|
Journal of Clinical Oncology, 2008 ASCO Annual Meeting Proceedings (Post-Meeting Edition).
Vol 26, No 15S (May 20 Supplement), 2008: 15009
© 2008 American Society of Clinical Oncology
Identification of 12 or more lymph nodes in resected colon cancer specimens as an indicator of quality performance
R. O. Dillman,
K. Aaron,
F. S. Heinemann and
S. E. McClure
Hoag Cancer Center, Newport Beach, CA; Hoag Hospital, Newport Beach, CA
15009
Background: The percentage of colon cancer resections, in which 12 or more lymph nodes were identified in pathology reports, was recently endorsed as an indicator of quality performance by various groups. However, who or what is actually being evaluated in such an analysis? Methods: We utilized the Hoag Cancer Center data base to examine several variables potentially associated with the number of nodes identified in colon cancer resections, including pathology procedures, surgeons, anatomic location, tumor stage, and patient age. Patients whose cancers were contained in a polyp were excluded. Statistical comparisons were two-tailed, and included the chi square test for paired or grouped proportions, t-test for means, and log-rank test for survival curves. Results: A change in tissue processing procedures in 1998 was followed by an increase in the number of nodes identified from 8 per specimen during 1989–1997 to 15 during 1998–2005 (p<.0001). For 571 patients who underwent surgical resection of colon adenocarcinoma at Hoag Hospital during 1998–2005, the rates of failure to identify at least 12 nodes varied from 17% to 43% by seven anatomic locations (p<.0001), from 25% to 35% by patient age (p=.027), from 27% to 41% by stage of disease (p=.004) and from 20% to 47% among 12 surgeons who performed at least 15 colon cancer resections (p=.014). Fewer than 12 nodes were identified in 49% of tumors resected by a cohort of 30 surgeons, each of whom performed less than 10 of the colon cancer resections. The greatest number of nodes (mean 19.5) were identified in patients with local extension node-negative (T3 and T4) disease. Identification of 12 or more nodes was associated with better survival for patients with local disease, and T3N0 disease. Conclusion: Analysis of this contemporary data set confirms the prognostic value of identifying at least 12 lymph nodes in resected colon cancer specimens. The variance among surgeons is of concern, but more extensive analysis is warranted before concluding that failure to identify 12 lymph nodes in colon cancer specimens is indicative of inferior quality of surgery.
No significant financial relationships to disclose.
|
