Colorectal cancer is the fourth most common malignant disease worldwide. In Korea, colorectal cancer makes up 11.2% of all the malignant disease and it is the fourth most frequent cause of cancer-related death (1). The incidence and mortality of this disease has increased in recent years (2). As a result, many topics concerned with colorectal cancer have been investigated. On the basis of those studies, it has been possible to predict the precise survival of colorectal cancer patients and to determine their prognosis.

The prognosis is related to the depth of tumor penetration into the bowel wall and the presence of both regional lymph node involvement and distant metastasis. These variables are incorporated into the various staging systems that have been recommended by Cuthbert Dukes and modified by Kirklin et al and Astler and Coller (3~5). In 1987 (6), the International Union Against Cancer (UICC) and the American Joint Committee on Cancer (AJCC) devised a worldwide taxonomy of cancer staging (TNM) based on the localized tumor characteristics, nodal involvement and the status of the metastatic implants (7). Since January 1, 2003, the sixth edition of the AJCC Cancer Staging Manual has been used to stage all the new cancer cases (8), and this has now become the standard (Table 1) (9).

However, it hasn't yet been determined whether the sixth edition AJCC staging system could be applied to Koreans, so we evaluated its value by performing retrospective analysis. The results will help us make a decision on whether the new system can be applicable to Korean clinical practice, and especially in the Youngdong district, which is the eastern coast area of Gangwondo.

Stage is the strongest predictor of survival for patients with colorectal cancer. Numerous different staging systems have historically been employed, but a single international system is required to ensure there is a common language for cancer that is understood by clinicians of all specialties. The TNM staging system has been widely used.

For the last two decades, agreement on cancer staging between the AJCC and the UICC has created a staging vocabulary that has been used to quantify the tumor burden of patients with colorectal cancer (6). The fifth edition system separated patients suffering with colon cancer into four stages, with somewhat wide spacing between the survival curves and particularly around stages II and III. The sixth edition of the AJCC Cancer Staging Manual brought about refinements of colorectal cancer staging, and these refinements were based on large data sets from the National Cancer Data Base (NCDB) (8,11); stages II and III were stratified into two and three subsets, respectively, and some of the gaps in the survival curves were filled in. Although the increased stratification of a staging system did not necessarily lead to different treatments (e.g., stage IIIa versus IIIb), it does provide important prognostic information that may help future clinical trials and improve the evidence-based literature.

Under the AJCC sixth edition staging system, it is interesting that stage IIIa shows rather better survival compared with that of stage IIb disease, which was similarity observed in our study (stage IIIa: 100% vs stage IIb: 84%). This finding may reflect the fact that adjuvant chemotherapy is recommended for stage III colon cancer patients, but not for the stage II patients (12). The three groups (stage IIIb vs. stage IIIc and stage IIIc vs. stage IV) showed statistic significance (p<.001), but the others didn't. So, we couldn't tell if the prognosis can be more precisely gauged by the revised 6th AJCC staging system. Yet we showed that the 5-year survival rates were much better for patients with stage IIIb disease (64%) than for the patients with stage IIIc disease (37%) (p<.001). Larger scale studies are needed for further improvements.

Our study has several potential limitations. First, the results are based on a small population. Moreover, the number of enrolled patients from stage IIb (n=6) to IIIa (n=5) was small, so this may have caused failure to produce statistical significance between each stage (except stage IIIb vs. stage IIIc and stage IIIc vs. stage IV, p<.001). But when compared with a previous large study, their 5-year survival rates for each stage are similar to ours except for stage IIb (72.2% vs 83%) and IIIa (83.4% vs 100%) (13). These findings may reinforce the reliability of this study. Second, the subjects were limited to only one hospital's inpatients. Although a single hospital's data was used, the data may well represent the Yongdong district. The patients who are diagnosed with cancer at the other hospitals in Yongdong have a strong tendency to visit Gangneung Asan Hospital to confirm the diagnosis and receive medical treatment afterwards (14). Third, we had difficulty in collecting the organized treatment data, including that on chemotherapy, radiotherapy and the pattern of treatment failure, which can not be overlooked. Fourth, we evaluated survival for colon and rectal cancer separately, but the results were short of statistical power due to the small study population for these two maladies.

This work is one of the trials that has evaluated the new AJCC sixth edition staging system for colorectal cancer. We observed that stage IIIb disease had better survival than stage IIIc disease. Yet we couldn't confirm that this new staging system is relevant in Korean clinical practice. Therefore, further study that's based on a larger scale (both colon and rectal cancer) and more refined cancer registry data that includes the treatment data is required to proceed, so that we can confirm the propriety of the new staging system in Korean medical practice.