Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), a new coronavirus discovered in 2019, resulted in the COVID-19 pandemic (1). Within a remarkably short time, however, research and clinical laboratories worldwide were able to gain knowledge on the natural immunological response to SARS-CoV-2 (2–4), in which several innate and adaptative mechanisms are implicated, and subsequently on the response to the newly developed vaccines. Estimating the immune competence of immunocompromised patients for an infection risk assessment constituted a challenge for clinical immunology units (5), which was particularly relevant for patients with inborn errors of immunity (IEI) (6), a large group of entities affecting a myriad of effector immune responses, with B-cell defects being the most prevalent (7).

Several immune mechanisms are involved at various levels in defending against SARS-CoV-2 (8). Antibody production by B cells is a critical component for adaptive immune responses (9). Naïve B cells recognize antigens from native proteins of infectious agents or proteins encoded in vaccine vectors, diversifying the response in secondary lymphoid organs. In the germinal centre (GC) reaction memory B cells (MBCs) are generated, with high affinity and specificity (due to the process of somatic hypermutation) and with different immunoglobulin heavy chain isotypes (through class-switch recombination) (10). MBCs differentiate into plasmablasts producing neutralizing antibodies. Thanks to the generation of MBC through subsequent exposure to the antigen by infection or vaccine boosters, the antibody responses are much faster and more efficient (11).

BNT162b1 is a lipid-nanoparticle-formulated, nucleoside-modified mRNA vaccine that encodes the trimerized receptor-binding domain (RBD) of the SARS-CoV-2 spike glycoprotein. In Spain, the first group to receive the SARS-CoV-2 mRNA vaccine (BNT162b2/Pfizer) was healthcare workers. Weeks later, immunocompromised patients began to receive the same vaccine.

IEI are a heterogeneous group of diseases caused by intrinsic defects of the immune system (5). Here, we report patients from various IEI categories (7), including combined immunodeficiency (CID) affecting humoral and cellular immunity: CD40 ligand deficiency (CD40L), CID associated with syndromic features: Wiskott-Aldrich syndrome (WAS) with congenital thrombocytopenia and DiGeorge syndrome (thymic defect with additional congenital anomalies). Predominant antibody deficiencies: Bruton tyrosine kinase (BTK deficiency or X-linked agammaglobulinemia), Common variable immunodeficiency with no gene defect specified (CVID), activated p110 d syndrome (APDS), CD19 deficiency, NFKB1 deficiency, NFKB2 deficiency, and isolated immunoglobulin (Ig) G subclass deficiency with IgA deficiency. Diseases of immune dysregulation: CTLA-4 haploinsufficiency (regulatory T-cell defects) and autoimmune polyendocrinopathy-candidiasis-ectodermal dystrophy (APECED). And congenital defects of phagocyte: X-linked granulomatous disease (CGD); as well as 2 patients with suspected IEI and 1 patient with a secondary antibody deficiency after rituximab therapy.

The aim of this study was to determine the humoral responses of the patients with IEI through a comprehensive analysis of specific RBD⁺ IgG⁺ memory B cells by flow cytometry, together with routine S-specific IgG antibodies and QuantiFERON SARS-CoV-2, before the vaccine and 3 weeks after the second dose, compared with healthy donors (HD).

Since 2020, laboratories worldwide have stepped up their efforts in the fight against the COVID-19 pandemic. Our laboratory, specialized in the B-cell dysregulation in IEI, enabled us to apply our knowledge of B-cell biology through multiparametric flow cytometry (15) to analyze the specific B cells generated by natural SARS-CoV-2 infection and after immunization with the SARS-CoV-2 mRNA vaccine (BNT162b2/Pfizer) in HD and patients with IEI, together with routine well-standardized techniques such as EliA SARS-CoV-2-Sp1 IgG and QuantiFERON SARS-CoV-2.

Our first attempt to establish a multiparametric flow cytometry panel to dissect the specific RBD⁺ IgG⁺ MBCs was run with healthcare workers and patients recruited in the internal medicine department for evaluation of past SARS-CoV-2 infection, together with unexposed healthcare workers. We found a higher percentage of specific RBD⁺ IgG⁺ MBCs in the participants with a history of recent SARS-CoV-2 infection compared with the non-infected HD, as previously reported (16).

