key: cord-0913486-hghzc3cd
authors: Al-Dury, Samer; Waern, Johan; Waldenström, Jesper; Alavanja, Marko; Saed, Hevar Hamah; Törnell, Andreas; Arabpour, Mohammad; Wiktorin, Hanna Grauers; Einarsdottir, Sigrun; Ringlander, Johan; Ringström, Gisela; Hellstrand, Kristoffer; Martner, Anna; Lagging, Martin
title: Impaired SARS-CoV-2-specific T cell reactivity in patients with cirrhosis following mRNA COVID-19 vaccination
date: 2022-04-27
journal: JHEP Rep
DOI: 10.1016/j.jhepr.2022.100496
sha: 938d840b224964fcb780946300a335c3af8451fe
doc_id: 913486
cord_uid: hghzc3cd

BACKGROUND & AIMS: Liver cirrhosis entails elevated risk of COVID-19-associated mortality. This study determined T cell-mediated and antibody reactivity against the spike 1 (S1) protein of SARS-CoV-2 among 48 cirrhotic patients and 39 healthy controls after mRNA COVID-19 vaccination. METHODS: SARS-CoV-2-specific T cell reactivity was measured by induced level of T cell-derived interferon-γ (IFN-γ) in blood cells stimulated ex vivo with multimeric peptides spanning the N-terminal portion of S1. S1-induced IFN-γ was quantified before and after the 1(st) and 2(nd) vaccination (BNT162b2, Pfizer-BioNTech or mRNA-1273, Moderna) alongside serum IgG against the receptor-binding domain (RBD) within S1 (anti-RBD-S1 IgG). RESULTS: T cell reactivity against S1 was reduced in cirrhotic patients after the 1(st) (P<0.001 vs controls) and 2(nd) (P<0.001) vaccination. Sixty-eight % of patients lacked detectable S1-specific T cell reactivity after the 1(st) vaccination vs. 19% in controls (OR 0.11, HR 0.03-0.48, P=0.003) and 36% remained devoid of reactivity after the 2(nd) vaccination vs. 6% in controls (OR 0.12, HR 0.03-0.59, P=0.009). T cell reactivity in cirrhosis remained significantly impaired after correction for potential confounders in multivariable analysis. Advanced cirrhosis (Child-Pugh class B) was associated with absent or lower T cell responses (P<0.05 vs. Child-Pugh class A). The deficiency of T cell reactivity was paralleled by lower levels of anti-RBD-S1 IgG after the 1(st) (P<0.001 vs. controls) and 2(nd) (P<0.05) vaccination. CONCLUSIONS: Cirrhotic patients show deficient T cell reactivity against SARS-CoV-2 antigens along with diminished levels of anti-RBD-S1 IgG after dual COVID-19 vaccination, highlighting the need for vigilance and additional preventative measures. LAY SUMMARY: T cells are a pivotal component in the defence against viruses. We show that patients with liver cirrhosis have impaired SARS-CoV-2-specific T cell responses and lower antibody levels after mRNA vaccination against COVID-19 compared with healthy controls. Cirrhotic patients with more advanced liver disease exhibited particularly inferior vaccine responses. These results call for additional preventive measures in these patients. CLINICAL TRIAL REGISTRATION: EudraCT 2021-000349-42

antigens along with diminished levels of anti-RBD-S1 IgG after dual COVID-19 23 Regardless of aetiology, end-stage liver disease is characterized by impaired 3 immunity. Cirrhosis-associated immune dysfunction (CAID) is believed to arise 4 secondary to injury of hepatic reticuloendothelial cells, reduced hepatic production of 5 proteins crucial for innate immunity [1] along with systemic inflammation [2] and 6 translates into a perturbing propensity for severe and life-threatening infections. 7

Although CAID is mostly associated with flawed innate responses [3] [4] [5] , recent studies 8 report that subsets of T cells in cirrhotic patients express markers of exhaustion, as 9 reflected by expression of TIM-3, CTLA-4, and PD-1, suggesting that T cell deficiency 10 may contribute to the observed susceptibility to infection [6, 7] . 11 12 SARS-CoV-2-infected patients with cirrhosis are at elevated risk of decompensation, 13 severe morbidity, and death [8, 9] . Thus far scarce data regarding the immunogenicity 14 of cirrhotic patients had suboptimal antibody levels [10] . Similarly, after 2 doses of viral 18 vector (AstraZeneca) or mRNA (Pfizer-BioNTech or Moderna) vaccines, Ruether et al. 19 detected T cell responses using a cytokine release assay in 17/26 (65%) of vaccinated 20 patients as compared with 19/19 (100%) of healthy controls. In the latter study anti-21 RBD-S1 IgG levels were similar among cirrhotic patients and controls [11] . 22

