A Novel Antibody-Drug Conjugate Targeting the CD179a Component of the Surrogate Light Chain for the Treatment of B-Cell Acute Lymphoblastic Leukemia/Lymphoma

Background: The surrogate light chain (SLC) differentiates the pre-B cell receptor (preBCR) from the B-cell receptor (BCR) in B-cell ontogeny. During the pro-, pre-BI, and pre-BI transitional stages of development, the preBCR is a critical step that supports B-cell maturation into BCR-expressing immature B-cells. Like CD22, the CD179a (VpreB1) and CD179b (lambda 5) components of the SLC are internalized when crosslinked by an external ligand. Since more than 90% of B-cell acute lymphoblastic leukemias/lymphomas (B-ALLs) are developmentally arrested during these critical transitional stages of B-cell development, we hypothesized that an antibody-drug conjugate (ADC) targeting the CD179a/VpreB1 component of the SLC would have potent anti-leukemia activity, adding to the repertoire of B-ALL immunotherapeutic options. To investigate this strategy, we systematically assessed SLC antigen densities from four normal humans, six B-ALL cell lines, and harvested cells from 10 B-ALL patient-derived xenograft (PDX) models that have well-characterized stages of B-cell maturational arrest, each driven by specific, somatic oncogenic alterations. We tested the efficacy of an anti-CD179a mAb conjugated to calicheamicin via an AcBut linker, hereafter “VpreB1 ADC”, for its activity in B-ALL cell lines and PDX murine models.

Methods: CD179a surface expression was measured by quantitative flow cytometry analysis (CD179a-PE, Biolegend) of normal B-cells, B-ALL cell lines, and PDX models. Half-maximal inhibitory concentrations (IC₅₀) for the VpreB1 ADC (prepared by Creative Biolabs) were determined via in vitro cell viability. For PDX selection criteria, we used <100 ng/ml versus 100-1000 ng/ml as our primary feature, and whether or not the B-ALL PDX expressed luciferase. Burkitt’s and multiple myeloma cell lines were used as negative controls. For each of five PDX evaluations, twelve NSG mice were used to test the in vivo activity of the VpreB1 ADC: six animals each for the control and experimental groups. Upon detection of >1% human B-ALL in murine peripheral blood, the mice in the control group received normal saline, while the mice in the experimental group received three, 2 mg/kg intraperitoneal doses of the VpreB1 ADC on days 1, 4, and 7, once per experiment. Among the five PDXs chosen for longer-term survival analyses, one each had IGH::CRLF2, ETV6::PDGFRB or ETV6::RUNX1 rearrangements (pre-BI), and two hadTCF3::PBX1 alterations (pre-BI transitional). At necropsy, bone marrow and splenic tissues were assessed for human B-ALL using standard immunohistochemical techniques.

Results: We found that the antigen densities for CD179a component of the SLC on normal B-cells in the pro-, pre-B1 and pre-B1 transitional stages were 142±38, 558±83, and 535±82 receptors/cell, respectively. Surface expression of the SLC across the leukemia cells ranged from 120 (pre-BI, CRLF2/JAK2) to nearly 30,000 (pre-BI transitional; TCF3::PBX1) receptors/cell. Identified VpreB1 ADC IC₅₀s ranged from 2 ng/mL (TCF3::PBX1) to >1,000 ng/mL (KMT2A-R) amongst the 10 B-ALL PDX models tested, but we found no correlation (Pearson r, P=0.9963) between SLC antigen densities and IC₅₀s across a spectrum of B-cell maturational stages. We observed robust early in vivo responses to the VpreB1 ADC therapy. The most durable responses were seen for ETV6::PDGFRB and TCF3::PBX1 B-ALL PDXs, for which 10/12 (83%) mice survived >100 days, without immunohistochemical evidence of disease in 11/12 (92%) mice, approximately 2 months after one treatment course.

Conclusions: We detected a wide range in SLC antigen densities for normal and malignant B-lymphoblasts. Within the B-ALL PDX cohort, the SLC antigen density levels were similar to those reported for CD22 (0 to 15,000 receptors/cell), but at least one order of magnitude lower than for CD19 (10,000 to 20,000 receptors/cell). As previously described for inotuzumab ozogamicin (a calicheamicin-conjugated CD22 ADC), we found no correlation between antigen densities and IC₅₀s, even in xenograft mice with robust therapeutic responses to the VpreB1 ADC. Long-term durable responses were seen in the pre-BI and pre-BI transitional B-ALL models, but less so in the pro-B group. These data support the further investigation of VpreB1-targeting immunotherapies in children and adults with relapsed/refractory B-ALL.