PD-1 Inhibition Enhances Blinatumomab Response in a UCB/PDX Model of Relapsed Pediatric B-Cell Acute Lymphoblastic Leukemia

Abstract

Immune therapies such as blinatumomab, CD19-directed bispecific CD3 T-cell Engager (BiTE), have resulted in significant improvements in outcomes for relapsed B-cell acute lymphoblastic leukemia (B-ALL). However, up to half of blinatumomab treated patients do not respond completely or relapse after therapy. As a result, there is a need to identify potential strategies to improve the efficacy of BiTE therapy. The anti-PD-1 antibody pembrolizumab has been shown to successfully activate T cells against a wide range of cancer types. Here, we tested the ability of umbilical cord blood (UCB) reconstituted mice to respond to blinatumomab therapy with or without concurrent pembrolizumab treatment. Humanized mice were engrafted with patient-derived xenograft (PDX) cells derived from pediatric and adolescent/young adult (AYA) B-ALL patients who had either failed to achieve remission with negative minimum residual disease (MRD negative) or experienced a relapse. Mock-treated humanized mice engrafted with PDX cells efficiently developed overt disease within 30 days of engraftment of B-ALL. However, single agent therapy with either blinatumomab or pembrolizumab reduced disease burden in engrafted mice, with some mice observed to be MRD negative after the 28-day treatment course. Combination therapy significantly improved the percentage of MRD negative mice and improved long-term survival and cure rates as compared to mice that were given blinatumomab alone. Importantly, no benefits were observed in treated mice that lacked human immune cell reconstitution. These results indicate that UCB-humanized NRGS mice develop activatable immune function, and UCB-humanized PDX leukemia models can be used in preclinical studies to evaluate specificity, efficacy, and cooperativity of immune therapies in B-ALL.

Keywords: B-ALL; PD-1; bispecific T-cell engager; blinatumomab; humanized mice; immune checkpoint inhibition; patient derived xenograft; pembrolizumab.

Figure 1

Modeling combined blinatumomab (Blina) and PD-1 inhibition with pembrolizumab (Pembro) in humanized mice. (A) Schematic outlining the experimental setup. (B) Non-humanized mice were treated with Blina or control (CNTL) immediately after engraftment (2 weeks, beginning on day 1) with B-ALL and were re-treated after disease was readily detectable in PB (beginning on day 48). (C) B-ALL engraftment (from 4 unique PDX models) was detected in the bone marrow of mice 30 days after therapy was initiated. Treatment was started on day 1 of B-ALL injection according to the schedules in Table 1 . *p < 0.05 by Mann-Whitney test. (D) Proportion of mice with lower than 0.01% B-ALL in the marrow (MRD-) at day 30 for each of the 4 samples in Blina or combo groups. *p = 0.0128 by paired tTest. (E) Flow cytometry measurement of B cells in the PB of mice after therapy. *p < 0.05 by Mann-Whitney test. (F) Flow cytometry measurement of T cells in the PB of mice before and after therapy. *p < 0.05 by Mann-Whitney test. (G) Survival of treated mice correlates with MRD flow detection. *p < 0.05 by log rank test.