Background
Colorectal cancer has gradually become one of the most significant leading causes of death from malignancies world-wide, especially in China. Surgical management is still the mainstay of the treatment [
1,
2]. Conventional open surgery is reported with significant morbidity and a long recovery period. With the laparoscopic techniques applied to the surgical field for colorectal diseases, laparoscopic colorectal surgery was first performed in Japan in 1992, very soon after its initial description by Jacobs et al [
3,
4]. In 1993, the first laparoscopic colectomy was successfully performed in China. Since then, laparoscopic surgery has been widely performed for various benign colorectal diseases [
5‐
8], and furthermore, colorectal cancer.
Laparoscopic colorectal surgery is technically complicated as it involves almost all advanced laparoscopic techniques, such as mobilization, intracorporeal division, dissection of major vessels, and anastomosis. There is a steep learning curve to achieve advanced laparoscopic skills. But when the learning phase has been conquered, the benefits of laparoscopic surgery have been suggested with respect to decreased morbidity, decreased pain, faster recovery, shorter hospital stay and possibly reduced immunosuppression, comparing with open surgery [
9‐
12].
However, behind the great success of laparoscopic colorectal surgery, there are still many questions that remain unclear, including whether laparoscopic colorectal cancer surgery is radical or not, seldom reported superior short-term outcomes. Laparoscopic colorectal surgery is still not considered standard treatment [
13]. There are also controversies with potential port site recurrence after curative resection of tumor, not to mention too much economical costs.
In this article, we investigated the short-term and medium-term clinical outcomes of laparoscopic surgery versus open surgery for colorectal cancer over a period of 5 years in our center, aiming at investigating whether the laparoscopic surgery has any advantages for the patients with colorectal cancer or not.
Methods
Patient selection
Between January 2006 and June 2009, patients who underwent radical colorectal surgery for colorectal cancer in Affiliated First People's Hospital, Shanghai Jiao Tong University were consecutively enrolled in this study. Both open surgeries (OP) and laparoscopic-assisted surgeries (LAP) were performed by two stable surgical teams, respectively. No selection criteria were used to allocate patients to either a laparoscopic or an open operation. Patients were assigned to each surgical team (open or laparoscopic) according to their target dates for treatment and operating theatre availability. A minority of patients who wished to be operated laparoscopically were accommodated whenever possible. The trial received approval from local research ethics committee, and written informed consent was obtained from all patients before the investigation.
All patients enrolled accepted preoperative laboratory examination including tumor markers screening, coagulation test, chest x-ray, abdominal ultrasound, colonoscopy and if necessary, CT scan of the abdomen and pelvis. Endoscopic applications of metal clips for tumor localization, as well as intra-operative colonoscopic orientations were performed selectively. All patients were confirmed to have a malignant tumor after postoperative pathologic examination. Postoperative clinical staging after pathologic examination was based on the UICC cancer staging manual (2007). None of the patients had accepted preoperative radiotherapy or chemotherapy; out of the patients who were pathologically diagnosed as stage III, all accepted adjuvant chemotherapy with oxaliplatin and 5-fluorouracil for 6 months postoperatively. Exclusion criteria were: in situ or metastatic disease, emergency presentation, morbid obesity (defined as body mass index, i.e. BMI > 35 kg/m2), a classification V physical status according to the American Society of Anesthesiologists (ASA), associated gastrointestinal disease that required extensive operative evaluation or intervention, pregnancy or malignant disease in the past 5 years (except superficial squamous or basal cell carcinoma of the skin or in situ cervical cancer).
Preoperative preparations and operation procedures
All patients had oral administration of gentamicin and metronidazole, 3 times a day for 3 days before surgery. Polyethylene glycol-electrolyte solution or magnesium sulfate was given one day before the surgery for bowel preparation. Other preoperative preparations were standardized, as followed for traditional abdominal surgeries.
