Retrospective accuracy analysis of major guidelines and factors for lymph node metastasis of pT1 colorectal cancer

Abstract

Background and study aims

The aim of this study was to validation the accuracy of major guidelines and their adopted unfavorable histologic factors for lymph node metastasis (LNM) in pT1 colorectal cancer (CRC). National Comprehensive Cancer Network (NCCN), Japanese Society for Cancer of the Colon and Rectum (JSCCR), European Society for Medical Oncology (ESMO), and French Intergroup Clinical Practice Guidelines (FICPG) were included.

Patients and methods

Retrospectively, 519 cases receiving oncological resection with lymphadenectomy were included. Factors and four pathological parameters--histological differentiation grades (HDG), lymphovascular invasion (LVI), depth of submucosal invasion (DSI)(>1000μm), and tumor budding (TB)--were analyzed for their impact on LNM incidence individually and in combination as defined by the guidelines.

Results

HDG, LVI, TB, and gender (female) are risk factors for LNM in multivariate analysis. All guidelines have significant ability to predict LNM (P<0.001). NCCN and ESMO have similar performance, in terms of sensitivity (63.8%/68.1%) and specificity (62.7%/69.5%). JSCCR and FICPG had similar performance in terms of good sensitivity (100%/100%) and low specificity (25.6%/25.2%). The JSCCR/FICPG group had higher sensitivity and lower specificity and accuracy than the NCCN/ESMO group.

Conclusions

HDG, LVI, TB, and gender (female) are independent risk factors for LNM of T1 CRC. DSI is an excellent negative predictor, although not an independent risk factor. The NCCN/ESMO guideline has medium sensitivity and requires improvement. The JSCCR/FICPG guideline has perfect sensitivity but low specificity, thus exposing patients to many unnecessary surgeries. There is revision potential for current guideline factors, including those beyond current pathological ones, to improve LNM prediction accuracy.

 

Scroll to top

Keywords

Endoscopy Lower GI Tract - Colorectal cancer - GI Pathology - GI surgery

Introduction

With the introduction of population-based screening programs and advancements in modern clinical tools, the detection rate for early colorectal cancer (CRC) has been increasing in Taiwan [1] and worldwide [2]. As the incidence of pT1 CRC rises, necessity for standard radical resection of the affected colorectum after local excision (LE), including endoscopic or transanal, has been debated, given that the regional lymph node metastasis (LNM) rate in pT1 CRC is reported to be approximately 10% to 13% [3] [4]. This suggests that surgical resection may not be required for the majority of pT1 CRC cases. Guidelines addressing LNM risk factors for pT1 CRC, including those from National Comprehensive Cancer Network (NCCN) Clinical Practice Guidelines in Oncology (NCCN Guidelines), Japanese Society for Cancer of the Colon and Rectum (JSCCR), European Society for Medical Oncology (ESMO), and French Intergroup Clinical Practice Guidelines (FICPG), primarily emphasize the histopathological characteristics of the primary tumor, with some variations ([Table 1]) [4] [5] [6] [7] [8].

Table 1 Unfavorable factors adopted by major guidelines (NCCN, JSCCR, ESMO, FICPG).

	NCCN	JSCCR	ESMO	FICPG
				Flat/sessile	Pedunculated
DSI, deep submucosal invasion; ESMO, European Society for Medical Oncology; FICPG, French Intergroup Clinical Practice Guidelines; HDG, histology differentiated grade; JSCCR, Japanese Society for Cancer of the Colon and Rectum; LVI, lymphovascular invasion; NCCN, National Comprehensive Cancer Network; PNI, perineural invasion; TB, tumor budding.
Unfavorable risk factor	Margin (+) or fragmented	Margin (+)	LVI	Margin (+)	Margin (+)
	HDG 3 or 4	DSI (> 1000 μm)	HDG 3	HDG 3	HDG 3
	LVI or PNI	LVI	TB Grade 2/3	DSI (> 1000 μm)	Haggitt 4
	TB Grade 3	Histology	Pedunculated, Haggitt 4	LVI	LVI
		HDG 3 adenocarcinoma		TB Grade 2/3	TB Grade 2/3
		Signet ring cell carcinoma
		Mucinous carcinoma
		TB Grade 2/3

These guidelines adopt the concept of unfavorable risk factors to predict LNM, which subsequently inform recommendations for additional surgical resection. Among these guidelines, JSCCR includes the most comprehensive list of unfavorable risk factors: positive resection margins following LE, histological differentiation grade (HDG), lymphovascular invasion (LVI), depth of submucosal invasion (DSI) (> 1000 μm) [9], and tumor budding (TB) [10].

