Preclinical efficacy of humanized, non–FcγR-binding anti-CD3 antibodies in T-cell acute lymphoblastic leukemia

T-cell acute lymphoblastic leukemia (T-ALL) comprises aggressive neoplasms characterized by the proliferation and accumulation in blood, bone marrow, and lymphoid organs of T-cell precursors abnormally arrested in differentiation. Current first-line chemotherapy regimens provide overall survival rates of ∼85% to 90% in children and ∼50% in adults.^1,2  Relapses have a dismal prognosis. T-ALL represents a heterogeneous group of malignancies classified into different molecular subtypes on the basis of aberrant expression of specific driver oncogenic transcription factors and global transcriptomic signatures.^3,4  Across subgroups, a number of additional genetic/epigenetic alterations are found in oncogenes (eg, NOTCH1) and tumor suppressor genes (eg, PTEN), which contribute to T-ALL development and clonal evolution.^3,4  However, this knowledge has so far not translated clinically into efficient targeted therapies.

Recent immune-therapeutic approaches including monoclonal antibodies (mAbs), bispecific Abs, and chimeric antigen receptor T cells have considerably diversified and improved the therapeutic arsenal against B-ALL.⁵  These therapeutic advances are so far missing in T-ALL. We previously reported that in NOD.Cg-Prkdc^scidIl2rg^tm1Wjl/SzJ (NSG) mice xenografted with T-cell receptor (TCR)–positive T-ALL diagnostic cases (T-ALL PDX), administration of the anti-CD3 mAb OKT3/muromonab rapidly induced leukemic cell death and leukemia regression and improved host survival.⁶  These antileukemic effects were observed irrespective of the nature of the molecular alterations driving individual T-ALL cases and involved both TCRαβ⁺ and TCRγδ⁺ T-ALLs. However, murine OKT3/muromonab is clinically unsuitable because of its immunogenicity and induction of life-threatening adverse events associated with the release of proinflammatory cytokines.⁷  In contrast, an immunoglobulin G1 (IgG1) humanized version of OKT3 that carries 2 point mutations (L234A and L235A) in its Fc CH2 region (hOKT3γ1[Ala/Ala]/teplizumab), thus impairing its binding to all Fcγ receptor (FcγR) by 2 to 3 orders of magnitude,^8-11  induces only mild adverse effects in humans.^12,13 

We report here that the non–FcγR-binding anti-CD3 mAbs hOKT3γ1[Ala/Ala]/teplizumab and foralumab¹⁴  have efficient antileukemic activity in T-ALL PDX.

T-ALL diagnostic samples were included in the GRAALL-2005 study or in the FRALLE protocol, and informed consent was obtained from all patients at trial entry. Studies were conducted in accordance with the Declaration of Helsinki and approved by local and multicenter research ethical committees. All samples used contained ≥80% blasts.

NSG female mice (age, 2 months; Charles River Laboratories) were maintained under specific pathogen-free conditions. Experiments were carried out in accordance with the European Union and French National Committee recommendations, under agreement APAFIS #7393-2016102810475144-v1.

Mice were tail vein injected (10⁶ cells per mouse) with leukemic cells (defined as FSC^hi, hCD45⁺, hCD7⁺) obtained from the bone marrow of T-ALL PDX NSG mice. In vivo leukemia cell apoptosis was analyzed by annexin V staining, as described.¹⁵  Leukemia wexpansion in blood was followed once per week, as described.⁶  For the survival curve, end point was defined either when intensity staining of leukemic cells with hCD45 and hCD7 sharply dropped or when mice showed signs of dyspnea. Chemotherapy for leukemic mice consisted of intraperitoneal administration of 0.5 mg/kg of vincristine (Sigma-Aldrich) on day 1 and 15 mg/kg of dexamethasone (Sigma-Aldrich) on days 1 to 5.

The IgG2a mouse OKT3 mAb to human CD3 and the isotypic controls mIgG2a and hIgG1 were from Bio X Cell (West Lebanon, NH), and foralumab was from Creative Biolabs (Shiley, NY). The control human IgG1 (hIgG1[Ala/Ala]) and the humanized anti-CD3, bivalent IgG1 construct (hOKT3γ1[Ala/Ala]) were built in a human IgG1 backbone in which the Fc domain contained 2 mutations (L234A and L235A) for decreased binding to FcγR,⁸  and sequence-confirmed IgG was expressed transiently in Chinese hamster ovarian E cells. Abs were subsequently purified using protein A affinity chromatography followed by analytical size-exclusion chromatography. The final purified Ab was >99% monomer. A subset of the reported experiments (Figure 1B) used teplizumab, obtained from Creative Biolabs, as an alternative source of hOKT3γ1[Ala/Ala]. Unless otherwise stated, Abs were injected when tumor burden reached 1% to 2% in blood, representing a leukemic burden of 10% to 20% in bone marrow in mature T-ALL PDX (high tumor burden).

