Can depth of tumour invasion predict lymph node positivity in patients undergoing resection for early rectal cancer? A comparative study between T1 and T2 cancers.

OBJECTIVE: The present study investigated the risk of lymph node metastasis according to the depth of tumour invasion in patients undergoing resection for rectal cancer.
METHOD: The histology of patients undergoing oncological resection with regional lymphadenectomy for rectal cancer at St Marks Hospital from 1971 to 1996 was reviewed. Of the total number of 1549 patients, 303 patients with T(1) or T(2) rectal cancers were selected. The tumour type, grade, evidence of vascular invasion, depth of submucosal invasion (classed into 'sm1-3') were evaluated as potential predictors of lymph node positivity using univariate and multi-level logistic regression analysis.
RESULTS: Tumour stage was classified as T(1) in 55 (18.2%) and T(2) in 248 (81.2%) patients. The incidence of lymph node metastasis in the T(1) group was 12.7% (7/55), compared to 19% (47/247) in the T(2) group. The node positive and negative groups were similar with regard to patient demographics, although the former contained a significantly higher number of poorly differentiated (P = 0.001) and extramural vascular invasion tumours (P = 0.002). There was no significant difference in the number of patients with sm1-3, or T(2) tumour depths within the lymph node positive and negative groups. On multivariate analysis the presence of extramural vascular invasion (odds ratio = 10.0) and tumour grade (odds ratio for poorly vs well-differentiated = 11.7) were independent predictors of lymph node metastasis.
CONCLUSION: Whilst the degree of vascular invasion and poor differentiation of rectal tumours were significant risk factors for lymph node metastasis, depth of submucosal invasion was not. This has important implications for patients with superficial early rectal cancers in whom local excision is being considered.