Lymph node metastasis is one of the key factors of poor outcomes in patients with colorectal cancer. However, recurrence and metastasis rates are high in stage II colorectal cancer patients even without lymph node metastasis, and it has been reported that stage IIB patients have a poorer prognosis than stage III patients [11, 21]. Inaccurate lymph node staging may be one of the reasons for the low survival rate. Methods examined to improve lymph node staging include increasing number of lymph nodes for postoperative examination and using micrometastasis detection; however, it is difficult to apply these techniques in clinical practice due to procedural complexity and costs.
Theoretically, SLN detection can help reduce understaging because the SLN is the first to receive drainage from the tumor. In our study, ex vivo SLN detection was used in 54 patients and we found an identification rate of 100% and 4 patients were upstaged. Our results are consistent with other studies that have shown SLN mapping results in a higher detection rate and upstaging when combined with immunohistochemical or RT-PCR techniques [1, 16, 17, 20, 22]. A prospective randomized study of SLN ultrastaging showed that SLNs were successfully identified in 82 of 84 patients (97.6%), and significant nodal upstaging occurred (38.7% to 57.3%, p = 0.019) . Another prospective multicenter trial demonstrated that at least one SLN was identified in 268 of 315 enrolled patients (detection rate, 85%) and 21% (30 of 141) of the patients, classified as pN0 by routine histopathological examination, were found to have micrometastasis or isolated tumor cells in the SLN . Interestingly, a study by Cheng et al.  of colorectal cancer patients and liver metastases showed that nuclear beta-catenin overexpression in metastatic lymph nodes was strongly associated with liver metastasis. Perin et al.  reported a case a 46-year-old female who underwent sigmoid colon resection for colon cancer who subsequently developed pathologically proven breast metastasis of the colon malignancy. Examination of the axillary SLN revealed metastasis consistent with the colon primary.
Despite the encouraging results, there remains controversy regarding false negative rates. Our false negative rate was 68.2%, higher than that reported by other authors, and our results indicated that the SLN did not predict overall nodal status. This is similar to the findings of a recent systematic review and meta-analysis performed by van der Pas et al. . The authors included 52 studies with 3767 SLN procedures (78.6% colon carcinoma, 21.4% rectal carcinoma) and found a mean detection rate of 94% and a pooled sensitivity of 76%. Retter et al.  reported the results of 31 patients who received surgery for colon carcinoma in which in vivo SLN mapping was performed, and found that although the SLN was identified in 28 of the 31 patients, the false-negative rate to identify stage III disease was 66% and the accuracy was 14%. Finan et al.  also reported that ex vivo SLN mapping did not improve staging after proctectomy for rectal cancer. In contrast, van Schaik et al.  reported no false negative SLNs in 44 patients. The reasons for the high false-negative rates in some studies are unclear, but may possibly include the T stage, tumor location, and learning curve for SLN mapping. Thus, we believe that the cumulative data indicate that routine SLN detection alone cannot replace the conventional method that examines all dissected lymph nodes because a high false-negative rate tends to downgrade tumor staging such that some patients who may benefit from adjuvant chemotherapy will not receive it.
Although almost all studies analyzing SLN mapping for colorectal cancer demonstrate the detection of micrometastasis, the clinical significance of lymph node micrometastasis, particularly that identified solely by IHC staining, is unknown. There have been extensive studies on lymph node micrometastasis using IHC or RT-PCR techniques, and some reports have indicated that micrometastasis detection by ICH was associated with poor prognosis of colon cancers [8–10]. However, other studies have reported conflicting findings [5, 11]. Dahl et al.  reported that only patients with metastatic lymph nodes detected directly or within a SLN died of metastatic disease. Saha et al.  demonstrated with a 2-year minimum follow-up of 153 patients who underwent SLN mapping, 7% had recurrences as compared with 25% of 162 patients with standard resection and nodal staging. Bilchik et al.  reported that no colon cancer patient with a negative SLN by HE and PCR had a recurrence at a mean follow-up of 25 months. Our results indicated that patients without lymph node metastasis by HE staining who are upstaged by ICH detection of micrometastasis have a significantly poorer OS and DFS. However, survival analysis showed no difference in OS and DFS between the stage III and stage N0(i+) groups. This may imply that patients with SLN micrometastasis have the same prognosis as patients with stage III disease, which suggests that focusing on the detection of SLN micrometastasis could become a useful factor in determining prognosis, although further study is warranted due to limited number of cases. In addition, in multivariate analysis upstaging by SLN micrometastasis identification was not an independent prognostic factor; the small number of study patients may be the reason.
The primary limitation of the study is the relatively small number of patients. However, the results do support those of other studies of the utility of SLN mapping and IHC detection of micrometastasis.