Humoral and cellular responses to SARS-CoV-2 in patients with B-cell haematological malignancies improve with successive vaccination

Abstract

Patients with haematological malignancies are more likely to have poor responses to vaccination. Here we provide detailed analysis of the humoral and cellular responses to COVID-19 vaccination in 69 patients with B-cell malignancies. Measurement of anti-spike IgG in serum demonstrated a low seroconversion rate with 27.1% and 46.8% of patients seroconverting after the first and second doses of vaccine, respectively. In vitro pseudoneutralisation assays demonstrated a poor neutralising response, with 12.5% and 29.5% of patients producing a measurable neutralising titre after the first and second doses, respectively. A third dose increased seropositivity to 54.3% and neutralisation to 51.5%, while a fourth dose further increased both seropositivity and neutralisation to 87.9%. Neutralisation titres post-fourth dose showed a positive correlation with the size of the B-cell population measured by flow cytometry, suggesting an improved response correlating with recovery of the B-cell compartment after B-cell depletion treatments. In contrast, interferon gamma ELISpot analysis showed a largely intact T-cell response, with the percentage of patients producing a measurable response boosted by the second dose to 75.5%. This response was maintained thereafter, with only a small increase following the third and fourth doses, irrespective of the serological response at these timepoints.

Keywords: B cells; T cells; antibodies; haematological malignancies; infection; vaccines.

FIGURE 1

Seroconversion rates and neutralisation titres low after three vaccine doses, but largely rescued by a fourth dose. (A) Percentage of participants at the timepoints indicated above each graph, divided into seronegative, non‐neutralising and neutralising, determined for the wild‐type virus and for the variants indicated. From left to right, data are shown for pre‐second dose (n = 48), 1 month post‐second dose (n = 61), 6 months post‐second dose/pre‐third dose (n = 39), 1 month post‐third dose (n = 33) and 1 month post‐fourth dose (n = 33). Samples with a positive anti‐spike titre but no available sample to test for neutralisation were excluded from this analysis. (B–E) Neutralisation titres for each viral strain, measured over time. Reciprocal ID₅₀ were calculated using data from duplicate serial dilutions where at least two data points exhibited >50% neutralisation. Samples which gave rise to only one data point >50% neutralisation are under the limit of quantification (LOQ) as curve‐fitting cannot be performed and were therefore assigned a value of 1:20 as this is the first dilution in the serial titration. Samples which showed neutralisation <50% at a dilution of 1:20 are below the limit of detection (LOD) and assigned a value of 1:5 to distinguish them graphically from neutralising samples below the LOQ. Seronegative (SN) samples were assigned a value of 1 for the purposes of the logarithmic scale. Patients with either a detectable anti‐N titre prior to the start of the study or with a recorded SARS‐CoV‐2 infection are indicated at all subsequent timepoints with black triangles

FIGURE 2

Intact T‐cell responses to vaccination, irrespective of seroconversion status. (A) Interferon gamma (IFNγ) responses to spike protein‐derived peptide pools, normalised to spot‐forming units (SFU) per 10⁶ peripheral blood mononuclear cells (PBMCs). Negative values were given a value of 1 SFU for the purposes of the logarithmic scale. ELISpot assays were run for all samples where viable PBMCs were available (n = 41, 53, 36, 30, 31, at each timepoint respectively). (B) IFNγ responses to the omicron mutation peptide pool pre‐ and post‐third dose, and post‐fourth dose (n = 35, 30, 31, respectively). (C–G) IFNγ responses to WT spike protein peptides for each timepoint, divided into seropositive and seronegative patients. ELISpot results without matched serological data were excluded from this analysis. Dashed lines indicate the threshold for a positive response, calculated as three standard deviations above the mean of all unstimulated wells. Significance is shown on each graph as calculated using the Mann–Whitney U test, with values of p = 0.0133 and p = 0.0297 for the Pre‐ and Post‐2nd dose timepoints respectively.

FIGURE 3

Correlations observed between serological data and flow cytometry analysis of B, NK and T‐cell subpopulations. (A) Summary table listing the correlations between each flow parameter and the Roche anti‐spike Ig, anti‐spike IgG ELISA and WT neutralisation titre at the post‐fourth dose timepoint. Significant (p < 0.01) correlations are listed with the Spearman's rho value (r _s), the p value and the number of samples that were compared. Patients with CLL were excluded from comparisons of serological data with total CD3, CD19 and CD56 proportions, since expanded monoclonal CLL B cells would artificially increase the CD19 population, and thus decrease the CD3 and CD56 populations. Non‐significant correlations are listed as ‘ns’. (B–G) Representative examples of Spearman correlations, with Spearman's rho and p values shown. Seronegative samples are not shown on the graph due to their ‘zero’ value on the log scale for the neutralisation titres. Correlations shown are between the post‐fourth dose neutralisation titre and the flow cytometry values for proportions of (B) total B cells, (C) total T cells, (D) total NK cells, (E) T_FH cells as a proportion of total CD4 T cells, (F) naïve CD8 cells as a proportion of total CD8 T cells and (G) T_REG cells as a proportion of total non‐T_FH CD4 T cells.