4 Surgeons generally remained very

selective in their use of these treatments. Peranal local excision could be technically demanding in all but the smallest, most distal, posterior tumors. Furthermore reports emerged of substantial rates of lymph node metastasis in tumors, which had not breeched the muscularis propria; in 1982 Hojo reported lymph node metastases in 18% of 28 T1 and 38% of 82 T2 rectal tumors.5 In most centers, local excision was generally limited to elderly, high risk patients who would otherwise require a permanent stoma. In this issue of the Journal, Nakadoi et al. relate the presence of regional lymph node metastases to the pathological features of the primary tumor in 499 surgically BMN 673 datasheet resected T1 colorectal carcinomas.6 Lymph node metastases, found in 8.2% of subjects, were mostly predicted by the presence of poor differentiation, lymphovascular invasion or high grade tumor budding. They found a low rate of lymph node metastasis (1.2%) if all such features were absent. All of the lymph node metastases selleck chemical occurring in tumors without these high risk features were in tumors with a depth of invasion ≥ 1800 µm. The authors present a case for endoscopic management of low-risk T1 colorectal carcinomas so selected. While the study

appears rigorous and the case well-argued. Caution should be exercised. First, the significance of lymph node metastasis and the biological processes by which this occurs needs consideration. Lymph node metastasis is an accepted surrogate of poor survival. A simplistic view of stepwise cancer progression leads one logically to the view that radical resection is appropriate and is essential for cure when lymph node metastases are present. In many cases, however, lymph metastases might be an indicator of disease behavior—the harbinger of poor outcome despite radical surgery. If one considers that the process of metastasis is a function of biological factors, time and the Glycogen branching enzyme area of tumor exposed to the vascular and lymphatic surfaces,

“early” tumors that spread to lymph nodes might be assumed to be biologically aggressive. If tumor grade, the presence of lymphovascular invasion and budding reflect this biological activity, it may be that in cases exhibiting such features, radical surgery is of little benefit since the disease is already a systemic one. An analogy with breast cancer might be appropriate: local treatment with aggressive systemic therapy producing the best outcomes. One might expect this hypothesis to become more deserving of investigation as the proportion of cancers detected by screening increases. Equally, failure to detect involved lymph nodes cannot be regarded as an assurance that there is no resectable disease beyond the submucosa.