Natural Killer-Cell Recovery in Patients Receiving CD19 CAR T-Cell Therapy: Dynamics and Clinical Significance

Background: Chimeric antigen receptor (CAR) T-cell therapy is an effective treatment for B-cell malignancies and systemic lupus erythematosus (SLE). However, due to the use of fludarabine and cyclophosphamide and the infusion of CAR T-cells, the immune-cell profile and immune reconstitution of patients will be affected. Current research on immune recovery mainly focuses on T cells, B cells, and neutrophils. The dynamics of natural killer (NK) cells, which are important anti-virus and anti-tumor cells, after CAR T-cell therapy remain unclear.

Aim and Methods: The study aimed to track the dynamic changes in NK-cell recovery after CD19 CAR T-cell treatment. Samples were collected from 52 patients with non-Hodgkin lymphoma, B-cell acute lymphoblastic leukemia, and systemic lupus erythematosus at pre-lymphodepletion (pre-LD) and during the first year after CAR T-cell therapy (NCT04008251 and NCT05765006), as well as from 10 healthy volunteers. Multiparameter flow cytometry was performed to longitudinally monitor the dynamics of the NK-cell profile.

Results: NK-cell number reached the nadir at day 0 and returned to normal reference at about 1 month after CAR-T cell infusion, and the growth of PB-NK cells was synchronized with BM-NK cells. The imbalance of maturation subpopulations of NK cells appeared at pre-LD and was most unbalanced at day 7, with a higher proportion of CD56^dimCD16^- and a lower proportion of CD56^dimCD16⁺.

For receptors, expressions of CD16 and NKG2D were impaired at baseline and after LD and recovered to normal reference levels at 2-3 months, while the levels of NKG2A and TRAIL were higher than the upper limit of the normal reference range throughout the entire follow-up period. NK cells were highly activated, characterized by high expression of activation indicators CD25 and CD69, as well as exhaustion markers PD-1, LAG3, and TIM3, especially on the day of CAR-T cell infusion.

For patients experiencing cytokine release syndrome (CRS), delayed NK-cell proliferation, CD16 recovery, and NKG2D expression were observed, and a lower percentage of memory-like NK cells was recorded during the observation. A significantly increased proportion of CD56^brightCD16^-NK cells appeared from day 7 to 2-3 months after CAR-T cell infusion in patients experiencing CRS when compared with those without CRS. Moreover, there were different differentiated profiles of CD56^dim NK cells with markedly higher proportions of NKG2A⁺CD57^-, NKG2A^-CD57^-, and a significantly lower proportion of NKG2A^-CD57⁺ in the CRS group. Furthermore, we found that NK cell counts and the expression of NKG2D and CD16 were negatively associated with peak CAR-T cells or inflammatory molecules (e.g., IL-10, IFNγ, or CRP) levels.

Clinically, patients with CRS occurrence or a low NK/T ratio (<median) at the end of the first month tended to be more likely to experience virus infection/reactivation. Of note, patients rich in the CD56^dimNKG2A⁺CD57^- subset seemed to be more likely to experience CRS, and those rich in the CD56^dimNKG2A^-CD57⁺ subset were at high risk of recurrence/progression.

Conclusions: In conclusion, the number, phenotype, and maturity of NK cells were changed during CAR-T cell therapy. Attention should be paid to the differentiated status of NK cells for CRS and inferior prognosis prediction and prevention. Measures taken to stimulate NK cell expansion and prevent virus infection are also necessary when NK-cell recovery is impaired during CAR T-cell therapy.

No relevant conflicts of interest to declare.