A Bispecific DLL3/CD3 IgG-Like T-Cell Engaging Antibody Induces Antitumor Responses in Small Cell Lung Cancer

Lung cancer remains the number one cause of cancer-related deaths in both men and women in the United States with over 220,000 new cases diagnosed annually (1). Out of these, small cell lung cancer (SCLC), a poorly differentiated neuroendocrine tumor, accounts for roughly 10% to 15% of cases (2). SCLC is the most lethal and aggressive subtype of lung carcinoma with 5-year survival rates below 7% characterized by highly chemotherapy-resistant recurrence within 12 months in the majority of patients (2).

Delta-like ligand 3 (DLL3) is a member of the Notch receptor ligand family that plays a critical role for Notch signaling during embryonal development and is functionally distinct from the related Notch family members DLL1 and DLL4, as demonstrated in animal studies in xenopus laevis and mus musculus. In mice, DLL3 is expressed by postmitotic cells during somitogenesis (3). However, DLL3 is not presented on the surface, but instead interacts with Notch1 in the late endocytic compartment preventing Notch1 from reaching the cell surface (4). This finding has been confirmed in in vitro studies by demonstrating that DLL3 does not activate Notch signaling through binding to Notch expressed on the cell surface, but rather acts as a Notch antagonist by suppressing Notch signaling in a cell-autonomous manner (5). DLL3 expression has been described in SCLC tissue samples (6). In normal tissues, DLL3 is expressed during embryonal development with highest expression in fetal brain, but is absent in adult normal tissues (6). Although in developmental processes DLL3 is expressed intracellularly, in SCLC tumors with DLL3 overexpression, DLL3 escapes to the cell surface which makes it targetable with antibody-based therapies (6). IHC studies indicate a high prevalence of DLL3 expression in SCLC tissues from Caucasian and Japanese patients, with 76% to 88% of tumors comprising at least 1% of DLL3-positive tumor cells and 32% to 67% of patients with tumors comprising more than 50% positive tumor cells (7–9).

Bispecific T-cell engagers represent a promising class of antibody-based immunotherapy. These engineered molecules are designed to induce the formation of a cytolytic synapse in an MHC-independent manner by binding concomitantly to a respective antigen on the cell surface of tumor cells and to CD3 on T-cells, and direct their cytolytic activity selectively to the tumor cells. After formation of the cytolytic synapse, the T-cells produce perforin and granzyme B, leading to apoptosis of the tumor cells. Activation of T-cells leads to transient release of cytokines, which engages other immune cells and broadens the immune response against the tumor tissue leading to conversion of a noninflamed (cold) to an inflamed (hot) tumor environment, infiltration and proliferation of T-cells, and serial killing of tumor cells (10–13). After recent clinical successes of bispecific T-cell engagers with a short half-life (BiTE) for the treatment of hematologic malignancies (14, 15), the next generation of T-cell engagers incorporating half-life extension for increased dosing convenience for patients for the treatment of solid tumors is emerging (13, 16–18).

In this article, we describe the preclinical profile of a novel half-life extended DLL3/CD3 IgG-like T-cell engager (DLL3/CD3 ITE). DLL3/CD3 ITE monotherapy treatment induces potent and strictly DLL3-dependent lysis of tumor cells and T-cell infiltration into tumor tissue leading to complete tumor regression and an inflamed tumor environment in vivo.

Single B-cell technology from mice immunized with the recombinant extracellular domain of human DLL3 generated the anti-DLL3 antibody. The murine anti-CD3 antibody described previously was the source of the anti-CD3 arm used in the test article (19). Humanization and further sequence optimization of both variable regions was performed using a Fab expression vector system according to methods previously described (20). In brief, closely matching human germlines identified in silico and libraries of Fab variants based on these germlines were evaluated for binding to the target, and percent human and Epivax (in silico predictive tool for potential immunogenicity) scores. Subsequently, the optimized variable regions with drug-like properties (high percentage of human sequence, minimal chemical liabilities, and reduced immunogenicity potential) were formatted in an ITE-bispecific expression construct using pTT5 expression vector (National Research Council, Canada) with common molecular biology techniques. The ITE is a two-chain heterodimeric bispecific antibody with Fab domains that contain polypeptide linkers to assure correct light-heavy chain pairing and an Fc domain engineered for heterodimerization and efficient purification (21). The DLL3/CD3 ITE produced in CHO-E (National Research Council, Canada) cells by transient transfection with DLL3:CD3 DNA plasmid mass ratio of 1:3 enabled efficient purification of heterodimeric species as follows. After 10-day culturing, the supernatant was harvested and purified by Protein A affinity chromatography using MabSelect column (GE Healthcare). The bispecific antibody was further purified to homogeneity by cation exchange chromatography using a Poros 50 HS column (Applied Biosystems). The purified protein was concentrated and stored in 50 mmol/L sodium acetate and 100 mmol/L NaCl, pH 5.0 buffer. Analytical characterization by intact mass spectrometry, SDS-PAGE, and size-exclusion chromatography showed that the purified antibody was the expected heterodimeric mass and contained minimal high-molecular-weight aggregate (less than 1%) and no contaminating homodimeric species.

