Outcome of laparoscopic colorectal resection

Transcription

1 Surg Endosc (2004) 18: DOI: /s y Ó Springer-Verlag New York Inc Outcome of laparoscopic colorectal resection M. Degiuli, 1 M. Mineccia, 1 A. Bertone, 2 A. Arrigoni, 2 M. Pennazio, 2 M. Spandre, 2 M. Cavallero, 1 F. Calvo 1 1 Department of Oncology, Division of Surgery, San Giovanni Battista di Torino Hospital, Via Cavour 31, Turin, Italy 2 Department of Oncology, Division of Gastroenterology, San Giovanni Battista di Torino Hospital, Via Cavour 31, Turin, Italy Received: 5 December 2002/Accepted: 16 April 2003/Online publication: 2 February 2004 Abstract Background: The aim of this study was to assess the feasibility and safety of laparoscopic surgery for colorectal diseases. Methods: A retrospective review was undertaken of all patients undergoing a laparoscopic colorectal procedure (LCP) for large bowel disease. All opertions were performed by a single experienced team. Patients were divided chronologically into three consecutive groups (G1, G2, and G3). Data collection included the incidence and cause of both proper and mandatory conversions to laparotomy, the incidence and type of early and late postoperative complications, incidence of operative mortality, and the length of hospital stay. The incidences of conversion to laparotomy and of early and late postoperative complications were also determined as related to diagnosis, type of LCP attempted, and chronological group. Results: Between January 1996 and December 2001, a total of 108 patients (49 men and 59 women) with a mean age of 65.1 years underwent an LCP for colorectal disease. Proper conversion to open surgery was necessary in five patients (4.6%), whereas a mandatory conversion was needed in 10 with patients advanced cancer (9.2%). The overall morbidity rate was 11.9%. There were no anastomotic leaks. In two patients (1.85%) developed a complication requiring reoperation. Postoperative mortality was nil. Mean postoperative hospital stay was 7.2 days. The rates of conversion and of early and late complications decreased through the three chronological periods. No trocar site recurrences were observed in the cancer patients. Conclusion: Laparoscopic colorectal surgery performed in experienced centers is safe; the observed morbidity and mortality rates are low and acceptable and compare favorably to those observed after standard open surgery. Correspondence to: M. Degiuli Key words: Laparoscopic surgery Laparoscopic colectomy Colorectal surgery Laparoscopic colorectal surgery Outcomes Complications In the last decade, the applications for laparoscopic surgery have broadened from cholecystectomy to include nearly all abdominal procedures, including solid organ extraction, gastroesophageal surgery, hepatobiliary surgery, and colorectal cancer. As with laparoscopic cholecystectomy, the benefits of decreased postoperative pain, early return of bowel function, shorter hospitalization, decreased disability, and better cosmesis are becoming evident for more advanced procedures. In fact, laparoscopic adrenalectomy, splenectomy, and Nissen fundoplication have become procedures of choice at centers where the technical expertise exists. Recently, laparoscopic procedures for the treatment of bowel disease have been performed with increasing frequency at different centers, but they are still rare and controversial [10]. Furthermore, although some prospective trials have confirmed that patients who undergo laparoscopic colorectal procedures (LCP) for benign disease receive benefits similar to those who have open procedures (OP) and that LCP compares favorably with OP [11], other authors are less sanguine about the advantages of LCP, particularly for malignant disease [6, 16]. The objective of this study was to assess the rate of intraoperative and postoperative complications incurred by patients who undergo LCP Materials and methods Between January 1996 and December 2001, a total of 108 patients underwent laparoscopic colorectal procedures performed by the same surgical team at our hospital. Data recorded for each patient included age, sex, body mass index (BMI), indications for surgery, procedure performed, intraoperative complications, conversion to laparotomy, early and late postoperative complications, postoperative mortality, and length of hospitalization. (Tables 1 4).

