Recent updates on the endoscopic treatment of rectal neuroendocrine tumor

Article information

Intest Res. 2025;.ir.2025.00141

Publication date (electronic) : 2025 November 27

doi : https://doi.org/10.5217/ir.2025.00141

Sunghyeok Ryou, Kwangwoo Nam

Department of Gastroenterology, Dankook University Hospital, Dankook University College of Medicine, Cheonan, Korea

Correspondence to Kwangwoo Nam, Department of Gastroenterology, Dankook University Hospital, Dankook University College of Medicine, 201 Manghyang-ro, Dongnam-gu, Cheonan 31116, Korea. E-mail: nambag1108@gmail.com

Received 2025 July 18; Revised 2025 August 18; Accepted 2025 September 8.

Abstract

INTRODUCTION

Gastroenteropancreatic neuroendocrine tumors (GEP-NETs) arise from enterochromaffin-like cells and can develop anywhere along the gastrointestinal (GI) tract, from the foregut to the hindgut. GEP-NETs encompass both GI and pancreatic NETs, and they account for approximately 2% of all GI malignancies. Among GI-NETs, rectal NETs (R-NETs) originated from the hindgut and are generally considered less aggressive than NETs arising from the small intestine or colon. The incidence of R-NETs has gradually increased in recent decades, especially in East Asia, largely due to the widespread use of colonoscopy for colorectal cancer screening program and better awareness of the disease.

Endoscopic resection is commonly performed for small, superficial, and well-differentiated GI-NETs in the stomach, duodenum, and rectum [1]. Therefore, a thorough understanding of the disease and careful selection of the patients based on appropriate indications are essential for optimal endoscopic management of R-NETs. In this manuscript, we reviewed the recent advances in the management and endoscopic treatment strategies for R-NETs.

EPIDEMIOLOGY OF R-NET

The incidence and prevalence of GEP-NETs have markedly increased over the last decades [2,3]. According to the Surveillance, Epidemiology, and End Results data, the age-adjusted incidence rate of GEP-NETs rose 6.4-fold from 1975 to 2015 (annual percentage change [APC], 4.98; 95% confidence interval [CI], 4.75–5.20; P<0.01) [3]. Among them, R-NETs demonstrated the most significant increase (APC, 6.43; 95% CI, 5.65–7.23; P<0.001), whereas colonic NET exhibited the least significant increase (APC, 2.87; 95% CI, 2.28–3.47) [3]. In addition, the 20-year limited-duration prevalence of GEP-NETs also significantly increased, from 0.00138% in 1996 to 0.03917% in 2015. Among all primary sites, R-NETs had the highest prevalence, reaching 0.01505% in 2015 [3]. These trends are largely due to increased awareness of the disease and increased use of screening colonoscopy. Most R-NETs are asymptomatic and detected in screening colonoscopy, and the overall survival for localized R-NET is excellent (>25 years), which is better than that of localized colonic or cecal NET [2].

DIAGNOSIS AND INITIAL WORKUP OF R-NET

Typical endoscopic features of R-NET are yellowish or whitish protruding subepithelial lesions in the rectum [1]. As most R-NETs originate from the deep mucosal layer, the diagnostic yield of forceps biopsy is about 40% to 70%. However, post-biopsy scarring and fibrosis might make further endoscopic resection more difficult.

Considering its malignant potential, a thorough imaging workup is recommended for all R-NET cases. European Neuroendocrine Tumor Society (ENETS) and National Comprehensive Cancer Network (NCCN) guidelines recommend endoscopic ultrasound (EUS) or rectal (pelvic) magnetic resonance imaging (MRI) for initial local staging [4,5]. EUS has advantages for assessing the characteristics of R-NETs and determining their depth of invasion and regional metastatic lymph nodes [4]. Rectal MRI has advantages in the detection and characterization of regional metastatic lymph node and is recommended for tumors ≥10 mm, World Health Organization (WHO) grade G2-G3, or lymph node involvement is suspected [4]. In addition, abdominopelvic and chest computed tomography (CT) is needed to assess distant metastasis. Somatostatin receptor-based positron emission tomography (PET)-CT, which offers superior sensitivity and specificity for detecting NETs, is used as a part of distant metastasis staging workup (Fig. 1) [4].

Fig. 1.

Suggested management algorithm for rectal neuroendocrine tumor. EUS, endoscopic ultrasound; MRI, magnetic resonance imaging; CT, computed tomography; PET, positron emission tomography; ER, endoscopic resection (modified endoscopic mucosal resection or endoscopic submucosal dissection); EFTR, endoscopic full-thickness resection; SR, surgical resection; CTx, chemotherapy; LVI, lymphovascular invasion; TAMIS, transanal minimally invasive surgery; FU, follow-up. Modified from Rinke A, et al. J Neuroendocrinol 2023;35:e13309. [4]

R-NETs are classified by pathologic grade according to the WHO classification, based on mitotic count and Ki-67 proliferation index. Tumors classified as G1 and G2 are considered well-differentiated tumors, whereas G3 is considered as poorly differentiated tumors and referred to as neuroendocrine carcinomas (Table 1).

