Our study demonstrated that incompletely resected small G1 grade rectal NET (≤ 10 mm) was safely followed up for at least 2 years without the need for further resection. Furthermore, ESMR-L showed the highest success rate in the complete resection of small rectal NET among all the other endoscopic resection methods.
A special focus was directed at the incidental detection of rectal NET during screening colonoscopy at a local clinic. Among all the previously reported prognostic factors contributing to incomplete resection, the consistent factors were tumor diameter larger than 10 mm, depth in the proper muscle, and the presence of lymphovascular invasion [7–9]. Since EUS is not available in most local clinics, the decision should depend on the tumor size and the knowledge that the tumor is not a simple polyp but rather a subepithelial tumor which could be present in areas deeper than the mucosa.
Upon comparing different endoscopic procedures in terms of the rate of complete resection, ESMR-L procedure showed the highest rate (108/113, 95.6%) followed by ESD procedure (10/21, 47.7%) (Fig. 1). Our group has previously reported that ESMR-L had a significantly higher complete resection rate than ESD (53/53, 100% vs 13/24, 54.2%, p < 0.001) in the resection of small rectal NET with significantly shorter procedure time (5.3 ± 2.8 min vs 17.9 ± 9.1 min, p = 0.000)[15]. Recently, Park et al. reported that there was not a significant difference between the resection rates of ESD and transanal endoscopic microsurgery in rectal NET with size < 10 mm in diameter, (83.5% vs 93.9%, p = 0.063) [16, 17]. Our results suggest that NETs ≤ 10 mm in diameter are better resected with ESMR-L than other endoscopic resection methods in terms of clinical efficacy, technical ease, and procedure time.
We included not only pathologically definite positive margin but also very small safe margin into incomplete resection. This is because there are not standard treatment guidelines for the histological findings that show a very small safe margin from the NET or near exposure to margin (‘possible’ remnant NET). Moreover, our results conform with a previous prospective study by Sung et al. which reported a long-term good prognosis in possible remnant NET post endoscopic resection of a tumor less than 15 mm in diameter [18].
Of the 15 patients who were excluded because of the additional endoscopic resection that was carried out, eight patients showed a suspicious residual lesion. However, pathological results after additional resection showed no remaining tumor. Moreover, among the 31 patients with long-term follow-up, 24 cases had a clean scar in the resection site and seven cases had a scar with a visible lesion. However, all seven cases showing a suspicious residual NET were also confirmed to be negative after a follow-up biopsy. A previous study reported discrepancies between complete endoscopic resection and complete histological resection (100% vs 75.3%) [18]. Stier et al. also reported that even in healthy scar, six out of 27 cases were confirmed to have residual NET after additional endoscopic resection. These results suggest that the precision of predicting local recurrence by endoscopy can be low. Nevertheless, in our study, pathologically incomplete or very small margin NET showed a good prognosis. This suggests that positive resection margin is relatively safe in small rectal NET.
The exact suitable follow-up duration post incomplete resection has not yet been determined. In a 10-year retrospective study of long-term follow-up of 13 cases for possible remnant NET, only one patient showed local recurrence after 56 months during surveillance colonoscopy [18]. Moreover, there was a case report of liver metastasis 5 years after complete resection of an 8 mm sized G1 rectal NET [19]. In our study, neither local recurrence nor distant metastasis was observed over the 2-year follow-up period. However, considering previous reports and the slow growth rate of NET, a 2-year follow-up period cannot be suggested to be sufficient.
One of the limitations of this study is that it is a retrospective single-center study. Second, due to the strict inclusion criteria, only 31 patients were included which may not render the results generalizable to larger populations. Third, to collect data of as many patients as possible with longer follow-up periods, at least 2 years of follow-up was a maximal margin which seems to be a short period considering the slow growth rate of rectal NET. Furthermore, longer follow-up period of incomplete resection without additional treatment is warranted for the assessment of long-term prognosis. Despite these limitations, to the best of our knowledge, our study is the first to demonstrate a good prognosis without performing additional resection in a pathological incompletely resected small rectal NET.
We can suggest that pathologically incompletely resected small G1 rectal NET can be observed without additional treatment. However, the exact follow-up duration and intervals remain to be clarified.