2 428 Table 1. Reasons for conversion to open surgery Reason for conversion n % Proper conversions Hemorrhage Adhesions and inflammations Increase of carbon dioxide Mandatory conversions Clinical T3 T4 cancer Table 2. Postoperative complications Complications Early complications Hemorrhage a 1 (0.9) Ileus/bowel obstruction b 2 (1.8) Wound infection 2 (1.8) Anastomotic leak 0 Urinary tract infection 0 Cardiac 0 Pulmonary 0 Ulcerative gastritis 1 (0.9) early complications 6 (5.5) Death 0 Late complications Trocar site hernia 1 (0.9) Suprapubic hernia 3 (2.7) Uretheral stenosis 2 (1.8) Thromboembolic 1 (0.9) Trocar site implantations c 0 late complications 7 (6.4) Death 0 complications 13 (12.0) reoperations 2 (1.8) deaths 0 a Requiring reoperation b One requiring reoperation c For cancer patients only Preoperatively, all patients had standard bowel preparation. Perioperative single-dose antibiotics and thrombosis prophylaxis were mandatory. Informed consent was obtained for all procedures. Postoperative complications were divided into early complications (occurred during hospitalization) and late complications (occurred after discharge). Patients were divided chronologically into three different groups. The first group (G1) comprised the first 35 patients treated; the second group (G2) and the third group (G3) comprised, respectively, the second 35 and the last 38 patients submitted to an LCP (Table 5). Definition and criteria for surgery Patients were assessed preoperatively for the possible use of laparoscopic techniques by the same surgical team; decisions regarding patient suitability for the procedure were made by the surgeon on the basis of the preoperative diagnosis and, in cases of cancer, intraoperative features. Informed consent was always required. In cases of cancer, only patients with clinical pt1 pt2 or Astlet-Coller stage D disease were enrolled in the study. At laparoscopy, clinical T3 and T4 cancer patients were always converted. In cases of nonmalignant disease, all patients suitable for laparoscopy were enrolled. Because we do not accept emergency cases, none of the operations was performed on an emergent basis. All phases of mobilization, vascular control, and anastomosis of a complete LCP are described later. An LCP was considered to have been converted to an open procedure if any unplanned incision was performed or if any planned incision was made either sooner or larger than planned. A conversion was defined as mandatory whenever, in cases of cancer, a clinical pt3 pt4 tumor was diagnosed intraoperatively, although the preoperative staging had diagnosed an earlier stage. Thus, we considered a proper conversion rate and an overall conversion rate, which also included mandatory conversions. Surgical technique All interventions were performed by the same surgical team, whose members had advanced experience in surgical oncology and laparoscopic techniques. Laparoscopic resection with curative intent was performed following the same rules as with open surgery. The pneumoperitoneum was established using an open laparoscopic technique; intraabdominal pressure values were then maintained between 12 and 14 mmhg. Surgical standards for colon cancer treatment, including a nontouch technique, lymphatic dissection, and oncological vessel ligation, were respected. Procedures to prevent port site metastasis were adopted routinely. Results Patients LCP were performed on 108 patients with a mean age of 63.4 years (range, ). There were 47 men and 61 women. The average BMI was Table 3 summarizes the indications for surgery. The most common indication was neoplasia (82.4%); precancerous lesions counted for 37% (33 patients); 56 patients (45.4%) had cancer (13 T1, 17 T2, 26 T3); and diverticular disease was the indication in 14.8%. In the same time period, 151 patients underwent open procedures for colorectal disease that were performed by the same surgical team. Because all patients with benign disease were treated laparoscopically, all of these 151 patients had a cancer that was not deemed suitable for laparoscopy (clinical T3 or T4 cancer, general contraindications to laparoscopy, or lack of consent). Procedure performed The surgical procedures are detailed in Table 4. A completely laparoscopic treatment was performed in 93 of 108 