Research on surgical outcomes and risk factors of recurrence, metastasis after curative surgery for colorectal cancer

erence. Two qualitative variables are compare by using Chi-squared test (X2) - when expected frequency is higher than 5, and when the frequency is <5 Fisher's exact test is be applied. For quantitative variables, Mann - Whitney test is applied. When p < 0,05, we calculate relative risk - RR. RR is calculated according to the following formula (applicable to 2x2 table): Risk factor of recurrence Yes No Group With recurrence a c Without recurrence b d Total a + b c + d p1 = aa + b and p2 = cc + d RR = p2p1 = e Explanation of RR: Risk of recurrence of the group without recurrent factor decreases e% as compared to the group with recurrent factor. In our research, we do not use OR (odd ratio), although OR and RR do not have significant difference. However, RR is more related to risks than OR. 2.3. Research criteria: 2.3.1. Research criteria for objective 1 and 2: Features and outcomes of surgical treatment for recurrent colorectal cancer. - Features of primary tumor and features of the first surgery: Location of primary tumor, method of the first surgery, anatomical pathological features of primary tumor, disease stage, method of combination therapy. - Clinical features: Age, gender, chief complaint, clinical symptom, physical examination - Subclinical characteristics: CEA, colorectal endoscopy, ultrasound, pulmonary X-ray, thoracic - abdominal CT, abdominal - whole body MRI, PET-CT - Time of recurrence (by months): Is calculated from the first surgery to the time of detection of local recurrence or distant metastasis (equivalent to the definition of Disease-free survival). - Preoperative diagnosis and intraoperative injury. - Surgical indication and surgical method: Curative surgery - R0; Resection - R1; Palliative surgery - R2; exploratory surgery: Operation for discovery purpose. - Operation time: time from skin incision to closure, in minutes. - Intraoperative complication: Bleeding, injury to other organs while removing adhesion, or revealing tumor lesion (duodenum, small intestine, ureter, common bile duct, large blood vessels, etc.): Injury, number of injuries and treatment. - Early results: Time for intestinal circulation recovery; drainage from abdominal space: Quantity, quality, time of removing drainage (days), abdominal conditions after surgery: Normal, distension, abdominal pain, abdominal guarding; Conditions of incision: Dry, wet, bleeding, having fluid. - Postoperative complications: Bleeding; anastomotic leakage; surgical site infection; retracted stoma; electrolyte disorder; disruption of abdominal incision; residual abscess, early postoperative bowel obstruction; postoperative pancreatitis. - Death after surgery. - Length of stay. - Remote results: Evaluation of recurrence, survival or death; overall survival; rate of survival at selected points of time (6 months, 12 months, 24 months, 36 months, 48 months); postoperative survival time of each surgical method. 2.3.2. Research criteria for objective 3: Factors affecting recurrence Criteria of patients: Age, gender Criterial of tumor: Disease stage; Differentiation and grade of tumor; Petersen Index evaluating risk of recurrence (0-5 score scale); Number of metastatic lymph nodes and ratio of positive lymph nodes and number of dissected nodes; rate of positive lymph nodes; lymphatic or vascular invasion; perineural invasion; Histopathological type: Adenocarcinoma, mucinous adenocarcinoma, Signet ring cell carcinoma; mucinous organ: ³ 50% and < 50%; Tumor growth based on Borrmann’s classification; location of primary tumor. * Criteria related to surgery: Number of dissected lymph nodes and metastatic nodes: ≥ 12 lymph nodes and < 12 lymph nodes; conditions of resection margin and total mesorectal excision (TME) - Pre-operative CEA before surgery and postoperative follow-up: Adjuvant treatment: Yes/No 2.4. Data analysis , processing: Using Microsoft Excel and SPSS 22.0. Quantitative variables are analyzed to calculate the average value (Descriptives), Qualitative variables are analyzed