DLL3 × CD3 Target Pair Landscape — The TCE Benchmark Pair

1. Executive Snapshot

DLL3 × CD3 is the flagship T-cell engager (TCE) benchmark pair — the clinically validated reference case for the BsAb class. Tumor side: DLL3 (Delta-like ligand 3), a Notch-family ligand, is highly expressed on SCLC and other neuroendocrine tumors (NEC/LCNEC, NEPC) with minimal normal-tissue expression ([Amgen](https://amgenoncology.com/resources/dll3_neuroendocrine_tumors-USA-ASCO-80036.pdf)). Effector side: the anti-CD3 arm is validated in three formats — BiTE (tarlatamab), HLE TriTAC (gocatamig), and IgG-like TCE (obrixtamig) ([Frontiers](https://www.frontiersin.org/journals/immunology/articles/10.3389/fimmu.2025.1674449/full)).

Tarlatamab (AMG 757, Imdelltra®) is the first and only approved DLL3×CD3 bispecific: FDA accelerated approval 2024-05-16, full approval 2025-11, EC 2025, NMPA approval 2026-04 (commercialized by BeOne) ([drugs.com](https://www.drugs.com/newdrugs/fda-grants-accelerated-approval-imdelltra-tarlatamab-dlle-extensive-stage-small-cell-lung-cancer-6271.html), [BeOne](http://www.aastocks.com/en/stocks/news/aafn-con/NOW.1526634/latest-news/AAFN)). ≥3 clinical competitors follow (gocatamig, obrixtamig, QLS31904) plus a Henlius tetra-specific TCE; DeLLphi-304 (Ph3, 2L vs chemo) hit positive interim OS with ~40% death-risk reduction ([Amgen PR](https://wwwext.amgen.com/newsroom/press-releases/2025/06/imdelltra-significantly-reduced-risk-of-death-by-40-in-small-cell-lung-cancer-patients)).

The negative history is load-bearing: Rova-T, the DLL3 ADC AbbVie bought for ~$5.8B via Stemcentrx, failed Ph3 and was terminated in 2019 — the textbook biomarker/heterogeneity lesson ([pharmaphorum](https://pharmaphorum.com/news/abbvie-axes-multi-billion-cancer-flop-rova-t-after-another-trial-fails)). Tool angle: DLL3 ECD (DSL + EGF1–6) is a known difficult-expression recombinant target, and the CD3 arm (CD3ε generic vs CD3δ-specific formats) is a real reagent/format gap — where an intelligence-led platform adds value.

2. Asset Landscape

Approved

Tarlatamab (AMG 757) — Amgen, BiTE (tandem scFv anti-DLL3 × anti-CD3). DeLLphi-301 (Ph2, 10 mg Q2W): ORR 40%, mOS ~14.3 mo ([NEJM 2023](https://www.nejm.org/doi/full/10.1056/NEJMoa2307980)).

Clinical

• Gocatamig (MK-6070, ex-HPN328) — Merck (via Harpoon, ~$680M, 2024-01). HLE TriTAC (DLL3 × CD3 × HSA); Ph1/2 (NCT04471727), ORR 44% ([ASCO EDBK](https://ascopubs.org/doi/10.1200/EDBK-26-524964)); + I-DXd combo Ph1b/II (NCT07227597); brain-met SCLC activity.
• Obrixtamig (BI 764532) — Boehringer Ingelheim, IgG-like TCE. Ph1 (DAREON-1/5/9) → Ph3 first-line (DAREON-Lung-1, NCT07472517); FDA Fast Track for SCLC + epNEC; Ph1 DLL3+ NEC response 29% ([JCO 2025](https://ascopubs.org/doi/pdf/10.1200/JCO-25-00363), [CancerNetwork](https://www.cancernetwork.com/view/bi-764532-earns-fda-ftd-in-sclc-and-extrapulmonary-neuroendocrine-carcinoma)).
• QLS31904 — Qilu Pharma, anti-DLL3/CD3, Ph1 (advanced solid tumors/SCLC) ([AACR 2022](https://aacrjournals.org/cancerres/article-split/82/12_Supplement/5550/701369/Abstract-5550-QLS31904-An-anti-DLL3-CD3-bispecific)).
• Henlius DLL3 tetra-specific TCE — entered the clinic for SCLC/NEC ([allsci](https://allsci.com/news/henlius-pushes-tetra-specific-t-cell-engager-into-the-clinic-for-sclc-nec/)).

