Clinical trial and institutional data now becoming available are providing new insights into the value of sentinel node (SN) biopsy as a staging procedure for patients who present with primary melanomas, and the consistency of results obtained by institutions around the world has been remarkable. The reported morbidity of SN biopsy has been very low, and the suggestion that the incidence of in transit metastasis is increased by performing a SN biopsy has been shown to be incorrect by the results of several large, retrospective single institution studies and by the interim results of a large prospective randomized trial (MSLTI). It is already clear that SN biopsy provides better prognostic information than any other parameter or combination of parameters studied to date, and identifies patients who may benefit from immediate complete regional node dissection and/or adjuvant therapy. Without knowledge of SN status, accurate staging is not possible, adjuvant therapy with interferon for node positive patients cannot be offered, and patients are denied the opportunity to enter clinical trials of other adjuvant therapies. They are also left in a state of uncertainty, often associated with considerable anxiety, if a SN biopsy is not performed, because of the knowledge that they have an approximately 20% chance of recurrence in the regional node field if their melanoma is >1 mm in thickness. The interim results of MSLTI make it clear that if recurrence does occur, involvement of more nodes is probable and the chance of cure is therefore likely to be reduced. Technical aspects of the SNB procedure are nowwell established and validated, and it is likely that the false negative rates observed in studies reporting early experiences with SN biopsy will fall as time goes by. The SN has become a focus of attention for researchers seeking to better understand the process of melanoma metastasis, and to find more effective ways of dealing with this problem. Major advances in our understanding of the immunology and molecular biology of the metastatic process in patients with melanoma have already been made by focused study of sentinel nodes, and this work is proceeding apace.Meanwhile, efforts to identifyways of predicting non-SN metastasis, and to find rapid, reliable and completely noninvasive methods of assessing SN status are continuing.

4047 Background: Lymph node (LN) status is the most important prognostic factor in colon cancer (Cca). In various trials, the average nodal positivity of conventional surgery in Cca is about 33%. Ultrastaging of SLNs results in higher and more accurate nodal staging of patients (pts) with Cca. However, some recent publications of SLNM in Cca have shown variable results with differing conclusions. Hence, prospective data from 3 continents were analyzed to study the international experience of SLNM in Cca. Methods: Only centers with experience of 40 or more cases of SLNM in Cca were included in the study. SLNM was performed by peri-tumoral injections of 1–3 ml of 1% lymphazurin. First 1–4 blue nodes marked as SLNs were ultrastaged by multilevel microsections for H&E and IHC. Data for calculating the success rate, accuracy, skip metastases (mets), sensitivity, negative predictive value; nodal positivity and upstaging were collected from each center. Results: Our study included a total of 1,216 Cca pts from 9 centers over 3 continents. SLNM was successful in 92.9% pts ( Table 1 ). The average number of LN/pt was 18.5 and the average number of SLN/pt was 2.7. The overall sensitivity, accuracy rate and negative predictive value were 78.3%, 89.4% and 82.8% respectively. Nodal mets were found in 52.9% pts. Of these, SLNs were the exclusive site for mets in 30.1% pts while 18.3% pts were upstaged by SLNM. Skip mets were seen in 21.7% pts (range 9.5% - 44.1%). Conclusions: SLNM is highly successful in Cca when performed by experienced surgeons worldwide. Nodal positivity was found to be much higher in pts undergoing SLNM compared to conventional surgery. Upstaged pts may benefit from adjuvant chemotherapy. Though the variation of skip mets was wide, the clinical impact of skip mets in Cca is negligible compared to that in melanoma and breast cancer, since all pts undergo standard lymphadenectomy and all node positive pts (true +ve & skip mets) are usually treated with adjuvant chemotherapy. [Table: see text] No significant financial relationships to disclose.

3621 Background: Clinical application of SLNM in CRCa patients (pts) is controversial due to the variable results in the literature for its success, skip metastases (mets) and accuracy. Hence prospective data from 5 institutions were analyzed to identify the factors associated with failure and skip mets in CRCa pts undergoing SLNM. Methods: SLNM was performed by peri-tumoral injections of 1–3 ml of 1% lymphazurin subserosally. First 1–4 blue nodes marked as SLNs were examined by 4 sections with H & E and 1 for cytokeratin. Rate of failure and skip mets rate along with age, sex, tumor site, size, grade, T & N stages were analyzed. Results: 549 consecutive CRCa pts underwent SLNM; 453 colon (C) and 96 rectal (R) pts. M: F ratio was 48%: 52% and median age of 72 years. The average no. of LNs was 14.5. SLNM failed in 1% of C and 8% of R pts ( Table ). Of the 8 R failures, 7 had neoadjuvant therapy. No correlation was found between failure, size or T-stage. Overall nodal positivity (+ve) rate was 48%. Of the 466 invasive CRCa pts with successful SLNM, rate of skip mets was 7% in C and 4% in R. Age, sex, grade and neoadjuvant therapy had no correlation with skip mets. Of 28 skip mets pts, 79% had tumors >3cm and 93% had advanced (T3, T4) tumors. Higher skip mets also was found in the transverse colon tumor. The success, accuracy, and sensitivity rates were 98%, 95%, and 88% respectively. Of node +ve pts, 47% had mets found only in the SLNs. SLNM upstaged 35% of C and 36% of R pts with micromets. Conclusions: In CRCa, SLNM is highly successful and accurate in predicting the presence or absence of nodal mets. Submucosal fibrosis of the lymphatics due to neoadjuvant therapy may result in higher failure rates for R pts. Skip mets are higher in transverse C tumors and increase in frequency as the T-stage increases. Pts upstaged by SLNM may benefit from adjuvant chemotherapy. [Table: see text] No significant financial relationships to disclose.