In 2/5 HDs without clinical evidences of natural SARS-CoV-2 infection, a small percentage of specific RBD⁺ IgG⁺ MBCs was detected. We cannot formally conclude the presence of some unspecific background binding in our assay, or if this could be reflecting some common coronavirus cross-reactive memory B cells generated in response to previous human common coronaviruses. T-cell specific responses have been described as directed mainly to the S2 domain (17), as reported elsewhere (18–20) although Rodda et al. (16) also detected in HD RBD+ IgM+IgD+ MBCs that could be cross-reactive MBCs generated in response to a previous human coronavirus. Moreover, the cross-reactivity to SARS-CoV-2 T-cell epitopes in unexposed humans could explain the high heterogeneity in infected individuals (21).

In the last days of 2020, a vaccination scheme was started in Spain, and the first group to be vaccinated were healthcare workers. We designed the recruitment of non-previously exposed healthcare workers and others with a history of past SARS-CoV-2 infection to analyze specific B cells by flow cytometry, generated after immunization with the SARS-CoV-2 mRNA vaccine (BNT162b2/Pfizer), together with S-specific IgG antibodies shown by ELISA. Our results indicated that the antibody levels in those individuals with a past infection were higher than those of infection-naïve individuals after the first dose, in line with several previous studies that reported the strongest humoral (20) and cellular T-cell responses (22). Although QuantiFERON SARS-CoV-2 was not commercially available during the vaccination study of healthcare workers, cellular response was measured by the secretion of interleukin-2 and interferon-gamma by the specific CD4⁺CD154⁺Th1 cells (22).

S-specific IgG antibody levels were similar after the second dose in those previously exposed and those unexposed to SARS-CoV-2 infection, as reported in another cohort of healthcare workers (23). After the second dose of the vaccine, we also detected a significantly higher frequency of specific IgG⁺ RBD⁺ MBCs in those participants with previous natural SARS-CoV-2 infection. However, both groups presented an increase in the percentage of specific IgG⁺ RBD⁺ MBCs 21 days after the second dose.

Spike-specific memory B-cell clones can be detected for up to 6 and 8 months in individuals with mild to severe SARS-CoV-2 infection (22), most of which are IgG, whereas IgA is severely reduced, and IgM clones tend to disappear earlier (1). In naturally infected individuals, polyfunctional SARS-CoV-2 antigen-specific B and T memory cells have been detected up to 6 months after the infection (24). B-cell biology also appears to be after the different outcomes in the infectious process, the greater frequency of IgM⁺ memory B cells specific to SARS-CoV-2 was related to less extend symptoms (25).

Humoral immune response is in continuous evolution, with memory B cells starting clonal turnover at approximately 6 months (9). From that moment on, the evolution of the response occurs in the GC where antigens are retained by dendritic cells, and RBD memory B cells undergo greater somatic hypermutation (SHM), being able to produce a broader neutralizing response capacity (9). Prior knowledge from the influenza vaccine indicates that the specific antibody levels are promoted by plasma cells in bone marrow after 4 weeks and decrease after 1 year (26). Another study analyzed the lymph nodes after influenza vaccination and found that the vaccine-binding B-cell clones during the following weeks after vaccination were similar to those in peripheral blood and secondary lymphoid organs, with a high frequency of SHM and broad cross-reactivity. However, the GC B-cell clones in lymph nodes are highly specialized and can target new epitopes and broaden the spectrum of vaccine-induced protective antibodies (11).

Our previous study of healthcare workers provided a reference to compare the patients with IEI, who were included in the vaccination schedule months later. Patients with IEI are characterized by impairment of one or more components of the immune system, depending on their intrinsic defect (7). In general, the risk factors affecting mortality and comorbidity in these patients are similar to those of the general population, and approximately 30% of patients with IEI presented with mild COVID-19 disease (5). In fact, the 3 confirmed patients included in our study had mild COVID-19, and the other 4 patients who had low titers of specific antibodies in the pre-vaccination study might have been asymptomatic. Other cohorts have found a greater comorbidity in patients with primary immunodeficiency and especially in patients with symptomatic secondary immunodeficiency (SID) (27).