for frequency observation (Frequency). Two qualitative variables are compare by using Chi-squared test (X2) - when expected frequency is higher than 5, and when the frequency is <5 Fisher's exact test is be applied. For quantitative variables, Mann - Whitney test is applied. - A difference between rates is considered as statistically significant difference when p < 0,05. Postoperative survival is illustrated by survival curve - using Kaplan–Meier estimator. Chapter 3: RESULTS OF RESEARCH In 2013 and 2014, there were 598 patients with colorectal cancer undergoing curative surgeries at Viet Duc University Hospital, of whom there were 53 cases of recurrence. 3.1. Features of recurrence - Average age is 56,53. There are 28 male patients, or 52,8%. Male/female ratio is 1,12. - Disease detection: During routine examination 13,2%; during examination after having symptoms 86,8%. Admission for emergency surgery due to bowel obstruction 13,2%. - Average time to recurrence is 23,1 months, 60,9% of recurrence in the first 2 years, 90,6% of patients have recurrence in the first 3 years after surgery. - Average time to recurrence of the group having adjuvant chemical treatment is 24,6 months, and that of the group not having adjuvant treatment is 21,8 months. - Average time to recurrence of each stage: stage I: 26,9 months, Stage II: 22,2 months, Stage III: 24,5 months, Stage IV: 18,4 months. - The rate of recurrence after surgery for rectal cancer of the group having repeated surgery is 58,5%, recurrence at ascending colon is 15,1%, sigmoid colon 16,9%, transverse colon 3,6%, and descending colon 5,7%. - 13,2% of the patients do not have any symptom, and disease is found during routine examination, 86,8% patients have clinical symptoms: abdominal pain (47,2%), weight loss (16,9%), blood in stool (11,3%), mucus in stool (13,2%), anal pain (13,2%), of whom 24,5% have complications caused by tumor, including 13,2% bowel obstruction, 5,7% hydronephrosis, 1,9% occlusion of the lower extremity, 3,7% biliary obstruction. - 25 patients have elevated CEA level of more than 5ng/ml, taking up 47,2%. Average value of CEA is 46,8 ng/mL. - Ultrasound finds 11 cases of liver metastasis (20,8%), chest Xray find 2 cases of lung metastasis (3,7%), 20 cases of recurrence on colon are detected by colonoscopy (37,7%), abdominal - whole body 64-Slice CT, MRI: Detected colorectal tumor in 20 cases (37,7%), Splenic metastases 1 case (1,9%), adrenal metastases, 1 case (1,9%), ovarian metastasis 2 cases (3,7%), abdominal lymph node 13 cases (24,5%), and PET CT detected 9 cases of recurrent lesion (16,9%). - Features of recurrence: may be isolated, invasive or combined with metastasis: 7,5% of isolated recurrence in colon, 15,1% in tumor bed, 26,4% of isolated recurrence in pelvis (including rectal anastomosis), 1,9% recurrence in colon with local invasion, 5,7% recurrence in pelvis with local invasion, 20,6% local recurrence with metastasis. Locations of metastasis: liver (20,8%), lung (3,8%), ovary (3,8%). 3.2. OUTCOMES OF TREATMENT FOR RECURRENT COLORECTAL CANCER - Rate of curative surgery for R0 is 71,7%, palliative surgery 26,4%, and exploratory surgery 1,9%. Rate of emergency surgery is 13,2%, and elective surgery 86,8%. There is no case of complication or death after surgery. - Surgical method depends on location and growth of tumor: Reresection of colon (40%), liver resection (9,5%), abdominoperineal resection (9,4%), ovary resection (3,7%), resection of adrenal gland (1,9%), resection of abdominal wall tumor (1,9%), extensive resection (including small intestine, spleen, diaphragm, bladder, ureter, uterus, vagina, oviduct, pelvic vessels) (13,2%). - The rate of postoperative complication is 15,1%, surgical site infection (3,7%), postoperative pancreatitis (3,7%), urine leakage (3,7%), partial intestinal obstruction (1,9%), electrolyte disorder (1,9%). Complication mainly occurs in the group undergoing curative surgery (11,3%). Mortality rate is 0%. - Median length of stay is 11,1 days. - Average postoperative survival is 17,1 months. Postoperative survival of the group undergoing curative