On the other hand, the local residual disease is responsible for local failure of treatment in rectal cancer. The sigmoid, rectosigmoid, and intraperitoneal (without gross invasion of a contiguous structure) rectal cancers have the low risk for local recurrence and treated with adjuvant chemotherapy as in colon cancers. However, rectum cancers located below peritoneal reflection, especially 12 cm or less from the anal canal are considered as true rectal cancer, and should be treated by multimodal approach, including surgery, chemotherapy and radiotherapy. The pelvis is the most common site of tumor recurrence which is a major cause of morbidity and mortality in patients with rectal cancer. The effectiveness of salvage treatment is usually limited and only offers temporary palliation[5]. The major tumor-related risk factors for local recurrence are the number of involved regional lymph nodes, the extent of transmural penetration, and the tumor grade[6]. The lateral (radial, circumferential, tangential) margin is the best predictor of local failure[7,8]. The surgeon's operative technique and ability to achieve a negative circumferential marginly total mesorectal excision are strong determinants of local control for rectal cancer. That's why the frequency of local recurrence varies among individual surgeons from less than 10% to more than 50%[9].

University hospitals mainly have well-educated and practiced physicians in Turkey as well as in other developing countries. Therefore, we retrospectively analysed the patients with stage II and III colorectal cancer treated with adjuvant chemotherapy and/or radiotherapy in our center between December 1988 and April 2000 for prognostic factors in this study. We also analysed the patients for local recurrence and survival, to delineate whether operation centers (university hospitals vs. community hospitals) had an impact on the outcome of patients with stage II and III colorectal cancer or not.

Location of tumor and surgical treatment
Rectal cancer was defined as the cancer within 12 cm from anal canal verge. The cancer beyond this border was defined as colon cancer. Therefore, sigmoid colon was included in colon group.

All patients with rectal cancer had low anterior resection or abdominoperineal resection in conjunction with extensive local and regional lymph node dissection. Patients with colon cancer had undergone segmentary colon resection and local and regional lymph node dissection. The patients treated by other surgical methods (i.e., only tumor resection, partial resection or palliative resection) as the treatment of colorectal cancer were not included in the study. All patients had tumor free surgical margins, including proximal, distal sides, and the side toward depth of invasion.

Prognostic factors, staging of the tumor and survivals
Patients were analysed for prognostic factors, considering age, gender, the location, stage and the differentiation of tumor, the number of resected and involved local and regional lymph nodes, preoperative carcinoembryonic antigen (CEA) level, operation centers and adjuvant chemotherapy schedules. The tumor size which had been previously reported not to be related to the outcome in patients with colorectal cancer[10] was not used as a prognostic factor in the analysis. The surgical resection margins, including proximal, distal sides, and the side toward depth of invasion were tumor free in all patients. CEA levels were measured by radioimmunoassay, and were considered to be high if more than 3.5 ng/mL in nonsmokers and 5 ng/mL in smokers.

TNM staging system (pathological staging) defined elsewhere was used for the tumor staging[11].

Overall survival (OS) and disease-free survival (DFS) were calculated from the date of operation to the date of death due to any cause and to the time of first evidence of disease recurrence, respectively.

Adjuvant treatments
All of one hundred fifty two patients received at least two cycles of 5-FU/Leu or 8 weeks of 5-fluorouracil and levamisole (5-FU/Lev) combinations[12,13]. The schedule of 5-FU/Leu was Leu 20 mg/m² intravenous bolus half an hour before 5-FU infusion for 5 days, and 5-FU 425 mg/ m2 intravenous bolus for 5 days. 5-FU/Leu was repeated every 28 days, and it was continued up to 6 cycles if white blood cell count was ¡İ4x10⁹/L, neutrophils were ¡İ2x10⁹/L, platelets were ¡İ100x10⁹/L, and patient had adequate renal and hepatic functions. One week delay of adjuvant chemotherapy due to hematological toxicity was allowed. The schedule of 5FU/Lev was 5-FU 450 mg/m² intravenous bolus 1-5 days, then weekly from day 29, and Lev 50 mg, per oral, three times a day for 3 days every other week. 5-FU/Lev was continued for 1 year if bone marrow, renal and hepatic functions were adequate as defined above. Hematopoietic growth factors such as granulocyte-colony stimulating factor or granulocyte/macrophage-colony stimulating factor were not used prophylactically. Patients with stage III or locally invasive rectum cancer were also given local and regional radiotherapy.

Statistical analysis
Student's t test and chi-square test were used in the comparison of parametric and non-parametric data among the groups, respectively. Kaplan-Meier curve was used to calculate the survivals, and differences were analysed using log rank test. Cox-regression analysis was used to determine independent prognostic factors. The p value less than 0.05 was considered as statistically significant. All statistics were done with the SPSS for Windows 9.05 statistics package.

