Biomarker Profiling Differentiates Sepsis from Cytokine Release Syndrome in Chimeric Antigen Receptor T-Cell Therapy for Acute Lymphoblastic Leukemia (ALL)

Chimeric antigen receptor (CAR)-modified T cells with CD19 specificity (CTL019) are a highly effective novel immune therapy for relapsed/refractory ALL. Cytokine release syndrome (CRS) is the most significant and in some cases a life-threatening toxicity. These patients are often neutropenic and immunosuppressed and at high risk for sepsis. In addition, there have been cases where concurrent infectious complications potentially fueled underlying CRS leading to hypotension and hypoxia refractory to anti-cytokine directed therapies (Frey et al. abstract 2296, ASH 2014). Discriminating between sepsis and CRS can be a significant clinical challenge in the critical time-window required to initiate effective therapy, especially given that the treatments for CRS, including anti-cytokine therapies and corticosteroids, may worsen severe infections. We recently published an extensive study of 43 cytokines and soluble cytokine receptors, and clinical biomarkers in a cohort of 63 CTL019-treated ALL patients. Peak levels of 24 cytokines in the first month after infusion were highly associated with severe CRS, and CRS predictive models were described (Teachey et al. Cancer Discovery, June 1, 2016 6; 664).

The object of these current studies was to compare biomarker profiles in CRS and sepsis, and attempt to identify a profile that would discriminate between the two clinical syndromes, particularly at the time of ICU admission. We evaluated 50 cytokines, soluble receptors, and other serum biomarkers in 66 adult and pediatric patients treated with CTL019 and who developed CRS (ALL-CRS cohort), 15 patients not treated with CTL019 but who developed sepsis leading to ICU admission (sepsis cohort), and 10 normal healthy children (control cohort). Of the 66 patients treated with CTL019 and who developed CRS, 30 (45%) required ICU admission. The biomarkers tested included the ones described in our published work plus a panel of seven additional biomarkers (Angiopoietin 2, [ANG2] sCD163, Pentraxin 3 [PTX3], sCD14, PAI-1, P-selectin, ICAM-1) that have previously been reported in the literature to be associated with sepsis.

After adjusting for multiple comparisons, the following biomarkers were significantly elevated in sepsis compared to normal subjects: ANG2, GCSF, IFNα, IL1RA, IL4, IL6, MIG, MIP1α, PTX3, TNFα, sCD163, sCD30, sIL-1RI, sIL-1RII, sIL-2Rα, sIL-4R, sRAGE, sTNFRI, sTNFRII, sVEGFR1, sVEGFR2, sVEGFR3, and VEGF, whereas IL13 and RANTES were significantly lower in sepsis. In subjects with ALL and CRS, within 72 hours of ICU admission following CTL019 therapy (N=29), 23 biomarkers were significantly different compared to sepsis after adjusting for multiple comparisons; 16 were elevated in CRS: GM-CSF, HGF, IFN-γ, IFN-α, IL-10, IL-15, IL-5, IL-6, IL-8, IP-10, MCP1, MIG, MIP-1β, sIL-2Rα, sTNFRI, and sTNFRII; whereas 7 were elevated in sepsis subjects: CD163, IL-1β, sCD30, sIL-4R, sRAGE, sVEGFR-1, and sVEGFR-2. Figure 1 shows representative box-plots for sCD163 (elevated in sepsis) and IP-10 (elevated in CRS). Detailed biomarker profiling and an algorithm for discriminating between sepsis and CRS in patients treated with CTL019 will be presented. These comprehensive profiling data provide novel insights into CRS biology and importantly, will help us discriminate between CRS and sepsis in patients who become critically ill after receiving CAR T cell therapy. These data have direct translational therapeutic relevance.

View largeDownload slide

View largeDownload slide

Close modal