Laparoscopic-assisted resection involved mobilization of the colon, visualization of critical structures, and intracorporeal vascular ligation. A standard total mesorectal excision (TME) procedure was essential for rectal cancer resection. A small abdominal incision was required to remove the specimen. Anastomosis was performed either through the small incision for right-hemi, transverse and left-hemi colon cancer, or laparoscopically with a double-stapling technique, for sigmoid and rectal cancer. If the tumor was located so close to the dentation line that anal-saving could not guarantee the radical standards and operation safety, the Hartmann & Bacon procedures were performed. In the majority of cases, the operation was performed utilizing a lateral to medial approach. In this study, an incision longer or different to that planned was used to determine a conversion. Conversion to open colectomy was at the discretion of the surgeon based on concerns regarding patient safety, technical difficulties, or associated unexpected conditions requiring treatment by laparotomy. Conversions were recorded and analyzed as part of the laparoscopic arm of the study, but were excluded for further analysis. Open procedures were performed according to the standard techniques followed by the operating surgeon. All operations achieved a standard D2 lymph node dissection according to the Guidelines of Radical Laparoscopic Colorectal Cancer Surgery (2006, 2008) established by the Study Group of Laparoscopic and Endoscopic Surgery Affiliated to Chinese Medical Association, as well as the General Rules for Clinical and Pathological Studies on Cancer of the Colon, Rectum and Anus, 5th edition (1994), by the Japanese Research Society for Cancer of the Colon and Rectum.
Perioperative surveillance, postoperative managements and follow-up evaluation
Demographic and operative data were obtained regarding age, gender, BMI, ASA score, comorbidities, history of previous abdominal surgery, tumor location, surgical intervention, operative time, blood loss, maximum incision length, sample length, proximal and distal margin length, number of retrieved lymph nodes and lymph node metastases, tumor size, pathological differentiation and clinical stage. Postoperative data included analgesic usage, Visual Analog Scales (VAS) score, peristalsis recovery time, time until flatus, time until off-bed, time until first liquid and semi-liquid intake, postoperative duration of hospital stay and total time of hospital stay, were recorded.
Patients enrolled in the present study were managed postoperatively by the same group of surgeons. Patients in both groups were supported by infusions in the very first several hours after surgery. After confirmation of the peristalsis recovery, liquid diet was supplied. Semi-liquid diet was considered suitable for patients after report of flatus. For pain control, patients were given patient-controlled anesthesia (PCA) or short-acting drugs according to their own aspirations. Prophylactic antibiotics were used during 72 hours after surgery; however, if there was any indication of infection, this time was prolonged. The catheter was removed as early as possible except for patients with tumors located in the lower region of the rectum. The peritoneal drain was removed on POD7, only if no leakage or hemorrhage was confirmed, as well as the patient had taken semi-liquid food and had reported a formed stool. In patients with postoperative complications, the management was almost the same in both treatments groups, respectively. All patients were followed-up after being discharged from the hospital, according to a pre-established protocol. This included recording of medical history, physical examination, and laboratory studies such as, serum carcinoembryonic antigen (CEA) and carbohydrate antigen 19-9 (CE19-9) levels, which were assessed 1 month after surgery and every 3 months thereafter. At each patient visit, symptoms were recorded and wound scars examined for subcutaneous metastasis. Either ultrasonography or CT scan of the abdomen, in addition to chest X-ray was performed every 6 months whereas total colonoscopy was performed every year. When colonoscopy was incomplete, a combination of sigmoidoscopy and barium enema methods was undertaken. Recurrences were histologically confirmed and classified as distant metastasis, locoregional relapse (tumor growth restricted to the anastomosis or the region of primary operation), and incision/port-site metastasis. The last date for follow-ups was March, 2011.
Statistical analysis
Data were collected prospectively using a computerized data base according to pre-study Power calculation. Quantitative data was given as a mean ± standard deviation, and analyzed using Student's t-test. Count data between LAP and OP groups was assessed by Mann-Whitney, Chi-Square or Fisher's exact test as appropriate. Comparisons between the two groups were made on an intention-to-treat basis; thus, patients in the LAP group converted to the open procedure were not excluded from the analysis. Time to: (1) last follow-up evaluation, (2) treatment failure or (3) death was measured from the date of operation.
Recurrence and overall survivals were evaluated using the Kaplan-Meier method and compared with the log-rank test. Analysis of predictive factors of survival was performed. Variables analyzed univariately were, age, gender, BMI, ASA scores, preoperative comorbidities, tumor location, intervention, surgical procedures, tumor size, pT, pN, pathological differentiation, lymph nodes metastasis, clinical stage, and postoperative complications. Variables associated with recurrence or survival, were then used for multivariate analysis using a stepwise Cox proportional-hazards regression model. Statistical significance was defined as P < 0.05. All calculations were performed by using the SPSS software package version 12.0 (SPSS Inc., Chicago, IL).
