Evolving therapeutic landscape of duodenal neuroendocrine neoplasms: From endoscopic resection to systemic strategies

Abstract

Duodenal neuroendocrine tumors (D-NETs), a subset of gastroenteropancreatic NETs, arise from the diffusely distributed endocrine system and represent a clinically heterogeneous entity with rising global incidence. Their etiopathogenesis remains elusive. The diagnosis of D-NETs is challenging due to their non-specific symptomatology; however, hormone hypersecretion syndromes must be considered. Contrast-enhanced computed tomography and endoscopy serve as cornerstone diagnostic modalities, while five distinct histopathological subtypes have been characterized. Therapeutic decision-making is stratified by tumor size, location, grade, and stage, encompassing a multimodal arsenal including endoscopic resection, surgery, chemotherapy, targeted therapy, and peptide receptor radionuclide therapy. Postoperative surveillance with serial biomarker assessment and imaging is critical for detecting recurrence. Through early diagnosis, meticulous staging, complete resection, and structured follow-up, long-term survival outcomes for D-NET patients have substantially improved.

Key Words: Neuroendocrine tumor of the duodenum; Diagnosis; Treatment; Surgery

Core Tip: Neuroendocrine tumors (NETs) arise from the body’s diffuse endocrine system. Duodenal NETs represent a heterogeneous and uncommon neoplasm, frequently diagnosed serendipitously during clinical evaluation. Significant conceptual ambiguity persists regarding their clinicopathological manifestations, diagnostic algorithms, and therapeutic paradigms. This review provides a lucid and structured synthesis of contemporary evidence, delineating current understanding and critical knowledge gaps in duodenal NETs’ pathophysiology and management.

INTRODUCTION

Neuroendocrine tumors (NETs) originate from the diffusely distributed endocrine system and can potentially arise in virtually any anatomical site, with the gastrointestinal tract [particularly gastroenteropancreatic NETs (GEP-NETs)] and respiratory system representing the most frequent locations[1,2]. The current World Health Organization Classification of Tumors (5^th edition, 2019) and the 2016 European Neuroendocrine Tumor Society (ENETS) guidelines establish critical diagnostic frameworks for digestive system neoplasms[3,4]. While gastric NETs are stratified into three distinct subtypes - well-differentiated gastric NETs and neuroendocrine carcinomas - based on differentiation status, duodenal NETs (D-NETs) present unique classification challenges. Based on cellular origin and underlying pathophysiology, gastrointestinal NETs are fundamentally categorized according to the detectability of hormone secretion and presence of related clinical manifestations, yielding “functional” and “nonfunctional” subtypes. Functional tumors potentially secreting detectable hormones with current methodologies may display more aggressive behavior with invasive characteristics[5]. Epidemiological data from the United States National Cancer Institute’s Surveillance, Epidemiology, and End Results (SEER) program reveal a striking acceleration in gastric NET incidence over the past four decades, reaching 0.62 per 100000 in 2016[6]. Current multinational epidemiological evidence consistently demonstrates a progressive increase in D-NET incidence across multiple populations[7,8].

DIAGNOSIS OF D-NETs

Etiological complexity and clinical heterogeneity of D-NETs

The pathogenesis of D-NETs remains incompletely elucidated, though emerging evidence identifies significant associations with diabetes mellitus, 2hypertension, low serum high-density lipoprotein cholesterol levels, and past/present Helicobacter pylori infection[9]. This multicenter case–control study enrolled 396 patients with gastric NETs and 193 patients with D-NETs, with comparisons made against 1725 healthy controls. Multivariable analysis revealed that D-NETs were significantly associated with diabetes mellitus [odds ratio (OR) = 1.80], hypertension (OR = 1.68), low serum high-density lipoprotein cholesterol levels (OR = 2.29), and past or current Helicobacter pylori infection (OR = 5.42). The underlying etiology and pathogenesis remain unclear, warranting further investigation through additional literature review and studies[9].

The clinical presentation of D-NETs demonstrates remarkable heterogeneity, with functional variants exhibiting diverse manifestations mediated by ectopic hormone production. Nonfunctional D-NETs typically manifest through mass effect-related complications, including acute pancreatitis or obstructive jaundice resulting from local tumor progression[10,11]. These tumors may also present concurrently with metabolic disorders such as diabetes and hypertension[12]. A significant proportion of D-NETs demonstrate association with multiple endocrine neoplasia type 1 syndrome[13], frequently exhibiting occult clinical courses. Aggressive variants often debut with metastatic manifestations, with retroperitoneal lymphadenopathy representing a common initial presentation[14]. Alternatively, patients may present with nonspecific gastrointestinal symptoms, including diarrhea, melena, or nausea/vomiting, leading to incidental diagnosis during subsequent evaluation[15,16].

