG auto-antibodies to type I interferons, which were not present in any patients with mild SARS-CoV-2 infection. Production of auto-antibodies inhibited the ability of type I interferons to fight SARS-CoV-2 infection in vitro, and appeared to precede disease in patients. Individuals producing auto-antibodies were nearly all (94%) male. These results suggest a mechanism contributing to severe disease in a sizable minority of patients, which may be X-chromosome-linked, and they may lead to screening and therapeutic implications if replicated elsewhere.

Study design

Case series, case-control

Study population and setting

This study included 987 patients hospitalized with severe “life-threatening COVID-19 pneumonia” (defined as pneumonia with critical disease involving any organ system, requiring ICU admission) from multiple cohorts in several countries, along with 663 patients with SARS-CoV-2 infection with mild or no symptoms, and 1,127 healthy controls. Plasma and serum samples were obtained from participants, and these samples were tested for the presence of IgG auto-antibodies (Ab) against the type I interferons (IFNs) IFN-ɑ2 and IFN-ω. The authors tested whether these auto-Ab were neutralizing in vitro using plasma from patients with auto-Ab and healthy controls. A subset of samples from patients with auto-Ab to IFN-ɑ2 (n=22) were tested for the presence of auto-Ab to all 15 type I IFNs. Another subset of samples from patients with auto-Ab to IFN-ɑ2 (n=8) were tested to determine if they impaired the ability of IFN-ɑ2 to block SARS-CoV-2 infection in vitro. Finally, plasma concentrations of all 15 type I IFNs were measured among patients who had auto-Ab to these type I IFNs (n=41).

Summary of Main Findings

Among the 987 patients with severe COVID-19 pneumonia, 135 (13.7%) had IgG auto-Ab to IFN-ɑ2 and/or IFN-ω; these were also present in the two available samples from patients prior to SARS-CoV-2 infection. The prevalence of these auto-Ab against type I IFNs was estimated to be 0.33% (95% CI: 0.015% to 0.67%) in the healthy controls. In vitro, these antibodies were neutralizing for at least one type I IFN in 101 patients (10.1%), 95 (94%) of whom were male. All of the tested patients with auto-Ab to IFN-ɑ2 also had auto-Ab to all 13 type I INFs. The 8 plasma samples tested from patients with auto-Ab to IFN-ɑ2 were able to block the ability of IFN-ɑ2 to protect against infection from SARS-CoV-2 in vitro. Among the 41 patients tested who had auto-Ab to all 13 type I IFNs, 40 had no detectable levels of any type I INFs, and one had low levels. Among individuals with neutralizing auto-Ab against type I INFs, there was a predominance of males (94%), which was higher than the proportion of men among severe COVID-19 patients without these Ab (75%) and among those with mild or no symptoms (28%). One of the 7 women with auto-Ab to type I INFs had a condition in which cells activate only one X chromosome, providing evidence for X chromosome linkage.

Study Strengths

This study considered evidence at multiple levels (prevalence in cases and controls, neutralizing effect in vitro, effect on ability of IFN-ɑ2 to block SARS-CoV-2 infection in vitro, plasma type I INF concentrations). The study population was large and included from diverse ethnicities and nationalities.

Limitations

Only two samples were available from patients before SARS-CoV-2 infection, which limits the ability to conclude with certainty that auto-Ab production was not a consequence of severe disease. Demographic and clinical characteristics of patients and controls were not well described. Methods for selecting asymptomatic or paucisymptomatic controls were unclear.

Value added

This, along with an accompanying study in the same journal (Zhang et al.), is an important and novel piece of evidence; it is among the first to identify underlying variations in immune response that may cause severe COVID-19.

Our take —

A subset (3.5%) of 659 patients with severe COVID-19 pneumonia had uncommon genetic variants that suppress the type I interferon response to SARS-CoV-2 infection; none had previous severe viral illness. Some of these mutations are known to increase susceptibility to influenza, while others were not previously identified. This study, along with an accompanying study in the same journal, suggests that some people may have inborn deficiencies in their immune response that render them more susceptible to severe disease should they become infected with SARS-CoV-2. More research is necessary to determine if provision of type I interferons could improve outcomes in these patients.

