DLL3 Expression: Diagnostic and Therapeutic Implications in Neuroendocrine Neoplasms
Delta-like ligand 3 (DLL3) has become a focal point in the evolving landscape of targeted therapies for neuroendocrine neoplasms (NEN), especially given its established role in pulmonary small cell neuroendocrine carcinoma (SCNEC). As a Notch pathway ligand aberrantly overexpressed on the surface of neuroendocrine tumor cells—yet largely absent from normal adult tissues—DLL3 offers a tantalizing opportunity for selective therapeutic intervention. While DLL3’s prominence in pulmonary SCNEC has driven the development of multiple investigational agents, data on its expression across the broader and heterogeneous family of NENs have been notably scarce, limiting the strategic expansion of anti-DLL3 therapies into other disease subsets.
To address this gap, a comprehensive study recently evaluated DLL3 expression in a large cohort of tumors, analyzing 1,294 NEN samples spanning multiple histologic subtypes and anatomical sites, alongside 479 non-neuroendocrine carcinomas for comparative context. This investigation not only mapped the prevalence of DLL3 across distinct NEN categories—ranging from extrapulmonary SCNEC and large cell neuroendocrine carcinomas (LCNEC), to mixed neuroendocrine-non-neuroendocrine neoplasms (MiNEN), gastroenteropancreatic neuroendocrine tumors (GEP-NET), and pulmonary carcinoids—but also examined the stability of DLL3 expression through metastatic progression and its relationship to overall survival.
The findings reveal striking heterogeneity in DLL3 expression across NEN subtypes. As anticipated, DLL3 was most frequently expressed in poorly differentiated NECs, with 80.4% of SCNECs and 62.6% of LCNECs exhibiting positivity. MiNENs demonstrated intermediate expression rates at 28.6%. Of particular interest, pulmonary carcinoids—often classified as well-differentiated neoplasms—showed a notable DLL3 positivity rate of 41.5%, suggesting a subset could be biologically amenable to DLL3-targeted approaches. In stark contrast, GEP-NETs displayed minimal DLL3 expression at just 5.1%, aligning more closely with the very low prevalence observed in non-neuroendocrine carcinomas (1.3%).
Another crucial dimension of this work was the analysis of metastatic concordance. Among 67 paired samples of primary NENs and their metastases, DLL3 expression was remarkably consistent, with a concordance rate of 92.5%. This stability across disease progression reinforces DLL3’s reliability as a therapeutic target and supports the rationale for pursuing DLL3-directed strategies even in metastatic settings.
When examining clinical outcomes, the study found that DLL3 expression in pulmonary carcinoids and GEP-NETs was associated with decreased overall survival on univariable analysis. However, these associations did not persist in multivariable models that adjusted for key confounding factors such as tumor stage and grade, underscoring that DLL3 expression alone is unlikely to serve as an independent prognostic biomarker in these groups. Moreover, DLL3 positivity in NECs—whether pulmonary or gastroenteropancreatic—did not correlate with differences in survival outcomes, further suggesting that its principal clinical utility may lie more in guiding targeted treatment strategies than in prognostication.
These insights carry meaningful implications for oncologists and the design of future clinical trials. The high prevalence of DLL3 in SCNEC, LCNEC, and even a subset of pulmonary carcinoids provides a compelling rationale for prioritizing these patient populations in DLL3-directed studies. In contrast, the sparse expression seen in GEP-NETs and non-neuroendocrine carcinomas indicates that DLL3-targeted therapies would likely offer limited benefit in these groups, warranting alternative molecular approaches.
Importantly, the concordance of DLL3 expression between primary tumors and metastases suggests that a single biopsy may reliably inform treatment decisions throughout the disease course. This could streamline patient selection for DLL3-targeted agents, sparing many patients from repeat invasive tissue sampling.
As DLL3-targeted therapeutics—including bispecific T-cell engagers and antibody-drug conjugates—advance through clinical development, these findings provide a crucial foundation for refining trial eligibility criteria and optimizing biomarker-driven therapy in neuroendocrine oncology. By illuminating the diverse landscape of DLL3 expression across NENs, this study lays the groundwork for more precise, histology-tailored strategies, potentially expanding the benefit of DLL3-directed treatments to new subsets of patients while ensuring efforts are concentrated where the biology is most compelling.