Humoral Responses Against SARS-CoV-2 and Variants of Concern After mRNA Vaccines in Patients With Non-Hodgkin Lymphoma and Chronic Lymphocytic Leukemia

Abstract

Purpose: Patients with non-Hodgkin lymphoma including chronic lymphocytic leukemia (NHL/CLL) are at higher risk of severe SARS-CoV-2 infection. We investigated vaccine-induced antibody responses in patients with NHL/CLL against the original SARS-CoV-2 strain and variants of concern including B.1.167.2 (Delta) and B.1.1.529 (Omicron).

Materials and methods: Blood from 121 patients with NHL/CLL receiving two doses of vaccine were collected longitudinally. Antibody binding against the full-length spike protein, the receptor-binding, and N-terminal domains of the original strain and of variants was measured using a multiplex assay. Live-virus neutralization against Delta, Omicron, and the early WA1/2020 strains was measured using a focus reduction neutralization test. B cells were measured by flow cytometry. Correlation between vaccine response and clinical factors was determined.

Results: Mean anti-SARS-CoV-2 spike immunoglobulin G-binding titers were 85-fold lower in patients with NHL/CLL compared with healthy controls, with seroconversion occurring in only 67% of patients. Neutralization titers were also lower and correlated with binding titers (P < .0001). Treatment with anti-CD20-directed therapies within 1 year resulted in 136-fold lower binding titers. Peripheral blood B-cell count also correlated with vaccine response. At 3 months from last anti-CD20-directed therapy, B-cell count ≥ 4.31/μL blood around the time of vaccination predicted response (OR 7.46, P = .04). Antibody responses also correlated with age. Importantly, neutralization titers against Delta and Omicron were reduced six- and 42-fold, respectively, with 67% of patients seropositive for WA1/2020 exhibiting seronegativity for Omicron.

Conclusion: Antibody binding and live-virus neutralization against SARS-CoV-2 and its variants of concern including Delta and Omicron were substantially lower in patients with NHL/CLL compared with healthy vaccinees. Anti-CD20-directed therapy < 1 year before vaccination and number of circulating B cells strongly predict vaccine response.

FIG 1.

Decreased humoral immune response against SARS-CoV-2 vaccine in patients with NHL/CLL. Maximum IgG-binding titers against (A) SARS-CoV-2 spike, (B) RBD, and (C) NTD after two vaccine doses in patients with NHL/CLL. Maximum IgA-binding titers against (D) SARS-CoV-2 spike, (E) RBD, and (F) NTD. Maximum IgM-binding titers against (G) SARS-CoV-2 spike, (H) RBD, and (I) NTD. For all graphs, samples from healthy vaccinees and prepandemic patients with NHL/CLL are included as controls. Nucleocapsid-positive samples (N+ NHL, purple) reflect patients who had prior SARS-CoV-2 infection, whereas nucleocapsid-negative samples (N– NHL, blue) reflect previously uninfected patients. Horizontal line = background antibody levels determined from prepandemic samples. Error bars = mean ± SEM. **P < .01, ****P < .0001 by using Kruskal-Wallis test and adjusting for multiple comparisons using a Dunn's test. CLL, chronic lymphocytic leukemia; Ig, immunoglobulin; NHL, non-Hodgkin lymphoma; NTD, N-terminal domain; RBD, receptor-binding domain.

FIG 2.

Kinetics of serologic response after mRNA vaccination. (A) Anti-SARS-CoV-2 spike IgG-binding titers among nucleocapsid-negative patients (blue) on the basis of days after vaccination compared with healthy vaccinees (red). *P < .05, ****P < .0001 by using Kruskal-Wallis test. Error bars = mean ± SEM. After dose 1 = samples obtained at least 10 days after initial vaccination and before second vaccine. (B) Anti-SARS-CoV-2 spike IgG-binding titers in nucleocapsid-negative patients with NHL (blue) over time after receipt of the initial vaccine dose versus healthy (red). The slopes of the two lines (0.02670 for lymphoma and 0.02372 for healthy vaccinees) are statistically significantly different (P = .04). (C) Anti-SARS-CoV-2 spike IgG-binding titers over time in nucleocapsid-negative patients with available longitudinal data. Prepandemic NHL patient samples (pink) are included as control. For applicable graphs, horizontal line = background IgG levels determined from prepandemic samples. CLL, chronic lymphocytic leukemia; IgG, immunoglobulin G; NHL, non-Hodgkin lymphoma.