This prospective cohort study was conducted between April and October 2021 at 4

Sahlgrenska University Hospital, Gothenburg, Sweden. Forty-eight subjects with 5 cirrhosis of various aetiologies were enrolled among patients attending the outpatient 6 clinic at the Department of Gastroenterology and Hepatology at this hospital 7 (Supplemental Figure S1 ). Patients were diagnosed and examined by a specialist in 8 clinical hepatology. Thirty-nine healthy controls were recruited among healthcare 9 personnel at the Sahlgrenska University hospital as well as their family and friends. 10

The baseline characteristics of patients and controls are detailed in Table 1 . Patients 11 or controls with PCR-verified COVID-19 at screening or presence of antibodies against 12 SARS-CoV-2 at initial sampling were not included. 13

The participants received two doses of intramuscular mRNA vaccine (BNT162b2, 15

Comirnaty, Pfizer-BioNTech or mRNA-1273, Spikevax, Moderna), at a median 36 16 (range 26 -62) day interval. Peripheral blood was collected at baseline, i.e., 0-10 days 17 before the 1 st vaccination, as well as after the 1 st (median 35 days (IQR 25-40 days)) 18 and 2 nd (median 89 days (IQR 67-96)) vaccine dose. Serum levels of anti-RBD-S1 IgG 19 and the magnitude of T cell-derived IFN-γ production in response to multimeric S1 20 peptides after vaccination were predefined primary study endpoints. with seronegative serum and reanalysed allowing for an upper detection limit of >5680 13 BAU/ml). 14 15

Vacutainer lithium-heparin tubes (BD, Plymouth, UK) were used to collect peripheral 17 whole blood for assessment of SARS-CoV-2-specific T cell reactivity. One ml of whole 18 blood was transferred to 10 ml-tubes (Sarstedt) and stimulated or not with 1 19 µg/ml/peptide of 15-mer peptides with 11-amino acid overlap spanning the N-terminal 20 S1 domain of the SARS-CoV-2 S1 (130-127-041, Miltenyi Biotec). After two days of 21 incubation at 37°C and 5% CO2, the tubes were centrifuged for 5 minutes at 1,500 rpm 22 and plasma recovered. Plasma was stored at -80°C until analysis of released IFN-. 23

The 15-mer peptides used as stimuli in this assay can be presented on MHC class I 1 and II to activate spike-specific CD8 + T cells and CD4 + T cells, respectively. 2 3

Plasma collected from blood samples with or without S1 peptide stimulation was 5 assessed for IFN- content by ELISA (DY285B, R&D systems) according to the 6 manufacturer's instructions. To minimize nonspecific reactivity, plasma was diluted 7

(1:2) in PBS containing 1% BSA and 10% mouse serum (Invitrogen). Plates were 8 analysed for optical densities at 450 nm and 570 nm using a FLUOstar Omega plate 9 reader (BMG, Ortenberg, Germany). Levels of IFN- induced in response to S1 10 peptides, with background IFN- production in unstimulated samples subtracted, are 11

presented throughout the manuscript. The limit of detection (LOD) of the assay was 12 10 pg/ml as reported elsewhere [13] , and thus this threshold was used in the study. 13

Liver cirrhosis was confirmed at baseline using acoustic radiation force impulse (ARFI) 16 measurement by the ultrasound system Acuson S2000 (Siemens Medical Solutions, 17

Erlangen, Germany). Values were documented in median and interquartile range to 18 median ratio (IQR: median). 19 20

All participants gave written informed consent before enrolment. This study was part 22

of the DurIRVac study approved by the Swedish Ethical Review Authority (permit nos. 23

The trial is registered at the European Union Drug Regulating Authorities Clinical Trials 1 Database (EudraCT no. 2021-000349-42). 2 3

