Abstract
AIM: We compared variability in number of examined lymph nodes between pathologists and analyzed survival implications in lung and esophageal cancer after standardized lymphadenectomy. METHODS: Outcomes of 294 N2 dissected lung cancer patients and 132 2-field dissected esophageal cancer patients were retrospectively examined. The primary outcome was difference in reported lymph node count among pathologists. Secondary outcomes were overall and disease-specific survival related to this count and survival related to the 50% probability cut-off value of detecting metastasis based on the number of examined lymph nodes. RESULTS: The median number of examined lymph nodes in lung cancer was 13 (IQR 9-17) and in esophageal cancer it was 22 (18-29). The pathologist with the highest median number of examined nodes had > 50% higher lymph node yield compared with the pathologist with the lowest median number of nodes in lung (15 vs. 9.5, p = 0.003), and esophageal cancer (28 vs. 17, p = 0.003). Survival in patients stratified by median reported lymph node count in both lung (adjusted RMST ratio < 14 vs. ≥ 14 lymph nodes 0.99, 95% CI 0.88-1.10; p = 0.810) and esophageal cancer (adjusted RMST ratio < 25 vs. ≥ 25 lymph nodes 0.95, 95% CI 0.79-1.15, p = 0.612) was similar. The cut-off value for 50% probability of detecting metastasis by number of examined lymph nodes in lung cancer was 15.7 and in esophageal cancer 21.8. When stratified by this cut-off, no survival differences were seen. CONCLUSION: The quality of lymphadenectomy based on lymph node yield is susceptible to error due to detected variability between pathologists in the number of examined lymph nodes. This variability in yield did not have any survival effect after standardized lymphadenectomy.