Table 1: Patient characteristics

Prognostic factors and survivals
Considering the median age, gender, TNM stage, differentiation and location of the tumor, adjuvant chemotherapy regimen and relapse site, preoperative CEA level, the number of resected locoregional lymph nodes and the number of involved local and regional lymph nodes, there was no statistically significant difference between the group A and group B (Table 1).

Median follow-up interval was 44 (6-232) months. Median DFS and OS have not been reached yet for all patients. DFS and OS at 5 year were 51.8 and 52.9 percent in all patients, respectively. There was no statistically significant difference in both DFS and OS between group A and B (Table 2, Figure 1A and 1B). It was also found that the location of tumor, age, gender, the number of resected regional lymph nodes, adjuvant chemotherapy regimen (5-FU/Leu vs. 5-FU/Lev), location of tumor and preoperative CEA level had no impact on both DFS and OS (Table 2). However, OS in patients with undifferentiated tumor at 5 year was 0% but the number of patients (4 patients) with undifferentiated tumor was quite small (p=0.0032, Table 2).

Table 2: Results of univariate analysis for survivals

Fig 1(A,B): Kaplan-Meier curves according to the operation center. (A): for OS and (B): for DFS. There was not statistically significant difference in both OS and DFS between Group A and B

TNM stage of tumor was a significant predictor for both DFS and OS. OS were 72.1% and 42.6% at 5 year for patients with TNM stage II and stage III tumor, respectively (p=0.0009, Table 2, Figure 2A). Likewise, the 5 year DFS rates were 70.9% and 39.7% in patients with TNM stage II and stage III tumor, respectively (p=0.0004, Table 2, Figure 2B). Like TNM stage of tumor, the number of involved local and regional lymph nodes had also impact on DFS and OS (Table 2). The 5 year OS in patients with 0 (N0), 1-3 (N1) and ¡İ4 (N2) involved local and regional lymph nodes were 71.58%, 54.98% and 40%, respectively (Table 2, p=0.004). Again, the 5 year DFS in patients with N0, N1 and N2 disease were 69.53%, 47.05% and 30.49%, respectively (Table 2, p=0.001).

Fig 2(A,B): Kaplan-Meier curves according to TNM stage of tumor. (A): for OS and (B): for DFS. TNM stage was the only independent prognostic factor for both OS and DFS in our patients group

Recurrences of tumor (both local and distant) were observed in 63 patients (41.4%). The rate of local recurrence was 28.3% (43 patients) in all patients. The rates of local recurrence in patients with colon and those with rectum cancer were 25.9% (15 of 58 patients) and 29.8% (28 of 94 patients), respectively. Considering the relapse rates, there were no significant difference between group A and B (Table 1).

Cox-regression analysis showed the TNM stage of the tumor as the only independent prognostic factor for both DFS and OS (p=0.044). Although the patients operated at the university hospitals tend to have better OS and DFS, the difference between group A and B was not statistically significant (Table 2 and Figure 1A and 1B).

The stage of the tumor at the time of resection remains the most important prognostic factor to predict the probability of residual disease[1,2]. While stage I colorectal cancer have 90 percent cure rate by surgery alone, stage II and III disease have high risk of microscopic residual or metastatic disease[1,2]. Therefore, those patients should be considered for adjuvant treatment. Five year DFS rates in stage II and stage III colorectal cancer is 78-90% and 20-50%, respectively, with surgical treatment alone. Although the efficacy of adjuvant chemotherapy in stage II colon cancer is controversial, it significantly improves the outcome in patients with stage III colorectal cancer[3,4].

Besides the tumor stage, many factors such as the presence of obstruction or perforation, vascular or lymphatic invasion (or both), per neural invasion, peritumoral lymphocytic infiltration, the character of invasive margin and tumor type, presence and the number of mast cells, age and gender, tumor grade, DNA content, increased mitosis, low Bcl-2 expression, low apoptosis rate, vascular endothelial growth factor levels and allelic loss of chromosome 18q have been studied as a prognostic factor[14,15].

In the present study, the TNM stage of tumor was found to be an independent prognostic factor which is consistent with the previous studies. However, the results of our study were somewhat contradictory to the literature. The outcome of the patients younger than 40 years were reported to be poor in the literature. Also, females were found to have a survival advantage when compared to males in early studies, but this association was not confirmed by other studies[6,15]. Interestingly, we found that patients younger than 45 years had better both DFS and OS, and although there was no statistically significant difference in OS between males and females, males had a borderline DFS advantage over females (p=0.054). It is likely that the retrospective nature of the study may be partially responsible for these results. Other explanation for the poor outcome in patients older than 45 years may be the insufficient treatment in this group of patients. The shorter DFS in patients older than 45 years may support this explanation. Females are less likely to undergo screening studies than males[15]. Therefore female patients may have more advanced disease than male patients.

