Introduction

Technical advances and the development of new instruments over the last decade have significantly improved minimally invasive surgery for colorectal cancer. Evidence from controlled trials has revealed numerous benefits of laparoscopic colorectal resection [1–5]. The reported long-term results suggest that laparoscopic surgery does not jeopardize oncologic outcomes, and this has led to the widespread acceptance of laparoscopic colorectal surgery [6].

In conventional laparoscopic-assisted colorectal procedures, the colon is brought out through a minilaparotomy incision so that resection to the proximal margin can be performed [3, 5]. In general, the left lower quadrant [6, 7], the suprapubic area [8, 9] or, occasionally, the transumbilical area [10] is selected as the minilaparotomy incision site.

The aim of this study was to compare the transumbilical incision to the left lower transverse incision used in laparoscopically assisted anterior resection or low anterior resection, in which the specimen is delivered extracorporeally, in patients with sigmoid colon cancer and rectal cancer.

Materials and methods

Collection of prospective data of consecutive patients who underwent laparoscopic resection with curative intent for colorectal adenocarcinoma between April 2010 and December 2010 was performed in the Department of Colon and Rectal Surgery, Chonnam National University Hwasun Hospital, Gwangju, South Korea. The study protocol was approved by the institutional review board.

Inclusion criteria were an age of at least 18 years, diagnosis of left-sided colon cancer or rectal cancer not involving the anal sphincter, and performance of laparoscopic surgery. Exclusion criteria were occurrence of multiple primary colorectal cancers, synchronous distant metastasis, synchronous other malignancy, rectal cancer involving the anal sphincter or with adjacent organ invasion, performance of intersphincteric resection or abdominoperineal resection, or conversion to an open procedure.

Patients with T3-4 mid and low rectal cancer or positive lymph nodes underwent preoperative neoadjuvant chemoradiotherapy with 5-fluorouracil/leucovorin and radiotherapy administered as 5,040 cGy over 5 weeks (28 fractions). The operation was performed 6–8 weeks after neoadjuvant therapy.

After informed consent was obtained, patients scheduled for elective laparoscopic colectomy for sigmoid colon and rectal cancer were allocated to two groups according to the site of the minilaparotomy determined by the surgeon’s preference: a transumbilical incision group and a left lower transverse incision group. In the present study, the operations were performed by two colorectal surgeons (KHR and HJW) with experience of over 400 laparoscopic operations for colorectal cancer.

Demographic data included the patient’s age, gender, American Society of Anesthesiologists (ASA) score, body mass index (BMI, kg/m2), history of previous laparotomy, type of operation, tumor location, and stage. The same standard perioperative protocols were used in both groups. Mechanical bowel preparation was carried out on the day before operation with polyethylene glycol oral solution. All patients received prophylaxis with antibiotics and prophylactic treatment for thrombosis with compression stockings were done in accordance with the standards at our institution. Left colon cancer patients had their urinary catheter removed on postoperative day 1 and rectal cancer patients on postoperative day 3. Postoperative antibiotic therapy with intravenous cefoxitin sodium (1.0 g) 3 times a day for 4 days was administered.

Surgical outcomes including operation time, estimated blood loss, transfusion requirements, tumor size, harvested lymph nodes, distal resection margin, time to first flatus, and length of postoperative hospital stay were assessed.

Pain was managed with patient controlled intravenous administration of morphine chlorhydrate and with narcotics for 3 days after surgery. In addition, non-steroidal anti-inflammatory drugs (NSAIDs), analgesic ketorolac tromethamine (30 mg), and opioid pethidine hydrochloride (25 mg) were given for pain control. The total dosages of postoperative analgesics and a visual analog scale at 6, 12, 24, 72 h, and 5 days postoperatively were used to assess postoperative wound pain (on a scale of 0–10, with 0 representing no postoperative wound pain and 10 representing the worst possible wound pain).

