The Correct Patent Family and Why WO2021127088A1 Is Not Tarlatamab
WO2021127088A1 is not the Tarlatamab patent. That filing is assigned to Janssen Biotech and covers CD8/TCR-targeting molecules — not DLL3-targeting BiTE antibodies. The primary patent family covering Amgen’s DLL3×CD3 BiTE construct (Tarlatamab/AMG 757) is WO2017021526, with a priority filing date of August 2016. Key international family members include CA2986848A1, CN114292336A, and JP2022191477A, all protecting the core bivalent BiTE architecture, CDR sequences, and manufacturing methods.
The WO2017021526 family is foundational to understanding the structural and SAR innovations that distinguish Tarlatamab from earlier DLL3-targeting bispecifics. Patent databases such as PatSnap Life Sciences Intelligence allow researchers to map the full family, track continuation filings, and identify which claims cover the approved drug versus earlier-generation constructs. According to WIPO, the PCT filing strategy used here is standard practice for biologics claiming priority from a US provisional application.
WO2021127088A1 is a Janssen Biotech filing covering CD8/TCR-targeting molecules. The correct Amgen patent family for Tarlatamab (AMG 757) is WO2017021526 (priority August 2016), with family members CA2986848A1, CN114292336A, and JP2022191477A.
DLL3-Binding Domain SAR: Epitope Selection and Affinity Optimisation
Tarlatamab’s DLL3-binding domains are engineered to target an extracellular epitope of DLL3 that is highly expressed on SCLC cells but minimally expressed on normal tissues, a selectivity prerequisite for any viable oncology bispecific. The patent family describes multiple DLL3-binding clones subjected to systematic CDR mutagenesis to optimise three interdependent parameters: binding affinity to human DLL3, species cross-reactivity to cynomolgus monkey DLL3, and biophysical developability (thermal stability, aggregation propensity).
Tarlatamab’s DLL3-binding domains achieve sub-nanomolar to low nanomolar binding affinity (KD approximately 0.5–2 nM) to human DLL3, engineered through systematic CDR mutagenesis across multiple clones described in the WO2017021526 patent family.
The final DLL3-binding affinity for the approved compound (AMG 757) is reported at a KD of approximately 0.5–2 nM — high-affinity binding chosen to ensure robust tumor cell engagement even at moderate DLL3 expression levels. SAR-guided clone selection also prioritised cross-reactivity with cynomolgus monkey DLL3 to enable toxicology studies in a relevant non-human primate model, a regulatory requirement tracked by agencies such as the FDA for bispecific antibody programmes.
The patent family also covers manufacturing and quality attributes: the CDR-optimised clones are evaluated for expression yield in CHO systems, glycosylation profiles, and forced-degradation stability — all factors that influence which clone advances to IND-enabling studies. For practitioners mapping this landscape, PatSnap Eureka can retrieve the full CDR sequence tables and clone comparison data from the WO2017021526 family.
CD3 Engagement Affinity Tuning: Balancing Efficacy Against Safety
The CD3-binding domain in Tarlatamab is engineered to a deliberately moderate affinity — KD approximately 5–10 nM to CD3ε — because the therapeutic window of any T-cell engager is determined as much by what the CD3 arm does not do as by what it does. At high CD3 affinity, T cells can be activated non-specifically in the absence of tumor antigen, driving cytokine release syndrome (CRS) and on-target, off-tumor toxicity.
Tarlatamab’s CD3-binding domain is tuned to a moderate affinity of KD approximately 5–10 nM to CD3ε, ensuring T-cell activation occurs only when the BiTE molecule is simultaneously cross-linked to DLL3-expressing tumor cells, thereby reducing cytokine release syndrome risk.
Monovalent CD3 engagement (a single scFv) reinforces this safety strategy: T-cell activation requires the BiTE to be physically cross-linked between a DLL3+ tumor cell and a CD3+ T cell. The CD3-binding domain is derived from humanised anti-CD3 antibodies with CDR optimisation for reduced immunogenicity — a key developability parameter tracked by regulatory bodies including the EMA for bispecific antibody approvals in oncology.
“Monovalent CD3 engagement ensures T-cell activation occurs only when the BiTE is cross-linked to DLL3+ tumor cells — the structural basis for Tarlatamab’s manageable CRS profile, with most events being Grade 1–2.”
The step-up dosing strategy (10 µg in cycle 1 → 100 µg in cycle 2 onwards) further mitigates CRS risk by allowing T-cell priming at sub-therapeutic exposure before escalating to the full pharmacodynamic dose. This clinical strategy is directly enabled by the moderate CD3 affinity built into the molecule’s structure — a higher-affinity CD3 arm would compress the therapeutic window and make step-up dosing insufficient to manage initial cytokine release.
Analyse CD3 affinity tuning strategies across the full bispecific antibody patent landscape with PatSnap Eureka.
Explore Bispecific Antibody Patents in PatSnap Eureka →Bivalency Architecture Data and Half-Life Extension Mechanism
Tarlatamab’s bivalent DLL3 / monovalent CD3 architecture is the defining structural innovation of the WO2017021526 patent family. Two DLL3-binding scFv domains are arranged in tandem at the N-terminus, connected by (G₄S)ₙ flexible linkers (typically n=3–4) to allow independent domain folding, followed by an Fc domain and a single CD3ε-binding scFv at the C-terminus.
Tarlatamab employs a tandem scFv-Fc-scFv molecular format: [DLL3-scFv₁]–[Linker]–[DLL3-scFv₂]–[Fc-silenced (L234A/L235A/P329G)]–[CD3-scFv], providing bivalent DLL3 engagement with an extended serum half-life of approximately 5–7 days via FcRn recycling.
