Introduction

Several guidelines, including those from the US, Europe, and Japan, recognize "tumor differentiation" as a risk factor for lymph node metastasis (LNM) in T1 colorectal cancer. Surgical resection with lymph node dissection is recommended if poorly differentiated adenocarcinoma, signet-ring cell carcinoma, or mucinous carcinoma are observed in histological findings from endoscopically resected specimens. Recent reports suggest that moderately differentiated adenocarcinoma (tub2) may also be a risk factor for LNM.

Aims & Methods

This study aims to investigate whether tub2 is a risk factor for LNM. We initially evaluated 1,742 patients with T1 colorectal cancer who underwent surgical resection with lymph node dissection at our institution from April 2001 to April 2024. Lesions were assessed based on the worst grade of differentiation. Exclusions were made for patients with (a) metachronous or synchronous advanced cancer, (b) invasion deeper than the muscularis propria, (c) Lynch syndrome, familial adenomatous polyposis, or ulcerative colitis, (d) preoperative chemotherapy or radiotherapy, (e) missing data, or (f) lesions without surgical resection. In addition, after excluding 177 patients with poorly differentiated, mucinous, or signet-ring cell adenocarcinoma, 881 patients were enrolled for analysis. We investigated the sensitivity, specificity, and area under the receiver operating characteristics (ROC) curve of tub2 for predicting LNM, alongside correlations between pathological factors and LNM using multivariate analysis.

Results

LNM was detected in 9.9% (87/881) of patients. The sensitivity, specificity, and area under the ROC curve for tub2 were 78.2%, 42.2%, and 0.60, respectively. In the univariate analysis, tub2 was associated with LNM (p<0.001). Multivariate analysis identified lymphatic invasion, vascular invasion, and tub2 as significant risk factors for LNM, with odds ratios of 7.49, 1.97, and 1.95 (p < 0.01, < 0.01, and = 0.02, respectively).

Conclusion

Moderately differentiated adenocarcinoma (tub2) is a risk factor for LNM in T1 colorectal cancer.

Disclosure

Conflicts of interest:None

Introduction

Accurate prediction of lymph node metastasis (LNM) in submucosal invasive (T1) colorectal cancer (CRC) is essential for determining the need for additional surgery following endoscopic resection. Conventional histopathological assessments, however, are limited by interobserver variability. This study aimed to develop and externally validate a stacking-based artificial intelligence (AI) model integrating histopathological whole slide images (WSIs) and clinical parameters to predict the risk of LNM in T1 CRC.

Aims & Methods

This retrospective study included 1,056 patients with histologically confirmed T1 CRC treated at Showa Medical University Northern Yokohama Hospital. Of these, 593 cases were used for model training and 100 for internal testing. External validation was conducted on 25 cases from two independent institutions.
The proposed stacking ensemble model incorporated the Multiple Instance Self-Training (MIST) score derived from WSIs, patient sex, and tumor location (colon vs. rectum). Base learners included logistic regression, XGBoost, and random forest, with a meta-learner synthesizing their outputs to generate the final LNM prediction. Model performance was evaluated using the area under the receiver operating characteristic curve (AUC) in both internal and external cohorts.

Results

LNM was observed in 69 (11.6%) training cases, 15 (15.0%) test cases, and 2 (8.0%) external validation cases. The model achieved an AUC of 0.6965 in the internal test cohort and 0.7500 in the external validation cohort, demonstrating stable and generalizable predictive performance.
A representative case involved a 64-year-old man with a 16-mm type 0-Is lesion in the sigmoid colon resected via endoscopic mucosal resection (EMR). Histopathology showed well-differentiated tubular adenocarcinoma (tub1 > tub2), submucosal invasion depth of 3,300 µm (T1b), tumor budding grade BD1, Ly0, V1a, and negative resection margins (HM0, VM0). Despite guideline-based recommendations prompting additional surgical resection, no LNM was found (N0, 0/16), and the patient remained recurrence-free five years postoperatively—consistent with the AI model’s prediction.

Conclusion

The stacking-based AI model exhibited robust and generalizable performance in predicting LNM in T1 CRC, both internally and across external institutions. This approach offers a complementary tool to conventional pathology and may enhance individualized decision-making regarding the necessity of additional surgical intervention.

Introduction

Recent advances in endoscopic technology have allowed many T1 colorectal carcinomas (CRCs) to be resected endoscopically. After endoscopic resection (ER), we often wonder if we should perform an additional surgical resection (SR) with nodal dissection. At present, the indication of additional SR is based only on the risk of simultaneous lymph node metastasis, with no consideration of prognosis in mind. Although in particular elderly people generally have some complications and are more likely to die from other diseases, it is not clear whether the criteria for additional SR after ER should be the same as for non-elderly patients.