Postoperative length of the minilaparotomy incision was assessed in the operation theater and at 3 months postoperatively and was assessed during the postoperative follow-up period. Using a questionnaire, we assessed patient satisfaction by telephone or in the outpatient visit 6 months to 1 year postoperatively. Patient satisfaction was scored using the visual analog scale ranging from 0 to 10 points (with 0 representing no satisfaction). A 5-point score was postulated as a scar at the midline incision after open colorectal surgery. Patients were asked whether they were satisfied with the minilaparotomy scar of laparoscopic surgery compared to the long midline incision of open colorectal surgery. Postoperative general complications and minilaparotomy-related wound complications were evaluated. Early wound complications included surgical site infection, hematoma requiring open drainage, and fascial dehiscence. Delayed wound complications included incisional hernia and wound infections requiring readmission.

Surgical techniques

Each patient was positioned in the Lloyd-Davies position, and the surgeon and the scopist stood on the right side of the patient. A nasogastric tube was not used routinely in either group. Intravenous single-dose cefoxitin sodium (1.0 g) was administered at the induction of anesthesia in both groups. Bair Hugger® patient warming devices were used (Arizant, Eden Prairie, MN, USA) during the operation.

Briefly, the abdomen was exposed using an open surgical method, and an 11-mm supraumbilical optic port was èpositioned. A 12-mm port was placed in both the right and left lower quadrants, and two 5-mm ports were placed about 10 cm proximal to each 12-mm port. The abdomen was insufflated with CO2 gas to a pressure of 12 mmHg. After a careful inspection of the liver and entire abdomen for the detection of carcinomatosis, a high ligation of the inferior mesenteric artery and vein was performed, and dissection was begun with the medial-to-lateral approach, taking care to identify and preserve the left ureter and gonadal vessels. Laparoscopic total mesorectal excision was accomplished in the cases of mid and low rectal cancer. Splenic flexure take-down was performed in all patients. A double-stapling technique was applied using endoscopic linear staplers, and a circular stapler was used for sigmoid and rectal transection after a thorough rectal irrigation with betadine solution and re-establishment of intestinal continuity with primary anastomosis in the usual manner.

A 3-5-cm minilaparotomy incision was made in the transumbilical area or left lower transverse area. In the case of transumbilical incision, a crescent-shaped incision prolonging the first trocar incision was made around the umbilicus by one of the surgeons (KHR), and the linea alba was divided and closed with a single layer of No. 1 Vicryl (Ethicon, USA). A left lower transverse incision was made at the lower port site on the left by the other surgeon (HJW), and the anterior and posterior sheaths were bluntly separated from the muscle and were closed in layers with No. 1 Vicryl. Every wound was closed with subcuticular skin closure using 3/0 monosyn (Covidien, USA).

The OCTO port wound protector and retractor system (Dalim Corporation, Seoul, Korea) and the cylindrical vinyl film covered the minilaparotomy for protection against cancer cell dissemination during specimen retrieval. All procedures were performed in keeping with oncologic principles. A diverting ileostomy was performed based on the decision of safety in anastomosis. A pelvic drain was inserted.

Statistical analyses

A two-tailed p value <0.05 was considered to be statistically significant. Data are presented as numbers of patients and percentages or as mean ± standard deviation. Groups were compared using the Student’s t test for noncategorical baseline data, and Fisher’s exact test and two-tailed Chi-square test were used for categorical baseline data. Statistical analysis was performed using SPSS version 17.0 (Chicago, IL, USA).

Results

A total of 147 eligible consecutive patients were enrolled. Transumbilical incisions were performed in 92 patients, and left lower transverse incisions were performed in 55 patients. The mean follow-up period was 20.0 ± 5.4 months (range, 1–31 months). Preoperative chemoradiation therapy was administered to 26 patients (17.7 %) with rectal cancer including 16 patients (17.4 %) in the transumbilical group and 10 patients (18.2 %) in the left lower transverse group (p = 0.903).

Demographic data including age, sex, American Society of Anesthesiologists (ASA) grade, body mass index (BM), tumor location, and pathologic stages were similar in the two groups (Table 1). A previous history of laparotomy was higher in the left lower transverse incision group than in the transumbilical incision group (32.7 vs. 17.4 %, respectively, p = 0.033).

