Humoral immune response
One month after the second BNT162b2 dose, anti-RBD SARS-CoV-2 specific antibodies were detected in 52.4% of patients (n=11/21) using the IB assay (574.4±454 U/mL, Fig 2A). The humoral immune response was confirmed using ELISA (Fig 2B). Other anti-SARS-CoV-2 specific antibodies, such as NCP and anti-Spike S2, were detected in two patients; however, they neither developed symptomatic infection nor had positive RT-PCR results for SARS-CoV-2, suggesting asymptomatic infection.
The humoral response was comparable to that of the HCs (870±225.0 U/mL). However, patients with CVID had significantly lower virus-neutralizing antibody titers (49.4±81.5 U/mL vs. 960±1093 U/mL,p <0.0001, Fig 2C), which persisted at month 3 in 44.4% (n=8/18) of patients; however, they were significantly lower than those of the HCs (436.3 ± 415.4 U/ml vs. 900.3 ± 232.7% U/mL,p =0.0002). On the other hand, the level of neutralizing antibodies correlated with the concentration of anti-RBD SARS-CoV-2 specific antibodies (r= 0.82, p= 0.0001). Three patients were lost to follow-up. The antibody titer decreased further at 6 months and was detected in only 33.3% (5/15) of patients. The mean anti-RBD-specific IgG level was 218.2 U/mL (±268.5) in patients with CVID and 1056 U/mL (±360.4) in HCs, with statistically significant differences (p <0.0001; Fig 2A). Three patients were excluded based on RT-PCR-confirmed SARS-CoV-2 infection during the study.
Responders and non-responders did not differ in sex, disease duration, proportion of ATB prophylaxis, IRT dose, or IgG trough levels; however, we observed significant differences between age and serum IgM concentration. The responder group comprised significantly younger patients with CVID (39.1±8.1 years vs. 51.36±8.13 years; p =0.003) who had significantly higher serum IgM levels (0.29±0.21 g/L vs. 0.1±0.06 g/L; p =0.002). The humoral-response level (serum anti-RBD SARS-CoV-2 specific antibody concentration) also negatively correlated with higher age (r=-0.61, p =0.003) as well as serum IgM (r=0.46, p =0.036). The humoral response was not influenced by ongoing or previous immunosuppression or the presence of non-infectious complications. We did not observe any significant differences in the T-cell phenotypes. However, responders had a significantly higher proportion of CD19+ class-switched (CS) B cells (7.58±3.09 vs. 3.39±2.09%; p =0.009; Table 1). The number of CD19+CS B cells correlated with the concentration of anti-RBD SARS-CoV-2 specific antibodies (r= 0.58, p = 0.04).
Additionally, we assessed APRIL, BAFF, and IFNɑ as potential markers of humoral response that contribute to B-cell maturation, survival, and class switch. However, we did not observe any significant differences in the serum BAFF concentration (1841±527 vs. 2496±1183 pg/mL;p =0.61), APRIL (8.4±5.39 vs. 7.54±10.7 pg/mL; p =0.21), and IFNɑ (25.06±8.41 vs. 21.72±6.22 pg/mL; p =0.3) between responders and non-responders at month 1.
We tested all used IRT products to exclude the presence of anti-RBD SARS-CoV-2 specific antibodies leading to a false interpretation of vaccination response. Specific-antibody levels did not exceed the positive cutoff value (>18 U/mL) in any of the products (Kiovig: 7.38±0.4; HyQvia: 6.22±1.1; and Hizentra: 7.95±0.66 U/mL).