Checkpoint Inhibitors Augment CD19-Directed Chimeric Antigen Receptor (CAR) T Cell Therapy in Relapsed B-Cell Acute Lymphoblastic Leukemia

Abstract

CAR T cell therapy in relapsed B-ALL can result in complete response (CR) rates of 80-90%, but relapse-free survival declines to 60% within the first 12-months due to both CD19-positive and negative relapses. CD19-positive relapses that occur during this time are largely due to early CAR T cell loss. We hypothesize that inhibiting the PD-1:PD-L1 (programmed cell death 1) checkpoint axis may decrease T cell exhaustion, thereby improving CAR T cell function and persistence. We report our single institution experience of the use of PD-1 inhibitors in patients with relapsed or refractory B lymphoblastic malignancies treated with CD19-directed CAR T cell therapy.

Methods: Patients treated with CD19-directed CAR T cell therapy (murine CTL019 or humanized CTL119) at the Children's Hospital of Philadelphia who demonstrated repeated early CAR T cell loss or partial/no response to CAR T cell therapy received a PD-1 inhibitor starting no sooner than 14 days after CAR T cell infusion and after resolution of cytokine release syndrome (CRS) symptoms, with the possibility of repeated doses up to every 3 weeks.

Results: Fourteen patients, ages 4-17 years, with heavily pretreated, relapsed B-ALL (n=13) or B lymphoblastic lymphoma (n=1), were treated with CD19-directed CAR T cell therapy (CTL019, n=4; or CTL119, n=10) in combination with pembrolizumab (n=13) or nivolumab (n=1). Three of 6 patients treated with CD19 CAR T cells in combination with a PD-1 inhibitor for early B cell recovery re-established B cell aplasia (a reflection of CAR T cell function) for 5-15 months, 2 of whom have persistent B cell aplasia with ongoing pembrolizumab therapy. Four patients started pembrolizumab for bulky extramedullary disease unresponsive to or relapsed after CAR T cells, with 2 partial and 2 complete responses seen. In one patient, significant CAR T cell proliferation was measured within days of starting pembrolizumab and in temporal correlation to radiographic disease response. In 4 patients who failed to achieve disease remission with initial CAR T cell infusion, no CRs were achieved with the addition of pembrolizumab, although partial responses were seen, and one patient progressed with CD19-dim/negative disease.

CRS symptoms and fever typical of CAR T cell proliferative responses were observed in 3/14 patients within 2 days of starting pembrolizumab. Other early and delayed adverse effects associated with PD-1 inhibition were tolerable or reversible upon discontinuation, and including 1 case each of acute pancreatitis, hypothyroidism, arthralgias, urticaria, as well as 4 patients with grade 3-4 cytopenias. No grade 5 toxicities or graft-versus-host disease flares occurred. Two patients discontinued pembrolizumab for delayed adverse effects after multiple doses; both patients relapsed/progressed with CD19+ disease a few weeks after discontinuation.

Discussion: T cell exhaustion or activation induced CAR T death (AICD) has been suspected to contribute to poor persistence of CAR T cells. We hypothesized that the PD-1 checkpoint pathway may be involved in CAR T cell exhaustion in some cases, which may be overcome by checkpoint inhibition. Here, promising responses were specifically seen in those with early B-cell recovery and bulky extramedullary disease. In contrast, PD-1 inhibition had partial, but no durable, effect in the four B-ALL patients with poor initial marrow response to CAR T cell therapy alone, suggesting a different mechanism such as AICD may be responsible for poor initial responses. No unexpected or fatal toxicities were seen. This cohort shows initial evidence that checkpoint inhibitors can be used effectively and safely with CAR T cell therapy in children with relapsed B-ALL, and that this strategy may augment CAR T cell effect and persistence.