Aims & Methods

The aim of this study is to investigate the overall survival (OS) in patients with T1 CRCs treated by ER alone or SR with nodal dissection. After excluding complications of advanced cancer or hereditary colorectal cancer and so on, we retrospectively analyzed 1461 patients with T1 CRCs resected ER or SR at our center from April 2001 to June 2020, and followed them up over a mean period of 77.1 months. We divided the patents into two groups: ER alone group and SR with lymph node dissection group, and compared the OS using the Kaplan–Meier method and the log-rank test. In addition, the same analysis was conducted by dividing the above subjects into those ≤ 75 years of the age and those > 75 years.

Results

In the overall T1 CRC patients, the OS was significantly shorter in the ER alone group than in the SR group (p < 0.01). Among those ≤ 75 years old, that was also shorter in the ER alone group (p < 0.01), while in those > 75 years, no significant difference was found in the OS (p = 0.08).

Conclusion

Among T1 CRC patients > 80 years of age, there was no difference in the remaining life expectancy, with or without SR with nodal dissection, even if additional SR was indicated after ER based on the current criteria. For elderly patients, T1 CRCs could be initially resected endoscopically, and the indication for additional SR with nodal dissection might be carefully considered.

Disclosure

No

Introduction

There are insufficient data to guide clinical decision-making regarding the implementation of additional surgery following endoscopic therapy in patients with deep submucosal invasive colorectal cancer (T1b CRC) and no additional risk factors for lymph node metastasis (LNM). This investigation sought to identify factors associated with LNM in cases of T1b CRC without other risk factors.

Aims & Methods

Consecutive cases with pathologically proven T1b CRC were retrospectively collected from 2014 to 2023 at a single tertiary hospital in Korea. These patients received either surgical or endoscopic treatment. In patients who received endoscopic treatment alone, no LNM was deemed to exist if there was no recurrence after at least 3 years of follow-up. Excluded were cases with other risk factors, including poor differentiation, lymphovascular invasion, perineural invasion. The cases were compared based on the presence of LNM.

Results

A total of 305 cases were enrolled and LNM was observed in 28 cases (9.18%). In univariate analysis, differentiation, sm3 invasion, tumor growth pattern, and peri-tumoral white spots were related with LNM. Multivariate analysis identified presence of non-polypoid growth parts (odd ratio [OR] 6.39; 95% CI 2.76-14.79), and peri-tumoral white spots (OR 2.85; 95% CI 1.22-6.64) as significant predictive factors for LNM.

Table. Multivariate analysis

Conclusion

Endoscopic findings including presence of non-polypoid growth and peri-tumoral white spots were related with increased risk of LNM in T1b CRC in the absence of other pathologic risk factors. These findings may be useful in determining further treatment strategy for patients with Tb1 CRC received endoscopic resection.

References

Liselotte W. Zwager,1,2,3 Barbara A. J. Bastiaansen, et al.
Deep Submucosal Invasion Is Not an Independent Risk
Factor for Lymph Node Metastasis in T1 Colorectal Cancer:
A Meta-Analysis. Gut 2022;163:174-189

Chihiro Yasue, Akiko Chino, et al.
Pathological risk factors and predictive endoscopic factors
for lymph node metastasis of T1 colorectal cancer: a single-center
study of 846 lesions. J Gastroenterol(2019) 54: 708-717

Introduction

In the carcinogenic pathway of colorectal cancer, the first concept is the "adenoma-carcinoma sequence" and the second concept is the " serrated pathway ". In these concepts, adenomas and serrated lesions gradually progress to cancer. The third concept is the "de novo pathway," in which cancer develops without adenomas or serrated lesions and is also characterized by high malignancy and rapid progression. We have endoscopically characterized depressed colorectal cancers considered to be of the "de novo pathway" and validated the usefulness of pit pattern and EC classification.

Aims & Methods

A total of 54819 colorectal neoplasms were resected endoscopically or surgically in our institution from April 2001 to December 2024. They were subjects of this study. Advanced cancers were excluded. Of these, 1694 lesions were submucosally invasive (T1) carcinomas, 7330 lesions were classified as high-grade dysplasia and 39977 lesions as low-grade dysplasia. Morphologically, they were divided into 3 types: depressed, flat, and protruded type comprising 1.0 (%), 46.4 (%),and 52.6（%), respectively. Specifically, the distinctive features of depressed type neoplasms concerning their pit pattern and EC classification were examined. EC classification is a classification for an endocytoscopy that can evaluate pathology, such as nuclear atypia and glandular structure, by magnifying stained cells. EC1a is diagnosed as normal mucosa, EC1b is diagnosed as hyperplastic polyps, and if slit-like glandular ducts or spindle-shaped nuclei are present, it is classified as EC2 and diagnosed as an adenoma or intramucosal carcinoma. If irregularly shaped glandular ducts or nuclei with a high N/C ratio are found, it is classified as EC3a, and diagnosed as intramucosal carcinoma or SM lightly invasive carcinoma. If the glandular ducts are indistinct and the nucleus is irregularly shaped and darkly stained, it is classified as EC3b and diagnosed as SM highly invasive carcinoma.

