Abstract
Delta-like ligand 3 (DLL3), a Notch-pathway inhibitor selectively expressed on the surface of small cell lung cancer (SCLC) cells with negligible expression in normal adult tissues, has become one of the most compelling precision-oncology targets in thoracic oncology. Over the period 2020–2026, the field has undergone a fundamental transformation—from the clinical failure of first-generation antibody-drug conjugates (ADCs) to the regulatory approval of tarlatamab, a bispecific T-cell engager (BiTE), and the rapid expansion of a diverse pipeline spanning CAR-T cell therapies, next-generation ADCs, trispecific antibodies, and radiopharmaceuticals. This review synthesizes the biological rationale, therapeutic modality landscape, clinical evidence, safety considerations, and future directions for DLL3-directed treatment in SCLC, with emphasis on the period from January 2020 through July 2026.
1. Biological and Clinical Rationale for DLL3 Targeting in SCLC
DLL3 functions as an inhibitory Notch-pathway ligand that is aberrantly upregulated in neuroendocrine malignancies. In SCLC, unlike most other solid tumors where Notch signaling is aberrantly activated, the pathway is inactivated—a permissive state for neuroendocrine transcription factors and survival programs driven by ASCL1 2. DLL3 expression is mechanistically linked to this neuroendocrine phenotype, making it both a biomarker of tumor identity and a rational therapeutic target 2. Critically, approximately 80–96% of SCLC tumor specimens express DLL3 on the cell surface, whereas expression in healthy adult tissues is negligible, establishing the favorable therapeutic window that distinguishes DLL3 from most solid-tumor targets 16.
Patient selection implications deserve attention. While DLL3 expression is broadly prevalent in SCLC, it is heterogeneous in related neuroendocrine tumors. In large cell neuroendocrine carcinoma (LCNEC), TTF-1-positive cases showed DLL3 positivity in all evaluated specimens, whereas only 20% of TTF-1-negative cases expressed DLL3 3. Across modalities, the predictive significance of DLL3 expression thresholds varies: the bispecific antibody obrixtamig (BI 764532) demonstrated an objective response rate (ORR) of 40% in DLL3-high versus 3.3% in DLL3-low SCLC tumors, whereas the ADC ZL-1310 showed activity across a broad H-score range of 5–260, suggesting modality-specific biomarker dependencies 12. Standardized assay development and validated expression cutoffs remain priorities for precision patient selection.
2. Historical Lessons: Rovalpituzumab Tesirine
Rovalpituzumab tesirine (Rova-T), an DLL3-targeting ADC developed by AbbVie, generated early biomarker-enriched signals in DLL3-high SCLC, including ORRs exceeding 50% in heavily pretreated cohorts. However, subsequent phase III evaluation in the MERU maintenance trial demonstrated no survival benefit versus placebo, and the program was discontinued 212. The failure of Rova-T is not attributable to DLL3 as a target but rather to platform-specific limitations: an ADC linker-payload combination with inadequate tolerability, insufficient direct T-cell engagement, pharmacokinetic liabilities, and potential intratumoral antigen heterogeneity. These lessons precipitated a decisive pivot to immune-activating modalities—principally bispecific T-cell engagers and CAR-based therapies—that harness the host immune system rather than delivering cytotoxic payloads directly 212.
3. Therapeutic Modality Landscape: 2020–2026
3.1 Bispecific T-Cell Engagers: The Dominant Mature Strategy
Bispecific T-cell engagers bridge DLL3-expressing tumor cells with CD3+ cytotoxic T lymphocytes (CTLs), forming an immunological synapse that results in CTL-mediated tumor cell lysis. Preclinical studies demonstrated T-cell infiltration, tumor regression, and upregulation of PD-1, PD-L1, and LAG-3 on tumor-infiltrating lymphocytes—the last observation providing a mechanistic rationale for combining BiTEs with immune checkpoint inhibitors (ICIs) 4.
