DLL3 Expression in GEP-NECs: A Marker of Poor Differentiation and Emerging Therapeutic Target

DLL3 Expression in GEP-NECs: A Marker of Poor Differentiation and Emerging Therapeutic Target

In an effort to address a critical gap in the diagnostic and therapeutic landscape for gastroenteropancreatic neuroendocrine carcinomas (GEP-NECs), a study published in January of 2025 in Scientific Reports examined Delta-like protein 3 (DLL3), an inhibitory ligand of the Notch signaling pathway that is minimally expressed in normal tissues but upregulated in neuroendocrine malignancies.

Gastroenteropancreatic neuroendocrine carcinomas fall within the broader category of gastroenteropancreatic neuroendocrine neoplasms (GEP-NENs), which account for approximately 65 percent of all neuroendocrine tumors (NETs) and rank as the second most common gastrointestinal cancer. Incidence has steadily increased due to improved imaging and greater clinical awareness. GEP-NENs are classified by histologic differentiation and proliferative activity into well-differentiated NETs, poorly differentiated neuroendocrine carcinomas (NECs), and mixed neuroendocrine-non-neuroendocrine neoplasms (MiNENs). Among these, NECs are the most aggressive, typically diagnosed at an advanced stage and associated with five-year survival rates below five percent.

Although DLL3 has been well characterized in small cell lung cancer (SCLC), its role in high-grade neuroendocrine neoplasms of the gastrointestinal tract has remained largely undefined. This retrospective analysis, which included 248 primary GEP-NECs, evaluated the diagnostic utility of DLL3 immunohistochemistry and examined its potential as a therapeutic target in this high-risk patient population.

Looking at the key findings, DLL3 expression was detected in 54.8 percent of GEP-NECs, with higher prevalence in small cell neuroendocrine carcinoma (SCNEC) at 68.1 percent compared to 38.6 percent in large cell neuroendocrine carcinoma (LCNEC). Expression was consistent in both lymph node and distant metastases, suggesting stability across disease progression. DLL3 was absent in Grade 1 and Grade 2 GEP-NETs, found in only 15.8 percent of Grade 3 NETs, and completely negative in gastric adenocarcinomas (GACs), supporting its utility as a marker of poor differentiation.

When using a one percent threshold to define DLL3 positivity, the assay identified just over half of NECs while correctly excluding most Grade 3 NETs (sensitivity 54.8 percent, specificity 84.2 percent). Raising the threshold to 50 percent improved specificity to 94.7 percent for NETs and one hundred percent for GACs. This provided greater diagnostic confidence for NECs. However, sensitivity dropped to 31.9 percent, meaning more DLL3-low NECs would be missed. The results illustrate the trade-off between sensitivity and specificity, which should be weighed based on clinical priorities.

In addition to its diagnostic application, DLL3 showed potential as a therapeutic target. While its expression did not significantly correlate with PD-L1 expression in GEP-NECs, co-expression was observed in 22.5 percent of patients, highlighting the possibility of combined DLL3-targeted therapy and immunotherapy. Further studies are needed to evaluate the relationship between DLL3 expression and clinical factors, such as tumor stage, lymph node status, and proliferation index to better define its stability as a biomarker in GEP-NECs.

The study also examined the relationship between DLL3 and achaete-scute homolog 1 (ASCL1), a transcription factor known to regulate DLL3 in pulmonary NETs. ASCL1-positive GEP-NECs exhibited significantly higher DLL3 expression, particularly among SCNECs. This correlation suggests overlapping molecular features between gastrointestinal NECs and SCLC and may help define subgroups more likely to respond to DLL3-targeted therapies.

While the findings are promising, the authors acknowledge several limitations, including the retrospective design, limited treatment-response data, and uncertain significance of rare nuclear DLL3 staining patterns. Despite these limitations, the study supports the integration of DLL3 immunohistochemistry into diagnostic workflows for GEP-NECs and points to its potential role in informing therapeutic decisions.

This study highlights the diagnostic utility and potential therapeutic relevance of DLL3 in high-grade GEP-NECs, particularly SCNECs. Its selective expression, absence in well-differentiated GEP-NETs and gastric adenocarcinomas, and association with ASCL1 support its role in identifying poorly differentiated tumors. Although DLL3 was not prognostic, its consistent expression in metastatic sites and increased prevalence in chemotherapy-exposed tumors support further exploration of DLL3 as a therapeutic target in digestive neuroendocrine carcinomas.

Reference

Yin L, Wang R, Ma X, et al. Exploring the expression of DLL3 in gastroenteropancreatic neuroendocrine neoplasms and its potential diagnostic value. Sci Rep. 2025;15(1):3287. Published 2025 Jan 26.