Results

Most of the flat (91.5%) and protruded type (94.9%) lesions showed type IIIL or IV pit pattern, which are common findings in adenomas. On the other hand, most of the depressed type lesions (92.6%) classified as type IIIs, VI or VN pit patterns, are typically seen in carcinomas. As for endocytoscopy, most of flat and protruded type lesions were identified as EC2 corresponding to adenomas. In contrast, relatively high rates of depressed type lesions were categorized as EC3a (40.3%) and EC3b (52.4%), consistent with invasive carcinomas.
The proportion of T1 carcinomas in depressed type lesions was 58.7%, whereas in flat and protruded type lesions it was 2.5% and 2.5%, respectively. Additionally, 12.0% of depressed type lesions less than 5mm in diameter were T1 carcinomas, whereas flat and protruded type lesions were 0.01% and 0%, respectively.

Conclusion

This study revealed that depressed lesions were classified as type IIIs, VI or VN pit patterns on magnifying endoscopy and as type EC3a or EC3b on endocytoscopy. Lesions with these features or pit patterns tend to invade the submucosal layer even if they are small size. Despite their high potential for malignancy, depressed lesions are often ignored. However, it is critical to recognize their features and accurately diagnose lesions based on morphology, pit pattern, and EC classification.

Disclosure

Shin-ei Kudo and Masashi Misawa received speaking honoraria from Olympus Corporation and have an ownership interest of the products of Cybernet Systems. Tetsuo Nemoto received a research grant from Olympus Corporation for other studies. Other authors have nothing to declare.

Introduction

According to the European Society of Gastrointestinal Endoscopy (ESGE) guidelines, the curative criteria of endoscopic resection (ER) for T1 colorectal cancer (CRC) are well/moderately differentiated adenocarcinoma or papillary carcinoma, no lympho-vascular invasion, submucosal invasion depth <1000μm, and budding grade (BD) 1. Although an additional surgery is recommended for non-curative ER, the incidence of lymph node metastasis (LNM) is less than 12%. Many reports on lymph node metastasis and recurrence in T1 CRC originate from academic centers, with limited reports of real-world data.

Aims & Methods

The aim of this study was to clarify clinical outcomes in patients with T1 CRC. A total of 686 consecutive patients (Male:447 Female:239) were enrolled from April 2009 to August 2019 at 10 institutions. Exclusion criteria included evidence of familial adenomatous polyposis, hereditary nonpolyposis colorectal cancer, or inflammatory bowel disease; presence of active, malignant diseases in any other organs; presence of synchronous or metachronous advanced CRC. Risk factors of LNM and recurrence, and survival were analyzed.

Results

The mean age of patients was 69.3±10.2 years old. The mean tumor size was 20.3±12.4mm. The location of the lesion was 439(64%) and 247(36%) in the colon and rectum, respectively. The histopathological types included 681 cases of differentiated type and 5 cases of undifferentiated type. A number of patients with ER without additional SR (ER alone) were 176(curative resection 86, non-curative resection 90), those with SR (SR alone) were 510, and those with ER and additional SR (ER + additional SR) were 116.
(i) Risk factor of LNM: LNM was found in 13(11.2%) patients with ER + additional SR and in 40(10.6%) patients with SR alone. In multivariate analysis, positive lymphatic invasion and tumor size were significant independent risk factor for LNM (OR 5.9, 95% CI [2.2-16], p<0.01, OR 1.04, 95%CI［1.00-1.07］p=0.028, respectively).
(ii) Recurrence: During the mean observation period of 45.5±34.3 months, recurrence was found in 27(3.9%) patients. No recurrence was found in the ER curative resection group. Although 43% (90/206) patients with ER non-curative resection did not receive additional SR, there was no significant difference in recurrence rate between ER alone with non-curative resection and ER + additional SR groups (5.6% (5/90) vs. 2.6% (3/116), respectively; p=0.3). In the SR alone group, recurrence was found in 4.8% (19/394). The significant independent risk factor for recurrence was tumor size (OR 1.12, 95% CI [1.05-1.2], p<0.01).
(iii) Survival: There were 50 deaths (7.3%, 76.2±6.6 years old) during the observation period of 45.5±34.3 months, but CRC-related death was found in only 6 cases (0.8%); one case was ER alone with non-curative resection and the other cases were SR alone.

Conclusion

This study suggests that the possibility of expanding the indication for ER of T1 CRC, because positive lymphatic invasion and tumor size were the only risk factor for LNM and better long-term outcomes were shown despite including significant number of patients with ER alone with non-curative resection. Fifty cases of death showed a higher average age and fewer deaths from CRC. Therefore, even in cases of non-curative resection for T1 CRC after ER, avoiding additional SR may be considered due to advanced age or comorbidities.