Multiple COVID-19 vaccine doses in CLL and MBL improve immune responses with progressive and high seroconversion

Abstract

Patients with chronic lymphocytic leukemia (CLL) or monoclonal B-lymphocytosis (MBL) have impaired response to COVID-19 vaccination. A total of 258 patients (215 with CLL and 43 with MBL) had antispike antibody levels evaluable for statistical analysis. The overall seroconversion rate in patients with CLL was 94.2% (antispike antibodies ≥50 AU/mL) and 100% in patients with MBL after multiple vaccine doses. After 3 doses (post-D3) in 167 patients with CLL, 73.7% were seropositive, 17.4% had antispike antibody levels between 50 and 999 AU/mL, and 56.3% had antispike antibody levels ≥1000 AU/mL, with a median rise from 144.6 to 1800.7 AU/mL. Of patients who were seronegative post-D2, 39.7% seroconverted post-D3. For those who then remained seronegative after their previous dose, seroconversion occurred in 40.6% post-D4, 46.2% post-D5, 16.7% post-D6, and 0% after D7 or D8. After seroconversion, most had a progressive increase in antispike antibody levels. Neutralization was associated with higher antispike antibody levels, more vaccine doses, and earlier severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) variants; neutralizing antibody against early clade D614G was detected in 65.3%, against Delta in 52.0%, and against Omicron in 36.5%. SARS-CoV-2-specific T-cell production of interferon γ and interleukin 2 occurred in 73.9% and 60.9%, respectively, of 23 patients tested. After multiple vaccine doses, by multivariate analysis, immunoglobulin M ≥0.53 g/L, immunoglobulin subclass G3 ≥0.22 g/L and absence of current CLL therapy were independent predictors of positive serological responses. Multiple sequential COVID-19 vaccination significantly increased seroconversion and antispike antibody levels in patients with CLL or MBL.

Graphical abstract

Figure 1.

Sequentialpostvaccination antispike protein IgG levels. Changes in antispike protein IgG levels are shown in patients with CLL (A) or MBL (B) and treatment history (C). An antispike antibody level >50 AU/mL is classified as a positive response and an antispike antibody level ≥1000 AU/mL is classified as a strong positive response. Red bars in panel C indicate the median antispike antibody level in each group. An antispike antibody level of 0 could not be displayed in this figure. Although Australia had low COVID-19 case numbers, consensus guidelines developed at the beginning of the pandemic recommended deferral of CLL therapy where possible during COVID-19 outbreaks.^, Hence, the number of patients on therapy (panel C box 4) is relatively low. pre-vacc, prevaccine.

Figure 2.

Anti–SARS-CoV-2 spike antibody level stratification in patients with CLL or MBL postvaccination. Antispike level stratification after sequential vaccination were shown in patients with CLL (A) and MBL (B). Only patients with known quantitative antispike antibody levels were included.

Figure 3.

Detection of neutralization antibodies against SARS-CoV-2 variants D614G, Delta, and Omicron. (A) Heatmap summarizes the distribution of antispike antibody levels of each sample tested and their corresponding neutralizing activities against D614G, Delta, and Omicron antibody. (B) Bar chart showing the proportion of samples with positive or negative neutralization, split into 4 stratifications of antispike (anti-S) protein levels. (C) Odds ratio (OR) indicates the differences among patients with antispike levels between <1000 and ≥1000 AU/mL (top), or the differences between strains for those with an antispike protein level >1000 AU/mL (bottom), that is, the odds of a positive neutralizing antibody response against Delta was only 0.186 times the odds of that against D614G. The OR was adjusted for number of doses. (D) Determinants of neutralization response (positive/negative). The overall P values accounted for different variants, number of doses, serological responses (positive/negative), and antispike antibody levels. Antispike antibody levels per variant were not significant; that is, for all 3 variants, the patients with antispike antibody levels ≥1000 AU/mL were always consistently better than those with antispike antibody levels <1000 AU/mL. All other terms are statistically significant (having adjusted for the others). LCL, lower confidence limit; UCL, upper confidence limit.

Figure 4.

T-cell function analysis in patients with CLL after multiple vaccination doses. The IFN-γ and IL-2 responses were tested against the Miltenyi spike peptide pool. The mean number of responding cells in negative controls were subtracted from stimulated samples to account for background responses and results were expressed as spot-forming units (SFU) per 10⁶ cells. The positivity threshold (dotted line) was set based on the average number of responding cells across all negative controls in the cohort plus 3× the standard deviation. Median (red) and 95% confidence intervals were plotted for each antispike protein antibody level strata.

Figure 5.

Univariate and multivariate analysis of serological responses. OR for positive serological response (antispike antibodies ≥1000 AU/mL) to COVID-19 vaccine calculated by univariate (A) and multivariate (B) logistic regression models. Biological and clinical factors, using univariate models, associated with zero responders (>4 doses but remain antispike antibodies <50 AU/mL) (C), late responders (antispike antibodies <1000 AU/mL after 3 doses but increased to >1000 AU/mL after the fourth dose) (D), high responders (antispike antibodies ≥20 000 AU/mL at any point after their first dose) (E). Any analysis of IgG levels (or IgG subclasses) excluded patients on IgRT. All patients were treated with Bruton tyrosine kinase inhibitor (BTKi) received ibrutinib. P < .2 was used as the cutoff value for the above models. Pre-vax, prevaccination.