We found that OKT3 administration induced a dose-dependent decrease of leukemic cells in T-ALL PDX recipient NSG mice and that this antileukemic effect translated into improved host survival (supplemental Figure 1A-B, available on the Blood Web site). We next investigated the antileukemic activity of hOKT3γ1[Ala/Ala] and compared it with that of OKT3. The same T-ALL PDX, rendered luciferase positive to allow live follow-up of leukemia burden by bioluminescence imaging, showed that both OKT3 and hOKT3γ1[Ala/Ala] elicited a similar antileukemic response (Figure 1A). Comparison of the antileukemic activity of OKT3 and hOKT3γ1[Ala/Ala] was next extended to PDXs derived from 7 additional TCR⁺ T-ALLs (supplemental Figure 2). These diagnostic cases harbor different combinations of mutations (supplemental Table 1) in oncogenic pathways that are frequently altered in T-ALL and that could also potentially interfere with TCR signaling.^16,17  Administration of OKT3 and hOKT3γ1[Ala/Ala] resulted in rapid and clear reduction of leukemia burden (supplemental Figure 1C), induction of leukemic cell apoptosis (supplemental Figure 1D), and similar delay of leukemic cell expansion (Figure 1B-C; supplemental Figure 3A). In all cases, these antileukemic effects translated into enhanced survival of recipient mice (Figure 1D; supplemental Figure 3B). A strong antileukemic response was also obtained after treatment of T-ALL PDX with foralumab,¹⁸  a first-in-class fully human IgG1 anti-CD3 mAb mutated in its Fc domain (L234A and L235E) to prevent FcγR binding and activation of innate immune cells (Figure 1A,C-D; supplemental Figure 3A-B). We conclude that non–FcRγ-binding anti-CD3 mAbs show antileukemic activity as efficient as OKT3 in T-ALL PDX. We next investigated the consequences of anti-CD3 administration together with chemotherapy. As shown in Figure 2 and supplemental Figure 4, although anti-CD3 monotherapy led to delay in leukemia recurrence, with relapsing leukemias showing variable levels of TCR expression depending on the case considered (supplemental Figure 5), anti-CD3 Abs cooperated with chemotherapy to strongly inhibit leukemia expansion (Figure 2A-C) and promote mice survival and even cure (Figure 2D).

Administration of hOKT3γ1[Ala/Ala]/teplizumab led to reversal of rejection in kidney transplant patients¹³  and improved pancreatic β-cell function in trials of type 1 diabetes mellitus ^7,19,20  and was associated with manageable adverse effects. Teplizumab demonstrated clear efficacy in a phase 2 trial in delaying type 1 diabetes progression in high-risk patients, with only mild adverse effects.¹²  Available evidence also shows that foralumab is clinically associated with minor adverse effects.¹⁸  Our results showing potent antileukemic activity of hOKT3γ1[Ala/Ala]/teplizumab and foralumab thus call for an immunotherapeutic approach using these mAbs in the treatment of CD3⁺ T-ALL. Because the safety profile of teplizumab seems to vary in different human pathologies,^12,13  the tailoring of teplizumab or foralumab dosage in T-ALL patients and the efficiency of corticoid coadministration to mitigate adverse effects¹³  and further improve antileukemic potential (this study) must be established in phase 1/2 clinical protocols. Investigation of the mode of action of therapeutic mAbs has pointed to an essential role of complement and FcγR-expressing immune cells and ensuing activation of Ab-dependent cell phagocytosis and cellular toxicity, including in NSG mice.^21-24  Our results show that the antileukemic properties of anti-CD3 mAbs are largely Fc independent in T-ALL PDX and rather rely on a direct, TCR-mediated cell death program.⁶ 

The authors thank V. Voynov and Y. Zhang (Boehringer-Ingelheim, Inc), C. Lasgi (Institut Curie cytometry platform), and E. Belloir, C. Alberti, and C. Roulé (Institut Curie animal facility) for expert technical assistance.

This work was supported by funds from Institut Curie, Centre National de la Recherche Scientifique, INSERM, Institut National du Cancer (grants 2015-1-PL BIO-06-ICR1 and PRT-K18-071), and Ligue contre le Cancer (Comité de l’Essonne). B.Z. was supported by predoctoral fellowships from the University Paris-Diderot and Ligue Nationale Contre le Cancer.

Contribution: C.T.Q. and B.Z. designed the study, carried out the experiments, performed data analysis and interpretation, organized the data, and wrote the manuscript; R.H. participated in the experimental work and data analysis; R.G., S.S., and J.M.S. generated and purified hOKT3γ1[Ala/Ala] and control hIgG1[Ala/Ala] antibodies; E.L. and M.E.D. provided primary patient material and associated immunophenotypes and molecular data; V.A. provided primary patient material and associated immunophenotypes and molecular data and participated in the study design; and J.G. designed the study, carried out data analysis and interpretation, organized the data, and wrote the manuscript.

Conflict-of-interest disclosure: C.T.Q. and J.G. have research agreements with Servier and Autolus, Ltd. R.G. and S.S. are employees of Janssen R&D. The remaining authors declare no competing financial interests.

Correspondence: Christine Tran Quang, UMR3348 CNRS-Institut Curie, Centre Universitaire, Bat 110, 91405 Orsay, France; e-mail: christine.tran-quang@curie.fr.