Intra-operative detection of lymph node involvement in carcinoma of the colon.

BACKGROUND/AIMS: We studied the accuracy rate of intra-operative lymph node assessment compared with pathological examination to determine whether surgeons could modify the extent of lymphadenectomy during the operation. METHODOLOGY: Intra-operative and pathological lymph node assessments were compared in 360 patients with carcinoma of the colon. RESULTS: A total of 6,431 lymph nodes were examined, mean number per patient was 17.9. The overall accuracy rate of intra-operative diagnosis was 56.1%, sensitivity was 93.2%, and specificity was 41.7%. The accuracy rate of the diagnosis of N1 and N2 was 43.7% and that of N3 and N4 was 78.3% (p=0.001). There was no significant difference in the diagnosis rates in the colonic region. These results indicated that diagnosis in < or = N3 was more accurate than that in > or = N2. There were 5 false-negative cases. All of the false-negative lymph nodes were located adjacent to the colonic wall. CONCLUSIONS: Intra-operative diagnosis of the positivity of < or = N2 lymph nodes was too poor to decide the extent of lymph node dissection of < D2. It is adequate to dissect according to at least the D2 criteria in all cases. If the surgeon observes N3 involvement, he should add the D3 dissection.