Tarlatamab (IMDELLTRA, Amgen) is the first DLL3-directed therapy to achieve regulatory approval. It received US Food and Drug Administration (FDA) accelerated approval in May 2024 based on phase II DeLLphi-301 data (ORR 40%, duration of response [DoR] 9.7 months), followed by traditional FDA approval in November 2025 and European Commission authorization in June 2026 612. The pivotal phase III DeLLphi-304 trial randomized 509 patients with relapsed SCLC to tarlatamab monotherapy versus investigator's choice chemotherapy (topotecan, lurbinectedin, or amrubicin). Tarlatamab achieved a median overall survival (OS) of 13.6 versus 8.3 months (hazard ratio [HR] 0.60; 95% CI 0.47–0.77; p<0.001), representing a 40% reduction in mortality. Median progression-free survival (PFS) was 4.2 versus 3.2 months (HR 0.72; p<0.001). DLL3 expression was reported in 96% of evaluable specimens, and activity was observed in patients with treated stable brain metastases 12.
The principal toxicity of the BiTE class is cytokine release syndrome (CRS). In DeLLphi-304, any-grade CRS occurred in 56.7% of tarlatamab-treated patients, with serious CRS in 19.7%; immune effector cell-associated neurotoxicity syndrome (ICANS) was observed in 4.7% 12. Critically, grade ≥3 treatment-related adverse events (TRAEs) were substantially lower with tarlatamab (27%) than chemotherapy (62%), underscoring an improved high-grade safety profile. Step-up dosing (1 mg day 1, 10 mg day 8, then every 2 weeks) is employed to mitigate CRS, along with mandatory post-infusion monitoring 12.
Obrixtamig (BI 764532, Boehringer Ingelheim) is an IgG-like DLL3/CD3 bispecific antibody that completed phase I dose escalation in 168 patients with SCLC, extrapulmonary neuroendocrine carcinomas, and LCNEC, reporting an ORR of 23% overall and 40% in DLL3-high SCLC, with a median DoR of 8.5 months 712. The DLL3-high versus DLL3-low ORR contrast reinforces the potential utility of expression-based patient enrichment for this agent. Phase Ib expansion is ongoing.
HPN-328 (MK-6070, Merck/Harpoon Therapeutics) is a trispecific T-cell activating construct (TriTAC) targeting DLL3, CD3, and albumin, under evaluation in a phase 1/2 study (NCT04471727) in SCLC and neuroendocrine prostate cancer 1112. Mature efficacy data from the retrieved materials were not available as of the review date.
Roche's RG-6524 is a trispecific agent incorporating DLL3, CD3, and the immune costimulatory receptor CD137, currently in phase I evaluation in the United States 11.
3.2 Antibody-Drug Conjugates: Next-Generation Platforms
Following Rova-T's failure, next-generation DLL3 ADCs have emerged with distinct payload and linker chemistries, particularly topoisomerase I (TOP1) inhibitor payloads.
ZL-1310 (Zai Lab/Medilink Therapeutics), a DLL3-targeting ADC with a TOP1 inhibitor payload, received FDA Fast Track designation in May 2025 13. In the ASCO 2025 phase I update of 89 treated/74 evaluable patients, second-line ORR across all doses was 67%, with disease control rate (DCR) of 97%. At the optimal dose of 1.6 mg/kg, uORR reached 79% and DCR 100%. Notably, ORR in patients with baseline brain metastases was 68%, rising to 86% in those without prior cranial irradiation 1213. At doses below 2.0 mg/kg, grade ≥3 TRAEs were only 6%, with no grade ≥3 interstitial lung disease (ILD); higher doses were associated with two grade ≥3 ILD events, informing dose optimization. A registrational trial in second-line ES-SCLC is planned 13.
SHR-4849 (IDE849, Hengrui/IDEAYA Biosciences) is another DLL3 TOP1 ADC, with a phase I dataset of 100 patients (87 SCLC, 13 other neuroendocrine carcinomas) presented at the IASLC 2025 World Conference on Lung Cancer. Among 71 evaluable SCLC patients treated at doses ≥2.4 mg/kg, ORR was 73.2% (confirmed ORR 47.9%), with median PFS of 6.7 months and ORR of 66.7% in patients with brain metastases 12. Grade ≥3 TRAEs occurred in 48%; there were no treatment-related deaths and a 2% discontinuation rate. The strength and depth of these early signals are clinically notable, though the limited sample size and absence of durability data necessitate cautious interpretation in light of the Rova-T precedent.