Preclinical

Public preclinical data are thin (QLS31904's AACR 2022 abstract is the main benchmark); the battleground is format engineering rather than new biology ([Expert Opin Drug Discov 2025](https://www.tandfonline.com/doi/pdf/10.1080/17460441.2025.2522088)).

Discontinued

Rova-T (rovalpituzumab tesirine), DLL3 ADC (Stemcentrx → AbbVie, ~$5.8B): Ph3 failures (TAHOE, MERU), terminated 2019 ([biopharmadive](https://www.biopharmadive.com/news/abbvie-ends-rova-t-research-after-another-study-setback/561962/)).

3. Companies

• Originators: Amgen (tarlatamab); Harpoon Therapeutics (HPN328 TriTAC, acquired by Merck); Boehringer Ingelheim (internal obrixtamig); Qilu; Henlius.
• Partners: BeOne (BeiGene) — China commercialization of tarlatamab; Zai Lab × Amgen global DLL3 ADC + TCE deal (2026-04) ([bizwire](https://investor.wedbush.com/wedbush/article/bizwire-2026-4-1-zai-lab-announces-global-clinical-trial-collaboration-and-supply-agreement-to-evaluate-novel-dll3-adc-zocilurtatug-pelitecan-in-combination-with-a-bispecific-t-cell-engager-therapy)); Merck–Daiichi I-DXd combo.
• Licensees / financial: BeOne sold Imdelltra royalty rights to Royalty Pharma (~$885M upfront, 2025-08) ([sinodrugwatch](https://www.sinodrugwatch.com/beone-medicines-sells-imdelltra-r-ex-china-royalties-to-royalty-pharma-for-885m-upfront/)). AbbVie is the cautionary counterexample.

4. Clinical & Deal Timeline

5. Risks & Failures

• Rova-T ($5.8B AbbVie/Stemcentrx, terminated 2019): Ph3 failures (TAHOE/MERU) sank a $5.8B acquisition — the canonical lesson in DLL3 heterogeneity, biomarker selection, and payload toxicity ([pharmaphorum](https://pharmaphorum.com/news/abbvie-axes-multi-billion-cancer-flop-rova-t-after-another-trial-fails), [biopharmadive](https://www.biopharmadive.com/news/abbvie-ends-rova-t-research-after-another-study-setback/561962/)).
• DLL3 expression heterogeneity: ~47% of surgical SCLC specimens were DLL3-high in HOT1702 under a sensitive IHC readout — patient selection depends heavily on assay and cutoff ([PMC6853124](https://pmc.ncbi.nlm.nih.gov/articles/PMC6853124/)); no consensus companion diagnostic post-Rova-T.
• CRS: the most common AE of tarlatamab in DeLLphi-301 — any-grade ~51% (few G≥3); ICANS/treatment-related deaths carry label warnings ([NEJM 2023](https://www.nejm.org/doi/full/10.1056/NEJMoa2307980), [Imdelltra HCP](https://www.imdelltrahcp.com/clinical-results/efficacy)); obrixtamig pre-specifies CRS/IRR dose rules.
• Manufacturability: multi-chain TCEs suffer pairing inefficiency, mispairing, and aggregation — a direct developability/CMC risk (and service opportunity) ([OncoImmunology 2026](https://www.tandfonline.com/doi/pdf/10.1080/2162402X.2026.2632421)).
• CD3-arm off-tumor activation: affinity attenuation is the standard lever, with no universal optimum.
• Crowding: Merck, BI, Qilu, Henlius are fast-following; differentiation windows (1L, maintenance, brain mets, NEC/NEPC) are closing.

