Serum Amyloid-a (SAA) Is an Optimal Acute Phase Protein for Cytokine Release Syndrome in CAR-T Patients Exposed to Tocilizumab

Patients treated with CAR-T and experiencing cytokine release syndrome (CRS) are often exposed to tocilizumab (anti IL6R). After the administration, IL-6 peaks in serum determinations and looses its predictive power to associate with severity of CRS.

This study aimed to explore the role of Serum Amyloid A (SAA) as a potential biomarker for monitoring early toxicities, particularly in patients treated with CAR-T cells. The analysis included 34 consecutive patients treated with CAR-T therapy at our institution from May 2023 to June 2024. SAA levels were measured at multiple time points: before lymphodepletion, immediately before CAR-T infusion, and every other day up to day 14 post-infusion. Additional acute phase proteins (APPs), including C-reactive protein (CRP), interleukin-6 (IL6), interleukin-2 receptor (IL2R), procalcitonin, ferritin, soluble suppression of tumorigenicity 2 (SST-2), and soluble urokinase plasminogen activator receptor (SUPAR), were also assessed at these intervals.

The study found that median baseline SAA levels were elevated above the upper limit of normal and exhibited wave-shaped kinetics post-CAR-T infusion, peaking on day 4 and declining thereafter. SAA levels were significantly higher before lymphodepletion compared to day 14 levels. When comparing early responders to non-responders, non-responders showed higher median SAA levels before lymphodepletion and at the day 4 peak. SAA may therefore mark tumor-associated inflammation also in this context.

SAA levels were significantly higher in patients experiencing CRS of any grade, with a clear correlation between elevated SAA and CRS severity. However, no such correlation was found for immune effector cell-associated neurotoxicity syndrome (ICANS). Strong associations were also observed between SAA and other APPs, particularly CRP, procalcitonin, ferritin, SST2, and SUPAR, with moderate trends observed for IL6 and IL2R.

In a focused analysis on patients previously exposed to tocilizumab, SAA emerged as the best-performing APP in identifying patients at risk of clinically relevant (grade 2 or higher) CRS, with a ROC analysis yielding an AUC of 0.86, being superior to CRP, IL6. IL2R, procalcitonin, and ferritin. A cut-off of 70 mg/L for SAA was identified, offering a sensitivity of 87% and specificity of 83% in distinguishing patients with grade 2-4 CRS.

SAA is a valuable biomarker for assessing inflammation and CRS in CAR-T treated patients, particularly in those previously treated with tocilizumab, where traditional markers like IL6 may be less reliable. SAA measurement could guide clinical monitoring and intervention strategies, especially in settings where early detection of severe CRS is critical. Further research is recommended to confirm these findings and explore the broader applicability of SAA in CAR-T therapy.

No relevant conflicts of interest to declare.