patients; an anastomosis was fashioned in 91 patients. All of the anastomoses performed after a rightside resection were fashioned extracorporeally through the right pararectal incision. Left-side colectomies and low anterior rectal resections were always followed by a laparoscopically fashioned anastomosis. The two anastomoses fashioned after a limited sigmoid resection were done extracorporeally through a short left pararectal incision. Intraoperative endoscopy was necessary in 34 patients to determine the correct location of the lesion. Three of the patients who required an intraoperative endoscopy had a right-sided lesion, usually close to the hepatic flexure; the other 31 patients had a left-sided lesion. Mean operating time was 183 min (confidence interval [CI] 5 95%, ). The longest operation was the total abdominal colectomy (305 min), followed by anterior resections (average, 198 min; CI 5 95%, 185

3 429 Table 3. Indications for laparoscopic surgery and outcome Diagnosis (n) Proper conversions conversions Early Complications Late complications complications Diverticulitis (16) 2 (12.5) 2 (12.5) 1 (6.2) 1 (6.2) 2 (12.5) Neoplasia (89) 3 (3.4) 13 (14.6) 5 (5.6) 6 (6.7) 11 (2.3) Colonic carcinoma (36) 1 (2.77) 5 (13.8) 2 a (5.5) 2 (5.6) 4 (11.1) Rectal carcinoma (18) 1 (5.55) 7 (38.8) 1 (5.5) 1 (5.5) 2 (11.1) Anal carcinoma (2) Colonic adenoma (23) a (4.3) 1 (4.3) 2 (8.7) Rectal adenoma (8) (12.5) 1 (12.5) Familia adenomatous poliposis (FAP) (1) Peutz-Jeghers (1) Others (3) Colonic volvulus (1) Cecum fistula (1) Anastomotic leak (1) (108) 5 (4.6) 15 (13.8) 6 (5.5) 7 (6.5) 13 (12) Table 4. Outcomes by type of laparoscopic procedure performed Procedure No. of patients Proper conversions overall conversions Early Late Postoperative stay (d) Right hemicolectomy 15 1 (6.6) 3 (20) Left hemicolectomy 24 1 (4.1) 3 (12.3) 1 (4.1) 2 (8.2) 3 (12.4) 6.9 Sigmoid colectomy (5.2) 2 (10.4) 1 (10.4) +3 (15.7) 6.8 Low anterior resection 24 2 (8.2) 6 (24.6) 2 (8.2) 2 (8.2) 5 (16.4) 7.1 abdominal colectomy with ileorectal anastomosis Abdominoperineal resection 4 1 (25) 1 (25) 1 (25.0) 0 1 (25.0) 8.5 Diverting colostomy Transverse colectomy (25.0) 1 (25.0) 5.8 Segmental colorectal resection (50) Other (4.6) 15 (13.8) 6 (5.5) 7 (6.4) 13 (12.0) ) and right hemicolectomy with intracorporeally blood supply division (average, 190 min, CI 5 95%, ). The average operating time for left or sigmoid resections was 175 min (CI 5 95%, ); the shortest operation was for stoma creation (75 min). The mean operative times for (G1 and G2) (208 and 196 min, respectively) were greater than the mean operative time for of 63 (171 min). Intraoperative complications Intraoperative complications were observed in two patients (2.1%). Both patients incurred an intraabdominal hemorrhage one because of a bleeding from the superior mesenteric vein during a right hemicolectomy and bleeding from the other due to an inferior mesenteric vein during a left colectomy. The superior mesenteric vein bleeding was controlled only after conversion to open surgery. There were no bowel and/or ureter injuries and no stapling device or other mechanical failures. Conversions The overall conversion rate was 13.8%. In ten patients, conversion to open surgery was mandatory due to the presence of an advanced cancer, clinically T3 or T4. Thus, the proper conversion rate was 4.6%. Reasons for conversion are listed in Table 1. Conversions were necessary mostly because of tumor fixation and inflammation (three of five patients). The decision to convert was made at the beginning of the operation. The proper conversion rate was higher after low anterior resection (8.2%) than after the other left- (2.3%) and right-sided (6.6%) procedures (Table 4). The first two chronological groups, G1 and G2, had the same rate of proper conversions, but this rate decreased by half in G3 (Table 5). Postoperative complications The overall rate of postoperative complications was 11.9%. We distinguished between early postoperative complications (ocurred during hospitalization) and late complications (occurred after recovery). Early postoperative complications were seen in six patients (5.5%) (Table 2). Two of them, abdominal bleeding and bowel obstruction, were major complications, that required repair during the postoperative course.