6. Discovery Questions

• DLL3 epitope & avidity: membrane-proximal vs distal epitopes (DSL + EGF1–6) — effect on killing/tumor penetration; are tarlatamab vs obrixtamig epitope choices traceable?
• CD3 arm format: CD3ε scFv generic vs CD3δ-specific/selective binders; 1+1 vs 2+1 geometry; affinity window for the CRS/efficacy tradeoff — no public comparative dataset exists.
• Half-life engineering: no-Fc BiTE (~55 kDa, short t½) vs HSA-binding TriTAC vs Fc-bearing IgG-like TCE — PK/toxicity tradeoffs.
• Patient selection: which IHC/RNA assay and cutoff (HOT1702: 47% DLL3-high); companion diagnostic needed; CNS efficacy in brain-met SCLC?
• Combinations: TCE + PD-(L)1, TCE + DLL3 ADC (Zai Lab–Amgen enrolling), TCE + chemo/atezo (DAREON-Lung-1).
• Resistance: DLL3 loss and lineage plasticity after TCE exposure ([Frontiers 2026](https://www.frontiersin.org/journals/immunology/articles/10.3389/fimmu.2026.1901163/full)).
• Indication expansion: NEC/LCNEC/NEPC (BI: epNEC Fast Track), melanoma, GBM — validation status?

7. Why This Matters Now

Three signals make this the moment to position: (1) tarlatamab converted accelerated → full approval (2025-11), proving regulatory durability ([drugs.com](https://www.drugs.com/newdrugs/fda-grants-full-approval-amgen-s-imdelltra-tarlatamab-dlle-extensive-stage-small-cell-lung-cancer-6671.html)); (2) DeLLphi-304 is the first positive Ph3 OS for a post-line TCE vs chemotherapy (~40% death-risk reduction, 2025-06) — class-defining ([Amgen PR](https://wwwext.amgen.com/newsroom/press-releases/2025/06/imdelltra-significantly-reduced-risk-of-death-by-40-in-small-cell-lung-cancer-patients)); (3) competitors are entering Ph3 — obrixtamig DAREON-Lung-1 (1L), Merck combos, and the NMPA approval (2026-04). Tool inflection: generic DLL3/CD3 proteins and DLL3–CD3 bridging ELISAs are commoditized ([ACRO](https://www.acrobiosystems.com/products/kits/dll3-cd3e-bis-a005)), but the differentiating demand — custom hard-to-express DLL3 ECD variants, CD3ε/CD3δ format reagents, and TDCC/CRS functional evaluation — grows with the copycat wave.