surgery (28,89 months) is longer than that of the group undergoing non-curative surgery (10,13 months), p<0,001. Graph 1. Postoperative survival time. 3.3. FACTORS AFFECTING RECURRENCE: Table 1. Analysis of risk factors of recurrence, metastasis between the 2 groups with and without recurrence (n = 598) Features Group with recurrence Group without recurrence Value Age (years) 60,3 ± 12,80 (23-89) 55,0 ± 12,98 (19-79) p=0,008 TNM stage: I II III IV 7 21 18 7 93 247 193 6 p=0,0001 Number of dissected lymph nodes 6,7 ± 4,45 8,3 ± 5,65 p=0,081 Histopathological type Adenocarcinoma Mucinous adenocarcinoma Signet ring cell carcinoma 37 8 1 495 36 2 p=0,008 Tumor growth based on Borrmann’s classification BI/II BIII/IV 12 34 425 108 p=0,0001 RR=0,11 Mucinous organ < 50% ³ 50% 38 8 497 36 p=0,009 RR=0,39 Differentiation Well and moderately Poorly and none 40 13 481 62 p=0,009 RR=0,44 Satellite tumor Yes No 1 52 2 543 p=0,243 Petersen Index Low risk group High risk group 41 12 522 23 p<0,0001 RR=0,21 Table 2. Multivariate analysis at stage I, II between the 2 groups with and without recurrence (n = 368) Features Group with recurrence (n=28, 7,6%) Group without recurrence (n=340, 92,4%) Value Tumor growth based on Borrmann’s classification BI/II BIII/IV 9 (0,03) 19 (0,25) 284 56 p=0,0001 RR=0,12 Lymphatic invasion Yes (n=14) No (n=354) 3 (0,21) 25 (0,07) 11 329 p=0,047 RR=0,33 Number of dissected nodes 6,03 ± 4,367 (1-19) 8,2 ± 5,63 (1-41) 0.03 - Multivariate analysis of cases in state I, II shows that: Factors of tumor growth, lymphatic invasion, number of dissected lymph nodes are the factors that significantly affects recurrence. Table 3. Multivariate analysis at stage III between the 2 groups with and without recurrence (n = 211) Features Group with recurrence (n=18) Group without recurrence (n=193) Value Age (years) 51,56 ± 14,813 (17-71) 60,13 ± 12,322 (26-89) 0.018 Tumor growth based on Borrmann’s classification BI/II (n=144) BIII/IV (n=67) 3 (0,02) 15 (0,22) 141 52 p=0,0001 RR=0,09 Injury damaging visceral peritoneum - T4 Yes No 10 8 63 130 0.048 (1 side) Number of dissected nodes 7,944±4,452 8,43±5,671 0.992 Postoperative adjuvant treatment: Yes No 13 (0,07) 5 (0,38) 185 8 p=0,002 RR=0,17 - Multivariate analysis of cases in state III shows that: Factors of age, tumor growth, T4 tumor, adjuvant treatment are the variables that significantly affects recurrence. CHAPTER 4: DISCUSSIONS 4.1. Features of recurrence * Age and genders: Mean age of patients with recurrence is 56,53. Mean age of other researches: Mean age in the research of Nguyễn Tiến Sơn is 54,1 years old, Pham Thái Anh 58,1 years old, Bethesa and Sugerbaker 63 years old, and this does not differ from mean age of patient with colorectal cancer in general. There are 28 male and 23 female patients. Male/female ratio is 1,12. The gender ratio in our research is > 1, which is almost not different from other researches in Viet Nam and in the world, for example: Phạm Thái Anh 1,71, COLOR 1,1, CLASSIC 1,2. Time to recurrence Mean time to recurrence in our research is 23,11, and most of recurrence occur in the first 3 years (90,6%). According to John P. Welch: About 2/3 of recurrence occur in the first 2 years after the first surgery. Therefore we recommend patients to have routine examinations within the first 3 years to early detect recurrence and lesion for curative surgery. * Disease stage at the surgery for primary tumor: Group at state II takes up the highest percentage (39,7%), in which IIA 17%, II B 18,9%, IIC 3,8%, and III 35,8%; 7 cases of state IV cancer, taking up 13,2%. Comparing with the research on outcomes of surgery for primary colorectal cancer of Nguyen Xuan Hung, Nguyen Cuong Thinh, Nguyen Tien Son, our research shows similar results. Our research also shows that the time to recurrence of colorectal cancer is related to stage of the primary tumor at the time of the first surgery: The mean time to recurrence of the group at stage I is 26,96 months, that of stage II and III is 22,26 months and 24,54 months respectively, meanwhile that of stage IV is 18,4 months. * Disease detection: 7 cases