In this trial 16 adult immunocompromised patients and 16 healthy control (HCs), who had been previously vaccinated twice with a SARS-CoV‑2 vaccine and displayed SARS-CoV‑2 RBD antibody levels below 1500 BAU/ml were enrolled. Among the patient cohort, 11 patients were diagnosed according to the recent IUIS (International Union of Immunological Societies) classification [3] and 5 patients suffered from severe mannose-binding lectin (MBL) deficiency defined by MBL levels below 30 ng/ml, as described in Supplementary Table 1. Of the patients three preferred to be boosted with mRNA-1273, all other patients were immunized using BNT162b2 (Supplementary Table 1). The average patient age was 44 years (± 12 years), with a male-to-female ratio of 0.33. The sex and age-matched HC cohort included 16 adults with an average age of 44 years (± 11 years). All HCs received a third vaccination with BNT162b2. Patient and HC characteristics are summarized in Table 1.

Anti-SARS-CoV‑2 RBD antibodies were measured at baseline (week 0) and after 4 weeks. The primary endpoint was defined as achieving antibody levels > 1500 BAU/ml. The percentage of individuals reaching this primary endpoint was lower in the patient cohort compared to HCs (IEI/MBL deficiency: 12, 75% vs. HC: 16, 100%; p = 0.1012). Specifically, one patient with XLA (X-linked agammaglobulinemia) and three patients with CVID (common variable immunodeficincy) did not meet the primary endpoint. At week 0 there was no significant difference in antibody levels against the SARS-CoV‑2 RBD of the spike protein between the healthy controls and the patient cohort (HC median: 455 BAU/ml, IQR 253.8–702 BAU/ml vs. IEI/MBL deficiency median: 375.5 BAU/ml, IQR 70.93–850 BAU/ml; p = 0.632). Of note, the interval between the second and third vaccination was significantly shorter in the patient cohort compared to the control group (p < 0.0001; Table 1). After vaccination, a significant increase in the production of anti-SARS-CoV‑2 RBD antibodies was detected for HC between week 0 and week 4 (week 0 median: 455 BAU/ml, IQR (303–678)) vs. week 4 median: 18,185 BAU/ml, IQR 12,853–25,000 BAU/ml; p = 1.5−6; Fig. 1a) as well as for the patient cohort (week 0: median: 375 BAU/ml, IQR 83–832 BAU/ml vs. week 4: median: 6390 BAU/ml, IQR 1481–14,275 BAU/ml, p = 0.00056; Fig. 1a), indicating a robust response to vaccination in both cohorts; however, the median antibody levels at week 4 were significantly lower in the IEI/MBLdef cohort (median: 6390 BAU/ml, IQR 1481–14,275) when compared to HC (median: 18,185 BAU/ml, IQR 12,853–25,000; p = 0.013; Fig. 1a). Following this, the absolute change of anti-SARS-CoV‑2 RBD antibodies (∆) between week 0 and week 4 was significantly lower in the IEI/MBLdef cohort (median: 6190 BAU/ml, IQR 1393–13,361) than in the HC cohort (median: 17,903 BAU/ml, IQR 12,385–23,739, p = 0.012; Fig. 1b). Additionally, we observed a lower, albeit not statistically significant, median fold change (FC) in anti-SARS-CoV‑2 RBD antibody levels in the patient cohort (median FC = 14.31, IQR 9.26–50.84) as compared to the healthy controls (median FC = 38.92, IQR 24.92–59.18, p = 0.102, Fig. 1c). Interestingly, the patient with X‑linked agammaglobulinemia (XLA) exhibited the highest fold change in antibody levels, reaching 15.7 BAU/ml after the third vaccination, which might indicate a minimal antibody production after repetitive vaccination. Furthermore, levels of anti-SARS-CoV‑2 RBD antibodies significantly correlated with titers of neutralizing antibodies against SARS-CoV‑2 (NT) at week 4 in the IEI/MBLdef cohort (R = 0.98, p = 3.4−11) as well as in the HC group (R = 0.96 p = 5.8−9), Fig. 2a. Both cohorts developed a robust increase in NT titers 4 weeks after vaccination when compared to baseline (HC: week 0 median: 15, IQR 10–16.25 vs. week 4 median: 480, IQR 160–960, p = 1.6−6; IEI/MBLdef: week 0: median: 12.5, IQR 0–20 vs. week 4 median: 100, IQR26.25–360, p = 0.0039; Fig. 2b). Levels of neutralizing antibodies against SARS-CoV‑2 were significantly lower in the patient group (median NT level: 100, IQR 26.25–360) when compared to HC at week 4 after vaccination (median NT level: 480, IQR 160–960; p = 0.023; Fig. 2b). The absolute change in the production of neutralizing antibodies between week 0 and week 4 was significantly decreased in the patient cohort (median absolute change (∆) of NT level: 90, IQR 18.75–340) in comparison to the control group (median absolute change (∆) of NT level: 472.5, IQR 148.75–940, p = 0.019; Fig. 2c). Overall, the humoral immune response to a third COVID-19 vaccination was markedly diminished in the patient cohort when compared to healthy controls.

