Specific antibody response of patients with common variable immunodeficiency to BNT162b2 coronavirus disease 2019 vaccination

On March 11, 2020, the World Health Organization declared that the coronavirus disease 2019 (COVID-19) was a pandemic.¹ Since then, the disease has reached a 1% to 3% estimated overall mortality rate.² COVID-19 severity ranges from asymptomatic to acute respiratory distress syndrome and possible death owing to multiorgan failure.² Therefore, to ameliorate the resultant poor health and social and economic consequences, prophylactic vaccines were developed. On December 11, 2020, the US Food and Drug Administration issued the first emergency use authorization of Pfizer-BioNTech (Pfizer Inc, New York City, New York) messenger RNA (mRNA) vaccine (BNT162b2) for COVID-19 prevention.¹ The vaccine was approved after a large randomized, placebo-controlled trial in approximately 44,000 participants aged 16 years or older and revealed that a 2-dose regimen of BNT162b2 conferred 95% protection against symptomatic COVID-19.¹ This novel lipid nanoparticle-formulated nucleoside-modified RNA vaccine encodes the full-length spike protein of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), which contains the receptor binding domain (RBD) within the S1 subunit.³ The RBD is a key functional component within the S1 subunit responsible for binding SARS-CoV-2 to angiotensin-converting enzyme 2 receptor, a critical initial step enabling SARS-CoV-2 to penetrate target cells.⁴

Among healthy adults, two 30 µg doses of BNT162b2 elicited robust antigen-specific CD8+ and T_H1-type CD4+ T-cell responses and strong specific antibody responses directed against RBD.⁵ Nevertheless, it is unknown whether patients having primary immunodeficiency disorders of humoral immunity affecting B-cell differentiation and antibody production are able to produce effective specific antibody levels after the 2-dose BNT162b2 regimen. Common variable immunodeficiency (CVID) is an antibody deficiency with variable clinical manifestations; although patients mostly experience recurrent infections, there is an increased prevalence of autoimmune diseases and malignancy secondary to immune dysregulation.⁶ A CVID diagnosis established after the fourth year of life requires a suggestive clinical history, a marked reduced total immunoglobulin G (IgG) serum concentration with low IgA or IgM, poor responses to vaccines (or absent isohemagglutinins), or low IgD⁻/CD27⁺/CD19⁺ switched memory B (smB) cells, and no evidence of profound T-cell deficiency; in addition, other causes of secondary hypogammaglobulinemia must be excluded.⁶

We observed retrospectively the ability of patients with CVID to produce SARS-CoV-2 spike-specific IgG in response to the 2-dose BNT162b2 regimen as part of the national vaccination program of Israel. Furthermore, we looked for a correlation with CVID subgroups based on flow cytometry B-cell immunophenotyping.⁷ All patients diagnosed as having CVID (n = 17) were treated with intravenous immunoglobulin (IVIG) every 4 weeks at Lin, Zvulun, and Carmel Medical Centers belonging to Clalit Health Services in Haifa, Israel. Revised European Society for Immunodeficiencies registry criteria⁶ were used for CVID diagnosis. Between December 23, 2020, and March 6, 2021, all patients with CVID were vaccinated with the 2-dose BNT162b2 regimen. Blood samples were taken at least 14 days after the second dose, before receiving IVIG to measure SARS-CoV-2 S1 IgG levels and obtain and updated flow cytometry analysis. Day 14 was chosen because mRNA vaccine-induced B-cell responses typically peak 2 weeks after the second dose and SARS-COV-2 neutralizing titers seem to follow this pattern.⁵ SARS-CoV-2 S1 IgG values more than 50 AU/mL were considered protective by the Abbott Architect SARS-CoV-2 S1 IgG assay (manufacturer's data: sensitivity, 98.1% [95% confidence interval, 89.9%-99.7%]; specificity, 99.6% [95% confidence interval, 99.2%-99.8%]) performed by the serology laboratory of Clalit Health Services. There were 2 patients who were excluded: COVID-19 was detected on prevaccination polymerase chain reaction testing in one patient, whereas the second was receiving ongoing immunosuppressive medication (rituximab). The remaining 15 patients were divided into the following 3 groups, based on their results: group B−, total circulating CD19⁺ B cells less than or equal to 1%; group B+/smB+, total circulating CD19⁺ B cells greater than 1% and smB cells greater than 2%; and group B+/smB−, total circulating CD19⁺ B cells greater than 1% and smB cells less than or equal to 2%.⁷