of admission are due to detection in routine re-examination, accounting for 13,2%, and the remained cases (86,8%) are admitting to hospital due to clinical symptoms, especially 13 cases due to tumor complications, 6 cases due to bowel obstruction, 3 cases due to biliary obstruction, 2 cases due to hydronephrosis caused by recurrent tumor, 1 case due to bowel obstruction combined with hydronephrosis, 1 case due to occlusion of the lower extremity. * Symptoms of recurrence: Not clear, clinical symptoms usually occur when the tumor progresses. Occurrence of clinical symptom depends on location of tumor and extent of invasion of tumor. Common symptoms are: abdominal pain (47,2%), weight loss (16,9%), blood in stool (11,3%), mucus in stool (13,2%), anal pain (13,2%), of whom 24,5% have complications caused by tumor, including 13,2% bowel obstruction, 5,7% hydronephrosis, 1,9% occlusion of the lower extremity, 3,7% biliary obstruction. According to ASCRS and ASCO, routine examinations are compulsory for early detection of recurrent lesions. Occurrence of symptom means the tumor has progressed. * CEA test there are 53 patients, of whom 25/53 patients have CEA elevated to more than 5 ng/ml (47,2%), and 8 patients have CEA elevated to more than 100 ng/ml (15,1%). Beart R.W. and O’connell M.J. report that elevated CEA is the first indicative sign which is valuable in detecting recurrence and metastasis in 69% of cases of recurrence. The research of Chau I. on changes of CEA after surgery in 139 patients undergoing surgery for colorectal cancer shows that in 46 patients with recurrence, a rise in CEA > 1 had a predictive value of 74% for recurrence or metastases * Diagnostic imaging: Ultrasound detects 41 cases (77,4%), colonoscopy 32 cases (60,3%), abdominal CT scan 38 cases (71,7%), whole body CT scan 1 case (1,9%), pelvic MRI 2 cases (3,8%), PET CT scan 9 cases (16,9%), pulmonary X-ray (53/53 patients). * Features of recurrence of colon cancer and rectal cancer: Recurrence of rectal cancer is usually local recurrence, with rate of recurrence being higher than colon cancer, meanwhile colon cancer usually has metastatic recurrence. Total mesorectal excision (TME) and new chemoradiotherapy protocol has recently reduce rate of recurrence of rectal cancer to 6%. Rate of local recurrence after surgery for rectal cancer is 77,3%, lung metastasis 4,5%, liver metastasis 18,2%, and peritoneal metastasis 4,5%. Rate of local recurrence of colon cancer is 77,4%, lung metastasis are 3,2%, liver metastasis 22,6%, and peritoneal metastasis 16,1%. 4.2. Outcomes of treatment for recurrent colorectal cancer * Surgical indication and curative surgery: Rate of curative surgery in our research is 70,4%. Being different from recurrent gastric cancer, rate of curative surgery after surgery for colon cancer is higher. According to Trinh Hong Son: Of the 24 patients have recurrent gastric cancer, there are 3 cases of gastric re-resection, 3,2% in the research of Nguyen Ham Hoi, meanwhile the rate of curative surgery for recurrent colorectal cancer in the research of Yamada K. is 72%, Hahnloser D. is 77%, Rodel is 80%, or up to 95% in the research of Wieser. Lesions may be local or invasive to other adjacent organs or metastasis in various organs. Therefore, surgical indication for recurrent and metastatic colorectal cancer is considered when: The tumor can be dissected and the recurrent lesion causes complications such as bowel obstruction, peritonitis due to perforation, compression of other abdominal organs (kidney, ureter) or bleeding. For recurrent colorectal cancer, the issue to be considered is early detection of lesion and fully assessment of extent of lesion and possibility of surgery for tumor dissection. * Post-operative early result: Median length of postoperative stay is 11,1 days, with the group undergoing radical surgery is 11,8 days, and the group undergoing non-curative surgery and exploratory surgery is 9,3 days. Within the first 30 days after surgery, there is no case of death, and according to Gosens, the rate of mortality is 0-5% in the first month after surgery, and 8% in