Safety, Technical Efficacy, and Long-Term Clinical Outcomes of Endoscopic Therapy for Foregut Neuroendocrine Tumors

Introduction: Early, localized neuroendocrine tumors (NETs) arising from foregut (stomach/duodenum/ampulla) are amenable to endoscopic treatment—low frequency of lymph node and distant metastases. Dearth of literature on technical feasibility, safety, and long-term outcomes of endoscopic resection modalities limits widespread adoption of endotherapy for foregut NETs.

Aims: To outline the technical challenges and long-term clinical outcomes of endotherapy for foregut NETs.

Methods: Prospective, single-center study, patients undergoing endo-therapy for foregut NETs (January 2012–January 2023). Demography, clinical presentation, location, lesion size, number, EUS-layer of origin, endoscopic treatment modality, complications, resection margins, and histology recorded. All patients followed up periodically (3, 6, 12 months and annually) for assessment of recurrence, need for re-intervention.

Results: A total of 120 NETs addressed endoscopically in n = 72 patients (30.6% >1 lesion, median age 57 years (50.3–65), 69.4% males). Distribution-duodenal bulb (66.7%), descending duodenum (17.5%), stomach (12.5%), ampulla (3.3%). Presentation: incidental (74.3%), symptomatic (GI bleed/anemia/chronic diarrhea (25.7%). Median size 12 mm (10–15.75), macroscopic-Yamada Type I/II (92.5%), Type III/IV (7.5%). Layer of origin: submucosa 73.3%, muscularis propria (MP) 6.6%. Endoscopic resection modalities: ESD (n = 47), hybrid-ESD (n = 6), Cap-EMR (n = 5), ligation-EMR (n = 21), conventional-EMR (n = 8), exposed-EFTR (n = 9), device-FTR (n = 8), Snare-ampullectomy (n = 4). Adverse events: inadvertent muscle defects (10.8%)—primary closure in all; delayed bleeding (4.1%): endotherapy. En bloc resection: 92.9%. R0: 85.9%, R1: 14.1%, Histology: G1 (75.6%), G2 (18.9%), perineural/lymphovascular invasion: none. Median follow-up: 25 months (11–63). Recurrence 11.1% (n = 8); new lesion 5.5% (n = 4). 5.5% had >1 episode of recurrence. n = 3–conservative management with depot octreotide; n = 1: partial response to PRRT, n = 5 managed endoscopically, n = 3 surgery ([Table 1]). MP involvement on EUS: higher incidence of R1 resection (OR: 7.8, 95% CI: 1.12–54.62, p = 0.037).

Conclusion: Endoscopic resection for foregut NETs using carefully selected modality is safe, effective, and demonstrates low recurrence rate on long-term follow-up. MP layer invasion is a high-risk factor for incomplete resection by EMR/ESD, EFTR is treatment of choice.

Table 1 Cohort characteristics and outcomes

Cohort characteristics and outcomes	Number, n (%)
Total patients	72
Total lesions	120
Lesion size (largest diameter)
• <10 mm	23 (21.9)
• 10–14 mm	50 (47.6)
• 15–19 mm	17 (16.2)
• >20 mm	15 (14.3)
Type of procedure
• ESD/hybrid ESD	47 (43.5)/6 (5.5)
• Conventional EMR/band/ligation EMR/cap EMR	8 (7.4)/21 (19.4)/5 (4.6)
• Exposed EFTR/device FTR	9 (8.3)/8 (7.4)
• Snare-ampullectomy	4 (3.7)
Prophylactic closure
• Endoclips alone	36 (36.3)
• Endoloop and clips	8 (8.1)
• OTSC/omental patch (FTRD)	13 (13.1)
• Endo-suture	2 (2.02)
• No closure	19 (19.2)
• Prophylactic CBD and PD stenting	4 (4.04)
• Not applicable (ligation/band EMR)	21 (21.2)
Adverse events
• Inadvertent muscle defect	8 (10.8)
• Bleeding	3 (4.1)
Type of resection
• En bloc	92 (92.9)
• Piecemeal (6 Hybrid ESD, 1 FTRD + EMR)	7 (7.1)
• Ligation without resection	21 (17.5)
Resection margins
• R0	85 (85.9)
• R1	14 (14.1)
Follow-up duration in months (median, IQR)	25 (11–63)
Recurrence	8 (11.1)
New lesions	4 (5.5)

Publication History

Article published online:
22 April 2024