Is salvage surgery necessary for non-R0 endoscopic resection of small rectal neuroendocrine tumors?

Abstract

BACKGROUND

The rate of R0 endoscopic resection for small rectal neuroendocrine tumors (r-NETs) is nearly 90%. However, there is still controversy about subsequent treatment options for patients with non-R0 endoscopically resected r-NETs.

AIM

To investigate the necessity of salvage surgery for non-R0 endoscopic resection of small r-NETs.

METHODS

A single-center retrospective study. Thirty-eight patients with r-NETs who underwent non-R0 endoscopic resection at Sun Yat-sen University Cancer Centre between 2007 and 2023 were retrospectively analyzed.

RESULTS

Thirty patients underwent salvage surgery while 8 received follow-up only. Among those 30 patients with salvage surgery, 12 cases underwent salvage endoscopic submucosal dissection, while 16 underwent salvage transanal endoscopic microsurgery, 2 underwent salvage radical resection (low anterior resection). Postoperative pathology following salvage surgery indicated an absence of tumor cells in all 30 patients. Among the 8 patients who received follow-up only, 3 received endoscopic biopsy of the scar during follow-up. Pathological evaluation of the biopsy specimens also showed no residual tumor cells. All 38 patients underwent regular endoscopic and imaging follow-up with an average follow-up duration of 46.4 ± 14.8 months. No evidence of local tumor recurrence or distant metastasis was observed.

CONCLUSION

These findings suggest that salvage surgery may not be essential for patients with non-R0 endoscopically resected r-NETs, with regular follow-up being a safe alternative.

Key Words: Rectal neuroendocrine tumors; Endoscopic resection; Non-R0 resection; Salvage surgery; Follow-up

Core Tip: Rectal neuroendocrine tumors (r-NETs) are diminutive but potentially malignant neoplasms. The rate of R0 resection for endoscopic resection of small r-NETs is nearly 90%. However, there is still controversy about subsequent treatment options for patients with non-R0 endoscopically resected r-NETs. This retrospective study investigated 38 patients with non-R0 endoscopically resected r-NETs. Of these, 30 underwent salvage surgery while 8 received follow-up only. Postoperative pathology following salvage surgery indicated an absence of tumor cells in all 30 patients. No evidence of local tumor recurrence or distant metastasis was observed in all 38 patients during an average follow-up duration of 46.4 ± 14.8 months. These findings suggest that salvage surgery may not be essential for patients with non-R0 endoscopically resected r-NETs, with regular follow-up being a safe alternative.

INTRODUCTION

Rectal neuroendocrine tumors (r-NETs) are diminutive but potentially malignant neoplasms. Their management depends upon the dimensions of the tumor[1]. Current consensus recommendations for treating r-NETs recommend the use of endoscopic resection for small tumors (< 1 cm) as these have a relatively low likelihood of metastatic dissemination[2,3].

The rate of R0 resection for endoscopic resection of small r-NETs is nearly 90%[4-6]. However, there is still controversy about subsequent treatment options for patients with non-R0 endoscopically resected r-NETs. The NCCN guidelines recommend further endoscopic or local surgical excision for G1/G2-grade r-NETs if endoscopic resection is incomplete[2]. However, some experts suggest that for r-NETs not undergoing R0 resection, surveillance may be an option[7,8]. This is because small r-NETs generally have a favorable prognosis, and due to postoperative fibrosis, salvage surgery is technically challenging and carries a higher risk of perforation. Thus, we conducted a retrospective analysis of the clinical and pathological data of patients with r-NETs who underwent salvage surgery after non-R0 endoscopic resection to investigate the need for salvage surgery.

