Why DLL3 Is the Right Target in ES-SCLC

Small cell lung cancer accounts for approximately 13–20% of all newly diagnosed lung cancers and is defined by near-universal progression to extensive-stage disease, where median survival without treatment is as short as 2–4 months. The disease responds initially to platinum/etoposide-based chemotherapy but rapidly relapses into a chemorefractory setting with no standard salvage options — a clinical reality that makes the identification of a reliable tumor-associated surface antigen critical.

Delta-like ligand 3 (DLL3) is a single-pass type I transmembrane protein and atypical Notch pathway ligand. Multiple Amgen patent filings establish that approximately 86% of SCLC tumors express DLL3 by RNA-seq — a figure sourced from Giffin et al., Clin Cancer Res 27(5):1526–1537, 2021 — with expression predominantly on the tumor cell surface. Critically, DLL3 expression in normal adult tissues is restricted to cytoplasmic compartments of neurons, pancreatic islet cells, and pituitary cells. This selective surface expression pattern is the foundation of the therapeutic window that DLL3-targeting agents exploit.

DLL3 expression is also documented in large cell neuroendocrine carcinoma (LCNEC), neuroendocrine prostate cancer, melanoma, and glioblastoma, according to Amgen and Chinese biotech filings. Filings from the Board of Regents, University of Texas System further contextualize DLL3 within a molecular subtyping framework — SCLC-A, SCLC-N, SCLC-P, and SCLC-I subtypes — with DLL3 expression correlating with the SCLC-A subtype, suggesting that patient selection may eventually move beyond simple IHC positivity toward molecular subtype stratification.

The clinical patient selection threshold operationalized in Amgen’s WO and AU patent filings is ≥25% of SCLC cells expressing DLL3 at ≥2+ IHC intensity — a translational biomarker criterion consistent with clinical trial eligibility criteria and with implications for companion diagnostics development. According to WIPO filing records, Amgen’s multi-jurisdictional prosecution strategy for these patient selection claims spans at least 12 distinct jurisdictions.

Tarlatamab: The BiTE That Reached the Clinic

Tarlatamab (AMG 757) is a DLL3 × CD3 bispecific T-cell engager developed by Amgen Research (Munich) GmbH and Amgen Inc. that has been confirmed in a Phase 1 clinical trial for the treatment of SCLC. The mechanism is T-cell redirection: one binding arm engages DLL3 on the tumor cell surface while the second arm binds CD3ε on T cells, forming an immunological synapse that drives T-cell-mediated cytolysis of tumor cells independent of MHC-peptide presentation.

“A bispecific protein (AMG 757) targeting DLL3 and CD3 was used in a Phase 1 clinical trial for the treatment of SCLC” — confirmed in an Amgen Japan patent filing covering dosing regimens including step dosing from 0.3 mg initial dose escalating to 3–100 mg doses, once every two weeks.

Patent filings from Amgen span at least 12 jurisdictions — WO, US, JP, AU, CN, TW, ID, BR, CL, AR, ES, IN — covering core bispecific construct claims, dosing regimen patents, combination therapy, patient selection criteria, route of administration, and specific indication extensions. A key 2021 ES-jurisdiction filing from Amgen Research (Munich) GmbH defines the precise CDR sequences of the DLL3-binding arm (CDR-H1 to CDR-H3 as SEQ ID NO: 31–33; CDR-L1 to CDR-L3 as SEQ ID NO: 34–36) and specifies that the DLL3 epitope is contained within a region depicted in SEQ ID NO: 258–260 of the extracellular domain.

Step dosing regimens are codified in Amgen JP and CL patents: an initial dose of 0.3 or 1 mg on Day 1 of Cycle 1, escalating on Day 8 and Day 15, with maintenance doses of 3–100 mg Q2W thereafter. These detailed regimens are characteristic of active clinical development informed by observed cytokine release syndrome (CRS) risk with T-cell engagers — a well-documented safety consideration for BiTE-class molecules as noted by the FDA in its guidance on T-cell engager therapies.

