Background: In patients with resected colorectal carcinoma, lymph node involvement has particular importance for patient prognosis and adjuvant therapy. The network of French cancer registries (FRANCIM) established a study aimed at analyzing the validity of lymph node harvest reporting in a population-based sample.
Methods: The study population was comprised of 1081 resected tumors without distant visceral metastasis and classified using the TNM system. Correlation between the number of examined lymph nodes and the staging of the tumor was examined by logistic regression analysis to establish an estimate of the minimum number of lymph nodes required to determine whether a tumor is lymph node negative.
Results: An average of 7.7 +/- 0.2 lymph nodes were examined per specimen in the 851 patients for whom the number of lymph nodes examined was known. The proportion of cases classified as N+ increased significantly with the number of examined lymph nodes (chi-square trend = 24.6; P < 0.0001). If the probability of correct lymph node status assessment is 1 in the reference group (comprised of pathology reports of specimens with > or = 16 examined lymph nodes), the probability of correct N+/N- dichotomization was significantly < 1 for the 1 to 3 lymph nodes group and the 4 to 7 lymph nodes group (i.e., 53.7% of cases).
Conclusions: To comply with current rules for adjuvant chemotherapy, surgeons must provide pathologists with at least eight lymph nodes for optimal N+/N- dichotomization to reduce the risk of misclassification and understaging.