4 430 Table 5. Relationship between chronological status of patients and outcome Chronological group (n) Proper conversions Overall conversions Early Late Reoperations G1 (35) 2 (5.7) 4 (11.4) 3 (8.6) 3 (8.6) 6 (17.1) 1 (2.9) G2 (35) 2 (5.7) 7 (20.0) 2 (5.7) 3 (8.6) 5 (14.3) 1 (2.9) G3 (38) 1 (2.6) 4 (10.5) 1 (2.6) 1 (2.6) 3 (5.2) There were no anastomotic leaks. In cancer patients, we did not see any trocar site implantations during the follow-up period, and none of the patients died after the operation. We had seven late complications. There were two ureteral stenoses, both 4 months after the operation, due to a retroperitoneal fibrosis. This fibrosis was well documented in one of these patients, who underwent a reoperation for ureteral reimplantation. The other patient was treated endoscopically by ureteral stent implantation. The morbidity rate was higher after low anterior resections (16.4%) than after the other left colectomies. No complications were observed after right colectomy (Table 4). The incidence of complications seemed to be unrelated to the diagnosis (Table 3). To the contrary, it decreased from 17.1% in the G1 group, to 14.3% in G2, and 5.2% in G3 (Table 5). Length of hospital stay Patients whose colorectal procedure was performed and completed laparoscopically averaged 6.8 days of hospitalization postoperatively (median, 6; range, 3 19). Patients converted to open laparotomy were discharged an average of 9.3 days postoperatively (median, 8; range, 5 32). Patients submitted to right colectomies had the earliest recovery (average, 5.8 days) among those receiving an anastomosis. Patients in whom no anastomosis was performed (diverting colostomies) were discharged after an average of 3.5 days (range, 3 4) (Table 4). Oncological results Patient follow-up is ongoing. At present, the median follow-up time is 44 months (range, 13 84). There were two cancer-related deaths among the 56 patients treated with curative intent. Overall, the actuarial 5-year survival rate is 91.6% (95% CI, %). Because none of the other patients died, the overall and cancer-related survival rates are the same. Local and distal recurrences were documented, respectively, in one and three patients during the first 18 months. The probability of tumor recurrence was 13.5% (95% CI, %). Discussion In spite of the rapid growth of laparoscopic operative experience, buttressed by vast improvements in the technical instrumentation, the role of minimally invasive treatment in both benign and malignant diseases of the colon has yet to be [14, 19]. For a variety of reasons, the benefits of laparoscopic colorectal surgery have not been realized as readily as they were with cholecystectomy or the treatment of gastroesophageal reflux disease (GERD) [5]. In Italy, <3% of large bowel surgery is nowadays performed through laparoscopy; in Germany this incidence is lower still (1%, as reported by Kockerling et al. in 1998) [10]. The redutance to apply laparoscopy to colorectal disorders can be attributed in the main to the long learning curve for these procedures and to fear that the incidence of intraoperative and postoperative complications, and the operative mortality rate will increase as compared to the well-standardized open operations [1, 18]. Furthermore, although Lacey et al. have recently shown the real effectiveness of laparoscopic colorectal resection in a randomized trial [11], the prolonged operating time usually needed to perform a totally intracorporeal operation and troubling questions about the oncological adequacy of resection and lymph node retrieval for cancer patients may seem to many pratitioners to nullify the advantages of minimally invasive surgery [6, 7, 16]. Before the laparoscopic technique can be accepted as one of the standard means of treating colorectal disease, two basic questions need to be addressed. First, is colorectal resection technically feasible by minimally invasive surgery? And second, is laparoscopic colorectal surgery as safe as the traditional open approach? Large bowel resection via laparoscopy was feasible in most of the cases attempted in the present series. Excluding the 10 patients who were converted because of the presence of a clinically advanced cancer despite a negative preoperative investigation, only five cases (4.5%) presented major difficulties requiring a conversion to on open procedure. The presence of adhesions and inflammation was the major cause of conversion. Low anterior resections had the highest conversion rate (8.2%), meaning that there is a relationship between major difficulties in laparoscopic dissection and the site of the disease. The proper conversion rate was also found to be related to the relative experience of the surgeon, because the incidence of conversion in the last chronological group (G3) was only half that in the previous two groups (2.6 vs 5.7). The observed incidence of conversion to laparotomy compares favorably to those reported for many recent series [1, 3, 4, 8, 10, 11, 13, 16, 17]; moreover, the incidence of proper conversions was one of the lowest among all these series. The decrease of this rate in the last chronological group of patients argues for the relative importance of the learning curve; in the present