Table 1 provides the cohort characteristics and their serologic results. Patients ranged from the age of 22 to 81 years (average, 49.8 years). Blood serology samples were taken 14 to 61 days after the second dose (average, day 31). In addition, 4 patients (26.67%) did not produce SARS-CoV-2 S1 IgG after both BNT162b2 doses, whereas 11 (73.33%) had protective titers ranging from 58 AU/mL to 9780.3 AU/mL (average, 1764.00; median, 307.3). Note that although the 2 patients in group B− had negative serology result, all 6 patients in group B+/smB+ had seropositive result. For group B+/smB−, 5 of 7 patients were seropositive. Interestingly, the 2 patients with negative serology had a total peripheral CD19⁺ B-cell percentage below the lower limit for the normal range (6%-19%),⁷ whereas that of the 5 seropositive patients was within the normal range. It has been found that patients with CVID with nearly absent total CD19⁺ B cells (≤1%) have severe defects of early B-cell differentiation, whereas severely reduced smB cells (≤2%) indicate defective germinal center (GC) development.⁷ Our results suggest that patients with both CD19⁺ B% cells lower than the normal range (6%-19%) and reduced smB cells (≤2%) have prominent GC generation impairment. In line with this idea, the GC has been found to play a pivotal role on protective antibody generation for SARS-CoV-2 mRNA vaccines and that GC responses are strongly correlated with neutralizing antibody production.⁸

Table 1.

Cohort Characteristics^a (n = 15) and Serologic Results

Group	Sex	Age (y)	Flow cytometry results	Second vaccination/serology interval (d)	SARS-CoV-2 S1 IgG (AU/mL)
B−	Mb	51	B% = 1	29	<21
	F	30	B% = 0	14	<21
B+/smB+	M	50	B% = 4, smB% = 9	14	307.3
	M	72	B% = 3, smB% = 14	34	300.4
	M	22	B% = 9, smB% = 3	18	4924.9
	M	81	B% = 2, smB% = 10	41	58
	F	28	B% = 9, smB% = 11	15	9780.3
	M	61	B% = 11, smB% = 7	28	2178.3
B+/smB−	Fc	44	B% = 8, smB% = 0	61	205.7
	F	62	B% = 20, smB% = 0	36	84.6
	F	48	B% = 8, smB% = 2	55	625.8
	F	40	B% = 17, smB% = 0	18	109.9
	M	54	B% = 6, smB% = 2	48	828.7
	F	38	B% = 4, smB% = 1	25	<21
	F	66	B% = 5, smB% = 0	30	<21

Abbreviations: B%, percentage of total circulating CD19⁺ B cells as fraction of lymphocytes; smB%, percentage of IgD⁻/CD27⁺/CD19⁺ switch memory B cells as fraction of total circulating CD19⁺ B cells; F, female; IgG, immunoglobulin G; M, male; SARS-CoV-2, severe acute respiratory syndrome coronavirus 2.

^aPhenotypes are not listed because no correlation was found between a specific phenotype and vaccination response; there are no comorbidities that might have affected patients’ BNT162b2 response.

^bReceiving 10 mg prednisone for inflammatory bowel disease.

^cReceiving 40 mg prednisone for autoimmune hemolytic anemia.

A possible study limitation may be that patients acquired protective antibodies from the IVIG. Nevertheless, all patients with CVID were on PRIVIGEN (CSL Behring, Bern, Switzerland; manufacture date: January 14, 2020). Hence, IVIG-stimulated cross-reactive antibodies cannot explain the wide differences between protective antibody levels after vaccination. The presence of considerable protective COVID-19 antibody levels in these products is doubtful. In addition, although 2 patients were receiving steroids, their serologic results (Table 1) indicate that steroid use was not responsible for the lack of response to BNT162b2 vaccination.

In conclusion, vaccination of patients with CVID with the 2-dose BNT162b2 regimen is important, because most of them will produce specific SARS-CoV-2 S1 antibodies in good titers. Nevertheless, our data indicate that total peripheral CD19⁺ B cells below the normal range (6%-19%) together with smB cells (≤2%) or total peripheral CD19⁺ B cells (≤1%) may predict unresponsiveness to BNT162b2. Our data require further validation in larger populations with CVID and subsequent research to detect the rate of postvaccination antibody decay compared with that of the general population.