Continuous variables were described as mean, median, and range of values, as 5 applicable. Categorical data were described with contingency tables including 6 frequency and percent. Mann-Whitney U-test was applied to calculate differences in 7 serologic/cellular response between groups. The association between various 8 continuous parameters was determined using Spearman's correlation. Logistic 9 regression was used to calculate the impact of various parameters on cellular and 10 serological immune responses. Parameters with univariate P-values below 0.1 were 11 included in multivariate analysis, and the magnitude of response presented as odds 12 ratios with 95% confidence intervals. For some figures, the data were log-transformed, 13 as indicated in the figure text. Values of BAU/ml and pg/ml below the limit of detection 14 (LOD) were set to 50% of LOD. Data analyses were performed using SPSS for MacOS 15 and GraphPad Prism 8 for macOS. Statistical significance was set to P<0.05. P-values 16 are designated as follows: *P<0.05, **P<0.01, and ***P<0.001. All indicated P-values 17 are two-sided. 18

Results 1

To determine the reactivity of SARS-CoV-2-specific T cells in cirrhotic patients after 4 COVID-19 vaccination, blood samples collected after the 1 st and 2 nd vaccine doses 5

were stimulated with multimeric peptides spanning the S1-region of spike 1 followed 6 by analysis of induced levels of T cell-derived IFN-γ. This assay was previously shown 7

to reflect presence of CD4 + and CD8 + T cells with specificity for S1-antigens [13]. The 8

induction of IFN-γ in response to SARS-CoV-2 S1 peptides was impaired in cirrhotic 9 patients after the 1 st (median <10 vs. 79 pg/ml in controls, P<0.001) and 2 nd 10 vaccination (median 63 vs. 243 pg/ml, P<0.001) ( Figure 1A) . Similarly, the proportion 11 of cirrhotic patients with IFN-γ levels below the level of detection (10 pg/ml) [13] was 12 higher after the 1 st (68 % vs. 19%, P<0.01 vs. controls) and 2 nd vaccination (36% vs. 13 6%, P<0.01; Figure 1C ). 14 15

Similar to the impaired T cell response, anti-RBD-S1 IgG levels were lower in patients 17 with cirrhosis as compared with healthy controls after the 1 st (median 31 vs. 151 18 BAU/ml, P<0.001) and 2 nd (median 514 vs. 1044 BAU/ml, P<0.05) vaccinations 19 ( Figure 1B) . Additionally, after the 1 st vaccination a higher proportion of patients with 20 cirrhosis (35%) lacked detectable levels of anti-RBD-S1 compared with controls (6%) 21 (P<0.05; Figure 1D ). The characteristics of participants achieving or not achieving 22 detectable cellular immune responses (≥10 pg/ml) and >100 BAU/ml of anti-RBD-S1 23

IgG after two vaccine doses are detailed in Table 2 .

The Child-Pugh classification (A-C) determines the severity and prognosis of cirrhosis 3

where patients with class A have less pronounced liver disease and more favourable 4 prospects of long-term survival [14, 15] . Levels of S1-induced IFN-γ were lower in 5 patients with Child-Pugh class B compared with class A after the 1 st (P<0.05) and 2 nd 6 (P<0.01) vaccination (Figure 2A) . Similarly, anti-RBD-S1 IgG levels were lower in 7 patients with Child-Pugh class B cirrhosis after the 1 st vaccination (P<0.05 vs. class A; 8 Figure 2B ). The sample size of patients with Child-Pugh class C (n=2) was insufficient 9 for analysis. No differences were observed regarding T or B cell responses among 10 patients with or without ongoing immunosuppressive therapy or with intercurrent 11 disease ( Table 2 ). The aetiology of cirrhosis was diverse and multifactorial, but 12 insufficient sample size prevented meaningful subgroup analyses. 13 14

Logistic regression was performed to determine the impact of potential confounders 16 on the observed differences of vaccine responses. The T cell-derived S1-induced IFN-17 γ levels were dichotomized based on above or below 10 pg/ml, which reportedly 18 discriminates infected and uninfected subjects with >95% specificity and sensitivity 19

