High efficacy of blinatumomab combined with pembrolizumab in patients with relapsed/refractory (r/r) B cell acute lymphoblastic leukemia (B-ALL): primary analysis of a Phase 1/2 study

Abstract

Abstract Blinatumomab (blina), a bispecific anti-CD19/CD3 T cell engaging antibody (Ab), provides improved outcomes in patients (pts) with r/r B-ALL but over half fail to respond (CR/CRh, 44%) and most relapse without consolidation therapy. We hypothesized that pembrolizumab (pembro), a PD-1 Ab, would enhance T cell effector function and augment blina anti-leukemia activity. Here we report the primary analysis from a single arm, phase 1/2 trial (NCT03512405) evaluating blina-pembro in r/r B-ALL. Patients ≥18 years with CD19+ r/r B-ALL received standard blina 9 µg/day on days 1-7 and 28 µg/day on days 8-28 in cycle 1 and on days 1-28 in subsequent cycles. Pembro 200 mg was given on day 15 in cycle 1 and on days 1 and 22 in subsequent cycles which was defined as schedule A. Blood and bone Marrow (BM) biopsy were collected at prespecified timepoints for disease assessment and coorelative analysis. High parameter immunohistochemistry (IHC) was used to evaluate T cell infiltration and phenotypic profiles using a supervised machine learning cell classifier. As of July 14, 2025, 7 pts enrolled to phase 1 (1 unevaluable) and 19 patients enrolled to phase 2 (1 unevaluable). All were treated at Schedule A dosing which was determined as recommended phase 2 dose. Median age was 49 years (range, 24-74), with 1 median prior regimens (range 1-5), and 49% (range, 10-95%) median BM blasts with 50% of evaluable patients with more than 50% BM blast. Poor risk features were common in 87% (21/24) including 50% Philadelphia chromosome like ALL. The study met its prespecified primary endpoint. Of 24 evaluable pts, 17 (71%) achieved CR/CRi after a median of 1 (range, 1-2 cycles). The CR/CRi rate was 71.4% (15/21) in cytogenetic poor-risk group, 58% (7/12) in high disease burden group, and 66.7% (9/12) in Ph-like group. Sixteen pts were measurable residual disease not detectable (MRD-) by flow cytometry, one achieved CR after cycle 2 (refractory after cycle 1). Thirteen CR pts underwent allogeneic transplant (alloHCT) with post-transplant cyclophosphamide (PTCy) as graft versus host disease (GVHD) prophylaxis. The median treatment cycles were 2 for responders who received alloHCT vs 3 for those who did not (P=0.237). Both groups received 3 median doses of pembro (P=0.442). Among 4 CR pts not undergoing alloHCT, 3 relapsed with median progression free survival (PFS) of 3.0 (range, 2.3-12.4) months; one remained in remission at month 28 after CAR T consolidation. Three pts developed ≥gr3 acute GVHD and all resolved with treatment, and 5 had NIH defined mild cGVHD. With a median follow-up of 14.7 months (1.8-20.9), 11 (65%) CRs are ongoing (10 post-transplant and 1 post-CAR-T). Sixteen of 24 evaluable pts were alive at data cut-off, with a 1-year OS rate of 74% (95% CI: 58-94%). In all cycles of treatment, cytokine release syndrome (CRS) occured in 50% patients (13/26) including 2 gr 3 events. All except 3 occured prior to first pembro dose. Neurologic toxicities were observed in 50% (13/26) patients, with 2 ≥ gr3 events; all reversible. Five pts (19%) experienced pembro-related immune AEs. All-non-hematologic gr3 toxicity were reversible. No dose limiting toxicity, ≥ gr4 non-hematologic toxicity, or treatment related deaths were reported. Immunoprofiling of tumor immune microenvironment (TIME) using high-dimensional IHC revealed significantly increased densities of memory CD45RO+CD8+ T cells from baseline to end of first treatment cycle (p=0.028, 48.36 vs 66.65 cells/mm2), and Tim-3+ CD4+ T cells (p=0.03, 86.89 vs 504.2 cells/mm2), with a significant rise in Tim-3+ cell frequency amongst total CD4+ T cells (p= 0.0006, 6.2 vs 26.5%). A trend towards higher densities of memory CD4+ T cell at baseline was seen in pts with CR/CRi compared to refractory pts (p= 0.0548, 735.4 vs 128.8 cells/mm2). In conclusion, our study showed that blina combined with pembro demonstrated a manageable safety and encouraging high efficacy in patients with poor risk r/r B-ALL, supporting its potential as a successful bridging therapy to alloHCT with high rate of durable remissions among those receiving alloHCT, without increasing the GVHD rate. Immune profiling revealed significant remodeling of the TME in the BM, suggesting mechanisms of response involving CD4+ and CD8+ T cell activation and infiltration.