Table 1 Clinicopathologic characteristics of patients
Full size table

There was no difference in operation time, estimated blood loss, and transfusion requirements between the transumbilical incision group and left lower transverse incision group (Table 2). No conversion to an open procedure with any intraoperative event occurred in the two study groups. There was no statistically significant difference between the two groups in regard to the number of patients who received a diverting ileostomy: 21 (14.3 %) patients overall, 12 patients (13.0 %) in the transumbilical group and 9 patients (16.4 %) in the left lower transverse group (p = 0.578). The size of the tumor, number of harvested lymph nodes, distal resection margins, time to first flatus, and length of hospital stay in the two groups were similar.

Table 2 Intraoperative data
Full size table

The data relating to postoperative pain and length of the minilaparotomy incision are listed in Table 3. A higher dose of nonsteroidal anti-inflammatory analgesics was required in the transumbilical group than in the left lower transverse incision group, but the difference did not reach statistical significance (56.2 mg vs. 46.9 mg, respectively, p = 0.541). Ten patients (6.8 %) required opiates for postoperative pain control. Postoperative pain scores were similar between the transumbilical incision group and the left lower transverse incision group at 6 h (mean, 2.6 vs. 2.5, respectively, p = 0.760), 12 h (mean, 2.1 vs. 1.2, respectively, p = 0.085), 24 h (mean, 2.0 vs. 1.3, respectively, p = 0.081), 48 h (mean, 1.9 vs. 1.4, respectively, p = 0.236), and 5 days (mean, 1.7 vs. 1.1, respectively, p = 0.178). The minilaparotomy incision was shorter in the transumbilical group than in the left lower transverse group at the time of surgery (mean, 4.6 cm vs. 6.2 cm, respectively, p = 0.000). After 3 months, the incision scars of previous minilaparotomy were 4.1 cm in the transumbilical incision group and 5.8 cm in the left lower transverse incision group. One hundred thirty-five patients (91.8 %) answered the questionnaire regarding satisfaction. The mean satisfaction score was higher in the transumbilical group than the left lower transverse group, but the difference did not reach statistical significance (7.6 ± 2.5 vs. 7.1 ± 2.3, respectively, p = 0.224).

Table 3 Comparison of outcomes in the two groups
Full size table

The postoperative complications are listed in Table 4. Except for fascial dehiscence, more postoperative complications were found in the transumbilical group compared to the left lower transverse group, but this did not reach statistical significance. Eleven cases of anastomotic leakage were detected, and they were managed with laparoscopic diverting ileostomy, although they were not included in the 21 ileostomies presented.

Table 4 Postoperative complications
Full size table

Five cases of anastomotic bleeding were observed. One case was managed with diverting ileostomy after hemostasis, and the others were managed conservatively. Four cases of postoperative ileus were noted. One of these cases was managed with re-exploration, and the remaining three cases were treated conservatively. Four cases of chylous fluid drainage, 2 cases of urinary tract infection, 5 cases of urinary difficulties, one case of fever of unknown origin, and one case of postoperative peptic ulcer were managed conservatively.

Surgical site infection, early wound-related complication, occurred in 19 cases (12.9 %), 12 cases in the transumbilical incision group, and 7 cases in the left lower transverse incision group, but no statistically significant differences were observed (p = 0.810). Ten cases of purulent erythematous wound discharge, 7 cases of hematoma requiring open drainage, and 2 cases of fascial dehiscence were noted. There were 2 cases of incisional hernia (2.2 %). One patient with fascial dehiscence developed incisional hernia during the postoperative follow-up period, and incisional hernia was found in another patient during outpatient follow-up in the transumbilical group.

High BMI (>25 kg/m2) and longer operation time (>180 min) were risk factors for wound complications in univariate analysis (Table 5), although they were not significantly associated with multivariate analysis.

Table 5 The risk factors for wound complications
Full size table

Discussion

Traditionally, in laparoscopic anterior or low anterior resection for colorectal cancer, a laparoscope is introduced through the umbilical optic port, and a small minilaparotomy incision is needed for specimen delivery, anvil insertion, and returning the proximal stump to the pelvis for subsequent intracorporeal end-to-end stapled colorectal anastomosis.

The decision about where to make the minilaparotomy incision is based on the following considerations, the mobilization of the bowel with no resistance to delivery, and the surgeon’s preference and experience. Several incision sites have been used for minilaparotomy, including a left lower transverse incision [6, 7], suprapubic incision [8, 9, 11, 12], and transumbilical incision [10]. The ideal location of the incision has not yet been determined.