We detected a proper humoral and cellular response in 39% of the patients with IEI when analyzing with presence of S-specific IgG antibodies together with specific IgG⁺ RBD⁺ MBCs and a positive QuantiFERON SARS-CoV-2 (with the exception of 1 patient with IgG and IgA hypogammaglobulinemia). Thirteen percent of patients with IEI studied, who had very low S-specific IgG antibodies, had insufficient events in the IgG⁺ RBD⁺ MBC category ( Figure 4 ); however, they had a positive QuantiFERON SARS-CoV-2 (except for 1 patient). T-cell response was positive in all but one of the study patients with CVID after vaccination, with similar results in other cohorts (28) and in studies reporting natural SARS-CoV-2 infection (29).

If we consider only the positive humoral response by measuring antibodies, 74% of the patients with IEI were able to generate antibodies (in low and high titers), results similar to those obtained by Delmonte et al., with 85% of patients with IEI producing antibodies after the second dose (30). A pilot study with 11 patients with IEI detected seroconversion after vaccination, except for 1 patient with concomitant immunosuppression therapy (31). Another cohort with 41 patients with IEI showed variability in humoral response in CVID, which was totally impaired in XLA patients but with almost normal T-cell response (32). Reports have highlighted the importance of studying patients with antibody deficiency in an isolated manner due to each patient’s variability and unique characteristics in terms of cellular and humoral immunity (32). In fact, our cohort had a patient with a leaky BTK mutation (in press) that generated a powerful B and T-cell response.

Within the IEI classified as predominant antibody deficiency, we detected specific IgG⁺ RBD⁺ MBCs in half of them, with the exception of 1 patient with CVID, 1 patient with NFKB1 deficiency, 1 patient with NFKB2 deficiency, and 1 patient with suspected IEI without confirmed genetic diagnosis. However, those patients without detectable specific IgG⁺ RBD⁺ MBCs, could produce antibodies and had an positive cellular response. Interestingly, there were 2 siblings within our cohort with an NFKB1 defect with the presence of positive S-specific IgG antibodies and positive cellular response; in whom, we detected no specific IgG⁺ RBD⁺ MBCs in one of them, similar findings to those reported in the cohort of Amodio et al. in which one of their 2 siblings with the NFKB1 defect lacked humoral response, and both of them presented impaired cellular response (33, 34).

Our study showed that the reduction in B-cell counts in peripheral blood (<100 CD19 B cells/µL) and a reduced class-switched memory B-cell compartment (<2% Sm B cells) are critical for a good humoral response in terms of S-specific IgG antibodies, as others have already reported (30, 34, 35). The presence of B cells in peripheral blood is sufficient to generate specific antibodies in most patients. In certain cases, however, it is insufficient for a proper detection of specific IgG⁺ RBD⁺ MBCs, given that we were able to detect them in only 55% of those patients. Other studies have also reported heterogenic results in their CVID cohort when analyzing specific IgG⁺ RBD⁺ MBCs (33). Interestingly, we found 2 CTLA-4 patients from our cohort without detectable B cells in peripheral blood. These patients were able to produce antibodies after natural SARS-CoV-2 infection or after vaccination. One possible explanation is that these patients had a tendency to a rapid response in generating antibodies by plasma cells in secondary lymphoid organs or bone marrow, without detectable B cells in peripheral blood (9, 11). Another possible explanation found by Salinas et al. who analyzed the B-cell phenotype and found atypical specific B cells in patients with CVID and suggested that it could be the consequence of impaired GC reactions (32).

However, other patients of our cohort without detectable specific IgG⁺ RBD⁺ MBCs, including 1 with X-linked agammaglobulinemia deficiency, 2 with CVID, and a patient with profound hypogammaglobulinemia after rituximab treatment, also presented impaired antibody production but positive T-cell response by using a ELISpot estimating IL-2 and IFNγ secretion (27, 33), as previously reported. This situation is slightly different from those patients that are treated continuously with B-cell-depleting therapies; despite, the separation between treatment and vaccine of almost 20 weeks, only 36% of the patients could generate specific antibodies following vaccination (36).