MATERIALS AND METHODS

The data of 38 patients with r-NETs of clinical stage 1 (cT1N0M0) who underwent non-R0 endoscopic resection at Sun Yat-sen University Cancer Centre between 2007 and 2023 were retrospectively analyzed. An R0 resection was operationally defined as a surgical procedure in which both the lateral and vertical margins are found to be negative when inspected microscopically. Non-R0 resection (also R1 resection) was operationally defined as a surgical procedure with evidence of either positive vertical or positive lateral margins upon microscopic inspection of the resected tumors. The criteria for inclusion in the study were: (1) The lesion is located in the rectum and measures less than 20 mm in size; (2) Prior to the endoscopic resection, examinations using endoscopic ultrasonography/magnetic resonance imaging/computed tomography did not reveal any evidence of invasion into the muscularis propria, involvement of local lymph nodes, or presence of distant metastases; (3) The lesion was removed endoscopically via endoscopic mucosal resection (EMR) or endoscopic submucosal dissection (ESD); (4) The diagnosis of r-NETs was confirmed by histopathology; and (5) The resected tumors demonstrated evidence of either positive vertical or positive lateral margins upon microscopic inspection. The exclusion criteria included: (1) Lesions were removed via endoscopic biopsy forceps or cold knife resection; (2) Patients combined with other types of malignant tumors; and (3) Patients with a follow-up period of less than 6 months. The study protocol conformed to the guidelines of the 1964 Declaration of Helsinki. This retrospective research was approved by the Ethics Committee of the Institutional Review Board of the Sun Yat-sen University Cancer Center (No. SL-B2023-722-01). The requirement for informed consent was waived.

For those patients who have undergone non-R0 endoscopic resection of r-NETs, salvage local resection, such as ESD or transanal endoscopic microsurgery, is recommended. However, some patients still declined salvage local resection due to personal preference and instead opted for salvage radical resection (low anterior resection) or regular follow-up only. For patients who underwent salvage surgery, annual colonoscopy and imaging evaluation (pelvic computed tomography or magnetic resonance imaging) were recommended. For patients who did not undergo salvage surgery, colonoscopy and imaging evaluation were recommended every three months during the first year, followed by annual colonoscopy and imaging evaluation thereafter.

Statistical analysis

Continuous variables were represented using means ± SD, or medians and interquartile ranges. Categorical variables were represented as proportions. Variables were compared via a variety of statistical tests, including t-tests, Kruskal-Wallis tests, χ² tests, and Fisher’s exact tests, as appropriate for the data. Statistical testing was conducted using SPSS (Version 31.0, IBM Corp., Armonk, NY, United States).

RESULTS

Table 1 presents a comprehensive overview of the clinical and pathological data of the 38 patients with non-R0 endoscopically resected r-NETs. The mean age of the patients was 45.7 ± 13.6 years, while the average lesion size was 7.2 ± 2.6 mm. Twenty-two (57.9%) of the patients were male, and 16 cases (42.1%) had tumors located within 5 cm of the anus. Eighteen cases (47.4%) underwent EMR, and 20 (52.6%) underwent ESD. All 38 patients had positive vertical margins after endoscopic resection, and 32 cases (84.2%) were G1 grade.

Table 1 The characteristics of these 38 patients with non-R0 endoscopic resection.

	Total (n = 38)
Age (year), mean ± SD	45.7 ± 13.6
Lesion size (mm), mean ± SD	7.2 ± 2.6
Gender (%)
Male	57.9 (22/38)
Female	42.1 (16/38)
Location (anus, %)
≤ 5 cm	42.1 (16/38)
6-10 cm	47.4 (18/38)
> 10 cm	10.5 (4/38)
Endoscopic resection method (%)
EMR	47.4 (18/38)
ESD	52.6 (20/38)
Positive margins (%)
Vertical	100.0 (38/38)
Lateral	0 (0/38)
Lymphovascular invasion (%)	0 (0/38)
Histologic grade (%)
G1	84.2 (32/38)
G2	15.8 (6/38)

EMR: Endoscopic mucosal resection; ESD: Endoscopic submucosal dissection.

Figure 1 and Table 2 show the salvage surgery and follow-up data of these 38 non-R0 resection cases. Twelve cases (31.6%) underwent salvage ESD, while 16 (42.1%) underwent salvage transanal endoscopic microsurgery, 2 (5.3%) underwent salvage low anterior resection, and 8 (21.1%) received regular follow-up only. Postoperative pathology indicated an absence of tumor cells in all the 30 cases that underwent salvage surgery. Among the 8 patients who did not undergo salvage surgery, 3 received endoscopic biopsy of the resection scar during follow-up. Pathological evaluation of the biopsy specimens also showed no residual tumor cells. All 38 patients received regular endoscopic and imaging follow-up over a median follow-up of 46.4 ± 14.8 months. No cases of local tumor recurrence or distant metastasis were detected.