3.3 CAR-T and Cellular Therapies
Chimeric antigen receptor (CAR) T-cell therapies represent a mechanistically distinct approach. DLL3-directed CAR-T cells have demonstrated potent and specific killing of DLL3-positive SCLC lines in vitro and tumor suppression in xenograft models 9. LB-2102 (Legend Biotech), the most clinically advanced CAR-T program in the retrieved dataset, reported phase I results in 9 patients with relapsed/refractory SCLC or LCNEC: dose-dependent activity was observed, with 1 partial response and 2 stable diseases at dose level 3; no dose-limiting toxicities were recorded, and only one grade 1 CRS event occurred 12. AMG-119 (Amgen) is in phase I in the United States 11.
DLL3-CAR NK-92 cells represent an off-the-shelf alternative strategy. Preclinical data showed significant in vitro cytotoxicity against DLL3+ SCLC lines and tumor regression in both pulmonary metastasis and subcutaneous xenograft models, within an acceptable safety threshold 10. Manufacturing advantages over autologous CAR-T are potentially significant for scalability and accessibility.
A preclinical observation of note: PD-1 checkpoint inhibition dramatically enhanced the efficacy of DLL3-targeted bispecific antibodies but not CAR-T cells in xenograft models, suggesting modality-specific immunological interactions relevant to future combination design 9.
4. Combination Strategies and Innovation Trends
The upregulation of PD-1, PD-L1, and LAG-3 observed in tumor-infiltrating lymphocytes following BiTE treatment provides a mechanistic rationale for combining DLL3-directed BiTEs with ICIs 4. DeLLphi-303, a phase Ib study, evaluated tarlatamab as maintenance therapy following first-line platinum-based chemoimmunotherapy combined with atezolizumab or durvalumab. Disease control rates of 62.5% were reported in both combination arms, with 9-month OS rates of 91.8% and 86.7%, respectively, though the nonrandomized design limits interpretation 12. Phase III studies are underway evaluating tarlatamab in the first-line setting in combination with durvalumab (DeLLphi-305) and following concurrent chemoradiotherapy in limited-stage SCLC (DeLLphi-306) 12.
Emerging modalities under investigation include radiopharmaceuticals (ABD-147, Abdera Therapeutics, phase I) and near-infrared photoimmunotherapy, representing niche but potentially differentiating DLL3-directed platforms 111. The trispecific architecture of RG-6524 (DLL3/CD3/CD137) and HPN-328 (DLL3/CD3/albumin) exemplifies the broader trend toward construct sophistication aimed at improving immune engagement and persistence.
5. Competitive Pipeline Overview and Future Outlook
As of July 2026, the DLL3-targeted SCLC pipeline comprises approximately 21 programs across all stages of development 11. The ecosystem is organized around three geographic development poles: the United States, China, and Europe/UK. China contributes substantial ADC pipeline depth (IBI-3009, SHR-4849, SYS-6040, ZL-1310) and has licensed or co-developed advanced BiTE programs. The US maintains the broadest modality diversity; Europe/UK is concentrated around later-stage assets, principally tarlatamab and obrixtamig.
Key near-term milestones include: tarlatamab's NMPA (National Medical Products Administration) review in China (application accepted July 2025); ZL-1310's anticipated registrational trial initiation and a forecasted US accelerated approval in approximately 2027 13; and phase III readouts from DeLLphi-305 and DeLLphi-306 in the first-line and limited-stage settings.
