Impaired Memory B-Cell Response to Influenza Immunization in Patients With Common Variable Immunodeficiency (CVID)

Abstract

Background: Common variable immunodeficiency (CVID) is a heterogeneous primary immunodeficiency characterized by low serum antibody levels and recurrent infections. The cellular response to immunization in patients with CVID has not been fully investigated. In this study, we aimed to characterize vaccination-induced influenza-specific memory B-cell responses in CVID. Methods: Eleven individuals affected with CVID and 9 unaffected control individuals were immunized with the 2010-2011 non-adjuvanted seasonal influenza vaccine. Blood samples were collected on the day of vaccination and at week 8 and week 16 after vaccination, and PBMCs were immunophenotyped by flow cytometry. Influenza specific serology was determined using hemagglutination inhibition and microneutralization against vaccine antigens. Influenza-specific memory B-cell responses were determined by ELISpot.  Results: Individuals with CVID showed wide variability in the frequency of CD19+ B cells in blood. The CVID group had significantly reduced frequencies of CD19+CD27+ memory B cells. Frequencies of circulating T follicular helper (CD4+CXCR5+) cells were similar between those with CVID and healthy controls. In terms of serology, compared to healthy controls, the CVID group overall showed significantly reduced boosting to vaccine antigens by hemagglutination inhibition and microneutralization assays at 8 weeks compared to controls and failed to maintain responses by 16 weeks compared to controls, resulting in a post-vaccination geometric mean titer (GMT) ≥ 40 to strain A/H1N1 in only 27% at 8 weeks, and 22% at 12 weeks for patients with CVID vs 78% and 75%, respectively for healthy controls. In addition, there was a GMT ≥ 40 to A/H3N2 in only 9% at 8 weeks and 22% at 12 weeks for patients with CVID vs 56% and 50%, respectively for healthy controls. Healthy participants showed significant increases in flu-specific IgM-secreting memory B cells after vaccination, whereas patients with CVID showed non-significant mild increases. Before vaccination, patients with CVID had significantly lower frequencies of background level influenza-specific IgG and IgA memory B cells. Half of the patients with CVID showed an increase in influenza-specific IgG-secreting memory B cells post vaccination, whereas the other half showed none. All control participants exhibited an increase in influenza-specific IgG-secreting B cells. None of the patients with CVID developed influenza-specific IgA memory B-cell response post vaccination, compared to 5/8 in healthy controls. At week 16, the frequency of influenza-specific memory B-cell responses decayed but to non-zero baseline in healthy controls and to zero baseline in patients with CVID.  Conclusions: Together, these data demonstrate that patients with CVID respond heterogeneously, but as a group poorly, to non-adjuvanted influenza vaccine, with a subgroup unable to generate influenza-specific memory B-cell responses. No patient with CVID was able to maintain memory response for prolonged periods. Together, our results suggest a defect in Ig class switching and memory B-cell maintenance in patients with CVID during a de novo vaccine immune response.