The survival rates of patients with colorectal cancers have been well documented in many studies. Some studies have shown that proximal colon cancers have inferior survival rates when compared with distal colon cancers. However, the prognostic significance of tumor location with respect to survival remains controversial. By using data from a single physician, we analysed patient survival rates based on colon cancer subsite location, including rectal cancers.
We retrospectively analysed 881 patients with colorectal cancers between 1987 and 2008. Colon subsite locations were defined as cecum, ascending colon, transverse colon, descending colon, sigmoid colon, and rectum. Subsite-specific survival analyses were performed using Kaplan-Meier analysis and Cox proportional hazards ratios. The median follow-up time was 93 months.
A total of 689 colorectal cancer cases were included in our analysis, of which 14 were cecum (2.0%), 95 were ascending colon (13.8%), 21 were transverse colon (3.0%), 25 were descending colon (3.6%), 129 were sigmoid colon (18.7%), and 405 were rectum (58.8%) cancers. The 5-year overall survival rates were 77.8% for all colorectal cancers, which consisted of 92.9% for cecal cancer, 69.5% for ascending colon cancer, 76.2% for transverse colon cancer, 84.0% for descending colon cancer, 82.2% for sigmoid colon cancer, and 77.5% for rectal cancer.
Ascending colon cancer was associated with the poorest survival outcome, whereas descending colon cancer was associated with the best survival outcome except cecal cancer. Moreover, the survival rate associated with left colon cancer was better than the survival for right colon and rectal cancer.
Colorectal cancer (CRC) is the third most common cancer in South Korea [
From January 1987 to December 2008, we retrospectively analysed 689 among 881 patients diagnosed with CRC and treated by a single surgeon. Patients who underwent palliative surgery, underwent surgery at another hospital, died within 30 days of surgery, or experienced hereditary colon cancers, such as familial adenomatous polyposis and hereditary nonpolyposis CRC, were excluded from this study. Variables for clinical characteristics included age, gender, anatomic tumor location, preoperative serum carcinoembryonic antigen (CEA) levels, tumor node and metastasis (TNM) stage, and lymph node ratio (LNR). Written informed consent was obtained from all patients before the operation. After Institutional Review Board approval of our hospital, all records and medical charts of patients were reviewed.
Subsite location of the tumors was defined as follows: cecum, ascending colon (includes ascending colon and hepatic flexure), transverse colon, descending colon (includes splenic flexure and descending colon), sigmoid colon (includes sigmoid colon and rectosigmoid junction), and rectum. Moreover, right or proximal colon included cancers of the cecum, ascending colon, hepatic flexure, and transverse colon. The left or distal colon included cancers of the splenic flexure, descending colon, sigmoid colon, and the rectosigmoid junction.
The chi-square test was used for comparing risk factor distributions between subsite groups. Overall survival rates were analysed using the Kaplan–Meier method, and statistical significance of the survival rates was evaluated using the log-rank test. Multivariate analysis was used to assess all of the clinicopathological factors associated with overall survival using Cox proportional hazards regression (i.e., age, sex, CRC subsite location, TNM stage, LNR, and preoperative serum CEA levels). A P-value of <0.05 was considered statistically significant. Statistical analysis was performed using MedCalc Statistical Software ver. 14.10.2 (MedCalc Software bvba, Ostend, Belgium).
A total of 689 CRC patients were included in the analysis with a median follow-up of 93 months (range, 1–193 months). Of these patient tumors, 14 tumors were located in the cecum, 95 in the ascending colon, 21 in the transverse colon, 25 in the descending colon, 129 in sigmoid colon, and 405 in the rectum. The TNM stage of the study population was I in 47 patients, II in 326 patients, III in 236 patients, and IV in 79 patients. The number of patients with tumors located at each subsite is listed in
Overall survival rates according to cancer subsite location are shown in
Survival analysis of stage I and II cancers showed no difference based on tumor location. However, stage III tumors showed statistically significant differences based on the location of the malignancy (P<0.05). Furthermore, stage III descending colon cancers were associated with a better survival rate than stage III tumors at any other location except cecum due to its rarity, whereas stage III rectal cancers were associated with the worst overall survival rate of all stage III tumor locations. In patients with stage IV tumors, no differences in the survival rates for any tumor location were observed. Additionally, right colon tumors were associated with worse survival when compared with left colon tumors (hazard ratio [HR], 0.31; 95% confidence interval [CI], 0.13–0.72) and rectal tumors (HR, 0.75; 95% CI, 0.35–1.61) (P<0.05) when analysing stage III tumors, but not when analysing stage I, II or IV tumors (
Due to their similar characteristics, colon and rectal cancers are often defined together as CRCs. Well-known prognostic factors for CRCs include age, sex, T stage, N stage, preoperative serum CEA level, tumor differentiation, lympho-vascular invasion, and perineural invasion [
Studies have suggested that the anatomical site of the tumor is an important factor in the management of CRCs. However, site-specific CRC survival data remain controversial. Many studies have classified colon cancers as right or left, of which right colon cancers are associated with poorer prognosis [
In our study, proximal colon cancer patients experienced poorer survival rates than distal colon cancer patients. Ascending colon cancer was associated with the worst survival of all locations analysed. Moreover, descending colon cancer was associated with the best survival rate out of all cancer analysed except cecal cancer, which was excluded due to the small sample size. Nawa et al. [
Several limitations are associated with this study. Some cancer locations, including the cecum, the transverse colon and the descending colon, were represented by very few patients. Moreover, accurate grouping of junctional cancers, such as hepatic, splenic flexure and rectosigmoid cancers, could have influenced our results. Finally, biomolecular analysis of the tumors for the mutational and expression level statuses of genes such as KRAS, BRAF, p53, and microsatellite instability could not be obtained because a large proportion of the data were obtained too long before the analysis.