All cell lines were obtained from the ATCC and cultured as described in the supplier's description. Recombinant human DLL1, DLL4, and cynomolgus monkey DLL3-expressing HEK293 cell lines were generated by transfection of HEK293 cells using Lipofectamine (Thermo Fisher Scientific) with pcDNA3.1⁽⁺⁾ (Thermo Fisher Scientific) containing the coding sequences including an N-terminal FLAG tag between the signal sequence and the extracellular domain. All cell lines were tested to exclude mycoplasma contamination and maintained in culture for a maximum of 20 passages.

RNA was isolated using the TRI Reagent (Sigma) and RNeasy Mini Kit (Qiagen). Five hundred nanograms RNA was subjected to library preparation using the TruSeq RNA Library Prep Kit v2 with poly-A selection (Illumina). The Library was then multiplexed and sequenced using a HiSeq device (Illumina) using paired end sequencing. All human SCLC tissues were obtained after receiving written-informed consent from the patients.

Fresh buffy coats from healthy volunteers were obtained from the Austrian Red Cross after receiving written-informed consent in accordance with the Declaration of Helsinki and with approval of the federal ethical committee in Austria. Peripheral blood mononuclear cells (PBMC) were purified from buffy coats by density gradient centrifugation. T-cell subsets were purified from PBMCs using the respective T-cell subset isolation kits (Miltenyi Biotech) according to the manufacturer's instructions. Catalog numbers of T-cell subset kits are listed in Supplementary Table S1.

Cells were incubated in the presence of DLL3/CD3 ITE on ice for 25 minutes at 4°C. Subsequently, cells were washed 2 times in FACS buffer and stained with phycoerythrin-conjugated goat anti-human IgG secondary antibody (Sigma-Aldrich) diluted 1:200 in FACS buffer for 25 minutes at 4°C. Samples were measured by flow cytometry on an FACS Canto II instrument (Becton Dickinson) and analyzed with FlowJo V10 software (Becton Dickinson).

Quantification of DLL3 cell surface molecules expressed on SCLC cell lines was performed by flow cytometry using the QIFIKIT (Agilent) according to the manufacturer's instructions.

Cytotoxicity, activation, degranulation, and proliferation of T-cells, and cytokine secretion were analyzed in multiparametric cytotoxicity assays.

Redirected T-cell cytotoxicity was assessed by quantification of lactate dehydrogenase (LDH) concentrations in the cell culture supernatant using the Cytotoxicity Detection Kit^PLUS (Sigma-Aldrich) according to the manufacturer's instructions. Monitoring of specificity of lysis in the presence of DLL3-positive and DLL3-negative cells was assessed by monitoring of depletion of cells labeled with carboxyfluorescein diacetate succinimidyl ester (CFDA-SE) or CellTrace Violet (CTV; both Thermo Fisher Scientific).

For analysis of T-cell proliferation, the PBMCs and T-cells were labeled with CFDA-SE (Becton Dickinson), and T-cells were additionally stained with anti-CD3 (BioLegend). Cytokine concentrations were analyzed from supernatants of cytotoxicity assays using the V-PLEX Assays (Mesoscale Discovery) following the manufacturer's instructions.

