Refractory Cytokine Release Syndrome in Recipients of Chimeric Antigen Receptor (CAR) T Cells

CTLO19 cells are CAR-modified T cells which recognize CD19 and produce high durable remission rates for pts with relapsed or refractory acute lymphoblastic leukemia (ALL). Cytokine Release Syndrome (CRS) has emerged as the major treatment related effect from CTL019, with symptoms that include high fevers and malaise but can progress to capillary leak, hypoxia and hypotension. CRS occurs hours to days after CTL019 infusion and correlates with rapid in vivo CTL019 expansion and marked elevation of serum IL6. In most cases, CRS is self-limited or rapidly reversed with anti-cytokine directed therapies. Here we report 3 cases of refractory CRS in adult pts with ALL. Our experience offers insight into clinical and investigational parameters describing this syndrome; highlights the variance of CRS across disease types and illustrates complexities of CRS management during concurrent infectious illness.

As of 7/1/14, 97 pts (30 pediatric ALL, 12 adult ALL, 41 CLL, 14 NHL) have been treated with CTLO19. To capture clinical manifestations of CRS across protocols, we developed a novel CRS grading scale which will be described. Severe CRS (Gr 3-5) occurred in 27 (64%) of ALL pts and only 16 (29%) of CLL/NHL pts (p=0.001).

12 adults with ALL received CTL019; 8/9 evaluable pts achieved CR (MRD negative) at 1 month and 1 pt with extramedullary disease had marked reduction of PET avid disease which is maintained at 1 yr. Severe CRS occurred in 11 of 12 adult ALL pts. CRS was self-limited in 2 pts, rapidly reversed with anti-IL6 directed therapy in 6 pts and was refractory to therapy, contributing to death in 3 pts who were not evaluable for disease response. No baseline attributes differentiate these 3 pts from the 9 adult ALL pts with manageable Gr1-4 CRS. We have shown however that ALL disease burden correlates with CRS severity (in press) and all 3 pts had significant disease burden at baseline. All received lymphodepleting chemotherapy with cyclophosphamide 300 mg/m2 q12h x 6 followed by infusion of CTLO19 cells. These 3 pts each received 6.50E+06, 6.70E+06 and 8.45E+06 CTLO19 cells/kg compared to median CTL019 dose of 3.62E+06 in the 9 adult ALL pts with manageable CRS.

Pt 21413-03 developed CRS 12 hrs after infusion and tested positive for influenza B on D3. Despite broad spectrum antimicrobials (including oseltamivir) and anticytokine directed therapy with tocilizumab (4mg/kg x 2) and steroids, he died with refractory hypotension on D5. Pt 21413-06 had extensive disease after 2 prior allogeneic SCTs and developed CRS within 12 hrs of infusion. In addition to broad spectrum antibiotics, she received tocilizumab 8mg/kg (D 3, 6 and 12); intermittent high dose steroids (D 4-15) and etanercept (D14). She died D15 with hypotension, hypoxic respiratory failure and concurrent MDR pseudomonas sepsis and pneumonia. Pt 21413-11 developed CRS within 24 hrs of infusion. He received tocilizumab 8mg/kg (D3&4); siltuximab (D5&15) and intermittent high dose steroids (D 4-15). After an initial response, he developed recurrent fever, pulmonary infiltrates and blood cultures positive for stenotrophomonas. He died D15 with refractory hypoxia and hypotension. All 3 pts’ clinical CRS correlated with marked in vivo CTL019 expansion and progressive serum cytokine elevations (data to be shown).

CONCLUSIONS: CRS is the major toxicity of CTL019 therapy and its clinical course varies depending on disease type (more frequent and severe in ALL) and disease burden (in ALL). The 3 refractory CRS cases described here (of 97 total pts treated) have all occurred in adult ALL pts with significant disease burden who received relatively high doses of CTL019 cells. In addition, all 3 had significant infectious complications which potentially fueled underlying CRS and/or were made more virulent due to impairment of immunity with administration of anti-cytokine directed therapies. Future protocol modifications will be made goal of limiting severity of CRS while maintaining high durable remission rates. Further exploration is planned to better correlate timing and choice of anticytokine directed therapy in relation to clinical and investigation parameters of CRS.