AUTHOR=Wang Yang , Song Zhiqiang , Geng Yuke , Gao Lei , Xu Lili , Tang Gusheng , Ni Xiong , Chen Li , Chen Jie , Wang Tao , Fu Weijia , Feng Dongge , Yu Xuejun , Wang Libing , Yang Jianmin TITLE=The risk factors and early predictive model of hematotoxicity after CD19 chimeric antigen receptor T cell therapy JOURNAL=Frontiers in Oncology VOLUME=12 YEAR=2022 URL=https://www.frontiersin.org/journals/oncology/articles/10.3389/fonc.2022.987965 DOI=10.3389/fonc.2022.987965 ISSN=2234-943X ABSTRACT=

Hematotoxicity is the most common long-term adverse event after chimeric antigen receptor T cell (CAR-T) therapy. Here, a total of 71 patients with relapsed or refractory (R/R) B-cell acute lymphoblastic leukemia (B-ALL) or large B-cell lymphoma (LBCL) were used to develop an early hematotoxicity predictive model and verify the accuracy of this model. The incidences of early hematotoxicity at 3 month following CAR-T infusion in B-ALL and LBCL were 45.5% and 38.5%, respectively. Multivariate analyses revealed that the severity of cytokine release syndrome (CRS) was an independent risk factor affecting early hematotoxicity. The analysis between the peak cytokine levels and early hematotoxicity suggested that tumor necrosis factor-α (TNF-α) and C-reactive protein (CRP) were closely associated with early hematotoxicity. Then, an early predictive model of hematotoxicity was constructed based on the peak contents of TNF-α and CRP. This model could diagnose early hematotoxicity with positive predictive values of 87.7% and 85.0% in training and validation cohorts, respectively. Lastly, we constructed the nomogram for clinical practice to predict the risk of early hematotoxicity, which performed well compared with the observed probability. This early predictive model is instrumental in the risk stratification of CAR-T recipients with hematotoxicity and early intervention for high-risk patients.