Deep Characterization of Immune Dysfunction in Patients with Multiple Myeloma (MM) and Identification of Cellular Biomarkers for Tailored Vaccination Strategies

Background: Infection is the leading cause of death in MM. However, the extent of immune dysfunction compared to other B-cell lymphoproliferative disorders (B-CLPD) and healthy adults remains poorly studied. Greater knowledge is needed for tailored vaccination strategies and to improve outcomes upon new pathogens or breakthrough infections.

Aim: Generate an atlas of the basal immune status and its response to vaccination in MM vs B-CLPD patients and age-matched health care practitioners (HCP), using the COVID-19 mRNA vaccines as a case study.

Methods: A total of 1,099 blood and serum samples were collected from 177 individuals: 28 MM, 53 B-CLPD and 96 HCP older than 50. They were studied before COVID-19 mRNA vaccination, at days 7 and 14 after the first dose, at days 7 and 62 after the second dose, as well as before and at day 17 after the booster. Immune profiling was performed using multidimensional and computational flow cytometry that systematically analyzed 56 immune cell types per sample and time point: 17 B, 30 T, 6 antigen-presenting cell (APC) and 3 granulocytic subsets. Serum levels of IgM, IgG and IgA against the receptor-binding domain (RBD) of the spike (S) glycoprotein, S glycoprotein, nucleocapsid (N) and main protease were quantified using a multiplex-microsphere-based flow cytometry assay. SARS-CoV-2-specific CD8 T cells were quantified using a dextramer panel of S, N, membrane, and ORF3 proteins.

Results: When compared to HCP and/or B-CLPD, MM patients showed abnormal distribution of 17/17 B, 22/30 T, 4/6 APC and 1/3 granulocytic cell subsets prior to vaccination. The most deviated cell types in MM were naïve CD21+ B-cells, naïve CD4 T-cells, PD1- and PD1+CD127low effector memory (EM) CD8 T cells, classical monocytes and neutrophils.

The B cell compartment of MM patients showed impaired response to COVID-19 mRNA vaccination. While B-CLPD patients and HCP displayed significant expansions of naïve CD21-, IgM+IgD+CD27+CD21-, IgG+CD27-CD21-, and IgG+CD27+CD21- B-cell subsets, as well as of IgA+ circulating plasma cells, such expansions were not observed in MM patients. Accordingly, anti-RBD IgM, IgA and IgG titers were significantly reduced in MM compared to B-CLPD and HCP after the second dose (P ≤.002). Importantly, the booster increased anti-RBD IgG levels in HCP (20,184 to 186,629 IU/mL, P <.001) but not in MM (1,991 to 5,998 IU/mL, P =.14) and B-CLPD (33,548 to 41,288 IU/mL, P =.19) patients.

The immune response of the T-cell compartment was also altered in MM patients. The significant expansion of CD127lowPD1+ and CD127+CD25+ EM CD4, naïve CD4 and CD127lowPD1+ EM CD8 subsets observed in HCP was unnoted in MM and B-CLPD patients. Accordingly, the percentage of virus-specific CD8 T cells after the second dose did not increase in MM and B-CLPD patients. The booster increased virus-specific CD8 T cells in HCP (0.08 to 0.14%, P =.02), but not in MM (0.08 to 0.09%, P =.94) and B-CLPD (0.23 to 0.16%, P =.85). Furthermore, the booster induced virus-specific CD8 T cell differentiation into an EM phenotype in HCP but not in MM and B-CLPD patients. In addition to the dysfunctional B and T cell response to vaccination, MM patients showed abnormal kinetics of APC such as intermediateclassical and non-classical SLAN- and SLAN+ monocytes, as well as of granulocytic subsets such as basophils and neutrophils.

Based on differences in immune-cell distribution using HCP as a reference, we calculated an immune dysregulation longitudinal cumulative score in each individual that included 48,496 immune parameters. Compared with HCP, up to 34% MM and 17% B-CLPD patients showed high immune dysregulation scores. Among them, 75% and 33%, respectively, had low seroconversion after the second dose. We found 7/30 T and 14/17 B-cell subsets associated with poor vaccine response in MM patients, namely CD127lowPD1+CXCR5+ EM CD8 T cells, as well as CD21- and CD21+ naïve, transitional, IgM+IgD+CD27+CD21+, IgM+CD27-CD21+, and IgG+CD27-CD21+ B cells.

Conclusion: We provide an atlas of the immune dysfunction in MM patients and how it affects the efficacy of vaccination strategies such as for COVID-19. The schedule of vaccine doses may thus benefit from individualization according to patients’ immune status, which could act as a surrogate of host, tumor and treatment-related immune dysfunction. Accordingly, we have identified key cell types for readily monitoring of patients’ immune status using routinely available flow cytometry.