8. Sources

• [drugs.com — FDA accelerated approval of Imdelltra](https://www.drugs.com/newdrugs/fda-grants-accelerated-approval-imdelltra-tarlatamab-dlle-extensive-stage-small-cell-lung-cancer-6271.html); [full approval](https://www.drugs.com/newdrugs/fda-grants-full-approval-amgen-s-imdelltra-tarlatamab-dlle-extensive-stage-small-cell-lung-cancer-6671.html)
• [Amgen PR — DeLLphi-304, ~40% death-risk reduction (2025-06)](https://wwwext.amgen.com/newsroom/press-releases/2025/06/imdelltra-significantly-reduced-risk-of-death-by-40-in-small-cell-lung-cancer-patients)
• [NEJM 2023 — DeLLphi-301](https://www.nejm.org/doi/full/10.1056/NEJMoa2307980); [Imdelltra HCP safety/efficacy](https://www.imdelltrahcp.com/clinical-results/efficacy)
• [BeOne — NMPA approval, China commercialization (2026-04)](http://www.aastocks.com/en/stocks/news/aafn-con/NOW.1526634/latest-news/AAFN)
• [pharmaphorum — Rova-T terminated after Ph3 failures](https://pharmaphorum.com/news/abbvie-axes-multi-billion-cancer-flop-rova-t-after-another-trial-fails); [BioPharma Dive — AbbVie ends Rova-T research](https://www.biopharmadive.com/news/abbvie-ends-rova-t-research-after-another-study-setback/561962/)
• [PMC6853124 — HOT1702: 47% DLL3-high](https://pmc.ncbi.nlm.nih.gov/articles/PMC6853124/)
• [pharmtech — Merck acquires Harpoon (~$680M)](https://www.pharmtech.com/view/merck-to-acquire-harpoon-therapeutics-in-deal-valued-at-680-million); [SEC 10-K](https://www.sec.gov/Archives/edgar/data/310158/000162828024006850/mrk-20231231.htm)
• [ASCO EDBK — gocatamig ORR 44%](https://ascopubs.org/doi/10.1200/EDBK-26-524964); [Merck trial NCT07227597](https://www.merckclinicaltrials.com/trial/nct07227597/)
• [JCO 2025 — obrixtamig Ph1](https://ascopubs.org/doi/pdf/10.1200/JCO-25-00363); [CancerNetwork — Fast Track](https://www.cancernetwork.com/view/bi-764532-earns-fda-ftd-in-sclc-and-extrapulmonary-neuroendocrine-carcinoma); [DAREON-Lung-1 NCT07472517](https://trialscreen.org/trials/phase-3-dareon-lung-1-people-advanced-small-cell-cancer-to-compare-obrixtamig-trial-nct07472517)
• [AACR 2022 — QLS31904 preclinical](https://aacrjournals.org/cancerres/article-split/82/12_Supplement/5550/701369/Abstract-5550-QLS31904-An-anti-DLL3-CD3-bispecific); [allsci — Henlius tetra-specific TCE](https://allsci.com/news/henlius-pushes-tetra-specific-t-cell-engager-into-the-clinic-for-sclc-nec/)
• [sinodrugwatch — BeOne–Royalty Pharma ~$885M](https://www.sinodrugwatch.com/beone-medicines-sells-imdelltra-r-ex-china-royalties-to-royalty-pharma-for-885m-upfront/)
• [bizwire — Zai Lab × Amgen DLL3 ADC + TCE collaboration](https://investor.wedbush.com/wedbush/article/bizwire-2026-4-1-zai-lab-announces-global-clinical-trial-collaboration-and-supply-agreement-to-evaluate-novel-dll3-adc-zocilurtatug-pelitecan-in-combination-with-a-bispecific-t-cell-engager-therapy)
• [Frontiers — TCE mechanism & resistance](https://www.frontiersin.org/journals/immunology/articles/10.3389/fimmu.2026.1901163/full); [Expert Opin Drug Discov 2025 — CD3 affinity attenuation](https://www.tandfonline.com/doi/pdf/10.1080/17460441.2025.2522088); [OncoImmunology 2026 — TCE manufacturability](https://www.tandfonline.com/doi/pdf/10.1080/2162402X.2026.2632421)
• [ACROBiosystems — DLL3 × CD3E bridging ELISA kit](https://www.acrobiosystems.com/products/kits/dll3-cd3e-bis-a005); [Amgen — DLL3 neuroendocrine tumor education](https://amgenoncology.com/resources/dll3_neuroendocrine_tumors-USA-ASCO-80036.pdf)

Research tools this landscape implies: custom hard-to-express DLL3 ECD variants (full-length/Domain3/mutants/multi-species) with SEC + SPR QC; CD3ε vs CD3δ format engineering (affinity-attenuated and selective binders, 1+1/2+1, Fc/HSA half-life designs); DLL3–CD3 bridging ELISA and TDCC/cytokine-release (CRS-prediction) functional panels benchmarked against tarlatamab/gocatamig/obrixtamig public data; TCE developability screening (stability, aggregation, chain-pairing efficiency) as the manufacturability-risk service entry point.