Lymph node metastasis risk factors in T2 colorectal cancer

INTRODUCTION

Advances in endoscopic techniques, including colorectal polypectomy, endoscopic mucosal resection, and endoscopic submucosal dissection, coupled with the use of magnifying endoscopes for pit pattern diagnosis, have enhanced the diagnostic and therapeutic management of T1 colorectal cancer (CRC). Given that approximately 10% of T1 CRC cases develop lymph node metastasis (LNM), the Japanese Society for Cancer of the Colon and Rectum guidelines advocate for additional surgical resection with lymph node dissection for high-risk LNM patients following endoscopic resection. Identified risk factors include depth of submucosal invasion, lymphovascular invasion, poor differentiation, and tumor budding.^{1, 2}

Conversely, T2 CRC, characterized by muscularis propria invasion, presents with LNM in about 25% of cases, with surgery and lymph node dissection being the standard of care.¹ However, emerging endoscopic techniques such as endoscopic full-thickness resection (eFTR), endoscopic intermuscular dissection (EID), and per-anal endoscopic myectomy (PAEM) have the potential to shift the treatment paradigms.^3-5 These new techniques offer a “resect and analysis” approach, in which T2 CRC is endoscopically resected, and the necessity for subsequent surgery is determined according to the histopathological evaluation of LNM risk.^6-8 This approach has shown promise in case reports and could be particularly beneficial for elderly patients or those ineligible for surgery due to comorbid conditions.^{9, 10}

Although some previous reports investigated risk factors of LNM in T2 CRC, most of them were limited to rectal or colon T2 cancer or small cases.^11-14 There are also no reports stratifying the risk of LNM combining these risk factors. This study aims to define the patients with very low risk of LNM to expand the number of cases in which endoscopic resection is a viable treatment option by discovering the important risk factors associated with LNM in T2 CRC.

METHODS

RESULTS

Study subjects

Figure 1 shows the study flow chart. Altogether, 731 patients with pT2 CRC were treated during the study period. Of these, 73 patients were excluded (synchronous invasive cancer; n = 3, transanal endoscopic microsurgery; n = 2, Lynch syndrome; n = 20, preoperative chemoradiotherapy; n = 46, and missing data; n = 2), and 658 were eligible and included in the analyses. The LNM rate was 25.8% (170/658; 95% CI, 22.5–29.4). Among these cases, 21.9% (144/658; 18.8–25.2), 2.7% (18/658; 1.6–4.3), and 1.2% (8/658; 0.5–2.4) were accompanied by N1, N2 and N3 metastasis, respectively. The mean number of retrieved lymph nodes per patient was 21, and the median was 19. The clinical characteristics of the patients are shown in Table 1. The median age of the patients was 69 years; 338 (51.2%) were male, with a median CEA level of 1.8 (1.1–2.7) and a CA19-9 level of 10.5 (6.2–17.7). The median tumor diameter was 28 mm (21–35), and 47.2% were located on the left side of the colon. Lymphatic and vascular invasion positivity presented in 297 (45.1%) and 388 (59.0%) cases, respectively.

TABLE 1. Characteristics of patients at baseline (n = 658).

Age (years)	69 (59–76)
Sex (male/female)	338 (51.4)/320 (48.6)
CEA	1.8 (1.1–2.7)
CA19-9	10.5 (6.2–17.7)
Tumor size (mm)	28 (21–35)
Location (left/right)	472 (71.7)/186 (28.3)
Histological differentiation Predominant (por/muc/sig)	19 (2.9)
Histological differentiation Highest-grade (por/muc/sig)	126 (19.1)
Lymphatic invasion (+)	297 (45.1)
Vascular invasion (+)	388 (59.0)
Adenomatous component (+)	22 (3.3)
Number of lymph nodes dissected	20 (13–28)
LNM-positive (+)	170 (25.8)

• Note: Results are expressed as median (interquartile range) or number of patients (%), as appropriate.
• Abbreviations: CA19-9, carbohydrate antigen 19-9; CEA, carcinoembryonic antigen; muc, mucinous carcinoma; por, poorly differentiated adenocarcinoma; sig, signet-ring cell carcinoma.

DISCUSSION

This study evaluated the clinicopathologic characteristics of T2 CRC to determine risk stratification and identify cases potentially treatable solely through endoscopic methods. The multivariate analysis identified lymphatic invasion, female sex, and elevated serum CEA (≥5.0 ng/mL) as statistically significant factors for LNM in T2 CRC at the time of surgery.

