In immunocompetent persons, the mRNA vaccines BNT162b2 and mRNA-1273 have been shown to be safe and efficacious against SARS-CoV-2 infection.1 However, most patients with B-cell malignancies, such as those with chronic lymphocytic leukemia (CLL), fail to develop a robust antibody response following vaccination.2, 3 This is considered to be largely due to the defects in innate and adaptive immunity in CLL which is observed even in the earlier phase of the disease.4 Furthermore, CLL treatment may exacerbate immune system dysfunction.3, 5 In agreement with this notion and previous reports, CLL patients are considered to have a heterogeneous and in most instances sub-optimal response to SARS-CoV-2 mRNA vaccines.3, 6-17

Up to now, the response to SARS-CoV-2 mRNA vaccines in CLL patients has been investigated in single-arm and case-control studies.3, 6-17 To assess the strength of the evidence and to better understand the current role of SARS-CoV-2 vaccination in CLL, we conducted a systematic review and meta-analysis using the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) statement (Figure S1).18 The analysis included all published studies assessing the seroconversion rate in CLL patients after SARS-CoV-2 vaccination consisting of two doses of mRNA vaccine. The literature search used PubMed to identify original full-text articles and research letters published in English on the efficacy of anti-SARS-CoV-2 mRNA vaccines in CLL patients reported before November 5, 2021. The search strategy used both medical subject headings terms and free-text words to increase its sensitivity. The electronic search yielded 28 records. A complementary manual search to screen the proceedings of the 2021 meetings of the American Society of Hematology identified three additional reports.15-17

The evaluation of different studies and data extraction were done by two reviewers (S.M. and D.G.). A cross-audit between reviewers allowed to include only studies that reported complete information on the seroconversion rate in the general population of CLL patients and in different treatment subgroups. Data, extracted in a standardized format, included patient demographics, disease status, treatment status, and type of therapy received. The heterogeneity of the data was evaluated by χ Q test and I2 statistics. For the Q test, p < .05 indicated significant heterogeneity; for the I2 statistics, an I2 value >50% was considered significant heterogeneity.

Altogether, we identified 13 studies involving 2082 patients.3, 6-17 The SARS-CoV-2 mRNA vaccine type was specified in 12 studies, including 1576 patients, with the BNT162b2 vaccine in 787 and mRNA-1273 and ChAdOx1 in 789.3, 6-16 The number of patients who received BNT162b2 or mRNA-1273 vaccination in the large study reported in abstract form by the French Innovative Leukemia Organization (FILO) was not available (n = 506).17

On pooled meta-analysis, the seroconversion rate in the general population of CLL patients was 52% (95% CI, 48%–74%), whereas the values of the Q or I2 test suggested a certain degree of heterogeneity across different studies (ie, Q = 93.25, p < .001; I2 = 87%) (Figure 1A).

Rate of seroconversion after SARS-CoV-2 mRNA vaccine on pooled meta-analysis in the general population of CLL patients (A), in patients treatment-naïve (B), in patients who had been given an anti-CD20 monoclonal antibody at any time before vaccination (C), patients vaccinated within 12 months from the last infusion of anti-CD20 antibody (D), patients on therapy with BTKi (E), and patients on therapy with venetoclax (F)

A potential confounder is that investigated studies included CLL cohorts heterogeneous concerning treatment status at the time of COVID-19 vaccination. Overall, information on treatment status was available in 1481 (71.1%) of 2082 patients.3, 6-8, 14, 16, 17

Among them, 533 (35.9%) were treatment-naive (TN). A pooled meta-analysis restricted to this patient subset showed a 73% seroconversion rate (95% CI, 64%–81%) but with heterogeneous results across different studies (ie, Q = 24.65, p < .001; I2 = 72%) (Figure 1B). The impact of anti-COVID-19 vaccination was also assessed in CLL patients receiving therapy at the time of anti-COVID-19 vaccination (n = 948 or 64.0%). Information on the type of therapy at the time of vaccination was available for 779 patients and allowed patients’ stratification in four different therapeutic groups;

Patients who had been given an anti-CD20 monoclonal antibody at any time before vaccination (n = 254). Patients vaccinated within 12 months from the last infusion of anti-CD20 antibody (n = 94). Patients on therapy with BTKi (n = 325). Patients on therapy with venetoclax (n = 94).

The seroconversion rate for patients who had received an anti-CD20 antibody at any time was 41% (95% CI: 23%–61%) (Figure 1C), with significant variability across different studies (Q = 20.32; p < .001; I2 = 85%). The heterogeneity degree was lower in patients who had received anti-CD20 antibodies within the year before vaccination. In this subgroup of patients, the seroconversion rate was 4% (95% CI: 1%–10%), this being consistent across different reports (Q = 7.11, p = .31; I2 = 16%) (Figure 1D).3, 6, 7, 13, 15-17 Finally, the seroconversion rate was similar in patients treated with either BTKi (29%, 95% CI, 20%–40%) or BCL2i (32%, 95% CI, 15%–43%), with a high level of heterogeneity across different studies (BTKi, Q = 19.22, p < .001; I2 = 69%; BCL2i, Q = 14.41, p = .01; I2 = 65%) (Figure 1E,F).

Besides randomized clinical trials, meta-analyses are considered to be the optimal tool to assess the robustness of clinical interventions. Our study is the first pooled meta-analysis investigating current results of SARS-CoV-2 vaccination in CLL patients. Despite the relatively high heterogeneity observed across different studies, the results of this meta-analysis confirm the overall lower seroconversion rate for SARS-CoV-2 vaccination in CLL patients as compared with the general population. According to our results, CLL treatment worsened the defective response to vaccination observed in CLL patients. Of note, among different treatment modalities, a short interval between anti-CD20 antibody exposure (ie, <12 months) and vaccination did abrogate seroconversion. Likewise, BTKis and BCL2is were found to similarly impair responses to COVID-19 vaccination, with a seroconversion rate ranging from 29% to 32% with both agents. Of note, the poor serologic response observed with SARS-CoV-2 vaccine in patients receiving a BTKi is in keeping with previous observations in the same patient population with hepatitis B vaccine.19 This suggests that exposition to BTKi is associated with a decreased immune response to de novo antigens.19 Further studies are needed for BCL2i therapy.

On the contrary, the immune response was partially preserved in CLL patients with asymptomatic disease who did not require therapy. In this patient cohort, about two-third of individuals experienced an almost optimal response to COVID-19 vaccination.

Identifying subgroups of CLL patients with different responses to SARS-CoV-2 vaccination and antibodies persistence over the time is important to decide treatment strategies and developing studies aimed at improving response to vaccination.20, 21 Recent preliminary results indicate that a third dose (“booster”) of SARS-CoV-2 mRNA vaccines induces seroconversion in 42% of the patients seronegative after the second vaccination dose.17 Finally, further research is required to understand the magnitude of the overall response to the SARS-CoV-2 vaccine and its effectiveness with newly SARS-CoV-2-related viruses. It is also essential to determine whether patients who have not achieved a serological response are unprotected from COVID-19 infection and to continue monitoring side effects.

CONFLICT OF INTEREST

The authors declare no competing financial interests.

AUTHOR CONTRIBUTIONS

SM designed the study, selected and evaluated studies, performed data extraction, evaluated and interpreted results, and wrote the manuscript; DG selected and evaluated studies, performed data extraction, performed statistical analyses, and evaluated results; EM evaluated and interpreted results and wrote the paper. All authors reviewed and approved the manuscript.