NCCN Guidelines do not explicitly include DSI as an unfavorable histologic feature, and TB criteria with only Grade 3 included is inconsistent with others [5] [6]. Furthermore, NCCN has an indefinite algorithm regarding endoscopic classification of sessile lesion without risk factors. In ESMO Guidelines, DSI is defined as level 4 (Haggitt) invasion for pedunculated polyps, representing invasion deeper than the stalk [11]. However, for sessile or flat malignant polyps, ESMO does not clearly define DSI; any presence of other unfavorable histologic features would prompt a recommendation for adjuvant surgery [7]. Recently, FICPG made separate criteria for pedunculated and flat/sessile lesion [8], aligning with ESMO for pedunculated lesions and with JSCCR for sessile/flat lesions.

Our study is based on the hypothesis that all pT1 CRC cases can be accurately identified, completely resected en bloc with clear margins using an endoluminal approach, adequately retrieved and fixed, and comprehensively evaluated by the pathologist. We hypothesize that LNM risk can be correlated with specific pathological parameters. Retrospectively, only cases that underwent oncological resection with lymphadenectomy were included for analysis. Pathological parameters, as defined by these guidelines, were analyzed for their individual and combined impacts on LNM incidence.

 

Scroll to top

Patients and methods

From 1999 to 2019, a total of 1,109 cases diagnosed with T1 CRC at Taipei Veterans General Hospital were retrospectively collected. Most patients with T1 CRCs would undergo colorectal resection once diagnosed during the study period. Only in recent years have JSCCR guidelines been adopted in our institution. Exclusion criteria included patients undergoing only local or endoscopic excision without colorectal resection, neoadjuvant chemotherapy or radiotherapy, synchronous CRC more advanced than the pT1 stage, familial adenomatous polyposis, Peutz-Jeghers syndrome, and ulcerative colitis ([Fig. 1]). The majority of cases further excluded were due to absent or poor specimen quality for pathology review and missing clinical data.

Fig. 1 Flow chart of enrolled patients.

On further examination, cases that were not confirmed as pT1 stage on review, those with absent or incomplete pathological materials, or cases with missing clinical data were excluded. Ultimately, 519 cases were included in the analysis. The study followed the TNM staging system of the American Joint Committee on Cancer (AJCC) (8th edition) [12].

The study protocol was reviewed and approved by the Institutional Review Board of Taipei Veterans General Hospital (IRB-TPEVGH: 2022–09–004CC). Available pathological slides with hematoxylin and eosin staining (H&E stain) were retrieved, digitally scanned, and reviewed by a dedicated gastrointestinal pathologist (WY Liang). Four pathological parameters outlined in the guidelines were meticulously reviewed and recorded, including HDG defined by the World Health Organization (WHO) Classification of Digestive System Tumors (5^th ed) (2019) [13], SMID defined by “Japanese Classification of Colorectal, Appendiceal, and Anal Carcinoma, third English edition” [9], LVI defined as either carcinoma cells identified within endothelium-lined channels or tumor emboli identified within endothelium-lined channels surrounded by a smooth muscle wall [14], and TB defined by International Tumor Budding Consensus Conference (ITBCC) [10]. The definition and evaluation method for HDG differs between JSCCR (predominant differentiation pattern) and NCCN/ESMO/FICPG (least differentiated component). Clinical information was reviewed and the endoscopic images were also reviewed and classified according to the Paris endoscopic classification of superficial neoplastic lesion [15].

Statistical analyses were conducted using IBM SPSS Statistics version 25. Categorical variables were analyzed using the Chi-square test and Fisher’s exact test. Variables with P < 0.1 were included in the Multiple Cox Regression analysis (Backward Wald) to identify independent factors. P < 0.05 was considered statistically significant.

 