Discussion
We report the Chinese experience of performing laparoscopic-assisted colectomy for colorectal cancer. We found that LAP for colorectal cancer is feasible and safe; patients had acceptable rates of complications and conversion to open laparotomy, as well as reasonably short postoperative durations of stay, a large number of lymph node retrieval, and finally, similar survival rates.
With the development of laparoscopic techniques, along with the improvement of laparoscopic instruments, a standard laparoscopic procedure for colorectal cancer surgery has gradually become widely accepted, and a radical cure resection seems feasible for laparoscopic surgeries. The present study showed that there were no differences in the outcomes between the two treatment groups. Apart from acquirement of a new skill the laparoscopic surgeon being a factor, which cannot be totally ignored, this was a straightforward comparison. Furthermore, there was no apparent deterioration in the quality of surgery associated with the introduction of laparoscopic resection, as stoma formation and APR rates in rectal cancer remained unchanged over time. The intraoperative comparison between the two groups in our study indicated almost similar operative time and complications, which was not in keeping with other randomized controlled studies [
14‐
20]. The mean operating time for the laparoscopic-assisted procedure was shorter in this study than in the Multicentre Randomized Controlled trial - Conventional versus Laparoscopic-Assisted Surgery In patients with Colorectal cancer (MRC CLASICC) [
14] trial but similar to the Colon Carcinoma Laparoscopic or Open Resection (COLOR) [
15,
17] trial.
The number of lymph nodes harvested during the surgical procedure influences clinical staging of the tumor and is not only influenced by the operative technique or the extent of lymphadenectomy, but to an even greater extent by pathological techniques involved in processing the specimens. Examining fewer than 12 lymph nodes in a specimen can result in under-staging [
21]. Since the specimens retrieved by either laparoscopic-assisted or open resection were processed in different ways, it has been difficult to compare the harvested lymph nodes in different studies. Nevertheless, since the standard D2 lymph node dissection was consistently followed for all operations, and the lymph node was always collected by a permanent surgeon and a permanent pathologist, a diminished bias during lymph node collection was assured. Our final analysis confirmed that there were no differences in lymph nodes harvested between the LAP and OP groups in this study, with the majority of patients having sufficient lymph nodes to be collected for accurate staging.
In previous reports with data on resection margins, none of the margins was found to be positive. Although this is a remarkable finding, it can be explained by the fact that most of these studies [
22‐
27] only reported distal and proximal margins. No data on circumferential margins were available from these studies. Results of the primary analysis indicated that laparoscopic procedure might have the ability to reach a better dissect field than the open procedure, assuring the radical cure resection.
Among our patients, those who underwent laparoscopic procedure had significantly faster recovery than those who underwent open surgery. LAP group patients definitely need a smaller dose of analgesic than their counterparts who received open surgery treatment. In fact, most laparoscopic procedures seem to cause less pain, so that analgesics are rarely necessary. It is reported that some centers are in favor of the epidural combined with general anesthesia during the operation. Thus, usually the PCEA (patient controlled epidural analgesia) and PCIA are both usual options of postoperative pain-control procedures. Some reports concluded that PCEA has greater advantages over PCIA [
28,
29]. In our center, general anesthesia is used routinely for laparoscopic surgery; PCIA is the choice only for patient controlled pain-control procedures. However, since the majority of LAP group patients did not require analgesia, the pain-control method did not seem to be an important parameter for laparoscopic colorectal surgery.
Total hospital stay and postoperative hospital stay are two important evaluation criteria for fast-recovery surgery. The postoperative hospital stay for LAP and OP group in the Multicentre Randomized Controlled trial - MRC-CLASICC Trial was 9 days and 11 days, respectively [
14]; however, the Clinical Outcomes of Surgical Therapy Study group (COST) Trial was 5.1 days and 5.6 days, respectively [
22,
23]. Length of hospital stay is an indicator which can be easily affected by different confounding factors, such as geographic locations, reflecting cultural and possibly financial reimbursement differences [
30]. In our group, all stage III patients accepted postoperative adjuvant chemotherapy. There was a set of patients in both groups who could not be discharged until the end of the first regimen of chemotherapy. This undoubtedly extended the length of hospital stay for these patients, thus introducing bias in the comparison of hospital stay between the two groups. Thus, we calculated the actual hospital stay after eliminating any such excess periods of stay during the investigation.
