Longitudinal SARS-CoV-2 mRNA vaccine immunogenicity in inflammatory bowel disease in relation to anti-TNF therapy and age: a prospective cohort study

This is the authors' abstract. We don't add key points for this paper.

BACKGROUND: Patients with inflammatory bowel disease (IBD) on immunomodulatory therapy may respond less well to vaccination. We followed humoral and cellular responses to SARS-CoV-2 vaccination over 12 months. METHODS: In this single-centre, prospective cohort study, adults with IBD received mRNA vaccine under the Austrian national programme. Two doses were administered 28 days apart and a third dose 6-12 months after the first. S1-RBD IgG, live-virus 50% neutralisation titres (NT50) against wild-type, Delta, Omicron BA.1 and BA.2, and IFN-γ ELISpot responses were measured at baseline and after 3, 6 and 12 months. Associations with anti-TNF therapy and age were analysed with rank-based longitudinal models and with mixed-effects models adjusted for age and sex. RESULTS: All 53 twice-vaccinated, infection-naive patients seroconverted at 3 months. Geometric mean (GM) S1-RBD IgG peaked at 3 months (465.7 BAU/ml) and fell by 6 months (124.8 BAU/ml). At 3 months, GM NT50 values were lower against BA.1 (9.0) than against wild-type (24.2) or Delta (26.9), and 13 patients (24.5%) never reached a wild-type NT50 of 1:16. Anti-TNF therapy (32/53, 60.4%) was significantly associated with faster waning. At 6 months, adjusted geometric mean ratio versus other therapy was 0.31 (95% CI 0.20-0.49) for S1-RBD IgG and 0.22 (0.12-0.41) for wild-type NT50. Each additional year of age was associated with 2.4% lower wild-type NT50 (0.976 per year, 95% CI 0.962-0.990). In 33 thrice-vaccinated patients, intercurrent infection (15/33) raised S1-RBD IgG 4.6-fold more than the third dose alone, without a measurable gain in neutralisation. In the adjusted model, IFN-γ ELISpot responses did not differ between treatment groups. DISCUSSION: Anti-TNF therapy and higher age within a relatively young cohort were associated with lower and shorter-lived humoral responses, while cellular responses appeared comparatively preserved. These exploratory associations need confirmation in larger cohorts with clinical endpoints.