Introduction:Cytokine release syndrome (CRS) and immune cell-associated neurotoxicity syndrome (ICANS) are both serious complications of CAR-T therapy associated with endothelial dysfunction, prompting prior use of a modified version of the endothelial activation and stress index (m-EASIX) to predict the occurrence of severe ICANS and CRS. Previous studies have linked both hypophosphatemia and elevated IL6 levels to CRS and ICANS. Our study aimed to enhance the early prediction of both syndromes by integrating phosphorous and IL-6 both together and separately into the m-EASIX score.Methods:Forty-two patients with non-Hodgkin’s lymphoma presenting for CAR-T treatment were used to generate three variations in the m-EASIX score, assessing performance for the clinically actionable time points of day +0 through day +3.Results:The addition of phosphorous through the P-m-EASIX improved the predictive capabilities for the occurrence of ICANS, most notably on day +1 (AUC 89.6%;p= 0.0090, OR of 2.23;p= 0.0096) compared to the m-EASIX (AUC 80.8%;p= 0.0047, OR 1.72;p= 0.0046). The P-m-EASIX also showed enhanced predictive capabilities for the occurrence of CRS, with peak discriminatory function on day +3 (AUC 92.0%;p= <0.0001, OR 2.21;p= 0.0014). The addition of IL6 in the IL6-m-EASIX showed the highest discriminatory capacity for the prediction of CRS progression to grade ≥ 2 with peak function on day +3 (AUC 89.7%;p= 0.0040, OR 1.57;p= 0.031).Conclusions:Incorporating phosphorus levels into the m-EASIX score offered a cost-effective and straightforward method to improve the prediction of CAR-T toxicities. Larger-scale studies assessing the effectiveness of including phosphorus and IL-6 in the m-EASIX score to mitigate complications associated with CAR-T therapy are warranted.
引言:细胞因子释放综合征(CRS)和免疫细胞相关神经毒性综合征(ICANS)是CAR-T治疗中与内皮功能障碍相关的严重并发症,这促使我们先前采用改良版内皮活化与应激指数(m-EASIX)来预测重度ICANS和CRS的发生。既往研究已证实低磷血症和IL-6水平升高均与CRS及ICANS相关。本研究旨在通过将磷和IL-6指标分别及联合纳入m-EASIX评分体系,以提升对这两种综合征的早期预测能力。 方法:本研究纳入42例接受CAR-T治疗的非霍奇金淋巴瘤患者,构建了三种m-EASIX评分改良模型,评估其在临床可干预时间点(第0天至第3天)的预测效能。 结果:通过构建P-m-EASIX模型将磷指标纳入后,对ICANS的预测能力显著提升,其中第+1天的预测效能最为突出(AUC 89.6%;p=0.0090,OR值2.23;p=0.0096),优于原始m-EASIX评分(AUC 80.8%;p=0.0047,OR值1.72;p=0.0046)。P-m-EASIX对CRS的预测能力同样增强,第+3天达到峰值判别效能(AUC 92.0%;p<0.0001,OR值2.21;p=0.0014)。而通过IL6-m-EASIX模型纳入IL-6指标后,对CRS进展至≥2级的预测显示出最佳判别能力,第+3天达到峰值效能(AUC 89.7%;p=0.0040,OR值1.57;p=0.031)。 结论:将血磷水平纳入m-EASIX评分体系为提升CAR-T毒性预测提供了一种经济高效且操作简便的方法。未来需要开展更大规模研究,以评估在m-EASIX评分中整合磷与IL-6指标对降低CAR-T治疗相关并发症的临床价值。