Study design

Case series, Case control

Study population and setting

This study included 659 patients with severe COVID-19 pneumonia who were admitted to the ICU from four cohorts in France, Italy, and the United States (mean age 51.8 years, range 0.1 to 99 years, 74.5% male). Patients with Kawasaki-like disease were excluded. A control group of 534 individuals with mild laboratory-confirmed or suspected SARS-CoV-2 infection was also enrolled. The authors sequenced the genomes of participants from whole blood, with the aim of identifying monogenic inborn errors (primary immunodeficiency) of SARS-CoV-2 immunity that were associated with severe COVID-19. In particular, the authors tested the hypothesis that errors of TLR3- and IRF7-dependent type I interferon (INF) immunity were associated with severe COVID-19, by comparing the frequency of variants at 13 loci known to be mutated in patients with severe influenza pneumonia (or connected to three core genes dictating susceptibility to pneumonia) between cases and controls. Cells from patients with the identified variants were then tested in vitro for their effects on type I INF production after stimulation with SARS-CoV-2. Finally, blood concentrations of the 13 types of IFN-ɑ were measured in 10 patients with the identified mutations.

Summary of Main Findings

Four patients had biallelic variants of IRF7 or IFNAR1, while 113 patients had an additional 113 monoallelic variants at 12 loci; 9 of these variants were predicted to be loss-of-function (pLOF). Compared to controls, patients with severe COVID-19 were enriched in the pLOF variants at the 13 loci associated with influenza susceptibility. Of the 118 total variants, the authors tested 113 in overexpression systems, and 24 (carried by 23 patients) were found to be deleterious. In all, 23 of the 659 patients (3.5%) carried a deficiency at one of eight loci among the 13 loci tested. In vitro stimulation with SARS-CoV-2 of cells from patients with selected genotypes found those with autosomal recessive IRF7 deficiency had impaired production of type I INF by plasmacytoid dendritic cells, and deficiencies of TLR3 or IFNAR1 impaired fibroblast-intrinsic type I INF immunity to SARS-CoV-2. Ten of the 23 patients with the identified mutations had blood samples available, and among them, IFN-ɑ levels were all below 1 pg/mL, much lower than previously published measures in other patients with severe COVID-19.

Study Strengths

This was a multi-level study; variants that were enriched in the patient population were tested for responses to SARS-CoV-2 in vitro, and concentrations of IFN-ɑ were measured in blood samples from a subset of patients. Patients were from multiple countries and were of multiple ethnicities.

Limitations

Although it is biologically plausible, this study cannot establish that impaired INF type I production is responsible for severe COVID-19 among the subset of patients with the identified mutations. Patient demographic and clinical characteristics were not well described. Concentrations of IFN-ɑ were only measured in blood samples from 10 patients, all of whom had variants at the loci investigated; this limits possible inference.

Value added

This carefully conducted study suggests a mechanism for inborn susceptibility to severe COVID-19 among a subset of the population.

Our take —

This case-control study aimed to assess community and close contact exposures associated with COVID-19. This study included 154 cases and 160 controls who presented with symptoms to healthcare facilities within a US network who were tested for SARS-COV-2. COVID-19 cases were 2.4 times more likely to have reported dining at a restaurant in the 2 weeks preceding illness onset than controls. No differences between cases and controls were observed among several other community exposures. These findings should be interpreted in the context of several limitations, especially that controls were not positive for SAR-CoV-2, but they may have been infected with a pathogen with similar transmission characteristics. Although reported mask use in public spaces was high amongst both cases and controls, cases were nearly 3 times more likely to have a known COVID-19 contact many of which would have been contacts in private spaces without masks and cases were significantly higher amongst those attending restaurants where masks were not used by participants. This study suggests that restaurants may be a greater exposure risk than other community activities that do not involve eating or drinking where mask use is uncommon.

