Lymph node count may not be best predictor of colon cancer survival

Major Finding: Lymph node count of 12 or higher was not associated with better 5-year survival rates in patients undergoing surgery to treat colon cancer.

Data Source: A secondary analysis of data from the COST trial comparing laparoscopic vs. open colectomy in 787patients with stages I-III colon cancer.

Disclosures: Dr. Mathis and colleagues stated that they had no conflicts of interest related to their findings.



Lymph node counts of 12 or higher, widely considered a key marker of surgical quality in colon cancer resection, had no significant effect on 5-year survival rates in settings where surgeons are audited and credentialed and where surgical techniques are standardized.

Lymph node (LN) count is one of several measures used to determine the extent of surgical resection and an indicator of clear surgical margins, which have been assumed to result in better survival outcomes for colon cancer patients. A new data analysis, however, of the COST (Clinical Outcomes of Surgical Therapy) trial suggests reevaluating the use of surgical surrogates such as 12 LNs and margins.

Dr. Kellie L. Mathis

The COST trial, a large, multicenter randomized trial of colon cancer procedures, compared outcomes for laparoscopic and open techniques in treating colon adenocarcinoma. The trial collected data on a number of surgical variables, including tumor location and LN count. A total of 787 patients were included: 267 with stage I disease, 284 with stage II, and 236 with stage III. Their median age was 70 years, and 50% were male.

In the current study, Dr. Kellie L. Mathis of the Mayo Clinic in Rochester, Minn., and her colleagues found that 5-year overall and disease-free survival were not influenced by LN count of above or below 12 (Ann. Surg. 2012;257:102-7 [doi:10.1097/SLA.0b013e318260a8e6]). When they adjusted for age and cancer stage, LN count was seen as not predictive of overall or disease-free survival (P = .60).

Other surgical surrogates, including total bowel length, margins, or mesenteric length, likewise did not have a significant effect on survival (P greater than .05 for all), nor did tumor location (right, left, or sigmoid), surgical technique (laparoscopic or open), and sex. Only patient age and cancer stage were found to be predictive of survival.

"On the basis of abundant literature and the acceptance of the 12 LN count as a surgical quality surrogate by National Quality Forum, most would expect the 12 LN count or other surgical variables to be predictive of survival," Dr. Mathis and her associates wrote. However, they hypothesized that procedural standardization, monitoring, and credentialing may provide a better strategy for quality control.

Overall 5-year survival results from the COST trial, they noted, were 77.2% – better than national rates for comparable patient groups in the same time period. All enrolling surgeons underwent pretrial credentialing and had performed a minimum of 20 laparoscopic colon resections, for which they had submitted operative and pathology reports. All laparoscopic resections were video recorded, and videos were randomly audited by an external review committee.

If the observations in the current study can be validated by others, Dr. Mathis and her colleagues said, "we submit that now is the time to invest in the development of technical quality control programs that directly measure and monitor surgical procedures."

Dr. Mathis and her colleagues stated that they had no conflicts of interest related to their findings.

Next Article:

   Comments ()