Within our cohort, we also studied other IEI classified as CID associated with syndromic features, such as Wiskott-Aldrich syndrome, in which a proper humoral and cellular response was detected, as similar results reported (31). In 2 patients with DiGeorge syndrome, we observed a complete humoral response in one and low titers of specific antibodies but with no cellular response in the other, similar results as those obtained by Squire et al. (31). Despite the absence of B cells in peripheral blood, the 2 patients with CTLA-4 in our cohort produced antibodies after natural SARS-CoV-2 infection and vaccination, respectively, and presented positive T-cell response, although other studies have observed a lack of T-cell response in patients with CTLA-4 by using other technique detecting antigen specific T cells (34). In APECED (another IEI disease caused by immune dysregulation), Delmonte et al. found that the percentage of positive humoral response could be reduced in those patients (30). Of note, the APECED patient of our cohort could also generate S-specific IgG antibodies, but we were unable to detect IgG⁺ RBD⁺ MBCs.

We observed a profound failure of B and T-cell response in 3 of the patients studied: one patient with CD40L deficiency classified as CID affecting humoral and cellular immunity; one patient with CGD classified as a congenital defect of phagocyte and one patient with APDS1, a predominant antibody deficiency. Other studies also found undetectable levels of specific antibodies in 3/21 patients with IEI and impaired T-cell response in 5/21 patients (33).

Several studies in general population have raised the importance of booster for a greater protection against SARS-CoV-2 infection and leads to reduce disease severity. Among IEI patients the third doses of BNT162b2 reduces the seroconversion failure from 45% to 26% (37), similar results from Shields et al, reducing the seroconversion failure from 39 to 34% in IEI patients (38). However, there are still some IEI patients that failed to produce antibody response or a T cell response.

IEI patients are a good example to understand molecular mechanism of the immune system. In general, several reports have shown that the risk factors, the severity of the disease and case fatality in patients with IEI was similar to that observed in the general population (39). Also, a review of Tangye et al. (40) nicely summarizes that the percentage of infected asymptomatic IEI patients is similar to general population (10-20%), the majority of IEI patients developed mild disease (30-50%). Although there are some differences with the general population, for example IEI patients have increased case fatality rate, increased admission rate to ICU, younger age of infection and younger age at death, and longer disease duration with longer viremia/viral shedding. Also, similar findings showed in a Brazilian cohort, with a direct relationship between the severity of SARS-CoV-2 infection in IEI patients and older age and higher number of comorbidities (41) and reported by the Italian Primary Immunodeficiencies Network (42). Cousins et al. reported that patients through different phases of the pandemic experiences mild clinical course with limited symptoms (43). And that the clinical impact of COVID-19 in IEI patients varies from mild symptoms to death. However, Cousins et al, also reported that the Mount Sinai cohort over time have been less ill and required less hospitalization, in accordance to general population. Another study in IEI pediatric to young adults, reported higher COVID-19 pneumonia rate that the general age-range population with lower humoral and cellular responses in the acute phase (44).

The burden of the SARS-CoV-2 pandemic for the whole population, with its many initial uncertainties, was even greater for immunocompromised patients. Clinical laboratories in reference centers for IEI immediately assumed the task of exploring and validating novel methodologies to asses antibody and cellular specific responses to natural disease and first vaccines. This was crucial to support, as much as possible, individual data-based clinical decisions and risk assessment for patients. In this article we describe an in-depth evaluation of vaccine responses based on solid laboratory methods. Our results and others suggest that IEI and more specifically PADs patients elicit T cell responses, and remarkably antibody responses to rather similar to healthy controls. These methods could be applied in prospective studies for further delineating the persistence and reinforcement of these responses upon subsequent viral exposures or vaccine boosts.

The original contributions presented in the study are included in the article/ Supplementary Material . Further inquiries can be directed to the corresponding authors.

The studies involving human participants were reviewed and approved by Hospital Universitario La Paz, Madrid, Spain (PI-4101). The patients/participants provided their written informed consent to participate in this study.

ELG conceived the project. ELG, LPM, YBG and PN designed the experiments. LPM, YBG, PN and YS-S performed the experiments. LPM, YBG, PN and ELG wrote the manuscript. All authors contributed to the article and approved the submitted version.