Stage Classification

Considering specific features of the primary cancer (particularly pT status, lymphovascular invasion and tumor grade) are directly associated with the risk of LNM, primary local staging of rectal cancer is of paramount importance for the selection of appropriate candidates for transanal local excision. Diagnostic biopsies may allow proper determination of tumor grade. On the other hand, determination of lymphovascular invasion often requires excisional biopsy specimens and is therefore almost impossible to accurately assess preoperatively. Ultimately, clinical/ radiological T and N classification (cT and cN) are frequently the only sources of information used for management decision of these patients.



Depth of the primary tumor may be accurately determined by the use of different radiological imaging modalities. Both endorectal ultrasound (ERUS) and high-resolution magnetic resonance (MR) have been extensively studied for this particular purpose. Both imaging modalities provide acceptable overall accuracy for each cT classification (>90 %) [6-8]. However, considering the risk of LNM amongst T3 and T4 rectal cancers are exceedingly high, these patients are not even considered for local excision except in extreme palliative situations. The distinction between T2 and T3 rectal cancers would therefore possibly provide a first filter for patients potentially suitable for a local procedure. A meta-analysis of accuracy rates for local staging of rectal cancer has been performed comparing different staging modalities [9]. Interestingly, ERUS was associated with higher sensitivity rates for the distinction between T2 and T3 cancers whereas specificity was nearly identical. In other words, MR may result in significantly more underestimation of T stage of these patients, potentially leading to inappropriate indication of local treatment for unsuspected T3 disease [9].



Distinguishing between cT1 and cT2 is perhaps the most relevant step in the assessment of these patients. The same meta-analysis of the results of rectal cancer staging with MR and ERUS suggests that specificity for the distinction of pT1 from pT2 was best for ERUS, even though sensitivity was similar between both modalities. Therefore, in contrast to the distinction between T2 and T3, this means that MR overestimates more frequently between T1 and T2 rectal cancer when compared to ERUS [9]. But the ideal patient for a transanal local excision is the one with a cT1 cancer, preferably restricted to sm1. This is due to the fact that the risk of LNM may also be correlated with the level of submucosal involvement. In fact, full-thickness excision allows better estimation of sm level invasion. In contrast, partial thickness endoscopic resections may not provide the entire submucosa for pathological review and therefore subdivision into thirds may be impossible. In this setting, absolute measurement of depth of tumor invasion (in specimens without entire submucosa available) may provide clinically relevant information as well. In non-pedunculated T1 cancers, invasion within the submucosa of <1,000 p is associated with no risk of lymph node metastases even in the presence of lymphovascular invasion. In addition, three-dimensional ERUS in experienced hands was able to correctly identify sm level of invasion with acceptable accuracy rates [10]. In this study, patients with pT0 and pT1 sm1 were correctly distinguished from pT1 with massive invasion or pT2 with excellent accuracy rates [10].

  • Contact
  • Category: Cancer