Prognostic factors in T1 colorectal adenocarcinoma
Posted Sep 07 2010 4:40am
A frequently encountered specimen is the adenocarcinoma with submucosal invasion arising in a dysplastic adenomatous polyp or sessile lesion that has been resected endoscopically. While much effort is directed at determining margin adequacy and depth of invasion, less attention has been given to other parameters that may be associated with risk of lymph node (LN) metastasis. Measuring the depth of invasion is frequnetly problematic given the morcellated nature of most specimens. Generally, if the lesion is felt to be adequately excised, a formal colectomy may not necessarily be needed since the risk of LN metastasis in a T1 tumor is felt to be too low to warrent the morbidity of a colectomy.
A recent article in the August 2010 Modern Pathology sought to detect patients at high risk for LN metastasis after endoscopic mucosal resection.
Tateishi Y, Nakamishi Y, Taniguchi H et al. Pathological prognostic factor spredicting lymph node metastasis in submucosal invasive (T1)colorectal adenocarcinoma. Mod Pathol 2010;23:1068-1072.
This is a single-institution study that examined 322 consecutive patients treated by colectomy (complete tumor excision plus lymphadenectomy) for T1 colorectal adenocarcinoma. The entire lesion was examined microscopically in all cases and criteria evaluated for the study included submucosal invasive depth, tumor size, lymphatic invasion, venous invasion, histologic grade, tumor budding at the submucosal invasion front, and status of muscularis mucosae.
The authors found LN metastases in 14.3% of patients (46/322)--which was a surprising and sobering finding in itself to me. By univariate analysis, tumor budding, histologic grade, lymphatic invasion, venous invasion, submucosal invasion depth > 1 mm, and completely disrupted (type B) muscularis mucosae were associated with LN metastasis. By multivariate analysis, however, only tumor budding, high grade tumor differentiation, and lymphatic invasion were associated with LN metastasis. Moreover, of the 46 patients with LN metastasis, 40 (87%) had at least 1 out of 3 factors by multivariate analysis associated with LN metastasis; the remaining 6 all showed type B muscularis mucosae status. Using these four criteria, 285 out of 322 patients were positive for at least one criteria and would have undergone colectomy after submucosal resection--for 100% sensitivity but only 13.4% specificity.
Certainly, the decision whether colectomy is the most appropriate definitive treatment following endoscopic resection must be individualized. But it seems in my experience that this is a gray zone for surgeons and oncologists and these criteria, while needing independent and prospective validation, are worthy of further study and consideration of being adopted for routine use now. The evaluation of muscularis mucosae status is not familiar in the U.S. and the reader is referred to the original article by Tominga et al in Dis Colon Rectum (2005) for more detail.
The practical points here are 1) that these parameters can be readily assessed in routine H&E sections and 2) using four histologic risk factors--tumor budding, lymphatic invasion, high-grade histologic tumor differentiation (moderately- to poorly-differentiated), and type B muscularis mucosae status--the authors were able to select all patients who were shown to have LN metastasis.