5 431 series, as also reported by other authors, >50 60 cases were needed for the operating team to gain full expertise and flatten the learning curve. The value of operative experience is also confirmed by data on the operative time; like Senagore et al. [18], Wishner et al. [20], and Aghacan et al. [1], we observed a decrease in the mean operating time from >200 min to 170 min over the first cases. The incidence of postoperative complications has been reported in large laparoscopic series to range from 6.8 to 36.6% [1, 3, 4, 8, 10, 11, 13, 16, 17]. We had 13 complications (11.9%), six of which occurred during the hospitalization period and seven of which occurred after discharge. The rate of postoperative complications seems to be higher after open surgery. For examples Bokey et al. reported a 37.2% morbidity rate after elective resection for colon cancer and a 40.2% morbidity rate after resection for rectal cancer [2]; similarly, the rate was 32.6% in the series reported by Franklin et al. [6] and >30% in the one reported by Lacey et al. [11]. One of the proposed benefits of laparoscopic colorectal surgery is a reduction in the rate of wound infection; in this series, the overall incidence of this complication was 1.8% a percentage considerably lower than that reported in recent open surgery series. McArdle et al. reported a wound infection rate of 17% in 176 patients undergoing open bowel surgery [12]. Franklin et al. [6] observed 17 wound infections in 224 colorectal resections (7.6%). The wound infection rate was 11.1% after 971 colonic resections and 12.5% after 757 rectal resections in the series reported by Bokey et al. [21]. In the Spanish trial reported by Lacey et al., it was 7% after laparoscopic surgery and 16% after open surgery [11]. Another important finding regarding complications is the relative incidence of postoperative hernias. We had three cases of hernia rising in the site of suprapubic laparotomy and one case of trocar site hernia; overall, the incidence of postoperative hernia was 3.6%, where it was 0 <1% in several recent series. The onset of postoperative clinical fistula due to anastomotic leak has been described in large laparoscopic series, with an incidence ranging from 0.8% to 36% [1, 3, 4, 8, 10, 11, 13, 16, 17]. After open colorectal procedures, this rate was also relevant; Phillips et al. reported an 8.1% rate of anastomotic dehiscence in a recent report summarizing the results of the Large Bowel Cancer Project [15]. In the series reported by Bokey et al., the rate of anastomotic complications was 14.8% after low anterior resection (LAR) and 2.8% after elective colon resection [2]. It was 1.7% in Franklin et al. s series [6] and <2% in Lacey et al. s randomized trial [11]. We observed no cases of anastomotic dehiscence. Because a tension-free anastomosis with a good blood supply is required to avoid anastomotic complications, during left-sided procedures and low anterior resections the inferior mesenteric vein was always ligated and sectioned, the splenic flexure was always mobilized, and whenever the inferior mesenteric artery was ligated at its origin the bowel was always sectioned at the level of proximal rectum to ensure an adequate supply of blood from the veins of the hemorrhoidal plexus vessels. Although a number of authors have reported the occurrence of death after laparoscopic colorectal surgery [4, 10, 11, 16, 17], on average, the postoperative mortality rate is <2%. Data available from laparoscopic series seem comparable to, if not better than, those reported after open surgery. A study by the German Study Group for Colorectal Cancer (SGCRC) found mortality rates of 2.7% and 3% following, respectively, colon an rectal surgery [9]. Bokey et al. reported mortality rates of 3.6% and 7.0%, respectively, after elective colon resection and rectal resection (low anterior resection or abdominoperineal excision) [2]. Peroperative mortality was, respectively 1% and 3% after laparoscopic and open resection in the recent randomized trial reported by Lacey et al. [11]. None of the patients in our series died after the operation. Although these data cannot be compared directly, they indicate that colorectal procedures can be completed laparoscopically in almost all cases of benign