Study design

Case-Control

Study population and setting

This case-control study aimed to assess community and close contact exposures associated with COVID-19. This study included adults who were symptomatic and tested for COVID-19 the first time at one of 11 Influenza Vaccine Effectiveness in the Critically Ill (IVY) Network sites across the U.S. between the period of July 1 to 29, 2020. The study included 154 cases and 160 controls. Cases were defined as individuals with evidence of SARS-CoV-2 RNA by RT-PCR in one the networks sites and were selected by random sampling of confirmed patients within a specific time period. Controls were symptomatic individuals from the network sites who tested negative for SARS-CoV-2. Multiple controls were randomly selected for each case, matching on age, sex, and study site. Information on demographic characteristics, medical history, exposure history, community activities, and mask wearing was collected through structured retrospective interviews with participants.

Summary of Main Findings

Close contact with one or more persons with known COVID-19 was reported by 42% of cases and 14% of controls. Cases were more likely to have reported dining at a restaurant in the 2 weeks preceding illness onset than were controls (adjusted odds ratio [aOR] 2.4; 95% CI 1.5–3.8), even after restricting the analysis to participants without a known close contact to be confirmed with SARS-Cov-2 (aOR 2.8, 95% CI 1.9–4.3). Restricted analysis also suggests that cases were more likely than control to have reported going to a bar/coffee shop (aOR 3.9, 95% CI 1.5–10.1) in the 2 weeks preceding illness onset. No differences between cases and controls were observed among those who reported shopping; gatherings with ≤10 persons in a home; going to an office setting; going to a salon; gatherings with >10 persons in a home; going to a gym; using public transportation; or attending church/religious gatherings.

Study Strengths

This study attempts to differentiate between different community activity exposure risks.

Limitations

These findings should be interpreted in the context of several limitations. Authors attempted to match cases and controls at a ratio of 1:2; however, due to ineligibility or refusal to participate the matching ratio was not possible, and matching based on age and sex was not maintained. Those who agreed to participate may differ from those who did not. Participation as well as recall of exposures may have been influenced by symptom severity, and therefore the results may be subject to bias. Controls included individuals who presented at the health facility and reported symptoms, however, did not test positive for COVID-19. These patients may have illness due to another pathogen with similar exposure risks and modes of transmission. Therefore, it may be difficult to differentiate between exposure risks between cases and controls. Finally, a significant limitation was that data were not able to distinguish between indoor and outdoor dining and indoor/outdoor bar/coffee shop attendance.

Value added

This study suggests that the restaurant environment may a particular concern for risk of exposure to SARS-CoV-2 compared to other community activities that do not involve removing masks alongside being in shared indoor spaces.

Our take —

Immune responses to COVID-19 are not well understood, but evidence suggests that they may be crucial to patient outcomes. This analysis considered a very large number of immune responses to COVID-19 among hospitalized patients, noting considerable heterogeneity in responses, but identifying several broad patterns. Most patients showed a strong plasmablast response, and subgroups of patients exhibited B cell and T cell activation and proliferation that persisted for at least 7 days. Although the authors identified three “immunotypes” associated with disease severity, more studies are required to determine whether these immune response clusters are present in the broader population of COVID-19 patients.

Study design

Case-Control

Study population and setting

This study included blood samples from 125 hospitalized patients (median age 60 years, 51% male, 68% African-American) with laboratory-confirmed COVID-19 at the University of Pennsylvania (collected 1-3 days after admission and 7 days after admission for those who remained hospitalized), along with samples from 36 non-hospitalized patients who had recovered from COVID-19, and from 60 healthy donors. Flow cytometry was used to analyze peripheral blood mononuclear cells (PBMCs) with respect to a large number of immunological markers.