Among all CRCs, patients with ascending colon cancer showed the poorest survival outcome. Descending colon cancer was associated with the best overall survival in CRC patients to the exclusion of cecal cancer. Moreover, left colon cancers were associated with better survival than right colon cancers and rectal cancers. Therefore, our results suggest that more careful approaches are needed for CRC treatment depending on the localization of the tumor.
No potential conflict of interest relevant to this article was reported.
This study was supported by a clinical research grant from the Pusan National University Hospital in 2014.
Five years survival rates according to the subsite of colorectal cancer.
Five years survival rates after right and left grouping colon and rectal cancer.
Baseline characteristics of the patients with colorectal cancer
Characteristic | Cecum | Ascending | Transverse | Descending | Sigmoid | Rectum | P-value |
---|---|---|---|---|---|---|---|
No. | 14 | 95 | 21 | 25 | 129 | 405 | - |
Age (yr) | 0.083 | ||||||
< 65 | 9 | 49 | 13 | 11 | 82 | 252 | |
≥ 65 | 5 | 46 | 8 | 14 | 47 | 153 | |
Sex | 0.122 | ||||||
Male | 9 | 46 | 11 | 9 | 58 | 173 | |
Female | 5 | 49 | 10 | 16 | 71 | 232 | |
TNM | 0.019 | ||||||
I | 0 | 3 | 2 | 0 | 10 | 32 | |
II | 7 | 44 | 12 | 13 | 54 | 196 | |
III | 5 | 25 | 4 | 10 | 52 | 141 | |
IV | 2 | 23 | 3 | 2 | 13 | 36 | |
CEA | 0.062 | ||||||
< 5 ng/mL | 9 | 49 | 10 | 13 | 68 | 251 | |
≥ 5 ng/mL | 5 | 46 | 11 | 12 | 61 | 154 | |
LNR | 0.850 | ||||||
0 | 7 | 48 | 14 | 15 | 66 | 234 | |
< 0.125 | 4 | 25 | 3 | 6 | 31 | 77 | |
< 0.25 | 1 | 7 | 2 | 3 | 15 | 46 | |
≥ 0.25 | 2 | 15 | 2 | 1 | 17 | 48 |
TNM, tumor node and metastasis; CEA, carcinoembryonic antigen; LNR, lymph node ratio.
Five years survival rate according to colonic subsite and rectal cancer with prognostic factors
Factor | Cecum | Ascending | Transverse | Descending | Sigmoid | Rectum | P-value |
---|---|---|---|---|---|---|---|
5-YSR | 92.9 | 69.5 | 76.2 | 84.0 | 82.2 | 77.5 | 0.012 |
Age (yr) | |||||||
< 65 | 88.9 | 79.6 | 92.3 | 72.7 | 81.7 | 81.3 | 0.66 |
≥ 65 | 100 | 58.7 | 50.0 | 92.9 | 83.0 | 71.9 | < 0.001 |
Sex | |||||||
Male | 88.9 | 73.9 | 81.8 | 88.9 | 79.3 | 79.2 | < 0.001 |
Female | 100 | 65.3 | 70.0 | 81.3 | 84.5 | 76.3 | 0.46 |
TNM | |||||||
I | - | 100 | 100 | - | 100 | 93.8 | 0.953 |
II | 100 | 88.6 | 83.3 | 92.3 | 92.3 | 88.8 | 0.586 |
III | 100 | 76.0 | 75.0 | 88.9 | 90.0 | 72.3 | 0.030 |
IV | 50 | 21.7 | 33.3 | 33.3 | 15.4 | 22.2 | 0.393 |
CEA | |||||||
< 5 ng/dL | 88.9 | 83.7 | 90.0 | 92.3 | 94.1 | 84.1 | 0.356 |
≥ 5 ng/dL | 100 | 54.3 | 63.6 | 75.0 | 68.9 | 66.9 | 0.026 |
LNR | |||||||
0 | 100 | 89.6 | 85.7 | 80.0 | 90.9 | 87.2 | 0.729 |
< 0.125 | 75.0 | 72.0 | 100 | 83.3 | 87.1 | 81.8 | 0.650 |
< 0.25 | 100 | 14.3 | 0 | 100 | 86.7 | 60.9 | < 0.001 |
≥ 0.25 | 100 | 26.7 | 50.0 | 100 | 35.3 | 39.6 | 0.041 |
5-YSR, 5-year survival rate; TNM, tumor node and metastasis; CEA, carcinoembryonic antigen; LNR, Lymph node ratio.
