As an alternative to radical surgery, local excision with or without adjuvant pelvic radiation therapy has been used in selected patients with rectal cancer. To determine which clinical and pathologic features can predict the presence of positive lymph nodes (LN+), a retrospective review of 168 patients who underwent potentially curative surgery for rectosigmoid and rectal cancer was performed. By univariate analysis, tumor penetration, grade, and histology were significant predictive features. This was confirmed by logistic regression analysis. The incidence of LN+ increased with increasing tumor penetration (T1, 0%; T2, 28%; T3, 36%; T4, 53%; P = 0.04), grade of adenocarcinoma (well-differentiated, 0%; moderately differentiated, 30%; poorly differentiated, 50%; P = 0.07, [borderline significance]), and the presence of any colloid histology (pure adenocarcinoma, 30%; total colloid, 52%; P = 0.04). Using 2 X 2 contingency tables, the presence of blood vessel invasion (BVI), lymphatic vessel invasion (LVI), vascular invasion (VI), total colloid histology, and high grade further increased the incidence of LN+ with increasing tumor penetration. Regardless of tumor size, grade, histology, BVI, LVI, or VI, none of the nine patients with Stage T1 tumors or the seven with well-differentiated adenocarcinomas had LN+. For this group, local excision alone is recommended. The incidence of LN+ was greater than or equal to 19% in all other categories. For this group of patients, if there is no evidence of gross tumor in the pelvis, then a local excision plus adjuvant pelvic radiation may be an alternative to radical surgery.