Summary of Main Findings

Among the 125 hospitalized patients, 83% had cardiovascular comorbidities, 30% required mechanical ventilation, and 14% died. Most hospitalized patients had elevated concentrations of inflammatory biomarkers such as CRP, d-dimer, and ferritin; troponin levels were also commonly elevated. Nearly half of hospitalized patients were clinically lymphopenic, but most had normal monocyte, eosinophil, and basophil counts. In principal components analysis of 193 immune parameters, the immune profile of hospitalized patients was clearly distinct from that in recovered and healthy donors, while recovered and healthy donors exhibited overlap. Approximately 80% of hospitalized patients exhibited CD8+ T cell activation above the levels seen in the control groups. There was a heterogeneous degree of CD4+ T cell response, with some distinct proliferating subpopulations of CD4+ T cells. In a subset of 8 SARS-CoV-2-negative blood samples, inflammatory cytokines and chemokines were elevated. Naive B cell counts were similar in hospitalized patients relative to controls, but there were significant changes in B cell subpopulation frequencies that did not appear to be related to systemic inflammation. Approximately two-thirds of hospitalized patients exhibited significantly elevated plasmablast (PB) frequency; declines in memory B cell frequencies and loss of CXCR5 expression were also commonly seen. In a subset of patients hospitalized for at least 7 days (n=48), changes in T cell and B cell responses between admission and day 7 were highly heterogeneous. More severe disease was associated with lower frequencies of CD4+ and CD8+ T cells, but there were no significant associations between temporal changes in B or T cell response and clinical severity. The authors used Uniform Manifold Approximation and Projection (UMAP) to identify clusters of immune response and correlate them with clinical outcomes; they defined three “immunotypes” having varying correlations with disease severity.

Study Strengths

The primary strengths of this study are the very large number of immune parameters considered, and the detailed manner in which variations in these parameters was assessed.

Limitations

Patients providing samples for analysis (n=125) were not large in number, and were limited to a single institution. Little information was provided about patient selection and eligibility criteria. Clusters of immune response observed, and their correlation with clinical severity, may not be generalizable to other populations. Although analyses attempted to adjust for age, sex, and race, they did not appear to adjust for comorbidities. In addition to comorbidity data, there may be other uncontrolled confounding variables associated with both immune parameters and clinical outcomes, such as (for example) cigarette smoking.

Value added

This study provided highly detailed measurements of a wide variety of immune responses to COVID-19, and outlined some broad patterns worthy of further research.

Our take —

This case-control study from Spain and Italy, available as a preprint and thus not yet peer reviewed, provides the first large-scale evidence of host genetic influence in severe COVID-19 disease, and identifies biological targets for further investigation. This analysis was only conducted in European populations, however these genetic variants do not vary widely between ancestries, and therefore should be similar across populations. Furthermore this study does not support a biological explanation for racial/ethnic disparities in COVID-19 morbidity and mortality both within the United States and globally. Finally, replication studies are needed with better control definitions and clinical characterizations to clarify the role of human genetics in COVID-19 severity, instead of overall frailty.

Study design

Case-Control

Study population and setting

A total of 1,610 patients with severe COVID-19 defined by hospitalization with respiratory failure were recruited from seven medical centers in Spain and Italy, and compared to 2,205 randomly selected blood donors. Population based-controls were significantly younger than cases; data were unavailable regarding pre-existing comorbidities or other relevant risk factors for COVID-19.

Summary of Main Findings

A genome-wide association study (GWAS) estimated the association between ~8 million single nucleotide polymorphisms (SNPs) across the human genome, adjusting for population substructure, age, and sex separately by country and then combined with a fixed-effects meta-analysis. Two regions were identified and met the threshold of genome-wide significance (P<10^-8): a region on chromosome 3 upstream of a solute carrier (SLC6A209) and a region on chromosome 9 overlapping the ABO blood group gene. Further analyses found a protective effect for blood group O, and an elevated risk for A positive individuals.

Study Strengths

By limiting cases to those hospitalized with documented COVID-19, the analysis is more likely to capture the host genetics underlying severe disease sequelae.

Limitations

This study uses population controls that are not documented to be exposed or tested for COVID-19, which could dilute the true effect size. Additionally, the combination of different age distributions in cases and controls (controls being significantly younger), and the lack of availability of additional clinical information such as comorbidities, makes the association inference unclear. The associated genetic regions may reflect a COVID-19-specific morbidity, an existing co-morbidity, or generally poorer health in an older population.