the third month. There are 8 cases with postoperative complication (15,1%): partial intestinal obstruction (1,9%), urine leakage (3,7%), surgical site infection (3,7%), postoperative pancreatitis (3,7%), electrolyte disorder (1,9%). After intensive treatment, all of these cases are stabilized and discharged from hospital. The rate of complication depends on method of treatment for tumor. For surgery for local cancer with invasion or for local recurrence with invasion, surgery for dissection of invaded organs is necessary for reaching cancer-related standards. Complications are anastomotic leakage, abscess after surgery or peritonitis due to anastomotic disruption. Causes of death are mainly related to infection, bleeding, multiple organ failure, and cardiovascular problems and pulmonary infarction. The method of supportive surgery or symptomatic treatment, with tumor still being left behind (not meeting oncological standards) takes up a relatively high rate in surgery for recurrent colorectal cancer, in our research, this rate is 28,3%, which is similar to that in researches of other author, being about 15-68% of cases. The rate of complication shall be low in the cases not undergoing tumor dissection surgery, but shall increase in the cases undergoing surgery for total dissection of tumor. Bleeding is the main cause and the most severe complication during operation, occurring in 0,2 - 9% of the cases, with rate of death being up to 4%, especially with surgery for removing tumor in pelvic space. However, in this research, we do not see any case of catastrophe, or bleeding complication during or after surgery. * Post-operative remote result: Of the 53 patients, we use patients’ information in their medical records to contact with 52/53 patients, and there is only 1 case (1,9%) that we cannot contact with the patient or their family so we do not have any information about their status after surgery. Mean postoperative survival time of the whole group is 17,1 months, of which the longest postoperative survival time is 50 months (currently the patient is still healthy), and the shortest is 3 months. A number of researches report the rate of 5 year survival time being 22-58% after curative surgery for R0 resection. Comparing the two groups undergoing curative surgery and non-curative surgery by Kaplan Meier graph, we see that the rate of survival after surgery of the group undergoing curative surgery is higher than that of the group undergoing non-curative surgery (mean: 24,9 months and 10,1 months, respectively, p<0,001). The research of Caricato and colleagues on prognostic factors after surgery for recurrent colorectal cancer for a total of 2204 patients (from 1960 to 2000) demonstrated a mean R0 rate of 41,2% (range 9,8 - 72%). A surgery completely removing recurrent tumor is a good prognostic factor to evaluate postoperative survival time as well as postoperative quality of life of a patient, and is the treatment goal for recurrent colorectal cancer. In order to reach that goal, early detection of recurrence by routine examination after surgery is necessary. The group having preoperative radiotherapy, combined with chemotherapy before or after surgery, has better rate of survival than that of the group undergoing surgery alone; the group undergoing palliative surgery or symptomatic treatment of R2 has outcomes of survival and quality of life being as bad as the group not undergoing surgery. As such, treatment of recurrent colorectal cancer requires a multimodal treatment protocol and good coordination of experts. 