Table 1.

Grade of Rectal Neuroendocrine Tumors

ENDOSCOPIC TREATMENT OF R-NET

Endoscopic resection is the standard treatment not only for early GI cancer but also for early-stage R-NET (T1N0M0) [6]. Small (<10 mm) and superficial R-NETs without high-risk features are generally good candidates for endoscopic resection [7]. Previous studies have demonstrated that the incidence of lymph node metastasis in R-NETs ≤10 mm is significantly lower than in R-NETs 11–20 mm (1.1%–2.5% vs. 6.6%–12.8%) [4]. Nonetheless, complete resection should be achieved given their malignant potential. Until now, various endoscopic resection techniques have been developed. Conventional forceps or snare polypectomy is generally not recommended due to their high incomplete resection rate, as enteroendocrine cells often extend into the subepithelial layer, and submucosal invasion has been reported up to 76% of R-NETs (Table 2) [1,8].

Table 2.

Variable Methods of Endoscopic Resection for Rectal Neuroendocrine Tumors

Conventional endoscopic mucosal resection (EMR) is technically simple, less invasive, and fast procedure. However, it has shown inferior outcomes in terms of en bloc and complete resection rates compared with modified EMR (mEMR) techniques. EMR with transparent cap (EMR-C), EMR with ligation (EMR-L), and EMR with precutting (EMR-P) are representative methods of mEMR (Table 2) [9]. Recently, anchored snare-tip (tip-in) EMR has been introduced, showing shorter procedure time with comparable efficacy [9,10]. The key principle of mEMR is to achieve deeper submucosal resection compared to conventional methods. For example, EMR-L demonstrated a significantly lower rates of deep margin positivity compared with conventional EMR (6% vs. 46%; P=0.029) [11]. Likewise, EMR-P showed favorable en bloc (98.6%) and complete resection rates (93.1%) [12]. In selected cases with sufficient expertise, endoscopic submucosal dissection (ESD) offers high complete resection rates which could reduce the risk of recurrence. However, ESD is associated with a higher risk of adverse events such as hemorrhage and perforation compared to mEMR. In previous study, mEMR and ESD showed comparable complete resection rate (90.0% vs. 82.3%) and post-procedural bleeding rate (5.5% vs. 2.8%) [13]. More recently, endoscopic full-thickness resection (EFTR) using a dedicated over-the-scope clip and snare was introduced. EFTR enables complete full-thickness resection around the lesion, which minimize the chance of residual tumor, and also useful for additional resection after incomplete resection of R-NET (Fig. 1) [14,15]. However, it is currently not reimbursed and is therefore rarely used in Korea.

According to the ENETS guideline, mEMR, ESD, or EFTR are considered as feasible treatment options for R-NET <10 mm [4]. For tumors measuring 10–20 mm without definite lymph node metastasis, a personalized approach including full imaging and multidisciplinary discussion is recommended to determine whether endoscopic or surgical resection is suitable. For lesions >20 mm without definite metastasis, radical surgical resection such as low anterior resection or abdominopelvic resection is generally recommended (Fig. 1) [5].

In cases of incomplete endoscopic resection of small R-NETs, additional salvage endoscopic resection can be performed to eliminate residual tumors. A recent study investigating salvage endoscopic resection for previous incomplete or inappropriate resection revealed that residual tumors were significantly more common in conventional forceps or snare polypectomy and conventional EMR compared with mEMR or ESD (41.7%–70.0% vs. 0%). Furthermore, prior incomplete resection was significantly associated with residual tumor risk compared to prior complete resection (65.7% vs. 0%) [8].

POST-RESECTION SURVEILLANCE AND LONG-TERM FOLLOW-UP OF R-NET

Surveillance strategy following initial treatment of R-NET is based on the tumor size and outcome of the resection. According to the recent NCCN guideline, no routine follow-up is recommended after complete resection of small (<10 mm) RNET, considering the very low risk of lymph node or distant metastasis [5]. However, lymph node metastasis has been reported even in a small, low-risk R-NET although it is rare and its progression speed was very slow [16]. Therefore, a no follow-up policy should be adopted with caution, and only after careful patient selection in this situation.

For tumors measuring 10–20 mm, the NCCN guideline recommends colonoscopy with rectal MRI or EUS at 6 months and 12 months after resection [5]. After 12 months from the initial resection, follow-up with biochemical markers and abdominopelvic/chest CT were recommended every 1–2 years up to 10 years to monitor the local recurrence or distant metastasis [5].