Scroll to top

Results

[Table 2] summarizes clinicopathological characteristics. pN+ was found in 47 cases (9.1%), including one case of pN2, which was also stage IV; no cases of N1c were observed. The percentage of male patients was slightly higher (57%). The median number of lymph nodes (LN) harvested was 12 (interquartile range [IQR] 9–17). The carcinoembryonic antigen level before any LE, available for 503 cases, was a median of 2.14 ng/mL (IQR 1.6–3.0). In terms of endoscopic morphology, pure early polypoid lesions (type 0-I) made up 45.1%, including pedunculated (type 0-Ip) 5.2%, semi-pedunculated (type 0-Isp) 9.6%, and sessile (type 0-Is) 30.2%. Early non-polypoid and mixed types made up 54.9% of findings. Most histological types were adenocarcinoma (99%). There were also one case of mucinous adenocarcinoma, three signet ring cell carcinomas, and one medullary carcinoma. In HDG (WHO 5^th), poorly differentiated (Gr3) comprises 27%. DSI was present in 76.1%, LVI in 6.6%, and TB Gr 2 (7.1%) and Gr 3 (12.9%) were observed. Among included cases, 98.7% (n = 512) was from cases undergoing primary colorectal resection, and only 1.3% (n = 7) was from receiving initial LE. It was because specimens of other cases receiving initial LE and following colorectal resection were absent or poor quality for pathology review, and therefore, were excluded. All parameters were analyzed for their association with LNM incidence. In univariate analysis, five risk factors—HDG, DSI, LVI, TB, and gender (female)—were significant for LNM and were included in further multivariate analysis. In multivariate analysis ([Table 3]), HDG, LVI, TB, and gender (female) remained significant independent risk factors.

Table 2 Clinicopathological characteristics of pT1 Cases (n = 519).

		Overall	LNM-	LNM+	P
AJCC, American Joint Committee on Cancer; CEA, carcinoembryonic antigen; IQR, interquartile range; JSCCR, Japanese Society for Cancer of the Colon and Rectum; LNM, lymph node metastasis; WHO, World Health Organization.
Gender, n (%)
Male		296 (57%)	276 (53.2%)	20 (3.9%)
Female		223 (43%)	196 (37.8%)	27 (5.2%)	0.035
Age, y, median (IQR) 65.8 (57.8–75.0)
< 65.8		259 (49.9%)	233 (44.9%)	26 (5%)
≥ 65.8		260 (50.1%)	239 (46.1%)	21 (4%)	0.44
Polyp size, mm, median (IQR) 20 (15–30)
< 20		280 (53.9%)	252 (48.6%)	28 (5.4%)
≥ 20		239 (46.1%)	220 (42.4%)	19 (3.7%)	0.42
CEA, ng/mL, median (IQR) (n = 503) 2.14 (1.6–3.0)
< 5.0		471 (93.6%)	429 (85.3%)	42 (8.3%)
≥ 5.0		32 (6.4%)	28 (5.6%)	4 (0.8%)	0.49
Lymph nodes totally retrieved number, median (IQR)		12 (9–17)
< 12		184 (35.3%)	229 (44.1%)	29 (5.6%)
≥ 12		335 (64.5%)	303 (90.4%)	32 (9.6%)	0.59
Tumor location
Right colon		139 (28.7%)	139 (26.8%)	10 (1.9%)
Left colon		173 (36.8%)	174 (33.5%)	17 (3.3)
Rectum		159 (34.5%)	159 (30.6%)	20 (3.9%)	0.37
Endoscopic morphology type
0-Ip	Pedunculated	27 (5.2%)	26 (5%)	1 (0.2%)
0-Isp	Semi-pedunculated	50 (9.6%)	48 (9.2%)	2 (0.4%)
0-Is	Sessile	157 (30.2%)	142 (27.4%)	15 (2.9)
Other or mixed types		285 (54.9%)	256 (49.3%)	29 (5.6%)	0.39
N Category, AJCC
N0		472 (90.9%)	472 (90.9%)	0
N1		46 (8.9%)	0	46 (8.9%)
N2		1 (0.2%)	0	1 (0.2%)	< 0.001
Stage, AJCC
Stage I		472 (90.9%)	472 (90.9%)	0
Stage III		46 (8.9%)	0	46 (8.9%)
Stage IV		1 (0.2%)	0	1 (0.2%)	< 0.001
Histology type
Adenocarcinoma		514 (99%)	468 (90.2%)	46 (8.9%)
Mucinous adenocarcinoma		1 (0.2%)	1 (0.2%)	0
Signet ring cell carcinoma		3 (0.6%)	2 (0.4)	1 (0.2%)
Medullary carcinoma		1 (0.2%)	1 (0.9%)	0	0.5
Histologic grade (Gr) (WHO 5th)
Gr1/Gr2, well/moderately differentiated		379 (73%)	355 (68.4%)	24 (4.6%)
Gr3, poorly differentiated		140 (27%)	177 (22.5%)	23 (4.4%)	< 0.001
Main histology (JSCCR)
Tubular/papillary		507 (99.7%)	462 (89%)	45 (8.7%)
Poorly differentiated/mucinous/signet ring cell		12 (2.3%)	10 (2.1%)	2 (0.4%)	0.35
Depth of Submucosal Invasion, ≥ 1000 μm, n (%)
Absence		124 (23.9%)	124 (23.9%)	0
Presence		395 (76.1%)	348 (67.1%)	47 (9.1)	< 0.001
Lymphovascular Invasion
Absence		485 (93.4%)	452 (87.1%)	33 (6.4%)
Presence		34 (6.6%)	20 (3.9%)	14 (2.7%)	< 0.001
Tumor budding grade (Gr)
Gr1, low	(0–4 buds/0.785 mm2)	415 (80%)	388 (74.8%)	27 (5.2%)
Gr2, intermediate	(5–9 buds/0.785 mm2)	37 (7.1%)	33 (6.4%)	4 (0.8%)
Gr3, high	(≥ 10 buds/0.785 mm2)	67 (12.9%)	51 (9.8%)	16 (3.1%)	< 0.001
Treatment
Local excision, then colorectal resection		7 (1.3%)	6 (1.2%)	1 (0.2%)
Primary colorectal resection		512 (98.7)	466 (%)	46 (8.9%)	0.63