It has been found that the tumor arising at or below the peritoneal reflection (rectosigmoid and rectum) have worse prognosis than those arising above the reflection. With regard to colon primary tumors, the prognostic value of primary tumor location is inconclusive[14,15]. We found that patients with colon cancer tend to have better outcomes than those with rectal cancer but the difference was quite small and was not statistically significant which may be related to small population size.

The degree of tumor differentiation is confirmed as a prognostic factor in several studies[16]. The patient with well or moderately differentiated tumor had a survival advantage when compared to those with undifferentiated tumor, but the number of undifferentiated tumor was quite small to make a firm conclusion in the present study.

Although the number of lymph nodes present in any given operation specimen may be limited by anatomic variation, surgical technique and pathological examination, it has been shown that a minimum 12-15 regional lymph nodes should be examined pathologically to accurately predict regional lymph node negativity[17,18]. We have also examined the impact of the number of resected regional lymph nodes on the outcome. There was no survival difference between the patients who had equal or less than ten and those who had more than ten resected regional lymph nodes. However, the number of involved local and regional lymph nodes had an impact on the outcome as in the TNM stage of tumor. The patients with N0 disease had significantly survival advantage over those with N1 and N2 disease. These findings are compatible with the knowledge of literature. No survival difference according to the adjuvant chemotherapy schedules (six months 5-FU/Leu, Mayo Clinic Regimen vs. 52 weeks 5-FU/Lev) was noticed.

Although CEA is the most reliable indicator for tumor recurrence in colorectal cancer after curative resection, the prognostic potential of preoperative CEA level is still unclear[15]. It is possible that an elevated CEA level may be the reflection of a more advanced colorectal cancer. In our study, it was found that preoperative CEA level was not a prognostic factor in both univariate and multivariate analyses. However, it was not measured preoperatively in the majority of patients in the present study and its level was elevated in only thirty three patients.

The distant metastases and local failure are important issues for the treatment failure in patients with colorectal cancer. The local failure is more frequent in patients with rectal cancer than in those with colon cancer. The inadequate surgical resection is a major cause of pelvic recurrence[9]. Therefore, the surgeon's ability to achieve a negative surgical margin is a strong determinant for local control. The early studies have shown that patients treated with abdominoperineal resection as a primary surgical treatment had a higher recurrence rate than those treated with low anterior resection, but later studies did not confirm these results[19,20]. Nowadays, total mesorectal excision in conjunction with low anterior resection or abdominoperineal resection is recommended as the optimal surgical treatment for rectal cancer[21]. The local recurrence rate in patients treated with total mesorectal excision ranges 3.5-13 percent in various studies, according to the locoregional lymph node positivity[21-23]. However, the local recurrence rate from one surgeon to another ranges from less than 10 percent to more than 50 percent[9]. It has been shown that removing rectal cancer surgery from routine surgical teaching and concentrating training in total mesorectal excision among specialized surgeons could drop the local recurrence rate to 7 percent when compared to historic control with a local recurrence rate of 23 percent[24]. In addition, the local recurrence rate and survival advantage in patients with rectal cancer is associated with surgeon colorectal surgery fellowship training[24]. The hospital size, hospital type (university vs. community), and experience of the surgeon are also important factors for improving survival and local recurrence rate in rectal cancer patients[25]. We found that the operation center had no impact on the outcomes in our patients group, though the patients operated at university hospitals had slightly better survival than those operated at other hospitals (i.e., community hospitals). Though there were no significant differences in patients' characteristics, in relapse rate and in other prognostic factors, local recurrence rate in our patients group was higher than those reported in the literature. Also, there was no significant difference in the rate of local relapse in patients with rectal cancer between group A and B. The colorectal surgery has been done as a routine surgical procedure in the majority of the hospitals, including many university hospitals in Turkey. It would be better that the individual surgeon's characteristics were analysed to delineate the impact of surgeon on the outcome of patients with operable colorectal cancer. Due to retrospective nature of our study, to clarify the surgeon's characteristics was very difficult. Therefore, we are planning a prospective study with a larger sample size to delineate the role of surgeon in the treatment of colorectal cancer.

In conclusion, TNM stage of the disease was the only independent prognostic factor in the current study. Operation at university hospitals had no impact on the outcomes in patients with stage II and stage III colorectal cancer, although patients operated university hospitals had slightly better outcome than those operated at community or other hospitals. Although our study does not represent the whole Turkish population, and it has relatively small sample size, it is informative. Prospective studies are necessary to clarify this issue in Turkey and World.

1) Willet CG, Tepper JE, Cohen AM, et al. Failure patterns following curative resection of colonic carcinoma. Ann Surg 1984;200:685-90.