Table 1. DLL3-Targeted Therapies in SCLC: Comparative Summary (2020–July 2026)
| Agent | Company | Modality | Mechanism | Phase | Population | Key Efficacy | Key Safety | Status |
|---|---|---|---|---|---|---|---|---|
| Tarlatamab (IMDELLTRA) | Amgen | BiTE | DLL3/CD3 T-cell engager | Phase III (pivotal) | ES-SCLC, post-platinum | Median OS 13.6 vs 8.3 mo (HR 0.60); ORR 40% (Ph II) | CRS 56.7% (any grade), 19.7% serious; grade ≥3 TRAEs 27% vs 62% with chemo | FDA accelerated (5/2024), traditional (11/2025); EMA approved (6/2026); UK CMA; China NMPA under review 612 |
| Obrixtamig (BI 764532) | Boehringer Ingelheim | IgG-like BiTE | DLL3/CD3 T-cell engager | Phase I/Ib | SCLC, epNEC, LCNEC | ORR 40% (DLL3-high) vs 3.3% (DLL3-low); DoR 8.5 mo | CRS 57% any grade, 3% grade ≥3; grade ≥3 TRAEs 24% | Phase Ib expansion ongoing 712 |
| HPN-328 (MK-6070) | Merck/Harpoon | Trispecific antibody | DLL3/CD3/albumin TriTAC | Phase 1/2 | SCLC, NEPC, HG-NET | Mature data not available in retrieved materials | Not available in retrieved materials | Ongoing (NCT04471727; est. completion 1/2028) 1112 |
| RG-6524 | Roche | Trispecific antibody | DLL3/CD3/CD137 | Phase I | SCLC | Not available in retrieved materials | Not available in retrieved materials | Phase I ongoing, USA 11 |
| ZL-1310 | Zai Lab/Medilink | ADC | DLL3-TOP1 inhibitor payload | Phase I | ES-SCLC, post-platinum | 2L ORR 67%; 1.6 mg/kg uORR 79%, DCR 100%; brain mets ORR 68% | Grade ≥3 TRAEs 23% all doses; 6% at <2.0 mg/kg; grade ≥3 ILD in 2 higher-dose pts | FDA Fast Track (5/2025); registrational trial planned; accelerated approval forecast ~2027 1213 |
| SHR-4849 (IDE849) | Hengrui/IDEAYA | ADC | DLL3-TOP1 inhibitor payload | Phase I | SCLC, NEC | 2L+ ORR 73.2% (confirmed 47.9%); median PFS 6.7 mo; brain mets ORR 66.7% | Grade ≥3 TRAEs 48%; no treatment-related deaths; 2% discontinuation | Phase I ongoing; global rights licensed to IDEAYA (12/2024) 12 |
| LB-2102 | Legend Biotech | CAR-T | DLL3-directed autologous CAR-T | Phase I | SCLC/LCNEC, R/R | DL3: 1 PR, 2 SD (n=9) | No DLTs; grade 1 CRS ×1; grade ≥3 cytopenias from lymphodepletion | Phase I, early feasibility 12 |
| AMG-119 | Amgen | CAR-T | DLL3-directed CAR-T | Phase I | SCLC | Not available in retrieved materials | Not available in retrieved materials | Phase I, USA 11 |
| DLL3-CAR NK-92 | Multiple institutions | CAR-NK | DLL3-directed NK-92 | Preclinical | SCLC | Tumor regression in xenograft models | Acceptable safety threshold (preclinical) | Preclinical 10 |
| ABD-147 | Abdera Therapeutics | Radiopharmaceutical | DLL3-targeted | Phase I | SCLC | Not available in retrieved materials | Not available in retrieved materials | Phase I, USA 11 |
| Rova-T (rovalpituzumab tesirine) | AbbVie | ADC (1st gen) | DLL3-DUBA toxin ADC | Phase III (MERU) | SCLC maintenance | No survival benefit vs placebo; early-line ORR signals not replicated | Tolerability concerns; program-level safety issues | Discontinued; historical reference 212 |
Table 2. Modality-Specific Advantages, Limitations, and Clinical Considerations
| Modality | Key Advantages | Key Limitations | Clinical Considerations |
|---|---|---|---|
| Bispecific/T-cell engagers (BiTEs) | Validated regulatory pathway (tarlatamab); direct immune activation; outpatient-oriented delivery; extended half-life enables intermittent dosing; combination signal with ICIs 46 | CRS and ICANS require monitoring infrastructure; adequate host T-cell numbers required; biomarker strategy not yet standardized across agents 12 | Step-up dosing, premedication, and 6–8h post-infusion monitoring protocols essential; consider DLL3 expression level for some agents (obrixtamig); combination with anti-PD-1/PD-L1 mechanistically rational 912 |
| Next-generation ADCs (DLL3-TOP1) | High early ORRs (67–79%); activity in brain metastases; potentially broader DLL3 expression threshold applicability; scalable manufacturing 12 | Rova-T precedent warns against over-relying on phase I signals; durability data immature; ILD risk at higher doses; payload-related hematologic toxicity 12 | Monitor for ILD at higher doses; dose optimization critical (1.6 mg/kg for ZL-1310); maintain vigilance for durability; await randomized confirmatory data 13 |
