Objectives:Patients with immune-mediated inflammatory diseases (IMIDs) treated with immunosuppressive therapies are at higher risk for infections, including those preventable by vaccines. Among these therapies, Janus kinase inhibitors (JAK-is) are increasingly used, though their association with varicella zoster virus (VZV) reactivation raises concerns. The recombinant zoster vaccine (Shingrix™) is recommended for immunocompromised individuals and has shown strong safety and efficacy; however, data on its immunogenicity in IMID patients—especially those receiving JAK-is, anti-TNF agents or MTX—remain limited. In the present study, we aimed to evaluate both humoral and cellular immune responses to the recombinant herpes zoster vaccine in IMID patients treated with different JAK inhibitors in comparison with IMID patients treated with ‘conventional’ therapies, i.e. anti-TNF and/or MTX. We also sought to assess whether immune responses differ between patients receiving pan-JAK-is vs selective JAK1 inhibitors, and to identify clinical factors associated with reduced vaccine-induced immunity.
Methods:This study investigated both humoral (seroconversion rate, titer of VZV-specific IgG antibodies) and cellular CD4 T cell (IL-2 plus IFN-γ) and CD8 T cell (Granzyme A and/or Granzyme B) immune responses following a two-dose regimen of the recombinant inactivated vaccine in two cohorts of IMID patients treated with JAK-i or anti-TNF and/or MTX. Immune responses were compared with responses in a cohort of healthy individuals matched by age and sex. We also sought to assess whether immune responses differ between patients receiving pan-JAK-is vs selective JAK1 inhibitors, and to identify clinical factors associated with reduced vaccine-induced immunity.
Results:A total of 176 participants (mean age 58.6 ± 9.9 years; 33% female) were included. Among them, 39 received non-selective JAK-is (14 baricitinib, 25 tofacitinib), 54 received selective JAK1 inhibitor (upadacitinib) and 46 were on anti-TNF and/or MTX; 37 healthy individuals served as controls. Seroconversion was lower in JAK-i-treated patients (74%) compared with anti-TNF/MTX (91%) and the controls (100%, P < 0.001). Median post-vaccination IgG titers were significantly lower in JAK-i-treated patients (1753 ± 1112 mIU/ml) than in the anti-TNF/MTX patients (2874 ± 1020 mIU/ml) and the controls (3185 ± 576 mIU/ml; P < 0.001). CD4 Th1 responses (IL-2 + IFN-γ) were observed in 28% of JAK-i patients, 67% of anti-TNF/MTX patients and 95% of controls (P < 0.001). CD8 T cell responses were similarly impaired: 29% (JAK-i) vs 41% (anti-TNF/MTX) vs 85% (controls) (P < 0.001). Non-selective JAK-is were associated with the poorest responses (baricitinib: 64% seroconversion; 0% CD4; 7% CD8), while upadacitinib-treated patients showed higher rates (83% seroconversion; 39% CD4; 39% CD8). Granzyme A/B, IL-2 and IL-6 levels post-vaccination were significantly higher in anti-TNF/MTX patients than in the JAK-i groups. Immunogenicity negatively correlated with cumulative glucocorticoid and MTX doses, longer JAK-i exposure and history of ≥2 DMARDs.
Conclusions:Our study highlights impaired immune responses to the recombinant herpes zoster vaccine in IMID patients treated with JAK-is. These patients showed significantly reduced humoral and cellular responses, suggesting lower protection against VZV reactivation. While anti-TNF/MTX-treated patients had mildly reduced responses compared with controls, their immunogenicity was relatively preserved. In contrast, JAK-i-treated patients had markedly lower antibody titers and T cell responses. Notably, this impairment was more pronounced with non-selective JAK-is than with selective JAK1 inhibitors, underscoring differences within the JAK-i class.