All animal studies were approved by the internal ethics committee and the responsible local governmental committee. Antitumor activity was evaluated in 8-week-old female NOD.CgPrkdcscidIl2rgtm1Sug/JicTac mice (Taconic). Note that 2.5 × 10⁶ SHP-77 cells were suspended in 1 × PBS/5% FCS and injected subcutaneously into the right dorsal flank of animals that had been sublethally irradiated with 2 Gy. Thirteen days later, when tumors reached a size between 71 and 118 mm³, mice were randomized in groups, and 2 × 10⁷ in vitro–expanded human T-cells (T-cell isolation, activation/expansion kits from Miltenyi Biotec) were injected into the peritoneal cavity. Treatment with DLL3/CD3 ITE was initiated 3 days after T-cell injection (Supplementary Fig. S1). Tumor size was measured by an electronic caliper and calculated according to the formula “tumor volume = length × diameter² × π/6.” Regressions were defined as a relative tumor volume < 1 when normalized to the tumor volume at the start of treatment. One-sided nonparametric Mann–Whitney–Wilcoxon U tests were applied to compare the treatment group with the vehicle group. The level of significance was fixed at α = 5%.

Mouse tumor tissues were harvested when animals were terminated, fixed overnight in 4% neutral-buffered formaldehyde, and embedded in paraffin. Note that 2-μm-thick serial sections were prepared on a microtome, put on glass slides, and subsequently dewaxed. SHP-77 tumor tissues were stained with anti-human CD3 (Roche Diagnostics), anti-human CD4 (Roche Diagnostics), anti-human CD8 (Roche Diagnostics), anti-human PD-1 (Cell Signaling Technology), anti-human LAG-3 (Novus Biologicals), anti-human PD-L1 (Abcam), anti-human cleaved Caspase 3 (CST), or anti-human DLL3 (Ventana SP347) antibodies. Two-sided nonparametric Mann–Whitney–Wilcoxon U tests were applied to compare the treatment group with the vehicle group.

Serum concentrations of DLL3/CD3 ITE were determined in an ELISA-based assay using anti-human IgG capture (Novus) and detection (MyBiosource) reagents. Data were treated as a naïve pool and fitted to a linear two-compartment model using a log-additive error model in Phoenix 64 WinNonlin 7.0.

The DLL3/CD3 ITE is based on a heterodimeric IgG scaffold that incorporates flexible peptide linkers between light and heavy chains and is designed to bind concurrently to DLL3 on tumor cells and CD3 on T-cells (Fig. 1A).

Binding specificity of DLL3/CD3 ITE for human DLL3 and CD3 was verified by using DLL3-positive SCLC cell lines with endogenous DLL3 expression, DLL3-negative cell lines, purified T-cells, and recombinant cell lines expressing human DLL1 or DLL4, the closest homologs of DLL3. Cell-bound DLL3/CD3 ITE was detected by flow cytometry. DLL3/CD3 ITE binds to DLL3-positive SCLC cell lines SHP-77 and NCI-H82, and CD3-positive purified human T-cells, but not to DLL3-negative cell lines RKO-E6, HL-60, and recombinant HEK293 cell lines expressing human DLL1 or DLL4 (Supplementary Fig. S2). DLL3/CD3 ITE–induced T-cell redirected lysis is strictly dependent on DLL3 expression and presence of T-cells (Fig. 1B). Donor dependency was observed in a cytotoxicity assay with PBMCs from 23 different human donors and the SCLC cell line SHP-77 with a geometric mean EC₅₀ of 5.5 ng/mL with a range from 0.7 to 20.8 ng/mL (Fig. 1C).

The potency of the DLL3/CD3 ITE–induced lysis of SHP-77 cells depends on the E:T ratio as indicated by the range of EC₅₀ values ranging from 1 ng/mL at an E:T of 30:1 to 129 ng/mL at an E:T of 2:1 with maximal lysis activity observed at an E:T ratio of ≥ 10:1 (Fig. 1D). The selectivity of the DLL3/CD3 ITE–induced lysis of DLL3-positive cells was further evaluated by coculture of PBMCs, CTV-labeled DLL3-positive SHP-77 cells, and CFDA-SE–labeled DLL3-negative HL-60 cells at a ratio of 20:1:1, and increasing concentrations of DLL3/CD3 ITE for 72 hours. The number of viable SHP-77 and HL-60 cells was determined by flow cytometry. The DLL3/CD3 ITE did not induce lysis of DLL3-negative HL-60 cells, when cocultured with DLL3-positive SHP-77 cells. The slight reduction of viable target cell numbers at high concentrations of DLL3/CD3 ITE is considered negligible compared with the strong reduction of SHP-77 cells (Fig. 1E). The DLL3/CD3 ITE lyses cells through induction of apoptosis of SHP-77 cells, as demonstrated by the increase of Annexin V binding and PI incorporation in CTV-labeled SHP-77 cells after cocultivation with PBMCs and increasing concentration of DLL3/CD3 ITE for 24 hours (Fig. 1F). Further analysis of the mode of action of the DLL3/CD3 ITE confirmed DLL3-dependent secretion of MCP-1 (Supplementary Fig. S3A) and IFNγ (Supplementary Fig. S3B), and proliferation of T-cells (Fig. 1G).