2) Minsky BD, Mies C, Rich TA, et al. Potentially curative surgery of colon cancer: patterns of failure and survival. J Clin Oncol 1988;6:106-18.

3) Andre T, Boni C, Mounedji-Boudiaf L, et al. Oxaliplatin, fluorouracil, and leucovorin as adjuvant treatment for colon cancer. N Engl J Med 2004; 350: 2343-51.

4) Efficacy of adjuvant fluorouracil and folinic acid in B2 colon cancer. International Multicentre Pooled Analysis of B2 Colon Cancer Trials (IMPACT B2) Investigators. J Clin Oncol 1999;17:1356-63.

5) Tschmelitsch J, Kronberger P, Glaser K, et al. Survival after surgical treatment of recurrent carcinoma of the rectum. J Am Coll Surg 1994;179:54-8.

6) Fielding LP, Phillips RK, Fry JS, Hittinger R. Prediction of outcome after curative resection for large bowel cancer. Lancet 1986;2:904-7.

7) Adam IJ, Mohamdee MO, Martin IG, et al. Role of circumferential involvement in the local recurrence of rectal cancer. Lancet 1994;344:707-11.

8) Wibe A, Rendedal PR, Svensson E, et al. Prognostic significance of the circumferential resection margin following total mesorectal excision for rectal cancer. Br J Surg 2002;89:327-34.

9) Heald RJ. Rectal cancer: the surgical options. Eur J Cancer 1995;31A:1189-92.

10) Compton CC, Fenoglio-Preiser CM, Pettigrew N, Fielding LP. American Joint Committe on cancer Prognostic Factors Consensus Conference: Colorectal Working Group. Cancer 2000;88:1739-57.

11) American Joint Committee on Cancer: Colon and Rectum. In: AJCC Cancer Staging Manual, Fleming ID, Cooper JS, Henson DE, et al, editors. Philadelphia: Lippincott Williams &Wilkins;, 1998;81-8.

12) O'Connell MJ. A phase III trial of 5-fluorouracil and leucovorin in the treatment of advanced colorectal cancer. A Mayo Clinic/North Central Cancer Treatment Group Study. Cancer 1989;63:1026-30.

13) Laurie JA, Moertel CG, Fleming TR, et al. Surgical adjuvant therapy of large bowel carcinoma: an evaluation of levamisole and the combination of levamisole and fluorouracil. The North Central Cancer Treatment Group and the Mayo Clinic. J Clin Oncol 1989;7:1447-56.

14) Compton CC. The pathology report in colorectal cancer: a user's guide. In: American Society of Clinical Oncology 2003 Educational Book. Perry MC(Ed). Alexandria: American Society of Clinical Oncology, 2003;502-16.

15) Skibber JM, Minsky BD, Hoff PM: Cancer of colon. In: Cancer: principles and practice of oncology. DeVita VT Jr, Hellman S, Rosenberg SA, editors. Philadelphia: Lippincott-Williams and Wilkins, 2001;1216-70.

16) Griffin MR, Bergstralh EJ, Coffey RJ, et al. Predictors of survival after curative resection of carcinoma of the colon and rectum. Cancer 1987;60:2318-24.

17) Tepper JE, O'Connell MJ, Niedzwiecki D, et al. Impact of number of nodes retrieved on outcome in patients with rectal cancer. J Clin Oncol 2001;19:157-63

18) Scott KW, Grace RH. Detection of lymph node metastases in colorectal carcinoma before and after fat clearence. Br J Surg 1989;76:1165-7.

19) Prolongation of the disease-free interval in surgically treated rectal carcinoma. Gastrointestinal Tumor Study Group. N Engl J Med 1985;312:1465-72.

20) Zaheer S, Pemberton JH, Farouk R, et al. Surgical treatment of adenocarcinoma of the rectum. Ann Surg 1998;227:800-11.

21) MacFarlane JK, Ryall RD, Heald RJ. Mesorectal excision for rectal cancer. Lancet 1993;341:457-60.

22) Enker WE, Thaler HT, Cranor ML, Polyak T. Total mesorectal excision in the operative treatment of carcinoma of the rectum. J Am Coll Surg 1995;181:335-46.

23) Dahlberg M, Glimelius B, Pahlman L. Changing strategy for rectal cancer is associated with improved outcome. Br J Surg 1999;86:379-84.

24) Skibber JM, Hoff PM, Minsky BD. Cancer of the rectum. In: Cancer: principles and practice of oncology DeVita VT Jr, Hellman S, Rosenberg SA, editors. Philadelphia: Lippincott-Williams and Wilkins, 2001;1271-319.

25) Simons AJ, Ker R, Groshen S, et al. Variations in the treatment of rectal cancer: the influence of hospital type and caseload. Dis Colon Rectum 1997;40:641-6.