A 2026 Amgen WO filing covers subcutaneous delivery of the anti-DLL3 agent with defined dosing schedules of Q2W, Q3W, or Q4W — signaling active development of a more convenient administration route potentially relevant for maintenance therapy settings. The breadth of geographic filings and the progression from core construct claims (2021) through dosing optimization (2023), combination therapy (2024), brain metastasis indication (2025), and subcutaneous administration (2026) is consistent with a program advancing through late-stage clinical development.

ADCs, CAR-T Cells, and Trispecific Proteins: The Broader DLL3 Pipeline

Beyond tarlatamab, the DLL3-targeting pipeline encompasses three distinct modalities — antibody-drug conjugates, CAR-T cell therapies, and trispecific proteins — each representing a different approach to exploiting DLL3’s selective surface expression on SCLC tumor cells.

Antibody-Drug Conjugates (ADCs)

AbbVie Stemcentrx LLC holds the original DLL3 ADC IP portfolio, covering rovalpituzumab tesirine (SC16LD6.5) — an anti-DLL3 antibody conjugated to a pyrrolobenzodiazepine (PBD) warhead — with filings in UY, BR, MX, IL, IN, and CO jurisdictions (2017–2018). An AbbVie Stemcentrx IN-jurisdiction patent includes a Phase I clinical study overview and results for SC16LD6.5 in recurrent SCLC and LCNEC, confirming clinical Phase 1 data. The ADC mechanism relies on DLL3-mediated internalization of the antibody-drug conjugate, enabling intracellular delivery of the cytotoxic payload. Multiple AbbVie Stemcentrx filings carry inactive legal status in several jurisdictions, potentially reflecting the known clinical development trajectory of rovalpituzumab tesirine.

New entrants have since emerged in the ADC space. Shanghai Fudan-Zhangjiang Bio-Pharmaceutical Co., Ltd. has filed patents (CN, BR, 2024–2026) disclosing novel anti-DLL3 antibodies with strong internalization activity conjugated to cytotoxic drugs for SCLC and neuroendocrine tumors. Sannai (Three Academies) Therapeutics Research Institute (CN, 2024) discloses anti-DLL3 antibody-PBD derivative conjugates. These next-generation entrants signal that the DLL3 ADC modality has not been abandoned, and potentially create opportunities for novel payloads — such as topoisomerase I inhibitors or RNA polymerase inhibitors — not covered by existing PBD-focused claims.

CAR-T Cell Therapy

Several patent families are directed at DLL3-specific chimeric antigen receptors (CARs) for adoptive cell therapy. Nanjing Legend Biotech Co., Ltd. (CN, 2023–2024) discloses anti-DLL3 CARs incorporating DLL3-binding domains derived from single-domain antibodies (sdAb) or scFv, with full CDR sequence coverage and application to SCLC. Allogene Therapeutics (CN, 2024) discloses DLL3-specific CARs and binding agents for immune cell engineering. Juno Therapeutics, a Bristol-Myers Squibb subsidiary (CN, 2025), provides DLL3-specific fully human antibodies, antibody fragments, and CARs for adoptive cell therapy. All CAR-T signals in this dataset are preclinical; no clinical data for DLL3 CAR-T approaches were recovered.

Trispecific Proteins

Harpoon Therapeutics (CN, 2021 and 2024) discloses DLL3-targeting trispecific proteins of less than 80 kDa comprising a DLL3-binding domain, a CD3-binding domain, and an albumin half-life extension domain. This format distinguishes itself from conventional BiTE constructs by addressing the short half-life of BiTE-class molecules while maintaining T-cell redirecting activity — a potential pharmacokinetic advantage that warrants monitoring for clinical progression signals. Daiichi Sankyo (WO, 2026) has additionally disclosed a combination of an anti-B7-H3 ADC with a DLL3-targeting trispecific protein, optionally with a PD-1/PD-L1 inhibitor and carboplatin, signaling awareness of DLL3-targeting agents as combination partners in a broader oncology portfolio.