Humoral immune response at week 4 after the third vaccination. a Antibody levels to the receptor-binding domain (RBD) of the viral spike (S) protein in IEI/MBLdef patients (n = 16) and healthy controls (HCs, n = 16) at screening (week 0) and at week 4 after the third vaccination were determined using an anti-SARS-CoV‑2 immunoassay. b Absolute change (∆) of anti-RDB antibody levels between week 0 and week 4 after the third vaccination. c Fold change (FC) of anti-RBD antibody levels between week 0 and 4 in patients and HCs. XLA X-linked agammaglobulinemia, WHIM warts, hypogammaglobulinemia, immunodeficiency, myelokathexis, MBL mannose-binding lectin, CVID common variable immunodeficiency, IFNγ interferon gamma

Neutralizing antibody response at week 4 after the 3rd vaccination. a Correlation of antibody levels to the receptor-binding domain (RBD) of the viral spike (S) protein and neutralizing antibodies (NT) against a SARS-CoV‑2 WT virus strain at week 4 after the 3rd vaccination of healthy controls (HCs, n = 16) and IEI/MBLdef patients (n = 16). b Neutralizing antibodies against the SARS-CoV‑2 at week 0 and week 4 after the 3rd vaccination in patients and healthy controls (HCs). c Absolute change (∆) of neutralizing antibodies against SARS-CoV‑2 between week 0 and week 4 after the third vaccination of HCs and patients. XLA X-linked agammaglobulinemia, WHIM warts, hypogammaglobulinemia, immunodeficiency, myelokathexis, MBL mannose-binding lectin, CVID common variable immunodeficiency, IFNγ interferon gamma

Before vaccination (week 0) and 1 week afterward, we conducted an ELISpot assay to evaluate the SARS-CoV-2-specific T‑cell response in 11 HC and 8 IEI/MBLdef patients. We confirmed a slight increase in SARS-CoV-2-specific spot-forming cells (SFCs) in the HC cohort between week 0 and week 1 after stimulation with omicron peptides (HC: median week 0: 34.5, IQR 19.75–155.0 vs. median week 1: 80.5, IQR 56.75–290.50 per 106 SFCs, p = 0.14). Analysis revealed a slightly greater increase for IEI/MBLdef patients (IEI/MBLdef: median week 0: 28.25, IQR 1.125–85.25 vs. median week 1: 154.5, IQR 79.125–235.86, p = 0.065). In parallel, we restimulated the equivalent samples of patients with WT peptides and observed an increasing tendency of SARS-CoV-2-specific T‑cell response (IEI/MBLdef: median week 0: 57.0, IQR 21.0–77.38.0 vs. median week 1:164, IQR 49.13–245.50 per 106 SFCs, p = 0.052, Fig. 3a), however, none of these results achieved statistical significance.