Table 3 Univariate and multivariate analyses of LNM risk factors.

	Sensitivity	Specificity	PPR	NPR	Univariate	Multivariate
					P	P	RC (95% CI)
CI, confidence interval; DSI, deep submucosal invasion; HDG, histological differentiation grade; LN, lymph node; LNM, lymph node metastasis; LVI, lymphovascular invasion; NPR, negative prediction rate; PPR, positive prediction rate; RC, regression coefficient; TB, tumor budding; WHO, World Health Organization.
HDG (Gr 3) (WHO 5th)	48.9%	75.2%	16.4%	93.7%	< 0.001	< 0.01	2.87 (1.42–5.78)
LVI (+)	29.2%	95.8%	41.2%	93%	< 0.001	0.03	1.87 (1.06–3.28)
TB Grade 2/3	41.7%	82.2%	19.2%	93.3%	< 0.001	< 0.001	3.58 (1.81–7.09)
DSI (+)	100.0%	26.3%	12.2%	100.0%	< 0.001	0.99
Gender (female)	57.4%	58.5%	12.1%	93.2%	0.035	< 0.001	2.55 (1.46–4.46)

[Table 4] shows the accuracy of guidelines in predicting LNM based on the unfavorable risk factors they adopt. All guidelines demonstrated significant ability to detect LNM (P < 0.001). NCCN and ESMO had similar performance, in terms of sensitivity (63.8%/68.1%), specificity (62.7%/69.5%), and other parameters. JSCCR and FICPG had similar performance, in terms of good sensitivity (100%/100%), low specificity (25.6%/25.2%), and other parameters. The JSCCR/FICPG group had higher sensitivity, lower specificity and accuracy than the NCCN/ESMO group.

Table 4 Accuracy of guidelines to detect lymph node metastasis.

	NCCN	JSCCR	ESMO	FICPG
ESMO, European Society for Medical Oncology; FICPG, French Intergroup Clinical Practice Guidelines; JSCCR, Japanese Society for Cancer of the Colon and Rectum; NCCN, National Comprehensive Cancer Network.
Sensitivity	63.8%	100.0%	68.1%	100.0%
Specificity	62.7%	25.6%	69.5%	25.2%
Positive prediction rate	14.6%	11.8%	18.2%	11.8%
Negative prediction rate	94.6%	100.0%	95.6%	100.0%
Accuracy	62.8%	32.4%	69.4%	32.0%

 

Scroll to top

Discussion

HDG, LVI, TB(Gr2/3), and gender are independent risk factors for LNM in multivariate analysis. All guidelines have significant ability to predict LNM. JSCCR/FICPG guidelines have better sensitivity, but lower specificity and accuracy than NCCN/ESMO guidelines. The main reason is because DSI criteria were adopted by JSCCR/FICPG. In this study, DSI is not an independent risk factor, but no tumors with submucosal invasion < 1000 μm had LNM, which is consistent with findings from Japanese studies [4] [16]. It is an ideal negative predictor. However, the inclusion of DSI in guidelines would result in a high false-positive ratio, and cause over-recommendation of the need for surgical resection. The low sensitivity of NCCN/ESMO appears to require improvement. Missing one-third of LNM cases is an issue.

Absence of DSI and inclusion of only Grade 3 TB (without Grade 2) are the main differences between NCCN and the other guidelines, which contributes to the low sensitivity of NCCN. However, the significance of DSI (> 1000 μm) remains controversial. A meta-analysis by Zwager et al. suggests that DSI (> 1000 μm, or two-thirds of submucosal invasion) is not an independent risk factor for LNM in T1 cancer [17]. They also reported, in the absence of other high-risk factor, that the ratio of LNM of T1 CRC with only DSI was 2.6%, whereas it was 6.7% (15/224) in this study. How DSI should be applied as an unfavorable risk factor for T1 CRC remains unclear, but excluding DSI as a factor does affect the detection rate and the sensitivity.