Half-Life Extension via FcRn Recycling
The Fc domain — even when silenced for effector function by mutations L234A/L235A/P329G — retains its capacity to bind the neonatal Fc receptor (FcRn) at acidic endosomal pH. This binding rescues the molecule from lysosomal degradation and recycles it to the cell surface, extending serum half-life from approximately 2 hours (conventional BiTE) to approximately 5–7 days. The practical consequence is a shift from continuous intravenous infusion to weekly or biweekly dosing — a transformative change in patient convenience and healthcare resource utilisation documented in the DeLLphi clinical programme.
The bivalent DLL3 engagement confers approximately 10–100× enhanced avidity compared to monovalent formats. This avidity gain is particularly important in SCLC, where DLL3 expression levels vary across patients and tumour heterogeneity can limit the effectiveness of low-avidity monovalent binders. Research published by Nature on bispecific antibody avidity effects supports the rationale that bivalent tumour-antigen engagement substantially improves target-cell selectivity in low-expression settings.
Tarlatamab’s tandem DLL3 scFv architecture provides approximately 10–100× enhanced avidity compared to monovalent DLL3 BiTE formats, enabling robust tumor cell engagement even at moderate DLL3 expression levels on SCLC cells.
Structural Differentiation from Earlier DLL3 Bispecific Formats
Tarlatamab’s structural innovations are best understood in contrast to earlier DLL3-targeting bispecific formats, which were limited by monovalent DLL3 binding and short serum half-lives requiring continuous infusion. The comparison below, derived from the patent family and clinical literature, quantifies the key differentiating parameters.
Map the full competitive landscape of DLL3-targeting bispecifics — patent families, structural formats, and clinical stage — in PatSnap Eureka.
Search DLL3 Bispecific Patents in PatSnap Eureka →| Feature | Earlier DLL3 BiTEs | Tarlatamab (AMG 757) |
|---|---|---|
| DLL3 Valency | Monovalent (1 scFv) | Bivalent (2 scFv) |
| CD3 Valency | Monovalent (1 scFv) | Monovalent (1 scFv) |
| Serum Half-Life | ~2 hours | ~5–7 days (Fc-mediated) |
| Dosing Schedule | Continuous IV infusion | Weekly or biweekly IV |
| Tumor Avidity | Lower (monovalent) | ~10–100× higher (bivalent) |
| DLL3 Binding KD | Variable (typically 2–20 nM) | ~0.5–2 nM (high affinity) |
| CD3 Binding KD | Variable | ~5–10 nM (moderate, tuned) |
| Fc Silencing Mutations | None (no Fc) | L234A/L235A/P329G |
| Clinical Stage | Preclinical / early Phase 1 | FDA Approved (May 2024) |
The combination of bivalent tumor binding (enhanced selectivity), monovalent CD3 binding (reduced off-tumor activation), and step-up dosing (10 µg → 100 µg) produces a manageable CRS profile in which most events are Grade 1–2. This safety architecture is not incidental — it is a direct consequence of the affinity and valency decisions encoded in the patent family claims, and it is what enabled Tarlatamab to complete the DeLLphi-300 Phase II trial and achieve regulatory approval where earlier formats did not progress.
The Recommended Clinical Compound: AMG 757 Efficacy and Approval Profile
AMG 757 (Tarlatamab-dlle) is the lead clinical compound from the WO2017021526 patent family, selected on the basis of its optimised DLL3 affinity (KD ~0.5–2 nM), moderate CD3 affinity (KD ~5–10 nM), extended half-life (~5 days), and bivalent DLL3 avidity advantage. The FDA approved Tarlatamab in May 2024 for the treatment of extensive-stage small cell lung cancer (ES-SCLC) after progression on platinum-based chemotherapy — making it the first DLL3-targeted BiTE antibody to reach the market.
In the DeLLphi-300 Phase II trial, Tarlatamab (AMG 757) achieved an overall response rate of 40% in patients with 2L+ extensive-stage small cell lung cancer, with a median duration of response of 9.7 months and median progression-free survival of 3.9 months.
DeLLphi-300 Phase II Clinical Outcomes
The DeLLphi-300 Phase II trial established the pivotal efficacy data supporting FDA approval. Key outcomes for the approved 10 mg → 100 mg step-up dosing regimen administered intravenously every two weeks are as follows:
- Overall response rate (ORR): 40% in 2L+ ES-SCLC
- Median duration of response (DOR): 9.7 months
- Median progression-free survival (PFS): 3.9 months
- CRS profile: Most events Grade 1–2, consistent with the moderate CD3 affinity and step-up dosing design
The approved dosing schedule — 10 mg in cycle 1 followed by 100 mg in cycle 2 onwards, administered every two weeks — directly reflects the structural decisions in the patent: the moderate CD3 affinity enables step-up dosing, and the ~5-day half-life supports biweekly administration without continuous infusion. For patent professionals and drug discovery teams mapping the IP landscape around SCLC immunotherapy, the PatSnap Life Sciences platform provides freedom-to-operate analysis across the full DLL3 bispecific patent space.
“Tarlatamab’s 40% ORR and 9.7-month median duration of response in 2L+ ES-SCLC — a historically treatment-resistant setting — validate the bivalent DLL3 / monovalent CD3 architecture as the optimal format for this target.”
The structural and clinical profile of AMG 757 sets the benchmark for next-generation DLL3-targeting programmes. Patent landscape analysis using tools such as PatSnap Eureka can identify white-space opportunities in DLL3 bispecific formats — including trispecific constructs, alternative Fc-extension strategies, and combination approaches — that build on or design around the WO2017021526 family claims.