Humoral and cellular immune responses to two and three doses of SARS-CoV-2 vaccines in rituximab-treated patients with rheumatoid arthritis: a prospective, cohort study

Abstract

Background: In rituximab-treated patients with rheumatoid arthritis, humoral and cellular immune responses after two or three doses of SARS-CoV-2 vaccines are not well characterised. We aimed to address this knowledge gap.

Methods: This prospective, cohort study (Nor-vaC) was done at two hospitals in Norway. For this sub-study, we enrolled patients with rheumatoid arthritis on rituximab treatment and healthy controls who received SARS-CoV-2 vaccines according to the Norwegian national vaccination programme. Patients with insufficient serological responses to two doses (antibody to the receptor-binding domain [RBD] of the SARS-CoV-2 spike protein concentration <100 arbitrary units [AU]/mL) were allotted a third vaccine dose. Antibodies to the RBD of the SARS-CoV-2 spike protein were measured in serum 2-4 weeks after the second and third doses. Vaccine-elicited T-cell responses were assessed in vitro using blood samples taken before and 7-10 days after the second dose and 3 weeks after the third dose from a subset of patients by stimulating cryopreserved peripheral blood mononuclear cells with spike protein peptides. The main outcomes were the proportions of participants with serological responses (anti-RBD antibody concentrations of ≥70 AU/mL) and T-cell responses to spike peptides following two and three doses of SARS-CoV-2 vaccines. The study is registered at ClinicalTrials.gov, NCT04798625, and is ongoing.

Findings: Between Feb 9, 2021, and May 27, 2021, 90 patients were enrolled, 87 of whom donated serum and were included in our analyses (69 [79·3%] women and 18 [20·7%] men). 1114 healthy controls were included (854 [76·7%] women and 260 [23·3%] men). 49 patients were allotted a third vaccine dose. 19 (21·8%) of 87 patients, compared with 1096 (98·4%) of 1114 healthy controls, had a serological response after two doses (p<0·0001). Time since last rituximab infusion (median 267 days [IQR 222-324] in responders vs 107 days [80-152] in non-responders) and vaccine type (mRNA-1273 vs BNT162b2) were significantly associated with serological response (adjusting for age and sex). After two doses, 10 (53%) of 19 patients had CD4⁺ T-cell responses and 14 (74%) had CD8⁺ T-cell responses. A third vaccine dose induced serological responses in eight (16·3%) of 49 patients, but induced CD4⁺ and CD8⁺ T-cell responses in all patients assessed (n=12), including responses to the SARS-CoV-2 delta variant (B.1.617.2). Adverse events were reported in 32 (48%) of 67 patients and in 191 (78%) of 244 healthy controls after two doses, with the frequency not increasing after the third dose. There were no serious adverse events or deaths.

Interpretation: This study provides important insight into the divergent humoral and cellular responses to two and three doses of SARS-CoV-2 vaccines in rituximab-treated patients with rheumatoid arthritis. A third vaccine dose given 6-9 months after a rituximab infusion might not induce a serological response, but could be considered to boost the cellular immune response.

Funding: The Coalition for Epidemic Preparedness Innovations, Research Council of Norway Covid, the KG Jebsen Foundation, Oslo University Hospital, the University of Oslo, the South-Eastern Norway Regional Health Authority, Dr Trygve Gythfeldt og frues forskningsfond, the Karin Fossum Foundation, and the Research Foundation at Diakonhjemmet Hospital.

Figure 1

Humoral response to two and three vaccine doses (A) Anti-RBD antibody concentrations in controls, patients who had received at least two doses, patients who had received two doses and would later receive a third, and patients who had received three doses. The violin illustrates the kernel probability density and the orange line indicates the median. Dots denote individual patients. (B) Time between last rituximab infusion and first vaccine dose according to response status in all patients after their second vaccine dose. The violin illustrates the kernel probability density and the orange line indicates the median. Dots denote individual patients. (C) Anti-RBD antibody concentrations after the second and third doses. Solid lines connect patients' two samples (circles). The horizontal dotted line indicates the cutoff for positivity (70 AU/mL). (D) Time between the last rituximab infusion and anti-RBD response after the third vaccine dose. AU=arbitrary units. RBD=receptor-binding domain.

Figure 2

T-cell responses after two and three vaccine doses (A) Anti-wild-type spike protein-specific T-cell responses in patients after two and three doses and in healthy controls after two doses. CD4⁺ T-cell responses and CD8⁺ T-cell responses are shown for all unstimulated and stimulated pairs. The p values from Wilcoxon matched pairs signed rank tests are shown, with * indicating p<0·001 and †indicating p<0·0001. Patients with a response (positive) and patients without a response (negative) are indicated. (B) Analysis of T-cell responses directed against wild-type and delta variant SARS-CoV-2 spike peptides in patients after two and three doses (Spearman correlation). Solid lines show simple linear regression of correlation and dotted lines represent 95% CIs. (C) Percentage of anti-spike protein CD4⁺ T cells versus anti-spike protein CD8⁺ T cells in patient responders to wild-type and delta variant spike peptides using combined data of the second and third doses. Spearman correlation is shown. (D) Percentage of anti-spike protein CD4⁺ T cells versus age in patient responders to wild-type and delta variant spike peptides using combined data of the second and third doses. Spearman correlation is shown. See the appendix (p 5) for supplementary data for gating and controls.

Figure 3

Adverse events following two or three vaccine doses in patients and controls (A) All patients. (B) Controls. (C) Patients who received three vaccine doses. Adverse events were reported for all patients and a subset (n=246) of healthy controls (health-care workers at Diakonhjemmet Hospital and Akershus University Hospital, Oslo, Norway). *Duration not measured. †No patients were hospitalised due to disease flares after vaccination.