The trusted source for
healthcare information and
Abstract & Commentary
Synopsis: This study involved a retrospective analysis of patients with T1 colorectal cancer. Using multivariate analysis, investigators found 3 variables that were statistically significant risk factors for associated lymph node metastasis in patients with sessile lesions. These factors were the presence of lymphovascular invasion, the depth of invasion into the submucosa, and tumor location in the lower third of the rectum.
Source: Nascimbeni R, et al. Dis Colon Rectum. 2002; 45:200-206.
Nascimbeni and colleagues from the mayo Clinic reviewed the records of more than 7000 patients in whom resection of a colorectal cancer was performed between 1979 and 1995. They identified 353 patients with sessile T1 adenocarcinomas after excluding cases with pedunculated tumors, synchronous tumors, hereditary predisposition, or inflammatory bowel disease. Patients treated with polypectomy or local excision alone were also excluded.
The 353 patients had a median age of 68 (range, 36-95). Eighty-seven patients had an attempt at local excision first, but a standard resection was subsequently performed because of unfavorable risk factors (57%), positive/uncertain margins (22%), or uncertain depth of invasion (13%).
Forty-six patients were found to have lymph node metastases (13%). A number of characteristics were studied as risk factors for lymph node metastasis. These included patient age, patient gender, size of the carcinoma, presence of mucinous carcinoma or grade (differentiation). Tumor grade was significant on univariate analysis (P = 0.004). However, only 3 factors were statistically significant on multivariate analysis: the presence of lymphovascular invasion (LVI), the location in the bowel, and the depth of invasion into the submucosa.
The location in the bowel yielded an interesting observation. Lesions of the lower third of the rectum had a higher risk of positive nodes (34%) than the middle (11%) or upper (8%) rectum (P = 0.007). However, no difference was seen when comparing tumors originating from the left colon vs. right colon vs. the rectum.
Although not part of the TNM staging system, the risk of lymph node metastasis was analyzed by the degree of invasion into the submucosa. Positive nodes were significantly more common if the cancer penetrated the lower third of the submucosa (23%) compared with the middle third (8%) or upper third (3%) (P = 0.001).
In their retrospective analysis, Nascimbeni et al make several interesting observations regarding the risk of lymph node metastasis in T1 colorectal cancer with sessile lesions. On a multivariate model, only the presence of LVI, the depth of invasion into the submucosa, and tumor location in the lower third of the rectum retained statistical significance. A poorly differentiated cancer was also statistically significant (P = 0.004), but only in the univariate model due to the association of these higher-grade tumors with deeper invasion into the submucosa (P = 0.001).
Lymphovascular invasion and tumor differentiation are well-recognized risk factors for lymph node metastasis in colorectal cancer. Less frequently discussed but previously reported1 is the depth of submucosal invasion. Compared with the study by Nascimbeni et al, remarkably similar rates of positive nodes were reported for each segment of submucosal invasion (25%, 10%, and 0% for the lower, middle and upper third, respectively).1 Although the depth of invasion makes biologic sense, it has not been shown to be reproducible.
In the multivariate analysis, a T1 tumor in the lower third of the rectum had a higher rate of lymph node involvement than tumors in the upper two thirds of the rectum or the colon. The reasons for this are unclear although one could hypothesize inherent biologic differences similar to the right-sided tendency of colon cancer in HNPCC. Nevertheless, the clinical significance of this is unclear because it is not reported how many of these node-positive, lower-third rectal cancers were well differentiated and without LVI.
Because patients treated with polypectomy were excluded, this study provides no information regarding the indications for resection after a polypectomy. Nascimbeni et al also point out that positive margins should be considered inadequate treatment rather than a risk factor. Their interesting study confirms the importance of tumor grade and LVI in early rectal cancer, and raises awareness of the potential importance of depth of submucosal invasion. Their results also suggest that there may be variations in the biologic behavior of rectal cancer depending on the site of origin within the rectum.
Dr. Kotz is of Hanover Medical Specialists, Wilmington, NC.
1. Kikuchi R, et al. Dis Colon Rectum. 1995;38: 1286-1295.