Antibody response to mRNA SARS-CoV-2 vaccine among dialysis patients—a prospective cohort study

Emerging data has established end-stage renal disease (ESRD) as one of the highest risk comorbidities for severe coronavirus disease 2019 (COVID-19), with short-term mortality above 20% [1, 2]. Based on both a large clinical trial and real-life data from Israel in the general populations, the effectiveness of mRNA-based BNT162b2 vaccine against severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) was estimated at 95% [3, 4]. Nevertheless, dialysis patients present diminished immune response following immunization with various vaccines, and the efficacy of the available SARS-CoV-2 vaccine among this vulnerable group is unknown [5]. Therefore, we performed a prospective study, evaluating antibody response among hemodialysis and peritoneal dialysis (PD) patients 2–6 weeks after receiving the second dose of BNT162b2. The study included hemodialysis and PD patients who were vaccinated with two doses of BNT162b2 vaccine, 21 days apart. Patients treated with current chemotherapy or immunosuppressive drugs, those who had only one vaccine dose, those who had been infected by COVID-19 before or those who were unable to give consent were excluded. To confirm sustained response, consenting participants were evaluated 2–6 weeks after receiving the second vaccine dose and were followed for up to 8 weeks. The study was approved by the ethics committee of Rabin Medical Center. Demographic data were collected by interview and medical records and blood samples for anti-spike (anti-S) SARS-CoV-2 antibodies were collected during the routine dialysis visits. SARS-CoV-2 immunoglobulin G (IgG) II Quant (Abbott©) assay was used for quantitative measurement of anti-S IgG antibodies of SARS-CoV-2. A test was considered as positive if IgG was above 50 AU/mL [6]. The primary outcome was the rate of seropositivity for anti-S antibodies. Univariate and multivariate linear regression analyses were performed to explore factors associated with higher log-transformed antibody titer. The results are presented as change of log-transformed antibody level per unit of explanatory variable (B). Overall, 122 hemodialysis patients were included in the study (Supplementary data, Figure S1). Patients’ characteristics are detailed in Table 1. At a median time of 36 days [interquartile range (IQR) 32–40, range 10–48 days] from the second immunization dose, 114 (93.4%) of the 122 hemodialysis patients were seropositive for anti-S IgG. The geometric mean of anti-S titer was 1190.8 AU/mL, median level was 1599 AU/mL (IQR 419.5–3976.9 AU/mL) and the mean log-transformed anti-S level was 3.08 ± 0.84 logAU/mL. Baseline characteristics of the hemodialysis cohort according to antibody response Response—above 50 AU/mL. Including: urinary tract obstruction (3), CACUT (3), cardio-renal syndrome (3), myeloma kidney (1), cholesterol emboli (1), obesity-related FSGS (2) and unknown (8). ADPKD, autosomal dominant polycystic kidney disease; nPCR, normalized protein catabolic rate; ESA, erythropoietin stimulating agents; CACUT, congenital anomaly of kidney and urinary tract; FSGS, focal segmental glomerulosclerosis. Data are presented as n (%), mean ± SD or median (IQR). Baseline characteristics of the hemodialysis cohort according to antibody response Response—above 50 AU/mL. Including: urinary tract obstruction (3), CACUT (3), cardio-renal syndrome (3), myeloma kidney (1), cholesterol emboli (1), obesity-related FSGS (2) and unknown (8). ADPKD, autosomal dominant polycystic kidney disease; nPCR, normalized protein catabolic rate; ESA, erythropoietin stimulating agents; CACUT, congenital anomaly of kidney and urinary tract; FSGS, focal segmental glomerulosclerosis. Data are presented as n (%), mean ± SD or median (IQR). Factors associated with non-response (≥50 AU/mL) were lower serum albumin (3.41 ± 0.56 versus 3.99 ± 0.35, P < 0.001) and higher intravenous iron sucrose dose [median dose (IQR) 150 mg/week (62.5–300) versus 50 mg/week (0–100), P = 0.003] in non-responders and responders, respectively. Younger age (B 0.021 per year decrease, 95% CI 0.011–0.031, P < 0.001), serum albumin above 3.5 g/dL (B 1.039, 95% CI 0.65–1.429, P < 0.001), lower intravenous iron dose (B 0.002 per mg/week decrease, 95% CI 0.00–0.004, P = 0.009) and body mass index (BMI) <30 kg/m2 (B 0.394, 95% CI 0.107–0.681, P = 0.008) were associated with higher log-transformed antibody titer in a multivariate analysis (Supplementary data, Table S1). Of the 23 PD patients, 22 (95.6%) were seropositive for anti-S IgG. The geometric mean of anti-S titer was 1515.14 AU/mL, median level was 1560 AU/mL (IQR 254.8–6423 AU/mL) and the mean log-transformed antibody level was 3.18 ± 0.83 logAU/mL. There was no significant difference between the mean antibody titer of PD and hemodialysis patients (P = 0.58). Younger age (B 0.034 per year decrease, 95% CI 0.009–0.06, P = 0.011) and serum albumin above 3.5 g/dL (B 1.147, 95% CI 0.536–1.758, P < 0.001) were associated with higher log-transformed antibody titer in a univariate analysis. Due to the small PD cohort, multivariate analysis was not performed. Only two major adverse events following any vaccine dose were reported among 145 cohort participants. Major adverse events included: one syncope event, 1 day following the first vaccination dose (hemodialysis patient), and one pericarditis event 2 days following second vaccination dose (PD patient). Seropositivity rates following BNT162b2 vaccination in both hemodialysis and PD patients were high and comparable to the seropositivity rates reported in healthy volunteers [7]. Overall, 114 hemodialysis patients (93.4%) and 22 PD patients (95.6%) were seropositive for SARS-CoV-2 anti-S IgG at 2–6 weeks following the second dose of BNT162b2 vaccination. Recently, similar high seropositivity rates were also reported by Grupper et al. [8] among 56 hemodialysis patients as well as in a control group of 95 health workers. Factors associated with higher log-transformed S1-binding antibodies titers in hemodialysis patients included younger age, BMI <30 kg/m2, serum albumin above 3.5 g/dL and reduced intravenous iron dose. In line with our findings, Grupper et al. [8] also demonstrated a significant inverse correlation between older age and antibodies levels in the hemodialysis group as well as in the control group. Nutritional status is a known risk factor for dampened antibody response to immunization in dialysis patients [5, 9]. Hypoalbuminemia as a marker of malnutrition and inflammatory state has been also shown to correlate with diminished seroconversion in response to hepatitis B (HBV) vaccine [10, 11]. However, higher BMI, which is often associated with improved nutritional state and outcome in dialysis patients, was previously reported to correlate with inferior response to HBV vaccine [10, 12]. Both iron deficiency and iron overload are associated with impaired immune reactivity. Intravenous iron supplementation was found to be associated with reduced antibody response in our study, as well as in a previous study which found that intravenous iron blunted the humoral response to HBV vaccination [13]. Our study has several limitations. First, correlation between antibody response to vaccine and protection against SARS-CoV-2 infection has not yet been proven. Nevertheless, evidence is accumulating to support antibody response as a potential correlate of disease protection. Furthermore, we did not include neutralization antibodies and cellular immunity assays in our study. However, a strong correlation has been reported between anti-S antibody titers and neutralization antibody levels [7, 14, 15]. In conclusion, in our cohort, high seropositivity rates above 90% were demonstrated among 122 hemodialysis and 23 PD patients following BNT162b2 vaccination, with few serious adverse events. Younger age, BMI <30 kg/m2, normal albumin level and reduced intravenous iron dose were associated with elevated anti-S antibody titers in hemodialysis patients. These findings strongly support the efficacy of the BNT162b2 vaccination among dialysis patients. Supplementary data are available at ndt online. B.R.-Z. reports a consulting fee from Fresenius Medicare and a payment for a lecture from AstraZeneca plc. B.Z. reports a consulting fee from Fresenius Medicare. All remaining authors have nothing to disclose.