The Impact and the Predictive Ability of Immunogenicity of Humanized Anti-CD19 CAR-T Cells for the Efficacy and Prognosis

Objective:

Chimeric antigen receptor T (CAR-T) therapy has achieved good results in hematological malignancies, but it's still limited by poor persistence of efficacy and resistance. Researchers generally believe that one of the major causes of the problem is the immunogenicity of CAR-T. In this research, we address the impact of CAR-T immunogenicity on the efficacy and prognosis of B-cell malignancies, as well as the potential mechanisms.

Methods:

We focuses on ADA (anti-drug antibody) of CAR-T, which is the indicator of immunogenicity, and secondary infusion, the clinical strategy that may lead to enhanced immunogenicity. Firstly, we collected plasma (pre-infusion to follow-up) of 58 NHL and B-ALL patients in our center's humanized anti-CD19 CAR-T clinical trial. Then the ADA levels of patients were detected with the method we developed pioneeringly for quantitative detection of ADA, which utilized CD19 protein as standard and used flow cytometry to detect ADA and standards after binding to CAR-CHO cells (CHO cells transfected with CAR). Afterwards, the relationship between ADA levels, secondary infusion and CAR amplification was analyzed by statistical methods, with ROC curves verifying the ability of ADA as a biomarker to predict CAR amplification. Subsequently, the laboratory indicators associated with increased ADA were explored, and Kaplan-Meier survival curves were used to analyze the correlation between immunogenicity and prognosis. Finally, two patients each with strong and weak immunogenicity were selected for pseudotime analysis and differential gene and pathway enrichment by single-cell transcriptome and T cell receptor sequencing.

Results:

All patients were allocated to the high-level ADA group(ADA_max≥3nM) and the low-level ADA group (ADA_max<3nM). ROC curves revealed that high-level ADA can serve as a practical biomarker of poor CAR-T cell expansion, and that CAR-T cell expansion was more significantly limited in patients with high-level ADA after receiving secondary infusion. ADA increasing was also correlated with B-cell reconstitution and increased IgM. Meanwhile, the increased risk of shortened PFS was significantly correlated with high-level ADA (HR=4.82; p=0.0023) and secondary infusion (HR=8.32; p<0.0001). In single-cell sequencing, a total of 20,490 cells that passed quality control were collected. The pseudotime analysis combined with TCR sequencing revealed that effector memory CAR-T cells were more in patients with weaker immunogenicity and had the ability to convert into cytotoxic CAR-T cells in vivo, meanwhile the persistence of CAR-T was stronger. The explore of B cells from patients with stronger immunogenicity indicated that genes related to cellular proliferation and activation of immune function were up-regulated and relevant pathways were enriched in, which may produce stronger immune effects.

Conclusions:

This study established a quantitative detection method for ADA levels, and demonstrated that the immunogenicity of CAR-T has important impact on both the amplification and prognosis of CAR-T therapy through experimental and bioinformatics validation. And ADA in plasma should be routinely monitored as a biomarker in the future, and can be further investigated in-depth in order to better guide the clinical strategy and to enhance the degree of diagnostic and therapeutic individualization.

No relevant conflicts of interest to declare.