The anastamotic leakage rate in LAR patients is significantly higher in laparoscopically treated cases (5/17, 29%) than in the OP group (4/26, 15%). However, after revisiting the data in the LAP group we discovered that all the leakage occurred in early cases; this may be explained as effect of learning curve. The investigation enrolled these patients when our laparoscopic surgeon was in the early stages of learning curve, which led to a higher rate of complications. However, for the open surgery, all enrolled patients were operated by a surgeon with experience of more than 500 cases of open colorectal cancer surgeries.
The conversion rate in this study was only 3%, which was far lower than that reported in other trials. The conversion rate from laparoscopic to open surgery was 17% in the COLOR trial [
15,
17], 25.4% in the COST [
22] and 29% in the MRC-CLASICC [
14] trial. The MRC-CLASICC trial included cases of both rectal cancer and advanced stage cancer; 34% of patients with cancer of the rectum compared with 25% of patients with cancer of the colon underwent conversion from laparoscopic to open procedure. The ALCCaS trial did not include patients with rectal cancer [
31]. It is also worth noting that in the MRC-CLASICC trial the rate of intraoperative conversions fell by the year of study from 38% in the first year to 16% in the sixth year [
14]. In our study, the laparoscopic procedures were all performed by a single surgeon and the conversion cases all reported in the early period. However, as time passed the experience in the procedure increased. With stabilization of the learning curve of the operating surgeon, the conversion rate significantly reduced. Furthermore, in our study stage IV patients were not included, and all patients were found in preoperative evaluation to be suitable for laparoscopic procedure, thus the conversion rate was lower than other trials. It was reported that there was no difference when comparing conversions to those completed in operative time, morbidity, length of stay, costs, and readmission [
32]. There was greater blood loss, longer time to first bowel movement, longer length of stay when converted cases were compared with the cases completed with the laparoscopic-assisted approach but no difference when compared with open surgery [
33].
Additionally, our study showed similar overall recurrence rate, as well as the survival rate between the two groups. The number of patients that developed a recurrence at the site of the primary tumor during the follow-up period of the study was similar to that after laparoscopic and open surgery in other trials respectively. Separate analyses for colon and rectal cancer showed no significant differences between laparoscopic and open procedures. No significant differences in the occurrence of port-site/wound metastases or peritoneal metastases were observed [
22,
34‐
39]. Long-term outcome data from three major multicenter trials are still awaited [
15,
22,
35]. Early results from the COST study, with a median follow-up of 4.4 years, also did not demonstrate any difference in tumor recurrence, disease-free survival, and overall survival rates between the two surgical techniques for treating potentially curable colon cancer [
22]. This study also confirmed that the laparoscopic approach can provide as good radical resection as an open approach for treating potentially curative colorectal cancer. Similar perioperative mortality in the two groups confirmed the clinical safety of the laparoscopic approach reported in earlier research [
14,
22,
34,
40]. The potential impact of laparoscopic surgery on survival is not clear in this study since the multivariance analysis only indicated the clinical staging as a high risk factor for overall survival. The role of immunosuppression has been proved because of mediators of immunologic response. This positive impact of the laparoscopic procedure is probably worth further investigation.
The present study was obviously limited in that the patients were partially randomized into the two treatments arms. However, these results were obtained by two different teams specializing in respective surgical procedures, operating at a high volume of cases. Moreover, since there were no differences in demographic data, and all observed biases have negligible impact on the results we believe our results are accurate. This study has confirmed the feasibility of laparoscopic procedures for colorectal cancer, advocating the fast recovery times, and demonstrating similar medium-term recurrence and survival between LAP and OP groups. Thus, in a dedicated laparoscopic center, laparoscopic procedures may result in a potential perioperative and follow-up survival benefit compared with open procedures, particularly in advanced cases.
Competing interests
The authors declare that they have no competing interests.
Authors' contributions
SJ, JT and QZ conceived the study, and participated in its design and coordination as well as drafting of the manuscript. CG, CJ, HK and CS carried out the perioperative management for all patients in this study, participated in the operation and contributed to the drafting of the manuscript. PY managed the equipment and instruments for the surgery as well as the intraoperative data collection. LH and HR participated in conducting the follow-up investigations. All authors have read and approved the final manuscript.