In further cytotoxicity experiments, we confirmed that CD4⁺ as well as CD8⁺ T-cell subsets contributes to the DLL3/CD3 ITE–induced T-cell redirected lysis of SHP-77 cells in a time-dependent manner. T-cell redirected lysis by PBMCs (Fig. 2A), pan T-cells (Fig. 2B), and CD8⁺ T-cells (Fig. 2D) became already apparent after 48 hours when CD4⁺ T-cells (Fig. 2C) still showed only minor lysis. After 72 hours, both CD4⁺ and CD8⁺ T-cells contributed comparably with the lysis of SHP-77 cells. In coculture experiments with purified subsets of T-cells, we demonstrated that CD4⁺ central memory (CD4⁺/CD45RA⁻/CD197⁺), CD4⁺ effector memory (CD4⁺/CD45RA⁻/CD197⁻), CD8⁺ memory (CD8⁺/CD45RA⁻), CD8⁺CD45RA⁺ (CD8⁺/CD45RA⁺/CD197⁺), as well as naïve T-cells (CD4⁺/CD45RA⁺ and CD8⁺CD45RA⁺) are involved in the DLL3/CD3 ITE–induced T-cell redirected lysis of SHP-77 cells (Fig. 2E–I).

DLL3 mRNA expression was analyzed by RNA sequencing in primary tumor tissues from 20 patients with SCLC and eight SCLC cell lines. The mRNA expression ranged from 30 to 370 TPM (transcripts per million) in tumor tissues and 20 to 230 TPM in the cell lines (Fig. 3A). Cell pellets of three representative SCLC cell lines with low (NCI-H2286), medium (NCI-H82), and high (SHP-77) DLL3 expression levels were generated and analyzed by an IHC assay (SP347) qualified for assessing DLL3 protein expression in early clinical trials with Rova-T (9). DLL3 protein expression was detectable for the SHP-77 and NCI-H82 cell pellets, but not the NCI-H2286 cell pellet (Fig. 3B). To analyze DLL3/CD3 ITE–induced T-cell redirected lysis of the eight SCLC cell lines, each cell line was cocultured with PBMCs and increasing concentrations of DLL3/CD3 ITE. The DLL3/CD3 ITE–induced T-cell redirected lysis of all tested SCLC cell lines in a dose-dependent manner, but the level of DLL3/CD3 ITE–induced T-cell redirected lysis depended on the tumor cell line and the DLL3 expression levels, indicated by the range of EC₅₀ concentrations from 5.9 to 24 ng/mL (Fig. 3C; Supplementary Table S3). Immunohistochemically detectable DLL3 levels correlated with higher DLL3/CD3 ITE potency in the SHP-77 and NCI-H82 cell lines. However, lysis activity could also be detected using the NCI-H2286 cell line for which no DLL3 protein could be detected immunohistochemically, indicating the potential of DLL3/CD3 ITE to induce lysis of cells with very low DLL3 expression levels.

In vivo efficacy studies were conducted in a subcutaneous human SHP-77 xenograft model in CD3⁺ T-cell humanized mice (Fig. 4A and B) where the DLL3/CD3 ITE was administered in a weekly (q7d) regimen supported by its long half-life of 20 days in C57/BL6 mice (Fig. 4C).

In a first efficacy study (Study I), the DLL3/CD3 ITE was administered at doses of 0.025 and 0.25 mg/kg i.v., and tumor growth was monitored until day 21. Although treatment with 0.025 mg/kg induced tumor regression in 3 out of 5 animals at day 21 (P < 0.05), a stronger antitumor response was observed with a dose of 0.25 mg/kg and tumor regression in all animals (5/5) already at day 9. Notably, four out of five tumors were still in regression at the end of the study (P < 0.05; Fig. 4A). In a second efficacy study (Study II) with a different human T-cell donor, DLL3/CD3 ITE monotherapy induced a statistically significant tumor growth inhibition (P < 0.05 at day 15) compared with the vehicle control group. Sustained tumor regressions could be observed for most of the tumors from day 15 (7/9 animals) until day 36 (8/9 animals). Although treatment was stopped after 4 cycles (day 22), tumor regression was maintained. Two out of the eight animals with tumor regression still showed complete tumor clearance (tumor not palpable anymore) at the end of the study (day 36), 14 days after the last dose (Fig. 4B).