Value added

This study provides evidence that genetic variants may be partially responsible for differences in COVID-19 severity.

Our take —

In a small autopsy study, lungs from COVID-19 patients demonstrated similar diffuse alveolar damage with immune cell infiltration as is typically seen with influenza. However, COVID-19 lungs also had evidence of notable blockage of the vessels with rampant new vascular development. This may be a key pathogenic feature of the disease, and may correlate with the large vessel disease observed in epidemiologic studies. Further research is needed to better describe the pathophysiology and its implications.

Study design

Case Series; Case-Control

Study population and setting

The study consisted of pulmonary autopsy specimens at the Hannover Medical School in Hannover, Germany, from seven patients who died from respiratory failure due to laboratory-confirmed SARS-CoV-2 infection. There were two sets of comparison specimens: lungs from seven age, sex, and severity matched patients who died of influenza A H1N1 virus infection, and lungs from ten uninfected patients. None of the COVID-19 patients received mechanical ventilation while five of seven patients with influenza received mechanical ventilation. Pulmonary specimens underwent detailed CT and electron microscopy imaging, staining, and gene-expression analysis.

Summary of Main Findings

All COVID-19 lung specimens had evidence of diffuse alveolar damage with widespread T-cell infiltration, as did the influenza-positive specimens. ACE-2 enzyme expression was elevated in both COVID-19 and influenza-positive lungs compared to uninfected controls. In both COVID-19 and influenza-positive lungs, four of seven specimens had evidence of precapillary thrombi. However, alveolar capillary microthrombi were nine times as common in the COVID-19 specimens compared to the influenza-positive specimens. Notably, there was much more evidence of distorted vasculature with significant new, nonsprouting vascular growth in COVID-19 specimens compared to both of the comparison specimens.

Study Strengths

The study provided a comprehensive assessment of both gross pathology and microvascular structure with gene-expression profiles.

Limitations

There was a very small sample size. The COVID-19 patients were considerably older than influenza patients (median age 78 versus 53 years). There appear to have been notable differences in the clinical course and clinical care between the influenza and COVID-19 patients, which may underlie some of the pathologic differences observed.

Value added

This is among the most detailed pathologic analyses of COVID-19 lung specimens to date. It highlights rampant and distorted angiogenesis that may be a key feature of disease pathogenesis.

Our take —

This study adds valuable evidence that ACE inhibitors and angiostatin receptor blockers, two antihypertensive medications, are not associated with a heightened risk of COVID-19 or severe clinical illness.

Study design

Case-Control

Study population and setting

The study included as cases 6272 patients in Lombardy, Italy with laboratory-confirmed SARS-CoV-2 infection diagnosed between February 21 and March 11, 2020. 30,759 beneficiaries of the region’s public health system without diagnosed COVID-19 were used as controls, and were matched to cases by age, sex, and residency location.

Summary of Main Findings

While the use of antihypertensive medications, including ACE inhibitors and angiotensin receptor blockers (ARBs), was more common among patients with COVID-19 than among controls, this association disappeared after accounting for relevant clinical factors such as pre-existing diagnoses. COVID-19 diagnosis was not associated with either ACE inhibitor use (Odds ratio [OR]=0.96, 95% CI: 0.87,1.07) or ARB use (OR=0.95, 95% CI: 0.86,1.05). Restricting cases to those with severe clinical disease produced similar results for both ACE inhibitors (OR=0.91, 95%CI: 0.69,1.21) and ARBs (OR=0.83, 95% CI: 0.63,1.10).

Study Strengths

The study used a large sample size across many healthcare settings. Sensitivity analyses were well-conceived, accounting for uncertainty in the ascertainment of medication use.

Limitations

Given that the fundamental question is whether certain antihypertensive medications modulate the risk of severe disease, it would have been more appropriate to limit the primary analysis to those with a hypertension diagnosis. Similarly, it would likely have been more appropriate to limit the analysis of predictors of severe disease to those with a COVID-19 diagnosis. Doing so mitigates the risk of confounding by hospital presentation, health status, and more.

