Characterization of the humoral and B-cell spike-specific immune response after mRNA SARS-CoV-2 vaccination in healthy and fragile subjects

Introduction. Vaccines against SARS-CoV-2 have shown good immunogenicity in healthy individuals, triggering both humoral and cell-mediated responses. In the real-world context, mass administration of SARS-CoV-2 vaccines has significantly reduced severe disease, hospitalizations, and deaths due to COVID-19. Nevertheless, the physiological decline of circulating anti-spike antibodies and the spread of variants of concern (VOCs) have led to a resurgence of infections, prompting the scheduling of additional vaccine doses. Fragile or vulnerable subjects, who are moderately or severely immunocompromised, exhibit a greater susceptibility to infections and relative reduced responsiveness to vaccination due to immunological disorders and pharmacological treatments, therefore they represent a priority category for vaccination strategies. Regrettably, fragile patients have been excluded from most clinical trials, resulting in scarce vaccine immunogenicity data for these cohorts, especially for the long-term persistence as immune memory. The generation of a memory response, in the forms of B cells (MBC) and T cells, is required to attain a long-term vaccine immunity. Upon second antigen encounter, MBC quickly undergo clonal expansion and produce new plasma cells and circulating antibodies to fight back the infection. To define these aspects, the induction and the persistence of spike-specific humoral and B cellular responses to SARS-CoV-2 vaccination in different cohorts of fragile subjects have been assessed and compared with healthy controls. Materials and Methods. People living with HIV (PLWHIV, 114 subjects) under antiretroviral therapy (ART), Myelofibrosis patients (MF, 42 subjects), hematopoietic stem cells transplant recipients (HSCT, 80 subjects), and healthy controls (HCs, 200 subjects) were enrolled in the context of the PATO_VAC and IMMUNO_COV longitudinal observational clinical studies conducted at Siena University Hospital. In collaboration with the San Paolo Hospital of Milan, we expanded our analysis to a larger cohort of PLWHIV (SPID-HIV-Vax; 496 subjects). Blood samples were collected before and after each vaccine dose, up to 2 years after the first vaccine dose. SARS-CoV-2-specific memory B cells (MBC) were characterized for surface markers expression with a multiparametric flow cytometry approach and for functionality, with enzymelinked immunosorbent spot assay. The humoral response was assessed with enzyme-linked immunosorbent assay and with a surrogate virus neutralization assay. Clinical and demographic data were collected and used for linear regression analysis and machine learning algorithms, in order to assess the effect of these parameters on vaccine immunogenicity. Data from PLWHIV SPID-HIV-Vax cohort were used to build a machine learning model capable of predicting antibody response to vaccination based on available demographic and clinical parameters. This machine learning algorithm was then applied to the PLWHIV cohort of PATO_VAC study. Results. In HCs significant spike-specific IgG and MBC responses were detected after the second vaccine dose which, after the decline in the first two months post vaccine administration, persisted for at least 9 months (CHAPTERS 3,4) (Ciabattini et al. 2021; 2023). The booster dose strongly enhanced the antibody titres and induced an Omicron-cross-reactive response (APPENDIX 1.1). PLWHIV showed a similar humoral response kinetic after the first vaccine cycle but a significant phenotypic difference in MBC compared to HCs. PLWHIV presented higher frequencies of double negative (DN) and lower frequencies of Ig-switched and resting cells (CHAPTER 5) (Polvere et al. 2023). The third and fourth doses induced a modulation of the antigen-specific B cell phenotype, becoming like that observed in HCs. The long-term analysis of the vaccine immunogenicity showed a significant spike-specific cellular and humoral response up to 2 years, although antibody titres against wild type and Omicron were slightly lower compared to HCs (CHAPTER 6). The Random Forest model applied to this cohort identified CD4 count and CD4/CD8 ratio as positively correlated to the IgG production while extremely positive or negative values of body mass index were negatively associated to antibody titres (CHAPTERS 6, 7). MF patients exhibited a reduced and slower humoral and MBC response following the first vaccination cycle, that was increased by the booster doses (CHAPTERS 8, 9) (Fiorino et al. 2021; 2022). The majority of HSCT patients developed a humoral response after the first vaccine cycle, while non-responders managed to develop an anti-spike response following the third and fourth doses. However, the MBC response consistently stayed lower than in healthy controls even after booster doses (CHAPTER 10; APPENDIX 1.2) (Pettini et al. 2022). Conclusions. These results contribute to demonstrate the high immunogenicity of anti-SARSCoV- 2 mRNA vaccination in HCs and the induction of MBC response that persists over time and is reactivated by subsequent boosters. Our data reveal a slower and lower spike-specific humoral and MBC response in some cohorts of vulnerable individuals after the initial vaccine cycle, significantly increased after the third and fourth doses. Fragile individuals may benefit from the administration of additional doses, which are able to induce humoral and memory responses even in subjects who did not respond to the initial vaccination cycle. These results offer practical guidance for implementing targeted vaccination strategies in fragile patients and are particularly relevant for improving possible epidemic or pandemic vaccination policies.