ESMO Congress 2026

DLL3 is a Notch pathway ligand, known in developmental biology to be regulated by Mash1 and involved in neurogenesis in the ventral telencephalon ^[3]; it also appears as a neuronal lineage-related gene in a prognostic model of high-grade glioma, forming a two-gene model together with PTEN ^[1]. As a therapeutic target, its appeal lies in its expression profile: DLL3 is highly expressed in SCLC and LCNEC PDX and primary tumors ^[2], with the protein present on the surface of tumor cells but not in normal adult tissues, and detected in more than 80% of patients with SCLC ^[6]. The evidence chain for DLL3-targeted ADCs unfolded from there: a DLL3-targeted ADC induced durable tumor regression across multiple PDX models, with serial transplantation experiments demonstrating effective elimination of DLL3-expressing tumor-initiating cells ^[2]; rovalpituzumab tesirine completed a phase 1 dose exploration of 0.05 to 0.8 mg/kg given every 3 weeks or every 6 weeks ^[6]. At the protein level, DLL3, alongside TTF-1, is regarded as a neuroendocrine phenotype-associated marker of SCLC ^[5], and is directly regulated by the transcription factor ASCL1 ^[4].

At ESMO26 there are two DLL3-related pieces of news. One is the rapid orals session in neuroendocrine tumors: an attending expert described Pb-based Radioligands and DLL3 driven drugs as promising new strategies in the field, with no specific drug names or numbers given in the post. The other comes from MSK: Salomon Tendler will present a poster on October 23 at 3:15 p.m. CEST showing that DLL3 PET identifies neuroendocrine differentiation and tumor heterogeneity in advanced prostate cancer; that post likewise provides no cohort size, tracer details, or diagnostic performance figures.

Compared with existing knowledge, the advance here lies in expanding the setting and the readout modality. DLL3 therapeutic evidence has so far centered on ADCs in SCLC/LCNEC ^[2][6], and protein-level characterization has been conducted within the ASCL1/NEUROD1 subtyping framework of SCLC ^[4][5]; at this meeting, DLL3 driven drugs appear in an oral session on neuroendocrine tumors, while DLL3 PET carries the target into identifying neuroendocrine differentiation in advanced prostate cancer. The property that DLL3 is enriched on the surface of tumor cells but absent from normal adult tissues ^[2] underpins its suitability for imaging and targeting; tumor heterogeneity previously describable only through immunohistochemistry and transcription factor subtyping ^[4][5] now gains an additional in vivo imaging readout.

The limitations should also be noted: one post is an expert's verbal comment and the other a poster preview, with no concrete numbers in either; the sample size, tracer information, and diagnostic performance of the DLL3 PET study, as well as the specific agents, doses, and efficacy data of the DLL3 driven drugs, are undisclosed or to be confirmed. In addition, the prevalence of DLL3 expression in prostate cancer populations is not provided in these materials, and the detection rate of more than 80% reported in SCLC ^[6] cannot be directly extrapolated.