3567 Background: CS and pathology understages 20% of patients (pts) with Cca. Adjuvant chemotherapy (CRx) is highly effective in pts with nodal mets. SLNM upstages up to 20% of Cca pts for nodal mets. Upstaged pts benefit from adjuvant CRx, reducing the chances of recurrence. A comparative analysis was done for nodal mets and recurrence in Cca pts undergoing SLNM vs. CS. Methods: Group (gp) A pts underwent SLNM with 1% Lymphazurin followed by standard oncologic resection. Gp B pts underwent CS. SLNs were examined by 4 sections for H&E and 1 for IHC. The nonSLNs in gp A and all LNs in gp B were examined by 1–2 section with H&E. The median follow-up was 54 months and the minimum follow-up was 27 months. Comparison was done for age, number of LNs, T-stage, nodal status and recurrence rates between gp A and B. Results: Gp A (SLNM) had 153 pts (median age =72 yrs); Gp B(CS) had 162 pts (median age =82 yrs). Mean number of LNs per pt in gp A was 15.5 vs. 11.8 in gp B (p=0.0001). T-stage distributions for gp A and gp B were as follows: 11.1% vs. 8%(T1), 22.2% vs. 14.2%(T2), 62.1% vs. 68.5%(T3) and 4.6% vs. 8.6%(T4) respectively. SLNM was successful in 100% of pts with an accuracy of 94.7% and a sensitivity of 89%. SLNM upstages 15% of pts with micromets which may have been missed by conventional pathological exam. For stage I (T1/T2, N0) pts, the recurrence in gp A was 2.5% vs. 12.5% in gp B (p=0.002). Overall, the recurrence rate in gp A was 7.1% vs. 24.6% in gp B (p=0.001). The distribution of nodal mets and recurrence between gp A and gp B are in Table 1. Conclusions: SLNM is highly feasible and accurate in determining nodal status. Compared with CS, SLNM upstages a significant number of pts whose disease may remain undetected by conventional exam. Upstaged pts received adjuvant CRx, leading to a significant decrease in recurrence and a possible increase in survival. Table 1 Distribution of nodal metastasis and recurrence rate between group A and group B A. Nodal metastasis between group A and B Group A (SLN) Group B (conv) Total No=153 Total No=162 p-value Node positive/total(%) Node positive/total (%) T1 2/17 (11.7%) 1/13 (7.7%) 0.72 T2 9/34 (26%) 3/23 (13%) 0.24 T3 64/95 (67.3%) 47/112 (42%) 0.0004 T4 4/7 (57.1%) 6/14 (42.8%) 0.55 Total 79/153 (51.6%) 57/162 (35.6%) 0.005 B. Recurrence rates by nodal status between group A and group B (7.1% vs 24.6%, p value=0.0001) Node Positive Node Negative Group A Group B p-value Group A Group B p-value (n=79) (n=57) (n=74) (n=105) T1 0 0 — 1 2 0.76 T2 0 0 — 0 2 — T3 8 19 0.001 1 13 0.009 T4 1 2 0.39 0 2 — Total 9(11.3%) 21(36.8%) 0.0004 2(2.7%) 19(18%) 0.0025 No significant financial relationships to disclose.

Current conventional surgical and pathological techniques substantially understage colon cancer. This is evidenced by the fact that a significant subset of patients who are stage I and II at the time of colectomy return with distant metastases and ultimately succumb to the disease within the next 5 years. The identification of more nodes within a specimen and the detailed analysis of lymph nodes with advanced pathological techniques such as immunohistochemistry and reverse-transcriptase polymerase chain reaction (RT-PCR) can improve the staging of colon cancer, but are also associated with tremendous financial, time, and labor constraints. Sentinel lymph node (SLN) mapping has provided an avenue of staging colon cancer with high success rates and accuracy rates, while maintaining cost- and time-effectiveness. The ability to reproduce these results is dependent on adherence to the technical details of the procedure, and thereby providing the pathologist with the true SLNs, upon which the advanced pathological studies can be applied. We report our experience of SLN mapping for colon tumors in 209 patients, elaborating on the materials used, technical details, pitfalls, and results of the procedure. Our results show a success rate of 100% (209/209) and an overall accuracy rate for predicting positive or negative metastatic disease of 96.2% (201/209). Nodal metastases were identified in 46.2% (85/184) of patients with invasive disease (stage T1 to T4). The SLN was the exclusive site of metastases in 38.8% (33/85) of these patients, and the nodal disease was detected only as micrometastases in 22.4% (19/85). The skip metastases rate (false negatives) was 9.4% (8/85). SLN mapping is a powerful tool for accurate staging of colon cancer with a high success rate. The upstaging associated with this procedure may reveal disease that might otherwise go undetected by conventional surgical and pathological methods. Those patients who are upstaged can then benefit from adjuvant chemotherapy, which has been shown to improve survival of colon cancer patients with nodal disease by at least 33%.