Diagnostic challenges and current modalities in D-NETs

The diagnostic paradigm for D-NETs remains challenging due to their clinical heterogeneity. In patients with established NETs, potential hormone excess syndromes - including paraneoplastic or carcinoid syndromes - must be systematically evaluated. Vilallonga et al[17] documented that among 2155 colorectal cancer patients, five were subsequently diagnosed with NETs, all presenting with paraneoplastic or carcinoid manifestations. However, the clinical presentation of D-NETs frequently lacks specificity, and not all tumors exhibit classical syndrome patterns. A retrospective analysis of 927 duodenal endocrine tumor patients revealed a mere 3.1% incidence of carcinoid syndrome[18], suggesting even lower prevalence in D-NETs specifically. Drawing parallels to gastric NET classification, types I and III typically lack paraneoplastic or carcinoid syndromes, further complicating diagnostic recognition.

Serological biomarkers, including chromogranin A, neuron-specific enolase, and relevant hormone panels, provide valuable diagnostic adjuncts. The multiplex serological monitoring strategy significantly enhances diagnostic precision by improving both the specificity and sensitivity for neuroendocrine tumor detection[19-21]. Regarding imaging modalities, multiple studies and clinical guidelines support the utility of contrast-enhanced computed tomography and endoscopic evaluation for GEP-NET detection[22-24]. In alignment with European Society for Medical Oncology (ESMO) recommendations, ⁶⁸Ga-DOTATOC positron emission computed tomography emerges as a pivotal diagnostic tool, offering superior sensitivity and specificity for gastrointestinal neuroendocrine tumor characterization[25].

TREATMENT OF D-NETs

Histopathological spectrum and therapeutic stratification of D-NETs

D-NETs demonstrate five distinct histopathological subtypes: Gastrinomas (50%-60%), somatostatin-producing tumors (15%), non-functional serotonin-containing tumors (20%), poorly differentiated neuroendocrine carcinomas (< 3%), and gangliocytic paragangliomas (< 2%). The majority localize to the duodenal bulb and post-bulbar region, with approximately 20% occurring in the periampullary area[26]. Therapeutic decision-making for D-NETs requires a comprehensive evaluation of tumor size, anatomical location, histological grade, staging, and specific subtype (Table 1).

Table 1 Concise classification and grading criteria for neuroendocrine neoplasms of the gastrointestinal tract and hepatopancreatobiliary organs.

Terminology	Differentiation	Grade	Mitotic rate1 (mitoses/2 mm2)	Ki-67 index2
NET, G1	Well differentiated	Low	< 2	< 3%
NET, G2		Intermediate	2-20	3%-20%
NET, G3		High	> 20	> 20%
NEC, small-cell type (SCNEC)	Poorly differentiated	High3	> 20	> 20%
NEC, large-cell type (LCNEC)			> 20	> 20%

¹Mitotic rates are to be expressed as the number of mitoses/2 mm² as determined by counting in 50 fields of 0.2 mm² (i.e., in a total area of 10 mm²).

²The Ki-67 proliferation index value is determined by counting at least 500 cells in the regions of highest labelling (hot-spots), which are identified at scanning magnification; the final grade is based on whichever of the two proliferation indexes places the neoplasm in the higher-grade category.

³Poorly differentiated neuroendocrine carcinomas are not formally graded, but are considered high-grade by definition.

LCNEC: Large-cell neuroendocrine carcinoma; NEC: Neuroendocrine carcinoma; NET: Neuroendocrine tumour; SCNEC: Small-cell neuroendocrine carcinoma.

Endoscopic resection

For small (≤ 10 mm), non-functional G1 tumors confined to the submucosa in the supra-ampullary region, endoscopic resection has been proven to be safe and effective. Endoscopic resection may be performed utilizing either endoscopic submucosal dissection or endoscopic mucosal resection techniques. The combined application of multiple endoscopic techniques - including cold snare polypectomy, endoscopic papillectomy, endoscopic mucosal resection, and endoscopic submucosal dissection - constitutes a viable therapeutic strategy for specific duodenal tumor subtypes, such as multifocal duodenal polyposis associated with familial adenomatous polyposis[27]. Lesions measuring 10-20 mm may be managed through either endoscopic or surgical approaches.