diseases. However, in cancer patients, close attention must be paid to avoid exacerbating advanced diseases by promoting cancer spread and adhesions. After completion of the learning curve, which can be particularly long (50 60 cases), the ranging of postoperative complications associated with laparoscopic bowel surgery can be expected to be lower than that reported after standard open surgery. Moreover, the length of hospital stay is always shortened, and the postoperative mortality rate should decrease significantly. Obviously, our successful survival results with cancer patients treated laparoscopically are not comparable to the results achieved with open treatment because these two subsets of patients are generally different in terms of stage of the disease (far advanced cases are excluded from laparoscopy). This findings should be helpful in confirming that laparoscopic bowel surgery is feasible and safe in experienced hands and at centers of demonstrated technical expertise. They should also help in the selection of cancer cases, at least until long-term results are available and large randomized series have shown that this technique is appropriate for any all patients with colorectal disease [5]. References 1. Aghacan F, Joo JS, Sher M, Weiss EG, Nogueras JJ, Wexner SD (1997) Laparoscopic colorectal surgery: do we get faster? Surg Endosc 11: Bokey EL, Chapuis PH, Fung C, Hughes WJ, Koorey SG, Brewer D, Newland RC (1995) Postoperative morbidity and mortality following resection of the colon and rectum for cancer. Dis Colon Rectum 38: Bruce CJ, Coller J, Murray JJ, Schoetz DJ, Roberts PI, Rusin LC (1996) Laparoscopic resection for diverticular disease. Dis Colon Rectum 39: Clinical Outcomes of Surgical Therapy (COST) Study Group (1996) Early results of laparoscopic surgery for colorectal cancer: retrospective analysis of 372 patients treated by Clinical Outcomes of Surgical Therapy (COST) Study Group. Dis Colon Rectum 39: Hazebroek EJ for The Color Study Group (2002) COLOR: A randomized trial comparing laparoscopic and open resection for colon cancer. Surg Endosc 16:

6 Franklin ME, Rosenthal D, Abrego-Medina D, Dorman JP, Glass JL, Norem R (1996) Prospective comparison of open vs laparoscopic colon surgery for carcinoma: five years results. Dis Colon Rectum 39 (10 suppl): S35 S46 7. Gray D, Lee H, Schlinkert R, Beart RW (1994) Adequacy of lymphadenectomy in laparoscopic assisted colectomy for colorectal cancer: a preliminary report. J Surg Oncol 57: Hartley JE, Mehigan BJ, Qureshi AE, Duthie GS, Lee PWR, Monson JRT (2001) mesorectal excision: assessment of the laparoscopic approach. Dis Colon Rectum 44: Hermaneck Jr P, Wiebelt H, Riedl S, Staimmer D, Hermanek P, German Study Group for Colorectal Cancer (SGCRC) (1994) Long time results of surgical therapy for colorectal cancer: results of the German Study Group for Colorectal Cancer (SGCRC). Chirurg 65: Kockerling F, Schneider C, Reymond MA, Scheidbach H, Konradt J, Barlehner E, Bruch HP, Kuthe A, Troidl H, Hohenberger W (1998) Early results of a prospective multicenter study on 500 consecutive cases of laparoscopic colorectal surgery. Surg Endosc 12: Lacey AM, Valdecasas JG, Delgado S, Castells A, Taura P, Piquè JM, Visa J (2002) Laparoscopy-assisted colectomy versus open colectomy for treatment of non-metastatic colon cancer: a randomized trial. Lancet 359: McArdle CS, Morran CG, Pettit L, Gemmell CG, Sleigh JD, Tillotson GS (1995) Value of oral antibiotic prophylaxis in colorectal surgery. Br J Surg 82: Ortega AE, Beart RW, Steele GD, Winchester DP, Greene FI (1995) Laparoscopic Bowel Surgery Registry: preliminary results. Dis Colon Rectum 38: Ota DM, Nelson H, Week JC (1994) Controversies regarding laparoscopic colectomy for malignant diseases. Curr Opin Gen Surg xx: Phillips RK, Hittinger R, Blesovsky L, Fry JS, Fielding LP (1984) Local recurrence following curative surgery for large bowel cancer. I: the overall picture. Br J Surg 71: Schiedeck THK, Schwandner O, Baca I, Baehrlehner E, Konradt J, Kockerling F, Kuthe A, Buerk C, Herold A, Bruch HP (2000) Laparoscopic surgery for the cure of colorectal cancer: results of a German five-center study. Dis Colon Rectum 43: Schlachta CM, Mamazza J, Seshandri PA, Cadessu M, Poulin EC (2000) Determinants of outcomes in laparoscopic colorectal surgery: a multiple regression analysis of 416 resections. Surg Endosc 14: Senagore AJ, Luchtefeld MA, MacKeigan JM (1995) What is the learning curve for laparoscopic colectomy? Am Surg 61: Stoker ME (1994) Laparoscopic colon surgery for cancer: controversy, caution and common sense. Int Surg 79: Wishner JD, Baker Jr JW, Hoffman GC, Hubbard 2nd GW, Gould RJ, Wohlgemuth SD, Ruffin WK, Melick CF (1995) Laparoscopic-assisted colectomy: the learning curve. Surg Endosc 9: