Humoral Immunity and Plasma Cell Changes in Patients Responding to CD19-Specific Chimeric Antigen Receptor (CAR)-Modified T-Cell Adoptive Immunotherapy

Introduction: T cells engineered to express chimeric antigen receptors (CARs) recognizing CD19 (CART19) can eliminate malignant cells in acute lymphoblastic leukemia (ALL) and chronic lymphocytic leukemia (CLL). We and other groups have shown that persistent tumor eradication by CD19-specific T cell immunotherapy is accompanied by normal B-cell aplasia. It is assumed that responding patients cannot make new antibody responses post-successful CART19 treatment; however, the status of previously established humoral immunity in these patients is currently unknown. Understanding the consequence of successful CART19 therapy on established humoral immunity has implications for both the clinical management of CART19-treated patients as well as the potential application of this therapy to non-malignant diseases such as autoimmunity and transplantation.

Methods: We performed a prospective, observational study of adult and pediatric patients with ALL and adults with relapsed/refractory CLL, who were enrolled in clinical trials of CART19 at our institution. Serum antibody titers to previously-generated vaccine or vaccine-related pathogens (Streptococcus pneumoniae, Tetanus toxoid, Hemophilus influenza type-B (HIB), Measles, Mumps, and Rubella) were determined along with a quantitative assessment of B-cell and plasma cell frequencies in blood and bone marrow aspirates. Specimens were collected during pre-established study assessments or additional time points when collected as required for clinical management. Due to the challenges of assessing plasma cells, multiple methods were employed for their quantification in fresh specimens including flow cytometry and immunohistochemistry (IHC). Flow cytometric assessment of plasma cells was performed on freshly obtained marrow samples. Only patients with at least 3 months of B-cell aplasia in the absence of regular intravenous immunoglobulin (IVIg) infusions were included in the study.

Results: All patients had no evidence of leukemia or peripheral B cells post-CART19 infusion at the time of this study. Compared to pre-CART19 serum titers, antibodies to S. pneumoniae remained stable or increased in 9 of 12 patients despite lack of circulating B-cells. Antibody titers to Tetanus toxoid were stable or increased in 13 of 14 patients. Anti-HIB levels were stable or increased in 9 of 11 patients and antibodies to Measles, Mumps and Rubella were stable or increased in 12 of 13, 11 of 13, and 12 of 13 patients, respectively. Flow cytometric analysis of bone marrow aspirates after CART19 infusion revealed three patients with persistence of CD38⁺ CD138⁺ plasma cells (at 1, 3 and 9 months post infusion, respectively) despite a complete absence of peripheral CD19⁺ B cells. In 9 patients, CD20 and CD138 IHC analysis of bone marrow core biopsies revealed a decrease in plasma cell (ranges: 1-5% pre-CART19, 0-<1% post-CART19), consistent with our previously published data. Finally, in another subset of patients, neither B cells nor plasma cells were detectable by flow cytometry of aspirate material or IHC of core biopsies collected either pre- or post-CART19 treatment.

Conclusions: The stable or increased titers of antibodies to previous vaccines are surprising and may, in part, reflect improved marrow function as a result of leukemia eradication. The demonstration of plasma cells in a subset of patients in the absence of detectable tumor or normal B cells provides strong evidence for the existence of a population of plasma cells that are resistant to lysis by CART19 cells. This is consistent with antibody titers to previously generated vaccine antigens, which remain stable despite effective CART19 treatment. The additional finding of a decrease in CD138⁺ cells in several patients by IHC suggests that some populations of plasma cells are either targeted directly by CART19 or have a short half-life (e.g. plasmablasts); CD138 is not sufficient to distinguish these populations. Overall, these results indicate that long-lived plasma cells are resistant to CART19, likely due to a loss of CD19 during plasma cell differentiation. Continued analysis of remaining plasma cells in the absence of ongoing B-cell maturation as a result of CART19 persistence may provide important information on turnover rates of these long-lived cells in humans.