Figure 1 The salvage surgery and follow-up data of these 38 patients with non-R0 endoscopic resection. NETs: Neuroendocrine tumors; ESD: Endoscopic submucosal dissection; TEM: Transanal endoscopic microsurgery; LAR: Low anterior resection.

Table 2 The salvage surgery and follow-up data of these 38 patients with non-R0 endoscopic resection.

	Total (n = 38)
Salvage surgery (%)
ESD	31.6 (12/38)
TEM	42.1 (16/38)
LAR	5.3 (2/38)
Follow-up only without salvage surgery	21.1 (8/38)
With endoscopic biopsy	37.5 (3/8)
Without endoscopic biopsy	62.5 (5/8)
Residual tumor cells (%)	0.0 (0/33)
Follow-up time (month), mean ± SD	46.4 ± 14.8
Recurrence (%)	0.0 (0/38)

ESD: Endoscopic submucosal dissection; TEM: Transanal endoscopic microsurgery; LAR: Low anterior resection.

The comparison results between the salvage surgery group and the observation group are presented in Table 3. We found that more cases underwent EMR resection in the observation group (87.5% vs 36.7%, P = 0.02). Additionally, all the 6 patients with G2 grade opted for salvage surgery, although no statistically significant difference was observed. There were no differences between the two groups in terms of age, gender, tumor size, margin status, or lymphovascular invasion.

Table 3 Comparison of these 38 patients with non-R0 endoscopic resection.

	Salvage surgery group	Observation group	P value
n	30	8	/
Age (year), mean ± SD	46.8 ± 14.1	41.8 ± 9.8	0.32
Gender (male, %)	56.7 (17/30)	62.5 (5/8)	1.00
Lesion size (mm)	7.0 ± 2.6	8.1 ± 3.5	0.09
Endoscopic resection method (%)
EMR	36.7 (11/30)	87.5 (7/8)	0.02
ESD	63.3 (19/30)	12.5 (1/8)
Positive margins (%)
Vertical	100.0 (30/30)	100.0 (8/8)	1.00
Lateral	0 (0/30)	0 (0/8)
Lymphovascular invasion (%)	0 (0/30)	0 (0/8)	/
Histologic grade (%)
G1	80.0 (24/30)	100.0 (8/8)	0.31
G2	20.0 (6/30)	0.0 (0/11)
Follow-up time (month), mean ± SD	47.5 ± 13.2	41.9 ± 20.3	0.10
Recurrence (%)	0 (0/30)	0 (0/8)	/

EMR: Endoscopic mucosal resection; ESD: Endoscopic submucosal dissection.

DISCUSSION

The NCCN guidelines recommend further endoscopic or local surgical excision for G1/G2-grade r-NETs if endoscopic resection is incomplete[2]. However, pathophysiological changes, such as inflammation, adhesions, and fibrosis in the local scar area following endoscopic resection, markedly increase the difficulty of performing subsequent local excision, along with higher risks of intraoperative bleeding and perforation.