In terms of TB, Grade 3 (High) is widely accepted as a significant risk factor for LNM in T1 CRC [2] [18]. However, the criteria for applying Grade 2/3 or Grade 3 to determine significance varies. Two major groups—ITBCC [10] and JSCCR[4]—advocate for applying Grade 2/3 TB criteria. The reported incidence of positive TB in T1 CRC ranges from 11.8% to 75%, depending on the criteria, tissue staining, or diagnostic methods used [19] [20] [21]. Typically, it falls between 10% and 20%, and our Grade 2/3 incidence (20%) is consistent with other reports.

While the NCCN guideline low sensitivity may seem concerning, it is important to note that the algorithm includes an indefinite decision (observation or colectomy) for sessile T1 colon cancers without unfavorable factors [5]. Its approach reminds clinicians that sessile lesions have a significantly higher incidence of adverse outcomes. However, our analysis was limited to cases with well-defined unfavorable factors and could not include the indefinite decision model. In addition, pedunculated T1 tumors without risk factors are not free of LNM, albeit with a very low incidence [18].

The major concern regarding these unfavorable risk factors is interobserver variability, particularly for DSI. In 2004, Ueno et al. introduced the SMID measurement method, which was later adopted by JSCCR. This includes two types of depth measurements [20]. The first uses muscularis mucosae (MM) as the upper reference line (Y1), whereas the second uses tumor surface (Y2) when MM is not recognized. The primary issue is that estimation of MM is observer-dependent, leading to variability even after training [22]. There is also the pathologist-dependent variation about Y1 or Y2 criteria usage. To minimize variability, multiple pathologists should be involved in reviewing pathological parameters. One of this study’s major limitation is that only one experienced pathologist conducted the review. We suggest that a study be conducted to verify the impact of these two measurement methods.

A comparison of validation results among relevant series is shown in [Table 5] [23] [24] [25]. The good sensitivity of JSCCR is confirmed in all studies from different areas. The reported sensitivity of NCCN ranges from 63.8% to 100%; ESMO is 68.1% to 100%, and our series is the lowest. In terms of specificity, NCCN is 44% to 82.4%, JSCCR 0% to 25.6%, and ESMO 0% to 69.5%. It is surprising that Ichimasa et al reported 0% specificity on JSCCR/ESMO [26]. We think our low LVI incidence is the main reason for the low sensitivity and high specificity of NCCN/ESMO. The incidence of LVI (6.6%) in this study is lower than previously reported (10% to 21%) [27] [28]. Bosch et al. reported 14.1% of LVI in their meta-analyses of 17 studies [2]. It is hard to explain our lower LVI rate, given to the same adopted criteria for LVI diagnosis. This lower LVI rate can probably be explained by the more conservative attitude of our pathologist in reaching LVI diagnosis. However, on reviewing the LVI rate in our whole CRC database, we found a reasonably progressive increasing LVI rate according to increasing staging. Special staining labelling lymphatic vessels can increase LVI detection. On the basis of confirmed LVI impact on LNM risk of T1 CRC, a comparative study of H&E and lymphatic vessel staining is recommended.

Table 5 Sensitivity and specificity of reported validation studies.

Study, Year			NCCN	JSCCR	ESMO		NCCN	JSCCR	ESMO
ESMO, European Society for Medical Oncology; JSCCR, Japanese Society for Cancer of the Colon and Rectum; NA, not available; NCCN, National Comprehensive Cancer Network.
Ichimasa et al. [23]	2018 (n = 690)	Sensitivity	100%	100%	100%	Specificity	44%	0	0
Piao et al. [24]	2023 (n = 651)		87.5%	100%	NA		82.4%	17.5%	NA
Tanino et al. [25]	2024 (n = 560)		98%	100%	98%		52%	19%	50%
This study	(n = 519)		63.8%	100%	68.1%		62.7%	25.6%	69.5%

The HDG evaluation criteria differ significantly between JSCCR and NCCN/ESMO. In JSCCR, it is defined as the predominant histological grade according to their guidelines [9]. JSCCR HDG criteria did not correlate with independent significance with LNM in either study. The HDG criteria for NCCN/ESMO are based on the WHO classification of tumors (5th) [13]: Presence of the least or poorly differentiated pathology. Prevalence of this HDG definition in this study is similar to the report by Tanino et al (27% vs. 20%). It correlates with LNM in the univariate and multivariate analyses in both studies.