Surgical management

For tumors exceeding 10 mm with suspicious features or positive resection margins, formal surgical resection is warranted. The surgical strategy - ranging from local excision to antrectomy or total gastrectomy - should be tailored according to pathological characteristics and invasive potential[4,28]. For ampullary D-NETs, characterized by elevated rates of lymphatic and metastatic dissemination, pancreaticoduodenectomy (Whipple procedure) with systematic lymphadenectomy remains the cornerstone of curative intervention, irrespective of tumor size or histological grade. Current evidence indicates comparable nodal metastasis rates between periampullary and non-periampullary NETs, with conventional imaging demonstrating limited sensitivity for metastatic detection[29]. The absence of reliable clinicopathological predictors for nodal involvement necessitates aggressive surgical approaches. Thus, pancreaticoduodenectomy represents the optimal strategy for achieving curative resection in medically fit patients with second-stage D-NETs. However, the impact of lymphadenectomy on long-term survival outcomes remains inadequately characterized[30].

Tumors originating from the minor duodenal papilla present unique management challenges. A comprehensive review of 44 cases revealed a mean tumor size of 14.0 mm (range: 2-27 mm), predominantly G1 grade (20/22). Notably, lymph node metastasis was documented in 58.3% (14/24) of lesions ≥ 10 mm, indicating substantially elevated metastatic potential in this subset[31]. A representative case documented an 8-mm G1 neuroendocrine tumor of the minor papilla, where robotic pancreaticoduodenectomy identified nodal metastases in the retro-pancreatic region despite a low Ki-67 index (< 1%)[32]. Thus, sentinel lymph node biopsy should be considered for patients with NETs located at the minor duodenal papilla.

Emerging data challenge conventional size-based treatment paradigms, with documented nodal metastases in multifocal NETs < 1 cm in diameter[33]. Consequently, pancreaticoduodenectomy with regional lymphadenectomy should be considered for multifocal non-functional D-NETs. Intraoperative frozen section analysis provides valuable guidance when malignancy remains uncertain during exploration[34]. For accurate pathological staging, retrieval of at least 8 regional lymph nodes is recommended during curative-intent procedures[35].

Systemic therapeutic strategies for GEP-NETs

While surgical resection remains the sole curative modality for localized D-NETs and other GEP-NETs[36,37], the propensity for early nodal dissemination and distant micrometastases in neuroendocrine carcinomas necessitates multimodal approaches. For locally advanced disease, radical resection followed by adjuvant chemotherapy represents the current standard. However, the majority of neuroendocrine carcinoma patients present with synchronous distant metastases, rendering them unsuitable for surgical cure. In such cases, systemic chemotherapy emerges as the primary treatment modality.

The etoposide-platinum regimen, originally developed for small-cell lung cancer, demonstrates 67% response rates in GEP-NETs[38]. First-line options further include epirubicin-cyclophosphamide and irinotecan-cisplatin combinations, while second-line alternatives encompass FOLFOX (5-fluorouracil/Leucovorin/oxaliplatin), FOLFIRI (5-fluorouracil/Leucovorin/irinotecan), and CAPTEM (capecitabine/temozolomide). Immune checkpoint inhibition with anti-programmed cell death protein 1 antibodies demonstrates efficacy in molecularly selected populations harboring microsatellite instability-high, mismatch repair-deficient, or high tumor mutational burden characteristics, warranting consideration as a third-line intervention[39]. Furthermore, immunotherapy combinations, which engage synergistic mechanisms to enhance antitumor immune responses and confer therapeutic benefits across multiple cancer types, represent among the most dynamic areas in oncology. Examples include immune-chemotherapy for advanced solid tumors[40,41], immune-targeted therapy for hepatocellular carcinoma[42], immune-bispecific antibody therapy for extensive-stage small cell lung cancer[43], and chimeric antigen receptor T-cell immunotherapy for relapsed/refractory acute lymphoblastic leukemia[44]. The evolution of programmed death-ligand 1 inhibitor-based combination strategies has progressed from initial immune-chemotherapy paradigms toward more precise, multi-mechanistic synergistic modalities, an approach that holds translational promise for the treatment of gastrointestinal neuroendocrine carcinoma.

For somatostatin receptor-positive NETs, subcutaneous or intramuscular administration of somatostatin analogs (octreotide long-acting repeatable 10-30 mg monthly or lanreotide 60-120 mg monthly) effectively controls hormonal secretion and associated clinical syndromes[39,45]. Beyond neuroendocrine carcinomas, conventional and targeted therapies generally occupy secondary roles in NET management. Combination regimens including cisplatin-etoposide or streptozotocin with doxorubicin/5-fluorouracil achieve partial responses in > 50% of patients based on radiologic and serologic criteria[46,47].

