Different anti-SARS-CoV-2 vaccine response under B- and T-cell targeted therapies versus anti-cytokine therapies in patients with inflammatory arthritides

BackgroundVaccination against SARS-CoV-2 is effective in preventing severe forms of COVID-19, but there remain concerns about a reduced vaccine response in patients suffering from inflammatory arthritides who are treated by immunosuppressive therapies.ObjectivesWe analysed the impact of bDMARDs on the humoral anti-SARS-CoV-2 vaccine response of patients followed in day hospitals.MethodsWe studied the vaccine response after a complete vaccine regimen followed in day hospital in 5 French hospitals and treated with an intravenous bDMARD between September 2019 and August 2021. After obtaining their informed consent, we included patients with an anti-SARS-CoV-2 serology. They were considered non-responders if the antibody level detected was inferior to the threshold of positivity of the kit used.Results205 patients were included (148 females/57 males). The median age was 64 years (Interquartile Range [IQR] 56-71). 25 were treated with tocilizumab (TCZ), 36 with abatacept (ABA), 53 with infliximab (IFX) and 91 with rituximab (RTX). When considering both patients after a complete vaccination schema (2 doses, or 1 dose in case of prior COVID-19) and those with 1 booster dose, 34 patients (16.6%) were non-responders (2 [5.9%] treated by IFX, none treated by TCZ, 9 [26.5%] treated by ABA and 23 [67.7%] treated by RTX). In multivariate analysis, the only characteristics that significantly and independently differed between responders and non-responders were last bDMARD and corticosteroid therapy at the time of 1st vaccination (Table 1). In RTX-treated patients, the delay from last infusion to 1st vaccine dose was significantly shorter in non-responders (median 4.3 IQR [2.9-6.1] months in non-responders versus 8.4 IQR [5.7-14.5] in responders, p=0.0007). Median survival, i.e. achieving a vaccine response in 50% of RTX-treated subjects, was achieved after 277 days (95CI [209-310]) in patients vaccinated with 2 or 3 doses (Figure 1). In ABA-treated patients, the delay from last infusion to 1st vaccine dose was not different between non-responders and responders.Table 1.Patients’ characteristics and comparisons between responders and non-responders.All patients (n=205)Responders (n=171)Non responders (n=34)Univariate p valueMultivariatep valueAge (median [IQR]), in years64 [56-71]64 [54-70]69 [57-75.5]0.070.40Female sex, n (%)148 (72.2)125 [73.1)23 (67.7)0.53Inflammatory arthritides, n (%)0.16**0.24 Rheumatoid Arthritis156 (78.0)128 (74.9)28 (82.4)0.51 Spondyloarthritis33 (16.1)31 (18.1)2 (5.9)0.12 Others*16 (7.8)12 (5.9)4 (1.9)0.31Last bDMARDs at time of 1st vaccination, n (%)0.0004ABA/RTX versus IFX/TCZ < 0.00010.00024 Infliximab53 (25.9)51 (29.8)2 (5.9) Tocilizumab25 (12.2)25 (14.6)0 Abatacept36 (17.6)27 (15.8)9 (26.5) Rituximab91 (44.4)68 (39.8)23 (67.7)Associated treatments at time of 1st vaccination CsDMARDs, n (%)126 (61.5)107 (62.6)19(55.9)0.56 Methotrexate91 (44.4)78 (45.6)13 (38.2)0.46 Median dose in users (mg /week) [IQR]15 [10-17.5]13.8 [10-15.6]15 [13.8-20]0.07 Corticosteroids, n (%)25 (12.2)19 (11.1)6 (17.6)0.29 Median dose (mg /day) [IQR]0 [0-0]0 [0-0]0 [0-2]0.0350.016Previous COVID-19 infection, n (%)23 (11.2)21 (12.3)2 (5.9)0.38Type of vaccine, n (%)0.62 Pfizer169 (82.4)142 (83.0)27 (79.4)0.62 Moderna14 (68.3)11 (6.4)3 (8.8)0.71 Astra-Zeneca17 (8.3)15 (8.8)2 (5.9)0.74 Janssen5 (2.4)3 (1.8)2 (5.9)0.19Vaccination, n (%) Complete167 (81.5)141 (82.5)28 (16.8)0.47 Complete + 1 booster dose56 (27.3)43 (25.1)13 (38.2)0.14Figure 1.Cumulative seropositive rate according to the interval (days) between the last course of rituximab administration and vaccinationConclusionABA and RTX alter the anti-SARS-CoV-2 vaccine response and were associated with nearly all vaccine non-responses in the present study. Corticosteroid therapy was associated with a lower vaccine response regardless of its indication or the concomitant use of bMARD.Disclosure of InterestsNone declared