Tumor tissues from mice treated with vehicle or 0.25 mg/kg DLL3/CD3 ITE in Study II were collected at the day of termination, and tissue sections were stained with anti-CD3, PD-1, and PD-L1 antibodies. Treatment with DLL3/CD3 ITE induced infiltration of CD3⁺ T-cells into the SHP-77 xenograft tissue. Infiltrating CD3⁺ T-cells upregulated PD-1 and LAG-3, and SHP-77 tumor cells upregulated PD-L1 (Fig. 5).

In a following study (Study III), we examined the ability of the DLL3/CD3 ITE to modulate the inflammatory environment in the tumor tissue. SHP-77 tumor tissues from animals were collected 8 days after administration of one dose of 0.25 mg/kg ITE i.v., and sections of formalin-fixed and paraffin-embedded tissues were immunohistochemically stained with anti-CD3, anti-CD4, anti-CD8, anti–PD-1, anti–PD-L1, and anti-cC3 (cleaved caspase 3) antibodies. A statistically significant increase in CD4⁺ and CD8⁺ T-cells (Fig. 6A–C) in DLL3/CD3 ITE–treated animals compared with the vehicle-treated animals was observed. In this study, we did not see an increase of PD-1 expression (Fig. 6D) and only a nonstatistically significant trend of increased PD-L1 (Fig. 6E) expression. In the tumor cells, a statistically significant increase of cleaved caspase 3 was detected (Fig. 4F). The proximity of the apoptotic tumor cells to the infiltrating T-cells was visualized by costaining of tumor tissues for cleaved caspase 3 and CD3 (Fig. 4G).

The impact of dexamethasone on the activity of the DLL3/CD3 ITE was assessed by adding increasing concentrations of dexamethasone ranging from 1 to 1,000 ng/mL to the multiparametric cytotoxicity assay simultaneously with 30 μg/mL DLL3/CD3 ITE. Treatment with dexamethasone led to a dose-dependent reduction of DLL3/CD3–induced T-cell redirected lysis of DLL3-positive SHP-77 cells with one out of three PBMC donors (Supplementary Fig. S4A). The DLL3/CD3 ITE–induced upregulation of CD25 on CD4⁺ T-cells was also reduced with one out of three donors, whereas dexamethasone did not have an effect on CD25 expression on CD8⁺ T-cells (Supplementary Fig. S4B and S4C). The DLL3/CD3-induced secretion of IFNγ (Supplementary Fig. S4D) and MCP-1 (Supplementary Fig. S4E), by PBMCs in presence of DLL3-positive SHP-77 cells, was reduced in the presence of dexamethasone with all three tested donors.

To support pharmacokinetic and safety assessment in nonhuman primates, the DLL3/CD3 ITE was designed to be cross-reactive to cynomolgus monkey DLL3 and CD3. The ability of the DLL3/CD3 ITE to induce T-cell redirected lysis of recombinant cynomolgus monkey DLL3-expressing HEK293 cells in presence of cynomolgus monkey PBMCs is shown in Supplementary Fig. S5. Pharmacokinetic studies showed that the half-life of the DLL3/CD3 ITE is 10 days.

Despite the high unmet medical need for new treatments for SCLC and the highly attractive tumor-specific expression of DLL3 for a targeted therapy, the development of such a therapy has yet to be realized. The first approach for targeting DLL3 with an antibody-based approach was Rova-T (rovalpituzumab tesirine), a humanized anti-DLL3 monoclonal antibody conjugated to a DNA-damaging pyrrolobenzodiazepine dimer toxin (6, 7). In a phase I trial including patients with small cell lung or large cell neuroendocrine tumors that have previously relapsed after platinum-based chemotherapy, 18% of patients had a confirmed objective response rate and a manageable safety profile (7). However, in a phase II study, Rova-T showed only modest activity [median overall survival (OS) of 5.7 months in patients with high DLL3 expression vs. 5.6 months in all patients] with associated toxicities of grades 3 to 5 in third-line and beyond (3L+) SCLC patients with DLL3-positive tumors (22). The toxicities observed with Rova-T, such as thrombocytopenia, nausea, fatigue, vomiting, edema, and effusions (7), have also been observed with vadastuximab talirine, an anti-CD33 antibody–drug conjugate (23), and are considered drug- but not DLL3-related.

