Bendamustine impairs humoral but not cellular immunity to SARS-CoV-2 vaccination in rituximab-treated B-cell lymphoma-affected patients

Abstract

Background: Patients with B-cell lymphoma are a fragile category of subjects, particularly exposed to infections and characterized by an impaired vaccination response due to the disease itself and, even more, to the chemotherapy regimen. For this reason, extensive knowledge of the immune response status of these subjects is of fundamental importance to obtain possible indications for a tailored immunization strategy.

Methods: We enrolled two cohorts of patients with B-cell lymphoma under rituximab treatment or 3-24 months after treatment. In all patients, we evaluated both humoral and cellular immunological memory toward SARS-CoV-2, after standard vaccination and upon one booster dose.

Results: We observed no Spike-specific IgG production in patients (n = 25) under anti-CD20 treatment, whereas patients (n = 16) vaccinated after the completion of chemotherapy showed a higher humoral response. Evaluating SARS-CoV-2-specific T-cell response, we found that patients in both cohorts had developed robust cellular immunity after vaccination. Of the 21 patients (51%) that experienced a breakthrough SARS-CoV-2 infection, only six patients developed severe disease. Interestingly, these six patients had all been treated with rituximab plus bendamustine. Notably, we observed that Spike-specific IgG levels in patients treated with rituximab plus bendamustine were absent or lower compared with those in patients treated with rituximab plus other chemotherapy, whereas Spike-specific T-cell response was not different based on chemotherapy regiment.

Discussion: Our results show that, in patients with B-cell lymphoma under rituximab therapy, anti-SARS-CoV-2 mRNA vaccination induces a weak or absent humoral response but a consistent T-cell response. In addition, chemotherapy regimens with bendamustine further reduce patients' ability to mount a Spike-specific humoral response even after a long time period from chemotherapy discontinuation. These results provide evidence that different chemotherapeutics display different immunosuppressive properties that could be taken in to account in the choice of the right drug regimen for the right patient. Moreover, they question whether immunocompromised patients, particularly those treated with bendamustine, need interventions to improve vaccine-induced immune response.

Keywords: B cell lymphoma; SARS-CoV-2; T cell response; anti-CD20mAb; bendamustine; humoral response; mRNA vaccine; rituximab.

Figure 1

Evaluation of Spike-specific antibodies and CD4⁺ T cells after COVID-19 vaccination in Ongoing-TP and Post-TP patients. Evaluation of IgG specific for Spike protein (A) or RBD (B) in TP-Ongoing patients (green dots) at T0 (n = 24), T1 (n = 23), T2 (n = 23), and T3 (n = 10) and in Post-TP (orange dots) cohort at T0 (n = 3), T1 (n = 9), T2 (n = 13), and T3 (n = 12). (C) Frequency of CD4⁺ T cells reactive to SARS-CoV-2, defined by expression of CD154 and at least one cytokine among IFN-γ, IL-2, and TNF-α (CD4⁺CD154⁺CK⁺) in TP-Ongoing patients (green dots) at T0 (n = 14), T1 (n = 15), T2 (n = 15), and T3(n = 10) and in Post-TP (orange dots) cohort at T0 (n = 3), T1 (n = 7), T2 (n = 13), and T3 (n = 12). (D) Representative flow cytometric plots of Spike-specific CD154⁺CD4⁺IFN-γ⁺ T cells, at each time point, in one selected TP-Ongoing individual (upper panels) and Post-TP individual (lower panels). Black lines in (A–C) represent mean values, and gray area delimited by a dotted line in (A) and (B) represents the cutoff value. The evaluation time points are T0 (before vaccination), T1 (before the second dose), T2 (1 month after the second dose), and T3 (1 month after the third dose). *p < 0.05 and ***p < 0.001 calculated with Mann–Whitney U-test.

Figure 2

RTX + bendamustine–treated patients showed a more severe disease course during SARS-CoV-2 infection. (A) WHO COVID-19 severity index of 21 lymphoma B-cell patients (TP Ongoing, n = 17, green dots) (Post TP, n = 4, orange dots) infected with SARS-CoV-2 after vaccination and (B) WHO COVID-19 severity index of 11 RTX + bendamustine–treated patients and 10 RTX + other chemotherapy–treated patients. Mean values are represented as black lines. *P < 0.05 calculated with Mann–Whitney U-test.

Figure 3

Evaluation of Spike-specific antibodies and CD4⁺ T cells after COVID-19 vaccination in bendamustine- and no-bendamustine–treated patients. Evaluation of IgG specific for Spike protein (A) or RBD (B) in Post-TP RTX + bendamustine–treated patients (red dots) at T0 (n = 2), T1 (n = 5), T2 (n = 6), and T3 (n = 6) and RTX + other chemotherapy–treated patients (gray dots) at T0 (n = 1), T1 (n = 4), T2 (n = 7), and T3 (n = 6). Correlation between IgG anti-S (C) or IgG anti-RBD (D) with time from last rituximab dose on total Post-TP RTX + other chemotherapy–treated group at T2 (n = 7; light green dots) or T3 (n = 6; light blue dots). Pearson’s correlation coefficients were used to calculate the correlations; statistical significance (*p < 0.05) is reported on each graph. (E) Frequency of CD4⁺ T cells reactive to SARS-CoV-2 in TP-Ongoing RTX + bendamustine–treated patients (red dots) at T0 (n = 7), T1 (n = 9), T2 (n = 10), and T3 (n = 8) and RTX + other chemotherapy–treated patients (gray dots) at T0 (n = ), T1 (n = 6), T2 (n = 5), and T3 (n = 3). (F) Frequency of CD4⁺ T cells reactive to SARS-CoV-2 in Post-TP RTX + bendamustine–treated patients (red dots) at T0 (n = 2), T1 (n = 3), T2 (n = 6), and T3 (n = 5) and RTX + other chemotherapy–treated patients (gray dots) at T0 (n = 1), T1 (n = 4), T2 (n = 7), and T3(n = 6). (G) Representative flow cytometric plots of Spike-specific CD154⁺CD4⁺IFN-γ⁺ T cells, at T2, in two selected TP-Ongoing individuals (upper panels) and in two selected Post-TP individuals (upper panels) RTX + bendamustine–treated or RTX + other chemotherapy–treated, left and right panels, respectively. Black lines in (A, B, E, F) represent mean values, and gray area delimited by a dotted line in (A) and (B) represents the cutoff value. *p < 0.05 calculated with Mann–Whitney U-test.