This study is based on the assumption that all T1 CRCs are well excised, fixed, and examined. However, this is not always the case, because factors such as tumor size, morphology, location, and procedure execution can affect outcomes. In our study, most specimens from initial LE cases were not suitable for review, which led to their exclusion. However, in terms of current clinical procedure, pathologies according to endoscopic excision first are common. This selection bias should be acknowledged in this study. This illustrates how diverse results can be, even with well-defined pathological factors. To ensure precise pathological interpretation for further decisions, a standardized procedure should be implemented.

Physicians must carefully consider both the benefits and risks of surgery when assessing LNM risk in T1 CRC. [Table 6] shows related results of previously published studies [1] [29] [30] [31] [32] [33] [34] [35]. Regarding surgical mortality for specific T1 CRC series, there were few previous reports. Two studies [30] [32] reported no surgical mortality; Belderbos et al. [31] reported a little higher 30-day mortality of 2.6%, whereas there is only case (1.9‰) in our study. The recurrence rate after curative resection, including primary surgical resection or additional resection after LE, was 1.9% to 4.3%, seemingly related to the LNM+ rate in each study except that of Belderbos et al. In our study, during a medium follow-up of 5.44 years (IQR, 2.45 to 8.23), there were 10 distant recurrences, without any local recurrence, and five of the patients died of CRC. With treatments according to JSCCR/FICPG guidelines, approximately 90% of resections will be over-done, although with very low surgical mortality rate. If with NCCN/ESMO guidelines, 81.9% to 85.4% of resections will be over-done, but will miss one-third of LNM cases. More validation studies of guidelines will be needed to examine these relevant results. Generally speaking, surgical resection risk is low. If LNM risk can be predicted more accurately, the benefit of surgery will be enhanced. Besides oncological outcomes, surgeons’ other concern is functional outcomes after rectal surgery, such as low anterior resection syndrome (LARS), which should be also considered and explained in detail to patients.

Table 6 LNM rate, surgical mortality, and recurrence rate of reported series of T1 colorectal cancers treated with curative resection.

Study, Year		Follow-up (year)	LNM+ n (%)	Surgical mortality rate	Recurrence, n (%)
IQR, interquartile range; LNM, lymph node metastasis; NA, not available; SD, standard deviation.
Kobayashi et al. [29]	2011 (n = 798)	Median (range), 1.9 (0.3–4.4)	84 (10.5%)	NA	18 (2.3%)
Yoda et al. [30]	2013 (n = 202)	Median (range), 5 (1–5.25)	23 (11.4%)	0	7 (3.5%)
Belderbos et al. [31]	2017 (n = 945)	Median, 6.6	82 (8.7%)	25 (2.6%)	39 (4.1%)
Liu et al. [32]	2020 (n = 481)	NA	55 (11.4%)	0	14 (2.9%)
Chang et al. [2]	2021 (n = 318)	Mean ± SD, 6.8 ± 2.62	50 (15.7%)	NA	11 (3.5%)
Ha et al. [33]	2022 (n = 466)	Median (IQR), 6 (4.92–8.08)	NA	NA	9 (1.9%)
Tamaru et al. [34]	2024 (n = 2438)	Mean, 5.53	226 (9.3%)	NA	88 (3.6%)
Nilsson et al. [35]	2024 (n = 1317)	Median (IQR), 5 (3.17–5.29)	154 (11.7%)	NA	48 (3.6%)
This study	(n = 519)	Medium (IQR), 5.44 (2.45–8.23)	47 (9.1%)	1 (1.9‰)	10 (1.9%)

There are limitations to this study. This was a retrospective, single-center series, and many pathological materials were excluded due to improper preparation, particularly from LE first cases. Occasionally, faded slide staining was not suitable for review, even after re-preparation from preserved paraffin blocks. This could affect accuracy of pathology. In addition, review of endoscopic morphology was limited, and final pathological and clinical reviews did not reach the initially designed sample size, reducing the study power. Selection and background biases including the fact that most LE specimens were not suitable for analyses, as well as use of a single pathologist, also limit study validity. Finally, differing algorithms used in the guidelines make direct comparisons challenging.

 

Scroll to top

Conclusions

In conclusion, HDG, LVI, TB(Gr2/3), and gender are independent risk factors for LNM of T1 CRC. DSI is an excellent negative predictor although not an independent risk factor. NCCN/ESMO guidelines have medium sensitivity and require improvement. JSCCR/FICPG has perfect sensitivity but low specificity, thus exposing patients to many unnecessary surgeries. There is potential for revision of current guideline factors, including those beyond current pathological ones, to improve LNM prediction accuracy.