SARS-CoV-2-specific T‑cell response. SARS-CoV-2-specific T‑cell responses were determined by IFN‑γ enzyme-linked immunosorbent spot (ELISpot) assay from peripheral blood mononuclear cells (PBMC) stimulated with omicron and with wild type peptide pools (WT) before (week 0) and 1 week after the third vaccination. ELISpot results from 11 HCs and 8 patients before and 1 week after the vaccination stimulated with omicron peptides (left panel). Dots show the sum of total responses from S1 and S2 peptide pools (covering the S1 and S2 domain of the SARS-CoV-2 spike protein). ELISpot results from 8 patients before and 1 week after the vaccination stimulated with WT peptides (right panel). XLA X-linked agammaglobulinemia, WHIM warts, hypogammaglobulinemia, immunodeficiency, myelokathexis, MBL mannose-binding lectin, CVID common variable immunodeficiency, IFNγ interferon gamma

Adverse events following vaccination were systematically assessed using a paper-based diary, coupled with daily body temperature recordings. Symptoms were graded on a scale from 0 to 3, indicating absence (0) to high severity (3). All healthy controls and 15 out of 16 IEI/MBLdef patients provided diary entries. Notably, 14 out of 15 patients (93.3%) and 15 out of 16 HCs (93.8%) reported experiencing at least 1 local or systemic symptom during this monitoring period. The predominant symptom over 7 days was localized pain at the injection site (HC: 87.5%; patients: 80%), followed by fatigue (HC: 62.5%; patients: 73.3%) and headache (HC: 56.3%; patients: 53.3%). Remarkably, IEI/MBLdef patients exhibited systemic symptoms, such as headache, muscle pain and fatigue persisting for up to 7 days, whereas HCs reported shorter durations of these symptoms (Fig. 4a). Moreover, patients demonstrated a heightened severity of fatigue within the first 7 days after vaccination when compared to healthy controls, as depicted in Fig. 4b. It is noteworthy that nausea was exclusively reported in the IEI/MBLdef cohort, with 5 out of 15 patients experiencing this symptom (Fig. 4a). Conversely, no discernible disparity in body temperature was observed between patients and healthy controls, as indicated in Supplementary Fig. 1a. A comprehensive summary detailing all assessed symptoms and the frequency of occurrence is provided in Supplementary Table 2. Importantly, no occurrences of serious adverse events were documented throughout the 4‑week observation period. Before and at week 4 post-vaccination, we examined the lasting impacts on disease activity and fatigue using a PGA with a numeric rating scale from 0 to 10. Among the patients, four individuals reported an escalation in disease activity (patients 1, 4, 9 and 15). Notably, as depicted in Fig. 4c, four patients noted an amelioration in the underlying condition: patients 6, 7, 11 and 13. Additionally, 4 out of 16 patients (patients 4, 11, 13 and 15) experienced heightened fatigue at week 4 post-vaccination compared to baseline. Conversely, six patients reported a decrease in fatigue (patients 1, 5, 7, 9, 14 and 16), as shown in Fig. 4d. These findings suggest minimal to negligible prolonged effects of vaccination on the progression of underlying disease and fatigue in patients suffering from IEIs or MBL deficiency.

Reactogenicity after the third vaccination. Manifestations at the site of injection and systemic symptoms were recoded over a period of 7 days after vaccination. a Relative occurrence of the indicated symptom in IEI/MBLdef patients (n = 15) and healthy controls (n = 16). (B) Mean symptom intensity. Symptom severity was assessed on a numeric rating scale between 0 and 3. c and d Heatmap of the dynamics of the activity of the underlying disease (c) and fatigue (d) evaluated by using a numeric rating scale ranging from 0 to 10 before the 3rd vaccination and at week 4. Vertical patient numbers refer to Supplementary table 1 with patient number: 1–6 CVID (common variable immunodeficiency), 7 XLA (X-linked agammaglobulinemia), 8 WHIM (warts, hypogammaglobulinemia, immunodeficiency, myelokathexis), 9 IFNGR1 (interferon gamma receptor 1) mutation, 10 MWS (muckle-wells-syndrome), 11 CAPS (cryopyrin-associated periodic syndrome), 12–16 MBLdef (mannose-binding lectin defficiency)