Five years survival rate after right and left grouping colon and rectal cancer
Factor | Right colon | Left colon | Rectum | P-value |
---|---|---|---|---|
5-YSR | 73.1 | 82.5 | 77.5 | 0.02 |
Age (yr) | ||||
< 65 | 83.1 | 80.6 | 81.3 | 0.859 |
≥ 65 | 61.0 | 85.2 | 71.9 | 0.002 |
Sex | ||||
Male | 77.3 | 80.6 | 79.2 | 0.011 |
Female | 68.8 | 83.9 | 76.3 | 0.582 |
TNM | ||||
I | 100 | 100 | 93.8 | 0.621 |
II | 88.9 | 91.0 | 88.8 | 0.557 |
III | 79.4 | 87.1 | 72.3 | 0.040 |
IV | 25.0 | 13.3 | 22.2 | 0.766 |
CEA | ||||
< 5 ng/dL | 85.3 | 93.8 | 84.1 | 0.100 |
≥ 5 ng/dL | 59.7 | 69.9 | 66.9 | 0.065 |
LNR | ||||
0 | 89.90 | 88.9 | 87.2 | 0.869 |
< 0.125 | 75.0 | 86.5 | 81.8 | 0.302 |
< 0.25 | 20.0 | 88.9 | 60.9 | < 0.001 |
≥ 0.25 | 36.8 | 38.9 | 39.6 | 0.228 |
5-YSR, 5-year survival rate; TNM, tumor node and metastasis; CEA, carcinoembryonic antigen; LNR, lymph node ratio.
Factors associated with survival
Factor | Hazard ratio on univariable analysis (95% CI) | P-value | Hazard ratio on multivariable analysis (95% CI) | P-value |
---|---|---|---|---|
Age | < 0.001 | < 0.001 | ||
< 65 | 1.00 | 1.00 | ||
≥ 65 | 1.72 (1.28–2.32) | 1.93 (1.43–2.60) | ||
Sex | 0.299 | - | ||
Male | 1.00 | - | ||
Female | 1.17 (0.87–1.56) | - | ||
TNM stage | < 0.001 | - | ||
I | 1.00 | 1.00 | ||
II | 3.53 (2.01–6.20) | - | ||
III | 7.73 (4.32–13.83) | 8.84 (1.81–43.16) | 0.002 | |
IV | 40.20 (18.09–89.35) | 30.67 (6.50–144.73) | < 0.001 | |
CEA | < 0.001 | < 0.001 | ||
< 5 ng/dL | 1.00 | 1.00 | ||
≥ 5 ng/dL | 2.78 (2.06–3.74) | 1.77 (1.29–2.43) | ||
LNR | < 0.001 | |||
0 | 1.00 | 1.00 | ||
< 0.125 | 1.62 (1.13–2.32) | - | ||
< 0.25 | 3.47 (2.09–5.78) | - | ||
≥ 0.25 | 7.01 (4.09–11.99) | 11.55 (1.44–92.59) | < 0.001 | |
Location | 0.012 | |||
Cecum | 1.00 | |||
Ascending | 6.94 (2.48–19.39) | 1.00 | ||
Transverse | 6.28 (1.79–22.09) | - | ||
Descending | 3.13 (0.94–10.41) | - | ||
Sigmoid | 3.23 (1.19–8.76) | 0.49 (0.29–0.82) | 0.006 | |
Rectum | 4.27 (1.63–11.14) | - | ||
0.020 | ||||
Right colon | 1.00 | 1.00 | ||
Left colon | 0.53 (0.34–0.84) | 0.60 (0.41–0.88) | 0.032 | |
Rectum | 0.71 (0.48–1.05) | - | - |
CI, confidence interval; TNM, tumor node and metastasis; CEA, carcinoembryonic antigen; LNR, lymph node ratio.