4.3. Risk factors of recurrence, metastasis: In 2013 and 2014, we followed up 598 cases of colorectal cancer patients undergoing curative surgery, the number of cases with recurrence was 53, being sorted into the group with recurrence, and 545 cases without recurrence into the group without recurrence. By using univariate and multivariate comparisons, we see that: Prognosis of recurrence comprises of various factors: Patients’ factors, treatment-related factors, and factors related to tumor pathology. * Age: Age is an independent prognostic factor for recurrence. Mean age of patients with recurrent cancer is 55,02 (17-79), which is lower than that of the group without recurrence being 60,35 (23 - 89) (p=0,008). The prognostic factors of age of patients are different between the two groups. The rate of old age patients with local recurrence or metastasis is lower. In contrast, the group of young patients who has colorectal cancer with a factor of family genetics has a very high risk of recurrence up to 80%, and the rate of 5 year survival of this group is 41%, as compared to 70% of the group of patients older than 60 years old. * Gender: Gender is an dependent prognostic factor for recurrence. A number of researches showed that male patents have worse prognosis than female patients, especially those with recurrent rectal cancer. However in our research, while comparing the 2 groups of with and without recurrence undergoing surgery in 2013 and 2014: We do not see any difference in male/female ratio between the two groups (p=0,757). * Disease stage: Disease stage by Duckes and TNM classifications is an independent and most important prognostic factor for postoperative outcomes, recurrence and 5 year survival. The rate of disease stages by TNM classification of the group with recurrence: stage I 13,2%, stage II 39,6%, stage III 34,0%, stage IV 13,2%. Meanwhile this rate of the group without recurrence is 17,3%, 45,8%, 35,8%, 1,1% respectively. Comparing using Chi-squared test, we see that disease stage between the 2 groups of with and without recurrence has a difference with p = 0,0001, which means that disease stage has an independent impact on recurrence. According to Micu B and colleagues, disease stage has huge statistically significant difference between the 2 groups of with and without recurrence. The risk of recurrence increases together with disease stage. In the research of Dziki, stage IIB and IIC have higher risks of local recurrence than stage IIIA, when tumor invades out of organ’s peritoneum into abdominal peritoneum or into surrounding organs (T4a – 4b), the possibility of cancerous cells spreading into abdominal space shall be higher. Therefore, a number of researches proves that cases of T4bN0 should be reclassified into IIIA stage (for colon cancer) and into IIIB (for rectal cancer). In contrast, cases in stage III are treated with adjuvant chemotherapy while there was no indication in stage II. Therefore, identification of high risk of recurrence together with TNM stage is necessary for determining whether a patent need postoperative adjuvant chemical treatment or not. According to Giovanni M. and Elias D., the risk of recurrence after surgery depends on extent of invasion to colon wall and lymphatic system, and according to Astler - Coller the risk of recurrent colorectal cancer by each stage of Dukes classification is 2% for stage A, 10-20% for stage B1, 20-35% for Stage B2, and up to 50% for stage C. TNM Staging System is the most widely used classification and has the highest degree of accuracy. According to American Joint Committee on Cancer (AJCC) and International Union Against Cancer (UICC), colorectal cancer has 4 stages. Adjuvant chemoradiotherapy treatment is mandatory for stage II or III rectal cancer. For cases of stage I rectal cancer there is no recommendation for adjuvant treatment, but with cases having high risks absence chemoradiotherapy shall reduce risk of recurrence and increase postoperative survival time. * Lymphatic metastasis and ratio of positive lymph nodes: Are the factors related to postoperative survival time, and as discussed in the disease stage by TNM classification, cases of lymphatic metastasis are in stage C by Dukes and stage III or higher (table 3.77). In researches on lymphatic metastasis, metastatic lymph nodes having diameters of less than 5mm can be considered as an important factors in evaluating disease stage. Cases with lymph nodes < 5 mm in diameter has a positive ratio of 50-78%. In the world, in order to accurately identify metastatic lymph nodes, pathologist often use cytokeratin to mark metastatic lymph nodes. In Viet Nam, in the recent ma