 

Scroll to top

 

Scroll to top

Contributorsʼ Statement

ChunChi Lin: Conceptualization, Formal analysis, Writing - original draft. Wen-Yih Liang: Formal analysis, Investigation, Software, Validation. Jui-Shen Chang: Data curation, Formal analysis, Software. Shih-Ching Chang: Conceptualization, Validation, Visualization. Shung-Haur Yang: Conceptualization, Data curation, Formal analysis, Funding acquisition, Investigation, Methodology, Project administration, Resources, Supervision, Writing - review & editing.

Conflict of Interest

The authors declare that they have no conflict of interest.

Acknowledgement

We thank Dr. Anna Fen-Yau Li (Department of Anatomical Pathology, Cheng Hsin General Hospital, Taipei, Taiwan) for her valuable advice.

• References

• 1 Chang LC, Shun CT, Lin BR. et al. Recurrence outcomes less favorable in T1 rectal cancer than in t1 colon cancer. Oncologist 2021; 26: e1548-e1554
• 2 Bosch SL, Teerenstra S, de Wilt JH. et al. Predicting lymph node metastasis in pT1 colorectal cancer: a systematic review of risk factors providing rationale for therapy decisions. Endoscopy 2013; 45: 827-834
• 3 Nascimbeni R, Burgart LJ, Nivatvongs S. Risk of lymph node metastasis in T1 carcinoma of the colon and rectum. Dis Colon Rectum 2002; 45: 200-206
• 4 Hashiguchi Y, Muro K, Saito Y. et al. Japanese Society for Cancer of the Colon and Rectum (JSCCR) guidelines 2019 for the treatment of colorectal cancer. Int J Clin Oncol 2020; 25: 1-42
• 5 NCCN Clinical Practice Guidelines in Oncology (NCCN Guidelines) Colon Cancer Version 5.2024 - August 22, 2024. https://www.nccn.org
• 6 NCCN Clinical Practice Guidelines in Oncology (NCCN Guidelines) Rectal Cancer Version 4.2024 - August 22, 2024. https://www.nccn.org
• 7 Argilés G, Tabernero J, Labianca R. et al. Localised colon cancer: ESMO Clinical Practice Guidelines for diagnosis, treatment and follow-up. Ann Oncol 2020; 31: 1291-1305
• 8 Lecomte T, Tougeron D, Chautard R. et al. Non-metastatic colon cancer: French Intergroup Clinical Practice Guidelines for diagnosis, treatments, and follow-up (TNCD, SNFGE, FFCD, GERCOR, UNICANCER, SFCD, SFED, SFRO, ACHBT, SFP, AFEF, and SFR). Dig Liver Dis 2024; 56: 756-769
• 9 Japanese Classification of Colorectal, Appendiceal, and Anal Carcinoma: the 3d English Edition [Secondary Publication]. J Anus Rectum colon 2019; 3: 175–195.
• 10 Lugli A, Kirsch R, Ajioka Y. et al. Recommendations for reporting tumor budding in colorectal cancer based on the International Tumor Budding Consensus Conference (ITBCC) 2016. Modern Path 2017; 30: 1299-1311
• 11 Haggitt RC, Glotzbach RE, Soffer EE. et al. Prognostic factors in colorectal carcinomas arising in adenomas: implications for lesions removed by endoscopic polypectomy. Gastroenterology 1985; 89: 328-336
• 12 Mahul B, Amin SBE, Greene FL. et al. AJCC Cancer Staging Manual 8 Edition. 8th ed.. Chicago: Springer Cham; 2017.
• 13 Nagtegal ID, Arenda M, Salto-Teliez M. Colorectal adenocarcinoma. WHO Classification of Tumors: Digestive System Tumours. 5th ed. Lyon, France: International Agency for Research on Cancer (IARC); 2019
• 14 Minsky B, Mies C. The clinical significance of vascular invasion in colorectal cancer. Dis Colon Rectum 1989; 32: 794-803
• 15 The Paris endoscopic classification of superficial neoplastic lesions: esophagus, stomach, and colon: November 30 to December 1, 2002. Gastrointest Endosc 2003; 58: S3-S43.
• 16 Kitajima K, Fujimori T, Fujii S. et al. Correlations between lymph node metastasis and depth of submucosal invasion in submucosal invasive colorectal carcinoma: a Japanese collaborative study. J Gastroenterol 2004; 39: 534-543
• 17 Zwager LW, Bastiaansen BAJ, Montazeri NSM. et al. Deep submucosal invasion is not an independent risk factor for lymph node metastasis in T1 colorectal cancer: a meta-analysis. Gastroenterol 2022; 163: 174-189