According to the recent ENETS guideline, for small ( <10 mm) R-NET G1 without lymphovascular invasion after complete endoscopic resection, no routine follow-up is recommended. After complete resection of small R-NET G1 with lymphovascular invasion or G2/3, follow-up with pelvic MRI every 6 months and sigmoidoscopy yearly for at least 5 years is recommended. After incomplete resection of small R-NET G1/G2(low), follow-up with endoscopy and EUS or MRI yearly for at least 5 years is recommended. In addition, PET-CT is recommended initially and after 12 months [4].

After complete resection of R-NET G1/G2 ≥ 10 mm, followup with MRI every 6 months and sigmoidoscopy yearly for at least 5 years is recommended. After complete resection of R-NET G3, follow-up with MRI every 3 months for 2 years, followed by 6 months for 5 years, and sigmoidoscopy 6 months for 2 years, followed by yearly for 5 years is recommended [4].

Chromogranin A (CgA) is a useful immunohistochemical marker in R-NET, although it is less commonly expressed than synaptophysin. A previous study reported that CgA positivity was more frequently observed in patients aged ≥50 years, male, those with incomplete resection, and more advanced stages. In addition, CgA expression was significantly associated with shorter survival outcomes in R-NET [17]. In the same study, patients with CgA positivity also showed elevated plasma CgA levels, suggesting its potential usefulness as a circulating tumor marker in R-NET [18]. However, other studies have reported no significant clinical association between plasma CgA level and the prognosis [19,20]. Therefore, further studies are needed to evaluate the clinical significance of plasma CgA as a biomarker in R-NET.

PROGNOSTIC FACTORS

Most R-NETs are small, localized, and superficial (confined to mucosa or submucosa) lesions, which carry a relatively low risk of lymph node metastasis. However, even in small size, high grade (G3), and positive lymphovascular invasion are important risk factors for lymph node metastasis. In these cases, additional surgical resection is generally warranted to decrease the risk of recurrence during surveillance (Fig. 1).

In addition, incomplete resection is another common reason for considering additional surgical resection. However, recent studies suggested that incomplete resection (R1) was not significantly associated with increased risk of metastasis or poor treatment outcomes, although the number of studied patients was very small [21,22]. It has been hypothesized that electrodiathermic burns themselves can cause coagulation necrosis at the remnant resection margin and may decrease the risk of local recurrence during follow-up, even in R1 cases [23]. Therefore, personalized management is warranted for the patients with R1 resection of R-NET and conservative management with close follow-up may be a valid option in selected patients (Fig. 1) [1].

Endoscopic resection is usually performed as a first treatment option in small ( <10 mm) R-NET. However, for more larger lesions measuring 10–20 mm, it is unclear whether endoscopic or surgical resection is more appropriate [4]. A previous study investigating the efficacy of endoscopic resection of 10–20 mm R-NET reported that lymph node or distant metastasis occurred in 2.9% of the patients during mean follow-up of 41.2 months. In this study, ESD was associated with significantly better long-term outcomes including a lower recurrence rate compared to mEMR or conventional EMR (0% vs. 6.9% vs. 12.5%; P=0.041) [21]. Although there are still limited data, endoscopic resection of large (10–20 mm) R-NET might be a feasible alternative option to decrease the unnecessary surgical resection in selected cases, especially in patients with older age and many comorbidities.

Multiple R-NETs at the time of diagnosis are not common, with an incidence of 2% to 5.7%. In some cases, it is difficult to differentiate multiple R-NETs from hyperplastic or adenomatous polyps [1]. The association between multiple R-NET and inflammatory bowel disease has been reported, and multiple tumors ( ≥ 8) were reported to be associated with a higher risk of lymph node metastasis [24]. Therefore, careful assessment is needed in cases of multiple R-NET.

To date, there was not enough data about molecular characteristics on R-NET. Although rare, BRAF V600E and D594G mutations have been reported in R-NET, especially in neuroendocrine carcinoma, suggesting the potential usefulness as viable molecular markers in R-NET [25]. Further studies are needed to investigate the potential molecular markers of R-NET.

SUMMARY AND CONCLUSION

R-NETs are often detected in screening colonoscopy as early stage with small and superficial tumors. Considering its malignant potential, a comprehensive initial evaluation, including EUS and cross-sectional imaging studies, is needed. Endoscopic resection is the main treatment option in small or some large R-NETs, and mEMR or ESD are recommended as proper methods. Achieving complete resection is crucial to minimizing the risk of metastasis during follow-up. Long-term surveillance, including colonoscopy and imaging, is recommended for up to 5–10 years, particularly in patients with high-risk features or incomplete initial resection.

Article information Continued