| CAR-T cell therapy | High target specificity; potential for durable response; multi-antigen constructs (bispecific CAR-T, BHP-01) under development 89 | Autologous manufacturing is time-intensive (weeks–months) and costly; lymphodepletion required; feasibility challenging in rapidly progressive SCLC; bridging therapy needed 12 | Best viewed as early translational development; allogeneic off-the-shelf approaches (ALLO-213) may improve accessibility; checkpoint inhibition may not additively enhance CAR-T in SCLC (preclinical finding) 9 |
| CAR-NK cell therapy | Off-the-shelf potential; manufacturing scale advantages over autologous CAR-T; reduced GvHD risk; rapid deployment 10 | Shorter persistence than CAR-T; limited clinical experience in SCLC; NK cell exhaustion mechanisms less understood 10 | Still preclinical for SCLC; promising as combination strategy given safety profile in xenograft models 10 |
| Radiopharmaceuticals / other | Potential for precision tumor delivery; complementary to T-cell–based approaches | Very early development; no clinical readouts in retrieved materials for SCLC-specific DLL3 programs | Monitor early phase I readouts from ABD-147 11 |
6. Translational and Practical Considerations
Toxicity management is central to clinical implementation. For BiTEs, CRS mitigation through step-up dosing, premedication with acetaminophen and antihistamines, tocilizumab availability for grade ≥2 events, and mandatory post-infusion monitoring represent the standard of care 12. ICANS, though less frequent, requires neurological surveillance. For ADCs, hematologic toxicity (anemia, neutropenia, thrombocytopenia) and ILD—particularly at supra-optimal ADC doses—necessitate dose-appropriate monitoring 12. CAR-T therapies require institutional manufacturing partnerships, lymphodepletion regimens, and bridging therapy for eligible patients, reflecting an operational complexity that is particularly challenging in rapidly progressive relapsed/refractory SCLC 12.
Resistance mechanisms remain incompletely characterized. Intratumoral DLL3 heterogeneity may select for antigen-low or antigen-negative clones under sustained immunological pressure. Bispecific targeting of multiple neuroendocrine antigens simultaneously—as explored in BHP-01 (DLL3/PD-L1/4-1BB)—is an emerging strategy to address antigen escape 11. The upregulation of PD-1 and LAG-3 observed after BiTE treatment suggests that combination ICI therapy may sustain T-cell effector function 4.
Manufacturing and access disparities represent a real-world implementation barrier. Off-the-shelf bispecific antibodies and ADCs confer clear scalability advantages, whereas autologous CAR-T and CAR-NK platforms require specialized infrastructure. Resource-constrained settings may have limited access even as global regulatory coverage of tarlatamab expands across the US, EU, and UK 12.
Conclusion
The period 2020–2026 has witnessed the transformation of DLL3 from a compelling but clinically unproven target into a validated precision-oncology axis in SCLC, anchored by tarlatamab's phase III survival benefit and multi-regional regulatory approval 612. The next wave of clinical activity is dominated by next-generation ADCs (ZL-1310 and SHR-4849) with encouraging early response rates and intracranial activity, alongside alternative BiTE platforms such as obrixtamig and innovative multispecific constructs 1213. CAR-T and CAR-NK therapies offer mechanistic complementarity but remain early-stage, and the Rova-T experience continues to serve as a necessary reminder that antigen validity alone is insufficient for clinical success. For SCLC specialists, the priority framework is clear: adopt and optimize tarlatamab-based treatment pathways now, monitor emerging ADC and BiTE readouts critically, and engage in first-line combination trials that may ultimately reposition DLL3-directed immunotherapy as a cornerstone of SCLC care from diagnosis rather than salvage 11112.