[13]. Anti-RBD S1-IgG levels were dichotomized above or below 100 BAU/ml [16] . The 20 S1-induced IFN-γ response remained significantly inferior in cirrhotic patients vs. 21 controls after the 1 st and 2 nd vaccination when taking gender, age, vaccine type, 22 intercurrent disease, immunosuppressive therapy, and time from vaccination to 23 sampling into account (Table 3) . Similarly, the antibody response after the 1 st 24 vaccination remained significantly reduced in cirrhotic patients in multivariable 1 analysis (Table 4) . The most commonly reported adverse events were reaction at the injection site (64%) 9

and fatigue (22%). The frequency or severity of adverse events did not differ between 10 patients and controls and no serious adverse events were reported or recorded. The main finding of this study was that patients with liver cirrhosis show significantly 3 abated antigen-specific T cell responses after COVID-19 vaccination. We thus 4 observed that 68% of cirrhotic patients lacked T cell reactivity against S1 antigens after 5 the 1 st vaccination and that 36% remained non-reactive after the 2 nd vaccination. showed that the observed T cell deficiency was independent of potential confounders, 9

including intercurrent disease or immunosuppressive therapy. However, the limited 10 sample size may have impacted these analyses. We also observed that T cell 11 dysfunction was significantly more pronounced in Child-Pugh class B cirrhosis than in 12 class A. Thus 9/9 of evaluable patients with class B cirrhosis were completely devoid 13 of T cell reactivity against S1 antigens after the 1 st The immune dysfunction in cirrhosis is primarily associated with flawed innate 24 responses leading to risk of severe and potentially life-threatening bacterial infections [3] [4] [5] . However, also T cell defects including impaired cytokine production elicited by 1 broad T cell stimulation of blood samples from cirrhotic patients have been reported 2 [6, 7] . These findings support that the here reported T cell deficiency against SARS-3

CoV-2 antigens may reflect a generic incapacity to mount T cell-mediated responses 4 to infectious agents in cirrhotic patients. Tables 3 and 4, likely  3 reflecting waning immunity. Further studies are required to clarify if the observed T cell 4 impairment is generic to cirrhosis or to distinct aetiologies of this disease and also if T 5 cell deficiency impacts on the susceptibility to SARS-CoV-2 infection or on the severity 6 of COVID-19. Also, to define an insufficient humoral response, we used a cut-off anti-7 RBD-S1 IgG level of 100 BAU/mL, which was previously utilized as a pre-specified Azathioprine, 3

Vedolizumab, 1

Total 5, (10)

Corticosteroids and azathioprine, 3 J o u r n a l P r e -p r o o f 24 Table 4 . Logistic regression of anti-RBD-S1 IgG ≥100 BAU/mL in the controls and patients with cirrhosis. from the SARS-CoV-2 S1 protein (B) for patients with cirrhosis and healthy controls.

Bar charts with percentages of cirrhotic patients and healthy controls with serum anti-RBD-S1 IgG levels below the limit of detection (C) and undetectable IFN-γ (<10 pg/ml) in supernatant plasma following stimulation with spike 1 peptides (D). Statistics were calculated by Mann-Whitney U test. *p<0.05, **p<0.01, ***p<0.001, ns = not significant. 

Cirrhosis-associated immune dysfunction: 2 distinctive features and clinical relevance

Cirrhosis-associated immune dysfunction

Causes and Consequences of

Immunodeficiency caused by cirrhosis

Immune Dysfunction in Cirrhosis

CD8(+)T cells from patients with cirrhosis display a phenotype that may contribute 13 to cirrhosis-associated immune dysfunction

Dysregulated Adaptive Immunity Is an Early Event in Liver Cirrhosis Preceding 16

Outcomes of COVID-19 in 18

Patients with Cirrhosis or Liver Transplantation

Analysis of antibody responses after

COVID-19 vaccination in liver transplant recipients and those with chronic liver 2 diseases

SARS-CoV2-specific Humoral and T-cell Immune Response After Second Vaccination 5

Liver Cirrhosis and Transplant Patients

WHO International Standard for anti-SARS-CoV-2 immunoglobulin

Rapid cytokine release assay for analysis of SARS-CoV-2-specific T cells in 12 whole blood

Assessment of the prognosis of cirrhosis: Child-Pugh versus 14

Superiority of the Child-Pugh 16 classification to quantitative liver function tests for assessing prognosis of liver 17 cirrhosis

Randomized Trial of a Third Dose of mRNA-1273 Vaccine in Transplant Recipients