In 2016, Moon et al[9] in South Korea retrospectively analyzed the clinical data of 407 endoscopically resected r-NETs collected in 16 endoscopy centers. Of these, 148 cases involved non-R0 resection, of which 14 underwent salvage surgery and 134 opted for follow-up observation. Over a median follow-up of 56.8 months, there was no significant difference in local tumor recurrence rates between the R0 and non-R0 resection groups (1.2% vs 2.1%). In 2018, Kwak et al[10] in South Korea retrospectively analyzed 99 cases of < 1 cm rectal submucosal masses removed using biopsy forceps. These were pathologically confirmed as r-NETs with indeterminate margins. Among these, 31 patients opted for follow-up observation only, while 68 underwent salvage endoscopic resection. Over a median follow-up of 6.5 years, no significant difference in the local tumor recurrence rates was observed between the observation and the salvage surgery groups. In 2019, a study by Cha et al[11] in South Korea retrospectively analyzed clinical data from 322 cases of endoscopically resected r-NETs, of which 76 cases involved non-R0 resection, with 38 undergoing salvage surgery and 38 receiving follow-up observation only. Over a median follow-up of 40.5 months, no local recurrence or distant metastasis occurred in any of the 322 cases. In 2021, Li et al[12] in China retrospectively analyzed the clinical data of 428 patients with r-NETs after endoscopic resection. Of these, 54 cases involved non-R0 resection, none of which received salvage therapy. Over a median follow-up of 38 months, no local tumor recurrence or distant metastasis occurred. In 2023, Zhuang et al[13] retrospectively analyzed the clinical data of 326 patients with r-NETs after endoscopic resection in southern China. Of these, 83 cases involved non-R0 resection. Eighteen (21.7%) cases received salvage treatment, while 65 cases (78.3%) refused salvage treatment. Over a median follow-up of 30.9 months, no death, metastasis, or recurrence was found. In 2025, Hayashi et al[8] focused on 47 patients with r-NETs who underwent ≥ 10 years of follow-up. The non-R0 resection rate was 15.7%. No recurrence occurred during the 10-year follow-up period. All of the above studies confirm that non-R0 endoscopic resection of r-NETs carries a favorable prognosis, and regular follow-up without salvage surgery is safe.

However, an article by Cheminel et al[14] in 2024 discussed the issue of the necessity of salvage procedures for non-R0 resected r-NETs. This multicenter retrospective study included 100 patients with non-R0 resected r-NETs. Among them, 86 patients underwent the salvage endoscopic procedure, and histological evaluation indicated residual r-NET tumor cells in 37 patients (43%). This is a meaningful but unexpected result, as the relatively high 43% rate of residual tumor cells in non-R0 resected r-NETs does not align with previous research in this field[7]. We hypothesize that the high rate of residual tumor cells may have been the result of most of the non-R0 resected r-NETs being removed using biopsy forceps (56.5% of cases) or cold snare (27.0% of cases). Only 16.5% of the non-R0 resected r-NETs were resected by EMR, one of the most commonly used electric resection procedures.

It has been suggested that endoscopic resection constitutes an electrosurgical intervention, in which intraoperative electrocautery/electrocoagulation can eradicate tumor cells sufficiently at the resection margin[15]. Thus, most “non-R0 endoscopically resected r-NETs” may represent only “pathologically positive margins” rather than true residual tumor cells. Furthermore, grade 1 r-NETs exhibit relatively benign biological behavior. Consequently, for non-R0 endoscopically resected r-NETs, especially those with a pathological grade of G1, salvage surgery may be unnecessary, with regular follow-up observation being an alternative approach[8].

A significant strength of our study is that we analyzed the pathological results from salvage surgery and follow-up biopsies for r-NETs with non-R0 endoscopic resection, confirming the absence of residual tumor cells. This finding provides clear guidance for clinicians that overtreatment is unnecessary for r-NETs with non-R0 endoscopic resection. However, as this study is only a single-center, small-sample retrospective investigation, future multi-center and prospective studies are still needed to validate these findings.

There are some limitations to be concerned. First, the follow-up period in this study was only 46.4 months, which is insufficient for r-NETs with a favorable prognosis. Consequently, we did not observe related endpoint events such as tumor recurrence. This has somewhat affected the reliability of our findings. Second, the sample size of this study was small, involving only 38 patients from a single center, increasing the risk of type II errors, which means there may be differences that the study fails to detect. Third, this study is a single-center retrospective analysis. Follow-up treatment plans for patients with non-R0 endoscopically resected rectal NETs were jointly determined by clinicians and patients rather than through randomization. Consequently, this study is subject to selection bias.

CONCLUSION

In conclusion, the issue of whether salvage surgery is necessary for non-R0 endoscopically resected r-NETs is worthy of discussion. This current research provides a new perspective on the management of non-R0 resected r-NETs resected using endoscopic electric resection. For non-R0 endoscopically resected stage 1 rectal NETs, especially those with a pathological grade of G1, long-term follow-up observation might be both safe and feasible when conditions for additional salvage resection are unavailable or the patient declines further surgery. Future multi-center and prospective studies are still needed to validate these findings.