Kawahara and colleagues reported that the incision should be made in the suprapubic area rather than the left lower abdomen [8]. Suprapubic incision is thought to provide adequate visualization for an extracorporeal anastomosis, better cosmesis, and with minimal exposure of the incision in daily life [11], although the approach is not used extensively in colorectal surgery. One possible reason is the potential difficulty with mobilization of the hepatic or splenic flexures, and fear of conversion to midline, which results in a T incision [11]. Suprapubic incisions are associated with an extremely lower risk of wound complications and if placed precisely in the suprapubic area, obesity plays less of a role in the outcome, and the complication rates are very low. Orcutt et al. [11] reported, in a retrospective cohort study, that the Pfannenstiel incision was associated with a decreased risk of short-term wound complications. According to De Souza, a Pfannenstiel incision is associated with the lowest rate of incisional hernia and should be the incision of choice for hand assistance and specimen extraction in minimally invasive colorectal resection wherever applicable [12].

There are potential advantages of the left transverse incision. Most importantly, the incision is located near the tumor, making it easy to extract the specimen. However, the splitting of the transversus and oblique muscles of the abdominal wall during the procedure may be accompanied by the potential risk of incisional hernia due to weakness of the abdominal wall, muscular hematoma, and seroma.

The transumbilical incision is often used in right-sided colon cancer and also has clinical benefits of a smaller incision in left-sided cancer, because it is not far from the specimen, and tension-free anastomosis is easily achievable with splenic flexure mobilization. Theoretically, making an opening in the abdominal wall for specimen delivery requires only the division of the linea alba, with avoidance of blunt splitting of the muscles that are dissected free from the incisional hernia and hematoma. The incision becomes smaller with shrinkage of the skin around the umbilicus, especially in a vertical direction.

In the literature, the length of reported minilaparotomy incisions for specimen extraction in laparoscopically assisted left-sided colon cancer and rectal resections was about 2 cm [13], 3 cm [14], 5 cm [2, 7], 7 cm [1, 3], or 5–7 cm [9]. In the present study, the length of the transumbilical incision was about 4.6 cm and that of the left transverse incision was 5.8 cm, which were values comparable to those reported in the literature. Patients are satisfied with the cosmetic effect that follows umbilical wound contracture, and the wound itself can be hidden along the umbilical contour.

Surgical site infection is a rather common complication of colorectal surgery, and its incidence ranges from 11 to 27 % [15–17]. The incidence rate of 14.3 % of wound-related complications in the present study was comparable to that in open surgery or conventional laparoscopic colorectal surgery. The rate of surgical site infection was not significantly higher in the transumbilical incision group than the left transverse incision group.

Incisional hernia is a common problem after both open and laparoscopic abdominal surgery, and its incidence ranges from 5 to 15 % [18]. Singh and colleagues reported that the midline extraction site resulted in a significantly higher incisional hernia rate (17.6) than the off-midline extraction sites in laparoscopic colorectal surgery [10]. In the present study, incisional hernias were found in 2 patients (2.2 %) in the transumbilical group during postoperative follow-up, but sufficiently longer follow-up evaluation is warranted to determine the incidence of incisional hernia.

The potential drawbacks of this study are that the procedures were not all performed by the same surgeon, which may affect the length of minilaparotomy incision. Although there was no statistically significant difference in tumor size and specimen length between the two groups, other variable factors, such as T4 advanced lesion, high BMI, surgeon’s individual preferences, and intraoperative findings that we did not analyze and that may be involved during extraction of the specimen during surgery, might affect the length of the incision site. Most of all, the surgeon’s individual preference may have biased the results, and further randomized controlled trials by a single experienced surgeon are warranted.

A major limit of the study is the lack of randomization. Another limitation is the short follow-up period. This affects the assessment of the incidence rate of incisional hernia. Long-term follow-up is needed for further, more structured investigation of this subject. Lastly, we did not assess other objective measures of data on quality of life following minilaparotomy to support our conclusions.

Conclusions

Transumbilical minilaparotomy for specimen retrieval in laparoscopic anterior and low anterior resection can be a good alternative to left lower transverse incision with the technique that appears to be associated with acceptable wound complications and possibly with better cosmetic outcomes.