T-cell engagers have a completely different mechanism of action to antibody–drug conjugates and have shown high potency and efficacy with a promising safety profile in hematologic malignancies (14, 15). However, the development of bispecific T-cell engagers for solid tumors presents new challenges. Firstly, given the potency of the mode of action, a tumor selectively expressed cell surface antigen is required to ensure optimal efficacy and safety. So far, bispecific T-cell engagers for solid tumors such as solitomab and MEDI-565 for the treatment of gastrointestinal cancers did not result in a sufficient therapeutic window in clinical trials (24, 25), as the antigens EpCAM (epithelial cell adhesion molecule) and CEA (carcinoembryonic antigen–related cell adhesion molecule 5), respectively, are widely expressed. Secondly, the BiTE molecules have a very short half-life and have to be administered by continuous intravenous infusion, for example. As such, opportunities to extend the half-life of T-cell engagers while maintaining the potent pharmacological effect have been sought to improve the dosing convenience for patients.

In this article, we describe a novel approach for targeting the tumor-selective DLL3 antigen and treating DLL3-expressing SCLC using T-cell redirection. We have designed a highly DLL3-selective novel IgG-like T-cell engaging bispecific antibody with an optimal pharmacologic profile, and antibody-like in vivo stability and half-life. In contrast to AMG 757 (26), a half-life extended BiTE consisting of a bispecific scFv fused to an Fc, the DLL3/CD3 ITE has an IgG-like architecture which may pose a lower immunogenicity risk in humans. DLL3/CD3 ITE induces potent T-cell redirected lysis of DLL3-positive SCLC cell lines in a dose, E:T ratio, and time-dependent manner, with EC₅₀ values in the low ng/mL range by redirecting CD4⁺ and CD8⁺ T-cells toward DLL3-postitive SCLC cells without lysing DLL3-negative target cells. As a result of the formation of the cytolytic synapse, T-cells secrete proinflammatory cytokines and proliferate. In vivo, DLL3/CD3 ITE monotherapy leads to infiltration of T-cells into tumor tissue and converts a noninflamed (cold) into an inflamed (hot) tumor environment, leading to tumor cell apoptosis and resulting in strong tumor regressions with complete tumor eradication in several mice in a human CD3⁺ T-cell humanized SHP-77 xenograft model.

In the phase II TRINITY study with Rova-T, cutoffs for analysis of DLL3 expression in SCLC tissues of patients were defined as DLL3-positive when tissues contained ≥25% of DLL3-positive tumor cells, and as DLL3-high when tissues contained ≥75% of DLL3-positive tumor cells (22). In our study, we could detect DLL3 expression in cell pellets of SCLC cell lines expressing more than 500 DLL3 molecules per cell using the same qualified IHC assay (Ventana, SP347; ref. 9) previously validated for the determination of DLL3 expression levels in the Rova-T phase II study (22). In in vitro cytotoxicity assays, cell lines with immunohistochemically detectable DLL3 levels correlate with higher DLL3/CD3 ITE potency, suggesting that the detection of DLL3 expression using the SP347 IHC assay may be a promising biomarker for patient selection that should be evaluated in clinical trials.

The FDA recently approved the PD-1 inhibitor pembrolizumab for the treatment of patients with SCLC who experienced disease progression after two or more prior lines of therapy based on two clinical studies showing a 19% overall objective response rate (27). In our preclinical studies, we observed an upregulation of PD-(L)1 and LAG-3 expression in vivo as a result of the inflamed environment induced by treatment with the DLL3/CD3 ITE, which is in line with observations from preclinical studies with other T-cell engagers, e.g., CEA BiTE, CEA TCB, AMG330, and Her2-TDB (16, 28–30). Furthermore, in a clinical trial, evidence of enhanced activity of RO6958688, a CEA-targeted T-cell engager, in combination with atezolizumab with a manageable safety profile has been reported (18). In future studies of the DLL3/CD3 ITE, it will be important to look at the impact on other immune checkpoint molecules.

