Incidence of covid-19 infection and hospitalisation according to vaccination status and dmard treatment in patients with rheumatoid arthritis: a nationwide matched cohort study from denmark

BackgroundPatients with rheumatoid arthritis (RA) may have impaired immunogenicity to COVID-19 vaccines.ObjectivesTo investigate the incidence of COVID-19 infection and hospitalisation in unvaccinated and vaccinated patients with RA compared with matched individuals; and secondarily in patients with RA according to DMARD treatment.MethodsDanish nationwide matched cohort study from January to October 2021. Patients with RA were identified in DANBIO and matched 1:20 with individuals from the general population on age, sex, and vaccination status (month and exact type of vaccination). Primary and secondary outcomes were COVID-19 hospitalisation (Danish National Patient Register) and positive SARS-CoV2 PCR test (Danish COVID-19 Surveillance Register), respectively. Stratified by vaccination status, incidence rates (IRs) per 1000 person years (PY) and comorbidity-adjusted hazard ratios (aHRs) in cause-specific Cox models were calculated with 95% confidence intervals. Using the Aalen-Johansen estimator, the cumulative incidence of COVID-19 hospitalisations was visualised according to RA and vaccine exposure status.ResultsRegardless of vaccination status, patients with RA had increased incidence of COVID-19 hospitalisation compared to matched individuals (Table 1). However, the absolute risk was 0.20% for unvaccinated patients at 60 days and 0.08% for comparators, whereas it remained below 0.05% at 180 days of follow-up in both groups when fully vaccinated (Figure 1). Increased SARS-CoV2 infection rates were seen only among unvaccinated patients with RA (Table 1). Unadjusted analyses showed increased incidence of COVID-19 hospitalisation among rituximab-treated compared with conventional DMARD treated: unvaccinated HR 4.71 (1.98 to 11.18) and vaccinated HR 11.69 (2.07 to 66.06). However, the proportions of patients with previous cancer and treated with prednisolone were higher among the rituximab treated.Table 1.UnvaccinatedPartially vaccinatedFully vaccinatedRAControlsRAControlsRAControlsN28 447568 94027 154542 61026 217523 826Women, %71.371.371.271.271.071.0Age in years, median [IQR]67.7 [34.2 to 88.3]67.8 [34.2 to 88.4]68.4 [36.4 to 88.6]68.4 (36.5 to 88.6)68.9 [40.9 to 88.7]68.9 (41.0 to 88.7)Methotrexate /55.5 /0.5 /55.4 /1.2 /55.7 /1.3 /Sulfasalazine /14.2 /0.1 /13.7 /0.3 /13.5 /0.3 /Hydroxychloroquine /10.4 /0.1 /10.3 /0.0 /10.3 /0.0 /Other csDMARD,11.0 /0.2 /10.7 /0.3 /10.6 /0.3 /Prednisolone,all in %12.52.012.20.512.20.5TNFi /16.9 /0.1 /17.2 /2.9 /17.1 /3.1 /abatacept /1.5 /0.0 /1.5 /0.5 /1.5 /0.5 /tocilizumab /3.0 /0.0 /3.0 /0.0 /2.9 /0.0 /rituximab, all in %2.20.12.10.12.10.1COVID-19 hospitalisationN65727119511131Median [IQR] days of follow-up102 [62 to 137]115 [88 to 146]28 [22 to 35]30 (21 to 39)150 [111 to 189]150 (111 to 189)Rate per 1000 PY10.4 (8.0 to 13.4)4.7 (4.3 to 5.1)5.5 (3.0 to 10.0)2.2 (1.8 to 2.7)0.9 (0.5 to 1.6)0.5 (0.4 to 0.6)Adjusted HRa1.88 (1.44 to 2.46)1 (Ref.)2.47 (1.25 to 4.89)1 (Ref.)1.94 (1.03 to 3.66)1 (Ref.)SARS-CoV2 infectionRate per 1000 PY37.8 (33.6 to 42.6)33.9 (33.1 to 34.8)27 (20.7 to 35.1)28.5 (27 to 30.2)11.3 (9.2 to 13.9)10.4 (9.9 to 10.9)Adjusted HRa1.22 (1.09 to 1.57)1 (Ref.)0.87 (0.95 to 1.74)1 (Ref.)1.09 (0.92 to 1.14)1 (Ref.)IQR, Interquartile range. a Adjusted for cancer history, cardiovascular disease, diabetes mellitus, chronic kidney disease, and chronic lung disease.Figure 1.Cumulative incidence of COVID-19 hospitalisation (%) as a function of follow-up time (days) for (A) unvaccinated, (B) partially vaccinated and (C) fully vaccinated patients and comparators.ConclusionThe incidence of COVID-19 hospitalisation was increased for both unvaccinated and vaccinated patients with RA compared with controls. Importantly, the parallel decreasing risk for patients with RA suggests a comparable relative benefit of vaccination. Less favourable outcomes among rituximab-treated warrant that this drug should be considered with extra care.AcknowledgementsThe authors wish to acknowledge The Danish Departments of Clinical Microbiology and Statens Serum Institut for carrying out laboratory analysis, registration, and release of the national SARS-CoV-2 surveillance data use in the present study. Further, the authors wish to thank all the Danish departments of rheumatology for reporting to the DANBIO register.Disclosure of InterestsRené Cordtz: None declared, Salome Kristensen: None declared, Rasmus Westermann: None declared, Kirsten Duch: None declared, Fiona Pearce Grant/research support from: Pearce reports a grant from Vifor Pharma outside the submitted work., Jesper Lindhardsen: None declared, Christian Torp-Pedersen Grant/research support from: Torp-Pedersen reports grants from Bayer and Novo Nordisk outside the submitted work., Mikkel Porsborg Andersen: None declared, Lene Dreyer Speakers bureau: Dreyer has received speakers bureau from Eli Lilly and Galderma., Grant/research support from: Dreyer has received research grant/support from BMS.