To the Editor:

We read with interest the article by Johnson et al.¹ These investigators analyzed data from the Surveillance, Epidemiology and End Results program to demonstrate the prognostic significance of the number of negative lymph nodes in patients with stage III colon cancer. As they indicate, this study is on the background of multiple other series reporting that total lymph node count is an important prognostic factor for patients with colorectal cancer (CRC). Johnson et al speculated that the number of lymph nodes might be a marker for the adequacy of surgical, pathologic, or institutional care. However, these investigators were unable to further analyze the data by individual surgeon or institution, and also were unable to reference any literature regarding the impact of quality of pathologic technique on outcome. We have addressed the knowledge deficit regarding the influence of pathologic service on lymph node yield by analyzing a series of our colorectal surgical databases.

Initially, we compared the lymph node yield after colon cancer surgery at two hospitals in Adelaide with separate pathology departments, and examined the period between January 1999 and December 2004. At pathology provider A, where 191 specimens were reported, the median lymph node count was 10 while at pathology provider B, where 197 specimens were reported, the median lymph node count was 19. The difference between the two hospitals was statistically significant (P < .001).

To ensure that this difference was not due to variations in the quality of the surgical practice at the two sites, the cases of an individual high-volume surgeon (N.A.R.) who had operated at both hospitals (76 specimens for pathology provider A and 54 specimens for pathology provider B) were examined. Again marked differences in lymph node yield were evident (median, 10 v 18; P < .001).

To determine if an intervention in pathology protocol could influence lymph node yield, we reviewed the experience of two Melbourne pathology providers. Provider C in 2003 instituted a standardized method of handling and reporting on CRC specimens in order to improve quality of care. For colon cancer patients before 2003 (n = 81), the median lymph node count was 7; after the improvements in practice (n = 81) the median lymph node yield increased to 12 (P ≤ .001). Over the same time period at a separate Melbourne institution (pathology provider D) where many of the same surgeons operate, and where there was no change in pathology practice, the median lymph node count for colon cancer surgery remained steady at 14 (n = 100 pre - and post 2003). These data suggest that a deliberate intervention to improve quality may lead to an increased number of analyzed lymph nodes.

As data emerges regarding the prognostic significance of lymph node yield in colon cancer, understanding the factors that contribute to the number of lymph nodes retrieved and analyzed is critical to improving patient outcomes. Our data demonstrates that the pathologist has a significant influence on the number of lymph nodes reported after colon cancer surgery, and that improved lymph node yields can be obtained with appropriate intervention in pathology practice. Further study of pathologic quality and ways that it may be optimized are warranted.

AUTHORS' DISCLOSURES OF POTENTIAL CONFLICTS OF INTEREST

The authors indicated no potential conflicts of interest.