Value added

This study adds valuable evidence that ACE inhibitors and angiostatin receptor blockers are not associated with a heightened risk of COVID-19 or severe clinical illness.

Our take —

Markers of liver dysfunction in COVID-19 patients were elevated slightly, but not to a clinically significant degree. Liver dysfunction was not more pronounced than in a control group with community-acquired pneumonia, and did not predict COVID-19 severity. The small sample size and single location limit generalizability.

Study design

Case-Control

Study population and setting

Cases were comprised of all 115 health care workers and family members (>14 years old without hepatitis B infection) with confirmed SARS-CoV-2 infection who were admitted to Zhongnan Hospital of Wuhan University, Wuhan, China, from 1/18/20 to 2/22/20. Controls were 119 patients with community-acquired pneumonia from the same hospital.

Summary of Main Findings

Serum markers of liver dysfunction and inflammation were not significantly elevated in COVID-19 patients relative to community-acquired pneumonia patients. Only the neutrophil-to-lymphocyte ratio was a predictor of COVID-19 severity.

Study Strengths

Patients were tested for a wide array of liver function indices. The presence of a control group improves the validity of the analysis.

Limitations

The sample size is small and comes from a single hospital. There were few (n=31) severe cases of COVID-19. The comparison groups may differ by much more than just disease etiology – healthcare workers and their families represent a different population than the broader patient population. Because the sensitivity of the SAR-CoV-2 diagnostic testing is, as yet, unclear, COVID-19 patients may have been incorrectly classified as having non-COVID pneumonia, thus biasing the results.

Value added

This is one of the first studies to focus specifically on liver dysfunction as a manifestation of COVID-19.

Our take —

This study compares the blood groups of those with diagnosed COVID-19 and those from a historical control cohort. While the results indicate a signal between blood group O and a decreased risk of disease, this should be interpreted with caution given that we cannot rule out other major differences in this study between cases with COVID-19 and historical controls (e.g. age distribution).

Study design

Case-Control

Study population and setting

This was an unmatched case-control study, in which ABO-typed blood samples were collected from infected hospital cases, and ABO blood group distribution was captured from recent surveys of “normal people” to serve as comparison controls. Cases were included from three hospitals in Wuhan, Hubei province and Shenzhen, Guangdong province, China.  No specific time period was provided for when cases were infected or selected, and information was not provided on controls. There were a total of 1775 patients infected with SARS-CoV-2 (COVID-19 cases), including 206 deaths, from the Jinyintan Hospital in Wuhan. An additional 113 COVID-19 cases were included from Renmin Hospital of Wuhan University, and 285 COVID-19 cases from Shenzhen Third People’s Hospital. There were a total of 3694 controls from Wuhan City and 23386 controls from Shenzhen City.

Summary of Main Findings

Blood group A was associated with an increased risk for COVID-19, whereas blood group O was associated with a decreased risk, suggesting that the ABO blood type is a biomarker for differential susceptibility to COVID-19.

Study Strengths

There are few studies that have examined differential susceptibility to SARS-CoV-2 infection; ultimately this study assesses susceptibility to COVID-19 (disease status, not infection), but generates future hypotheses for testing around blood type and infection, as well as further analyses of progression to disease.  While multivariable analyses were not possible and therefore the role of chronic medical conditions is still unclear, this study made use of existing hospital data and surveys in order to do a rapid and methodologically replicable examination of a potential biomarker for COVID-19 susceptibility.

Limitations

The control group appears to have been sampled years before the study, is from a different source population and appears to have differences in at least age composition from the target population. Further, because controls are historical, we do not know if they ultimately were infected during the time period of this study.

Value added

This is one of the first studies of risk factors for COVID-19 susceptibility. This has potential implications for future prevention efforts (e.g. biomarker for who might need extra protection or who may need to take extra precautions against infection), however underlying methodological issues suggest caution in terms of interpretation. Importantly, a biological mechanism was hypothesized, though it requires further assessment. Ongoing and future serological studies with case reporting ascertainment may also consider assessment of blood type.