Statistical Evaluation of Number of Lymph Node in Dukes C Colorectal Carcinoma

Kim, Jong Hyun; Jung, Bong Wha

Abstract

PURPOSE: Colorectal surgery, there are many important prognostic factors---depth of invasion, lymph node metastasis and distant metastasis. In there, the involvement of the lymph nodes by metastatic colorectal carcinoma may depends on several factors such as the sex, the age of the patient, tumor site and size, the symptomatic duration of the disease, tumor cell differentiation, and operating methods. In that point of view, we want to know how many lymph nodes are dissected for enough to determine statistically considerated Dukes C stage.
METHODS
We had studied 128 operation cases of colorectal adenocarcinoma from our hospital admission during the period of May, 1988 to October, 1997 to determine to provide an accurate assessment of the presence of nodal metastases. Patients status, tumor site and size, symptomatic duration of disease, tumor cell differentiation, and operating methods of 128 cases were analyzed. We calculated the probability to find at least one positive node in Dukes C, by binomial distribution from SPSS (version 7.5).
RESULTS
Eighty-three specimens (65%) were classified as Dukes B. Forty-five specimen (35%) had lymph node metastases (Dukes C) with a mean of 4.1+/-4.1 positive lymph nodes per specimen. The mean total number of lymph nodes identified per specimen was 11.6+/-7.4 (range 1~41), Dukes B was 10.9+/-7.1 (range 1~29) and Dukes C was 12.8+/-7.9 (range 3~41). Applying Students t-test to compare two independent average means, the result was the absence of significant differences in the number of nodes for the specimens defined as Dukes B and Dukes C, sex, age, symptom duration, and operationmethods whereas significant differences did exist for the specimen depending on the tumor differentiation, tumor size, and location. Poorly differentiation cancer was more prominent meaning than well or moderately differentiation in Dukes B (p<0.05). In stage Dukes C, if tumor size was below 2cm it was differences in other sizes (p<0.05). In tumor location, if tumor sited ascending colon was more prominent than sigmoid and rectum (p<0.05). According to our result, minimum 6 lymph nodes per specimen were optimal Dukes C staging assessment.
CONCLUSION
The minimum 6 lymph nodes provided an accurate assessment of the presence of nodal metastases (95 percent confidence interval) in Dukes C stage.