Combination Strategies: Checkpoint Inhibitors, Chemotherapy, and Beyond

The most clinically advanced combination strategy pairing tarlatamab with checkpoint inhibition is codified across multiple Amgen jurisdictions. Amgen filings across US, WO, AU, ID, CL, and JP jurisdictions specifically claim combination of an anti-DLL3 agent with anti-PD-1 (480 mg Q4W) or anti-PD-L1 antibody — reflecting the mechanistic rationale that T-cell engager-induced immune activation may be enhanced by simultaneously relieving PD-1/PD-L1-mediated immune suppression within the SCLC tumor microenvironment.

Several Amgen filings additionally claim triplet combinations adding chemotherapeutic agents — carboplatin, cisplatin, and etoposide — suggesting investigation of DLL3-targeting in the first-line or maintenance setting alongside the current standard of care. According to ASCO guidelines, atezolizumab plus carboplatin/cisplatin plus etoposide constitutes the first-line standard for ES-SCLC — the treatment backdrop against which these DLL3-targeting combinations are being developed, as also reflected in separate Genentech and AstraZeneca filings within this dataset.

Harpoon Therapeutics’ trispecific proteins (DLL3 × CD3 × albumin half-life extension domain, <80 kDa) represent a next-generation engager format that may address the short half-life and CRS risk of conventional BiTE constructs. Daiichi Sankyo’s 2026 WO filing discloses a multi-target combination of an anti-B7-H3 ADC with a DLL3-targeting trispecific protein, optionally with PD-1/PD-L1 inhibitor and carboplatin — a strategy targeting both B7-H3 (a neuroendocrine tumor-associated antigen) and DLL3 to potentially address tumor antigen heterogeneity in SCLC. Research on combination immunotherapy strategies in solid tumors, including SCLC, is also tracked by the National Cancer Institute.

The University of Texas System’s molecular subtype-guided approach adds another dimension: filings propose DLL3-targeting agents specifically for SCLC-A subtype patients, while SCLC-N subtype patients may be better addressed with aurora kinase, JAK, or c-Met inhibitors. This stratified approach signals a direction toward precision patient selection beyond simple IHC DLL3 positivity — a development that has implications for clinical trial design and companion diagnostics.

Patent Landscape and Strategic Implications for DLL3-Targeting in SCLC

Amgen’s IP position in DLL3 × CD3 bispecific therapeutics is exceptionally broad in this dataset, spanning core construct claims, dosing regimen claims, combination therapy claims, patient selection criteria (IHC thresholds), route of administration, and specific indication extensions including brain metastases and ES-SCLC. Entrants seeking freedom-to-operate in this space face a dense patent landscape across major jurisdictions — US, WO, JP, AU, CN, TW, ID, BR, CL, ES — that must be carefully navigated.

The patient selection threshold (≥25% DLL3 IHC 2+ expression) claimed in Amgen’s WO and AU filings has direct implications for diagnostic test development and companion diagnostics IP. Developers and CDx partners should monitor these claims closely, as they codify both eligibility criteria and potentially patentable assay parameters. The companion diagnostics regulatory pathway, as outlined by the EMA, requires co-development and co-submission of CDx with the therapeutic — a consideration relevant to any DLL3-targeting program.

ADC development for DLL3 appears to have stalled at the AbbVie/Stemcentrx level — multiple inactive legal statuses in retrieved results — but new Chinese entrants (Shanghai Fudan-Zhangjiang, Sannai Therapeutics) and the distinct PBD payload approach suggest the ADC modality has not been abandoned. This potentially creates opportunities for next-generation ADC payloads not covered by existing claims.

Emerging DLL3 CAR-T and trispecific protein platforms from Chinese biotechs (Nanjing Legend, Allogene, Harpoon) represent potential differentiation from the BiTE paradigm, particularly for patients who progress on tarlatamab. The half-life-extended trispecific format (DLL3 × CD3 × albumin) may address CRS and pharmacokinetic limitations of standard BiTE constructs. The PatSnap Life Sciences intelligence platform tracks emerging assignee activity across all these modalities in real time.