Premium
Factors influencing lymph node retrieval in colorectal cancer and its effect on survival
Author(s) -
Evans M. D.,
Rees A.,
Stamatakis J. D.,
Karandikar S. S.
Publication year - 2006
Publication title -
colorectal disease
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.029
H-Index - 89
eISSN - 1463-1318
pISSN - 1462-8910
DOI - 10.1111/j.1463-1318.2006.01102_2.x
Subject(s) - medicine , colorectal cancer , lymph node , stage (stratigraphy) , lymph , log rank test , survival rate , survival analysis , cancer , surgery , gastroenterology , pathology , paleontology , biology
NICE has set standards for lymph node (LN) harvest in surgery for colorectal cancer (CRC). We report the effect of the operating surgeon and reporting pathologist on lymph node yield and the role of LN yield on survival. Method: Data on all patients with CRC treated in a single unit between 1999 and 2004 were collected in the ACPGBI database. Lymph node harvest was identified for each of three surgeons and three pathologists. Actuarial survival of all patients with Dukes’ stage B with greater than or less than 8 reported LNs and Dukes’ stage C were compared using log‐rank test (Kaplan–Meier method). Results: A total of 380 cases had a curative resection with a unit median LN retrieval of 13. There was no difference in survival between Dukes’ stage B <8 LN (mean 42 months, 95% CI 29–55) and Dukes stage C (mean 45 months, 95% CI 29–51), log‐rank P = 0.7618. Survival between Dukes stage B (8 LN and Dukes C was significantly different (mean 58 months, 95% CI 53–64 vs 45 months, 95% CI 29–51), log‐rank P = 0.006.Colorectal surgeon Non colorectal Pathologist 1 n = 183 2 n = 64 3 n = 86 n = 47 1 n = 200 2 n = 106 3 n = 57 Locum n = 17Median lymph node observed (Range) 12* 1 – 37 14* 0 – 25 14* 2 – 42 14* (1–32) 15** (2–42) 12** (1–32) 11** (0–29) 11** (2–39) 95% CI 12.5–14.5 11.7–14.6 13.3–16.6 12.–16.3 14.1–16.0 12.0–14.4 9.7–12.8 7.7–16.9* P = 0.498 (Kruskal–Wallis). ** P < 0.0001 (Kruskal–Wallis).Conclusions: This study demonstrates that within a single unit the reporting pathologist significantly influences LN yield in CRC treatment and that a minimum of eight lymph nodes needs to be examined to stage CRC as Dukes’ C accurately. In this unit the surgeon was not a significant variable.