• 18 Backes Y, Elias SG, Groen JN. et al. Histologic factors associated with need for surgery in patients with pedunculated T1 colorectal carcinomas. Gastroenter 2018; 154: 1647-1659
• 19 Morodomi T, Isomoto H, Shirouzu K. et al. An index for estimating the probability of lymph node metastasis in rectal cancers. Lymph node metastasis and the histopathology of actively invasive regions of cancer. Cancer 1989; 63: 539-543
• 20 Ueno H, Mochizuki H, Hashiguchi Y. et al. Risk factors for an adverse outcome in early invasive colorectal carcinoma. Gastroenterol 2004; 127: 385-394
• 21 Koelzer VH, Zlobec I, Lugli A. Tumor budding in colorectal cancer--ready for diagnostic practice?. Hum Pathol 2016; 47: 4-19
• 22 Karamchandani DM, Westerhoff M, Arnold CA. et al. interobserver variability in assessment of depth of submucosal invasion for colonic endoscopic resections among subspecialized gastrointestinal pathologists. Arch Pathol Lab Med 2022; 147: 534-545
• 23 Ichimasa K, Kudo SE, Mori Y. et al. Artificial intelligence may help in predicting the need for additional surgery after endoscopic resection of T1 colorectal cancer. Endoscopy 2018; 50: 230-240
• 24 Piao ZH, Ge R, Lu L. An artificial intelligence prediction model outperforms conventional guidelines in predicting lymph node metastasis of T1 colorectal cancer. Frontiers Oncol 2023; 13: 1229998
• 25 Tanino F, Yamashita K, Morimoto S. et al. Comparative prediction of lymph node metastasis in pT1 colorectal cancer among Western and Japanese guidelines. Frontiers Oncol 2024; 14: 1475270
• 26 Ichimasa K, Kudo SE, Yeoh KG. Commentary: An artificial intelligence prediction model outperforms conventional guidelines in predicting lymph node metastasis of T1 colorectal cancer. Frontiers Oncol 2024; 14: 1337576
• 27 Dawson H, Bokhorst JM, Studer L. et al. Lymph node metastases and recurrence in pT1 colorectal cancer: Prediction with the International Budding Consortium Score—A retrospective, multi-centric study. United Europ Gastroenterol J 2024; 12: 299-308
• 28 Ebbehøj AL, Smith HG, Jørgensen LN. et al. Prognostic factors for lymph node metastases in pT1 colorectal cancer differ according to tumor morphology: A nationwide cohort study. Ann Surg 2023; 277: 127-135
• 29 Kobayashi H, Mochizuki H, Morita T. et al. Characteristics of recurrence after curative resection for T1 colorectal cancer: Japanese multicenter study. J Gastroenterol 2011; 46: 203-211
• 30 Yoda Y, Ikematsu H, Matsuda T. et al. A large-scale multicenter study of long-term outcomes after endoscopic resection for submucosal invasive colorectal cancer. Endoscopy 2013; 45: 718-724
• 31 Belderbos TD, van Erning FN, de Hingh IH. et al. Long-term recurrence-free survival after standard endoscopic resection versus surgical resection of submucosal invasive colorectal cancer: A population-based study. Clin Gastroenterol Hepatol 2017; 15: 403-411 e401
• 32 Liu Y, Li R, Chang W. et al. The effect of non-curative endoscopic resection on cT1N0M0 colorectal carcinoma patients who underwent additional surgery. Surg Endosc 2021; 35: 2862-2869
• 33 Ha RK, Park B, Han KS. et al. Subpopulation analysis of survival in high-risk T1 colorectal cancer: surgery versus endoscopic resection only. Gastrointest Endosc 2022; 96: 1036-1046 e1031
• 34 Tamaru Y, Kuwai T, Kajiwara Y. et al. Long-term outcomes of additional surgery after endoscopic resection versus primary surgery for T1 colorectal cancer. Am J Gastroenterol 2024; 119: 2418-2425
• 35 Nilsson E, Wetterholm E, Syk I. et al. Risk of recurrence in high-risk T1 colon cancer following endoscopic and surgical resection: registry-based cohort study. BJS Open 2024; 8: zrae053

Publikationsverlauf

Eingereicht: 16. Februar 2025

Angenommen nach Revision: 02. September 2025

Accepted Manuscript online:
08. September 2025

Artikel online veröffentlicht:
09. Oktober 2025