Molecularly targeted agents have demonstrated significant survival benefits. Continuous daily administration of everolimus (37.5 mg) improves progression-free survival, overall survival, and objective response rates in advanced NETs, extending median progression-free survival by 6.4 months compared to placebo[48,49]. For isolated hepatic metastases, locoregional approaches including radiofrequency ablation, transarterial embolization, and chemoembolization provide effective disease control[50]. Selective internal radiation therapy with ⁹⁹Yttrium or ¹⁷⁷Lutetium-labeled analogs, along with ¹³¹I-MIBG, presents additional therapeutic options for neuroendocrine tumor liver metastases[51-53], expanding the armamentarium against advanced disease.

PROGNOSTIC REFINEMENT AND SURVIVAL OUTCOMES

The evolving therapeutic landscape - characterized by aggressive surgical intervention and the emergence of second-line therapies including long-acting somatostatin analogs and molecularly targeted agents - has progressively improved clinical outcomes and long-term survival for patients with D-NETs. Nevertheless, multicenter studies with larger cohorts are warranted to definitively characterize long-term survival outcomes, as existing datasets remain limited. Circulating tumor DNA (ctDNA), a cornerstone of liquid biopsy, is fundamentally reshaping the comprehensive management of tumors, including duodenal neuroendocrine neoplasms, by serving as a real-time molecular surveillance tool that informs therapeutic decisions and prognostic evaluation. It represents the most robust prognostic biomarker for detecting minimal residual disease post-operatively, where its detection within 2-10 weeks after curative resection serves as an independent early warning signal of a significantly elevated risk of recurrence[54-56]. For instance, in stage III colon cancer, patients with post-operative ctDNA positivity had a 3-year recurrence risk of 64%, compared to only 7% in those with negative ctDNA[57].

In guiding adjuvant therapy, minimal residual disease assessment is pivotal for risk-adapted “treatment escalation or de-escalation” strategies. The Circulating Tumor DNA to Guide Adjuvant Therapy in Stage II Colon Cancer demonstrated that in stage II colon cancer, restricting adjuvant chemotherapy to patients with post-operative ctDNA positivity is a safe approach that significantly reduces unnecessary treatment[55]. Conversely, for high-risk patients with persistent ctDNA positivity post-operatively (e.g., as explored in the ALTAIR study) or after adjuvant chemotherapy, intensified therapeutic strategies are under investigation[54].

All patients diagnosed with D-NETs require a comprehensive baseline evaluation and structured clinical surveillance during hospitalization and throughout follow-up to detect disease progression or recurrence. Current consensus recommends clinical assessments at 3-month to 6-month intervals during the initial three years post-diagnosis, transitioning to 6-month to 12-month intervals thereafter (Figure 1). Each evaluation should incorporate conventional laboratory parameters, targeted tumor marker profiling (including carcinoembryonic antigen, carbohydrate antigen 19-9, chromogranin A, and 5-hydroxyindoleacetic acid), gastrointestinal endoscopy, abdominal and nodal basin ultrasonography, thoracic radiography, and cross-sectional imaging via computed tomography or whole-body positron emission computed tomography[58].

Notably, patients with localized, well-differentiated D-NETs undergoing complete surgical resection demonstrate a favorable prognosis, with 5-year survival rates approaching 70%[2]. Several clinicopathological factors have been consistently associated with improved survival, including R0 resection of the primary tumor, absence of hepatic metastases, management of metachronous liver metastases, and aggressive treatment strategies for synchronous metastatic disease[59]. These prognostic determinants provide a valuable framework for risk stratification and long-term outcome prediction in D-NET management.

CONCLUSION

D-NETs represent a heterogeneous and uncommon neoplasm, frequently diagnosed serendipitously during clinical evaluation. Significant conceptual ambiguity persists regarding their clinicopathological manifestations, diagnostic algorithms, and therapeutic paradigms. Figure 1 comprehensively delineates the diagnostic and therapeutic algorithm for D-NETs and is intended to serve as a valuable reference for clinicians. Our review provides a lucid and structured synthesis of contemporary evidence, delineating current understanding and critical knowledge gaps in D-NET pathophysiology and management.

Figure 1 Comprehensive diagnostic and therapeutic algorithm of duodenal neuroendocrine tumors. CT: Computed tomography; MRI: Magnetic resonance imaging; NET: Neuroendocrine tumor; NEC: Neuroendocrine carcinoma; EP: Etoposide + cisplatin; IP: Irinotecan + cisplatin; EC: Etoposide + carbopla.