FIG 3.

Decreased live-virus neutralization antibodies after mRNA vaccination in patients with NHL. (A) Plasma neutralizing antibody titers against WA1/2020 strain in nucleocapsid-negative patients with NHL compared with healthy vaccinees. ****P < .0001 by Mann-Whitney test. Plasma neutralizing antibody titers correlated with higher (B) anti-SARS-CoV-2 spike and (C) RBD IgG-binding titers. Correlations were statistically significant using extra-sum-of-squares F test (P < .001). CLL, chronic lymphocytic leukemia; NHL, non-Hodgkin lymphoma; RBD, receptor-binding domain.

FIG 4.

Clinical correlates to the immune response against the SARS-CoV-2 vaccine. Maximum IgG-binding titers against SARS-CoV-2 spike protein on the basis of (A) age, (B) sex, and (C) race using Mann-Whitney test. (D) Maximum anti-SARS-CoV-2 spike IgG-binding titers by underlying NHL diagnosis. Statistical comparisons shown are versus healthy vaccinees using Brown-Forsythe and Welch ANOVA test and adjusting for multiple comparisons using a Dunnett T3 test. (E) Maximum anti-SARS-CoV-2 IgG-binding titers on the basis of receipt of prior lymphoma-directed therapy as compared using Kruskal-Wallis test. For all graphs, horizontal line = background IgG levels determined from prepandemic samples. *P < .05, **P < .01, ****P < .0001. Error bars = mean ± SEM. ANOVA, analysis of variance; DLBCL, diffuse large B-cell lymphoma; FL, follicular lymphoma; IgG, immunoglobulin G; MCL, mantle cell lymphoma; NHL, non-Hodgkin lymphoma; ns, not significant; SLL, small lymphocytic lymphoma.

FIG 5.

Correlates of lymphoma-directed therapy and B-cell counts with vaccine response. (A) Significant decrease in maximum anti-SARS-CoV-2 IgG-binding titers is observed in patients who received anti-CD20 antibody therapy (eg, rituximab and obinutuzumab) within 1 year of vaccination (orange). (B) Maximum anti-SARS-CoV-2 IgG-binding titers on the basis of other lymphoma-directed therapies. Any chemotherapy includes patients who received any cytotoxic chemotherapy at any point in their lymphoma treatment course. Post-autoSCT includes all patients who received high-dose chemotherapy with an autologous stem-cell rescue as part of their lymphoma treatment course at any point. For (A and B), comparisons were made using Brown-Forsythe and Welch ANOVA test and adjusting for multiple comparisons using a Dunnett T3 test. (C) Antibody responses after mRNA vaccination increased within the first 12 months with increased time between last anti-CD20 therapy and initial vaccination (P = .002). (D) Maximum anti-SARS-CoV-2 spike IgG-binding titers correlated with higher CD19+ CD20+ CD5− CD3− B cells at first available time point (P < .001). For (C and D), correlations were statistically significant using extra-sum-of-squares F test. For all figures, orange = patients who have also received anti-CD20-directed therapy within 1 year before vaccination, and horizontal line = background IgG levels determined from prepandemic samples. *P < .05, ****P < .0001. Error bars = mean ± SEM. ANOVA, analysis of variance; AutoSCT, autologous stem-cell transplantation; Bcl-2i, Bcl-2 inhibitor therapy (ie, venetoclax); BTKi, Bruton's tyrosine kinase inhibitor therapy (eg, ibrutinib and acalabrutinib); IgG, immunoglobulin G.

FIG 6.

Antibody binding and neutralization against SARS-CoV-2 variants of concern. (A) IgG-binding against spike proteins of SARS-CoV-2 variants of concern was reduced and showed increased variations compared with healthy vaccinees. (B) Plasma-neutralizing antibody titers against WA1/2020, Delta, and Omicron strains in select nucleocapsid-negative and nucleocapsid-positive patients with NHL. Comparisons were made using Brown-Forsythe and Welch ANOVA test and adjusting for multiple comparisons using a Dunnett T3 test. (C) Neutralizing antibody titers against WA1/2020, Delta, and Omicron strains in paired NHL samples. For all graphs, blue = nucleocapsid-negative, and purple = nucleocapsid-positive. Horizontal lines = limit of detection. Error bars = mean ± SEM. ANOVA, analysis of variance; IgG, immunoglobulin G; NHL, non-Hodgkin lymphoma.