Hypogammaglobulinemia and Infection Risk in Chronic Lymphocytic Leukemia (CLL) Patients Treated with CD19-Directed Chimeric Antigen Receptor T (CAR-T) Cells

INTRODUCTION

Anti-CD19 CAR T-cell immunotherapy is promising for patients with relapsed/refractory CLL. Hypogammaglobinemia can result from normal CD19+ B-cell depletion by CAR-T cells. CLL patients are already at risk for infections due to impaired immune function, lymphodepletion prior to CAR-T cells infusion, and immunosuppressive therapies. Intravenous immunoglobulin (IVIG) is used to manage hypogammaglobulinemia, although standard criteria for IVIG administration in this setting has not been established. We studied the incidence of hypogammaglobinemia and report infectious complications, risk factors, IVIG use, and clinical outcomes for CLL patients treated with anti-CD19 CAR-T cells.

METHODS

Adult CLL patients who received CD19-directed CAR-T therapy in 3 clinical trials (NCT01029366, NCT01747486, NCT02640209) from July 2010 to February 2020 were included. We reviewed demographics, available IgG levels, IVIG use, and clinical outcomes with a particular focus on infectious complications. Hypogammaglobulinemia was defined as IgG <4.5 g/L. Infections during the first 28 days after CAR-T were excluded to account for chemotherapy-induced neutropenia. CLL response was based on the 2008 International Workshop Group on CLL as per study protocols. The use of IVIG for hypogammaglobulinemia was at the discretion of investigator or treating physician. Fisher’s test, Wilcoxon test and logrank were used for data analysis.

RESULTS

Records of 71 adult patients with CLL were reviewed; 4 were excluded from further analysis as they did not have IgG levels after CAR-T. The median age at time of CAR-T was 63 years (range 43-78 years). 31 patients (46%) were alive, 31 (46%) were deceased, and 5 (7%) were lost to follow up at time of review. Median follow up for all patients was 33 months (range 2-114 months).

Of the 55 patients with an IgG level prior to CAR-T infusion, 24 (44%) had hypogammaglobulinemia at baseline After CAR-T infusion, 54 of 67 patients (81%) developed new or persistent hypogammaglobulinemia, and 40 of these patients (74%) received IVIG (Table 1). Forty-eight patients (72%) received at least one infusion of IVIG after CAR-T. Median time to initiation of IVIG after CAR-T infusion was 2.8 months (range 0.1-71.2 months). IVIG was used in 29 of 35 (83%) responders (defined as PR or CR) vs 19 of 32 (59%) in non-responders (defined as NR or PD) (p=0.056). There was no difference in survival observed based on whether or not patients had hypogammaglobulinemia (Figure).

42 patients (63%) had documented infections not related to chemotherapy-induced neutropenia. Fifteen (22%) had one documented infection, and 27 (40%) had more than one documented infection. There were 13 infections documented in patients who did not have hypogammaglobulinemia; the most common were ENT/sinus infections (5), bacteremia (2), URI (2), and other (2). There were 94 infections documented in patients who developed or had persistent hypogammaglobulinemia; the most common were lower respiratory tract infection/pneumonia (21), URI (20), and ENT/sinus infections (13). Complete and partial responders had more infections compared to non-responders or those who progressed (p = 0.01). Patients with hypogammaglobulinemia after CAR-T had an average of 1.74 (range 0-15) infections vs 1 (range 0-3) in patients without hypogammaglobulinemia. Patients who received IVIG had more infections (p=0.003); without IVIG the average number of infections was 0.58 (range 0-3), and with IVIG, the average number of infections was 2.00 (range 0-15).

CONCLUSION

Evaluating clinical outcomes with infections after CAR-T and potential strategies to minimize infection risk may improve morbidity and mortality in CLL patients. Use of IVIG was driven by individual practice, with heterogeneity regarding indication, frequency, and duration of treatment, though there was no difference in patient characteristics or response in patients who did or did not develop hypogammaglobulinemia. Patients who responded to CAR-T had more frequent infections, as might be expected in the setting of transient or persistent B- cell aplasia or hypoplasia. Patients who had more infections were more likely to receive IVIG. Further studies to define criteria for IVIG repletion in CLL patients treated with CD19-directed CAR-T cells may be incorporated in a standard clinical management algorithm.