The development of endoscopic techniques such as eFTR, EID, and PAEM has expanded the range of lesions treatable endoscopically. Current indications for eFTR and EID include the primary treatment of optically diagnosed T1 CRC, non-lifted lesions, severe fibrosis, and secondary treatment after incomplete endoscopic resection. Recent reports have documented the successful endoscopic resection of T2 CRC with negative margins, particularly in patients who are poor surgical candidates.⁹ Given the overall postoperative mortality rate for CRC surgery of 1.3%–8.0%, with specific rates increasing with patient age, endoscopic treatment may offer a viable alternative for low-risk LNM patients with T2 CRC.¹⁷ In this study, the LNM rate in patients without lymphatic invasion was 3.3% (95% CI, 1.7–5.7), a rate that may be considered acceptable when compared to the postoperative mortality rate. In contrast, the LNM rate in patients with lymphatic invasion—the factor most strongly associated with LNM—exceeded 53.2% (95% CI, 47.3-59.0), making surgical intervention highly recommended. Although the present study is limited by a small sample size, further stratification of LNM risk may be possible by incorporating female sex and CEA levels, which are independent risk factors for LNM in this study as the sample size increases. However, there is no evidence regarding overall or disease-free survival rates. In T1 CRC, the low-risk group for LNM has been well-defined, with oncologic safety as the highest priority, and an acceptable long-term prognosis has been documented. As new treatment strategies for T2 CRC are introduced, it is imperative to establish stricter criteria. Drawing from previous findings, potential endoscopic treatment indications for T2 CRC have been suggested: (1) eFTR for lesions smaller than 20 mm, regardless of location, or (2) EID/PAEM for rectal tumors, irrespective of size.⁷ Preoperative diagnosis of the depth of invasion will be essential for these strategies to distinguish T2 cancer from T1/T3 cancer. There is no systematic pit pattern classification or narrow band imaging classification for diagnosing the depth of invasion of T2 CRC, but endoscopic ultrasound may be effective.¹⁸ Previous reports have shown that the diagnostic accuracy of the T stage was 94% with no under-staging.¹⁹

Previous studies have consistently reported lymphovascular invasion as a reliable risk factor for T1 and T2 CRC.^20-22 This study supports those findings and highlights additional common risk factors, such as female sex. This result agrees with findings from previous reports of T1 CRC, which identified female sex as a risk factor for LNM.^{23, 24} Although the underlying reasons for this trend are unclear, it may be associated with hormone levels, particularly the role of estrogen and its receptors.^{25, 26} This hypothesis suggests that hormonal differences could affect tumor behavior and metastatic potential and warrants further investigation to understand the mechanisms and potential implications for treatment strategies. Despite the known correlation of CEA levels with prognosis in advanced CRC, we show a correlation with LNM, specifically in T2 CRC.^{21, 27} In the future, the use of circulating tumor DNA is expected to enable the development of predictive models that are more accurate than those based on non-invasive methods, leading to a more precise delineation of the treatment indications.²⁸

This study has some limitations. First, it is a single-center, retrospective analysis. Second, while the degree of invasion has been recognized as a risk factor in T1 CRC, it was not assessed in this study since it was not assessed routinely in T2 CRC. In addition, tumor budding and perineural invasion were not included in this study because they had been identified as prognostic factors only since 2013 in the Japanese guidelines.¹⁵ This may result in potential risk factors for LNM being overlooked. Third, the degree of vascular invasion-positivity was subdivided from three to four levels in 2018 in the Japanese guidelines. Therefore, we stratified the degree of vascular invasion into two categories, positive or negative, for analysis and lymphatic invasion.

In summary, lymphatic invasion, female sex, and elevated CEA levels were identified as independent risk factors for LNM in T2 CRC. Combining these risk factors may enable stratification of LNM risk.

CONFLICT OF INTEREST STATEMENT

None.

ETHICS STATEMENT

This study protocol was approved by the institutional review board of Showa University (approval No.21-106-B).

PATIENT CONSENT STATEMENT

Written informed consent was obtained from all patients before treatment.

CLINICAL TRIAL REGISTRATION

This study was registered in the University Hospital Medical Network Clinical Trials Registry (UMIN000046992).

ANIMAL STUDIES

N/A.