As activation of T-cells and cytokine secretion is a characteristic of the mode of action of T-cell engagers, which in some cases can also cause clinical side effects such as fever and hypotension, it might be desirable to reduce cytokine secretion as much as possible without negatively influencing the cytotoxic potential of T-cells. Glucocorticoids have been described to inhibit inflammatory responses such as cytokines (31), and a preclinical study using the glucocorticoid dexamethasone in combination with blinatumomab reported that dexamethasone can qualify as a potential comedication for TcE therapies (32). The data obtained in this study reveal that dexamethasone can reduce DLL3/CD3 ITE–induced secretion of IFNγ and MCP-1, and therefore support the use of dexamethasone to treat potential infusion reactions or cytokine release syndromes associated with DLL3/CD3 ITE treatment, but the treatment could result in reduced cytotoxic potential of DLL3/CD3 ITE.

According to the DLL3-expression profile, the DLL3/CD3 ITE has the potential to be developed in additional indications with high unmet medical need, such as glioma (33), medullary thyroid cancer (34), gastrointestinal neuroendocrine malignancies (35), dispersed neuroendocrine tumors of the pancreas (34), small cell bladder cancer (36), Merkel-cell carcinoma (37), and neuroendocrine prostate cancer (38).

Taken together, the DLL3/CD3 ITE provides a novel drug candidate for the treatment of SCLC as monotherapy with upside potential in other neuroendocrine tumor types.

S. Hipp reports grants from Basisprogramm grants of the Austrian Research Promotion Agency (FFG; 860968, 869530, and 875923) and personal fees from Boehringer Ingelheim (employment) during the conduct of the study; in addition, she has filed a patent application for DLL3-CD3 bispecific antibodies. The filed patent application is owned by Boehringer Ingelheim and relates to generation and use of DLL3-CD3 bispecific antibodies. V. Voynov reports personal fees from Boehringer Ingelheim (employment) during the conduct of the study; in addition, he has filed a patent application for DLL3-CD3 bispecific antibodies. The filed patent application is owned by Boehringer Ingelheim and relates to generation and use of DLL3-CD3 bispecific antibodies. B. Drobits-Handl reports grants from Basisprogramm grants of the Austrian Research Promotion Agency (FFG; 860968, 869530, and 875923) and personal fees from Boehringer Ingelheim (employment) during the conduct of the study. C. Giragossian reports personal fees from Boehringer Ingelheim (employment) during the conduct of the study. F. Trapani reports grants from Austrian research Promotion Agency (FFG; 860968, 869530, and 875923) and personal fees from Boehringer-Ingelheim (employment) during the conduct of the study. J.M. Scheer has filed a patent application for DLL3-CD3 bispecific antibodies. The filed patent application is owned by Boehringer Ingelheim and relates to generation and use of DLL3-CD3 bispecific antibodies. P.J. Adam reports grants from Basisprogramm grants of the Austrian Research Promotion Agency (FFG; 860968, 869530, and 875923) and personal fees from Boehringer Ingelheim (employment) during the conduct of the study; in addition, he has filed a patent application for DLL3-CD3 bispecific antibodies. The filed patent application is owned by Boehringer Ingelheim and relates to generation and use of DLL3-CD3 bispecific antibodies. No potential conflicts of interest were disclosed by the other authors.

S. Hipp: Conceptualization, supervision, methodology, writing-original draft, writing-review and editing, devised and designed the main conceptual idea of the study and wrote the manuscript. V. Voynov: Supervision, methodology, designed the molecule and established the production process. B. Drobits-Handl: Supervision, methodology, writing-review and editing, designed and supervised in vivo studies, and analyzed, assembled, and interpreted data. C. Giragossian: Supervision, methodology, designed and supervised PK studies, and analyzed, assembled, and interpreted data. F. Trapani: Supervision, methodology, designed and supervised IHC studies, and analyzed, assembled, and interpreted data. A.E. Nixon: Supervision, methodology, supported the molecule design principles. J.M. Scheer: Supervision, methodology. P.J. Adam: Supervision, writing-review and editing, supported the design and conceptual idea of the study and contributed to the writing and critical evaluation of the manuscript.

We thank Christoph Albrecht, Ilse Apfler, Kathrin Bauer, Oliver Bergner, Erica Bolella, Richard Liedauer, Irene Schweiger, Abdallah Souabni, and Christian Walterskirchen for their experimental support and help for this study, and all colleagues from Biotherapeutics Discovery for all aspects of molecule discovery.

This study was supported by the Basisprogramm grants of the Austrian Research Promotion Agency (FFG; 860968, 869530, and 875923).

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.