Cardiac Profile of Chimeric Antigen Receptor T Cell Therapy in Children: A Single-Institution Experience

Immunotherapy with chimeric antigen receptor (CAR)-modified T cells targeting CD19 for pediatric acute lymphoblastic leukemia (ALL) has demonstrated significant efficacy. The principle toxicity is cytokine release syndrome with resultant hypotension. However, the spectrum of cardiovascular effects associated with CAR T cell therapy has not been systematically evaluated. We reviewed all patients who received CD19-directed CAR T cells at the Children's Hospital of Philadelphia between April 2012 and September 2016. The primary endpoint was hypotension-requiring inotropic support. Secondary endpoints included echocardiographic dysfunction at discharge and 6-month follow-up. Descriptive and univariate analyses were performed, and 98 encounters were included (55% male patients; mean age, 11.8 years [range, 1.7 to 27.1]); 98% had B-ALL. Before infusion 10 had cardiomyopathy and 1 had single-ventricle physiology. Primary endpoint occurred in 24 patients with mean onset 4.6 days (range, 1 to 9) after CAR T cell infusion, including 6 patients receiving milrinone. Worsened systolic function occurred in 10 patients; there were no cardiac-related deaths. Pretreatment factors associated with primary endpoint included higher pretreatment blast percentage on bone marrow biopsy (blast > 25%: odds ratio, 15.5; 95% confidence interval, 5.1 to 47.1; P < .001) and baseline lower ejection fraction (P = .019) or diastolic dysfunction (P = .021); neither pre-existing cardiomyopathy (P = .062), total body irradiation (P = .629), nor anthracycline dose (P = .444) were associated. At discharge, 7 patients had worsened echocardiographic function, but persistent dysfunction by the 6-month follow-up was rare. Pretreatment factors were not associated with persistent dysfunction at discharge. This is the first study to describe the cardiovascular effects of pediatric CAR T cell therapy. Although 10% had new systolic dysfunction after treatment, persistence was rare. Pretreatment blast count > 25% or pre-existing cardiac dysfunction increased the risk for hypotension-requiring inotropic support; these patients may warrant close observation.