Abstract
Abstract
Background:
Human natural orifice transluminal endoscopic surgery (NOTES) is slowed down by technical hurdles. Concomitantly, single-incision laparoscopy has been increasingly reported as an alternative. By reducing the invasiveness of standard laparoscopy, we may further reduce postoperative pain, decrease morbidity, preserve abdominal wall, and enhance cosmesis. Such techniques have been widely applied, including in colorectal surgery. The aim of this multicenter study is to compare the results of single-incision right colectomy (SIRC) with the results of the standard laparoscopic right colectomy (SLRC) in patients with colon cancer.
Methods:
The files of patients who underwent right colectomy for cancer in five hospitals between January 2010 and December 2013 have been reviewed. Exclusion criteria were open surgery, emergency setting, and American Society of Anesthesiologists (ASA) score >3. Patients were distributed in Groups A (SIRC) or B (SLRC).
Results:
Five hundred ninety-two patients were included in this study, 336 in Group A and 256 in Group B. Mean operative time was 129.0 minutes (range 65–245) in Group A and 168.1 minutes in the Group B (range 70–290), respectively (P < .001). No mortality occurred in either group. The overall 30-day morbidity rates were 21.4% in Group A and 25% in Group B, respectively (P = .64). The median length of hospital stay was 4.95 days (range 3–14) in Group A and 5.5 days in Group B (range 3–12), respectively (P = .28). Conversion to laparotomy occurred in four patients in each group (P = 1). Length of skin incision was significantly shorter in Group A than in Group B (2.99 ± 0.63 cm versus 4.94 ± 0.65 cm, P < .001). Histological analysis of the operative specimens showed no significant differences.
Conclusion:
SIRC is feasible and sure for patients with colon cancer. As compared with SLRC, SIRC may offer some advantages, including lower operative morbidity, shorter hospital stay, and better cosmoses, without compromising the oncological quality of the resected specimen.
Introduction
L
More recently, natural orifice transluminal endoscopic surgery (NOTES) and single-incision laparoscopic surgery (SILS) inspired the development of more minimally invasive techniques in colorectal surgery.15–19 With the aim of further reducing the invasiveness of surgery, these techniques are supposed to lower postoperative pain, decrease abdominal wall complications, and provide better cosmesis.
The reported short-term results of NOTES and SILS colorectal resections seem to be similar to those found with more standard forms of laparoscopy. However, NOTES and SILS procedures are thought to be technically challenging, associated with long learning curves, and yet to be proven oncologically relevant.
In our group of minimally invasive surgery (The Intercontinental Society of Natural Orifice, Laparoscopic, and Endoscopic Surgery [iNOELS]), we started performing NOTES and SILS procedures in January 2007. Initial indications included appendectomy, cholecystectomy, and bariatric surgery. Starting on April 2008, we extended the indications to colorectal surgery. After a period of accomplishment of the learning curve (i.e., at least 30 colonic resections either right sided or left sided and 20 rectal resections), we started on January 2010 to include our patients in a multicenter, prospective database.
In this study, we compared the results of single-incision right colectomy (SIRC) to those of standard laparoscopic right colectomy (SLRC) in patients with cancer. We analyzed the short-term results, emphasizing the technique, the operative parameters, the early postoperative outcome, the pathological indicators of the specimen, and the cosmetic result.
Materials and Methods
We retrospectively analyzed the clinical records of all patients who underwent right colectomy in four different centers specialized in minimally invasive surgery between January 2010 and December 2013. Only patients with histologically proven diagnosis of adenocarcinoma located from the cecum to the proximal third of the transverse colon were eligible for this study. Exclusion criteria were nonelective procedures (bleeding, obstruction, or perforation), open surgery, or patient's American Society of Anesthesiologists (ASA) status >3. Two Groups A and B were defined depending on the procedure, SIRC, or SLRC, respectively. Conversion from SIRC to SLRC was defined by the addition of one or more trocars.
History of open surgery was not considered as a formal contraindication to either SIRC or SLRC. Patients' preoperative affectation for either technique depended mainly on the discretion of the surgeon.
No mechanical bowel preparation was performed preoperatively in either group. However, patients were put on very low fiber diet for 7 days before surgery. Thirty minutes before incision, one dose of intravenous antibiotics was administered in all patients (2 g of cefoxitin or a combination of 250 mg of tobramycin and 600 mg of clindamycin). 20 All patients received venous thrombo-embolic prophylaxis with subcutaneous, low-molecular-weight heparin for a total of 4 weeks. 21
Surgical techniques
Two surgeons in each center performed the entire procedures in both groups. In case of SIRC, a 3-cm midline incision was created after everting the umbilicus. The Gelpoint device (Applied Medical, Orange County, CA) was inserted through the wound. Three 5–12 mm trocars were placed through the Gelpoint. We used straight rigid instruments in all patients, including a 30°, 5- or 10-mm laparoscope. The specimen was retrieved through the umbilical incision.
In SLRC patients, four trocars were used, of which a 12-mm left para-umbilical port for the endoscope, a 12-mm suprapubic trocar, and two 5-mm diameter trocars in the left flank and upper quadrant. We used the same straight and rigid instruments in all patients. The specimen was retrieved through a transverse incision, performed either in the suprapubic area or in the right lower quadrant at the site of a previous appendectomy.
All procedures, either in the SLRC or in the SIRC group, were performed according to the same technical principles. The patient was then placed in the Trendelenburg position with tilting of the right side up. Regardless of the approach (SIRC or SLRC), the procedure was standardized in seven steps:
• Medial to lateral and down to up dissection into the right Toldt fascia after identification of the ileocecal vessels, the duodenum, and the right ureter. • Up to down division of the mesentery until the terminal ileum was reached a few centimeters away from the ileocecal valve. • Division of the ileocecal vessels using an energy source device or clips. • Division of the right colonic vessels (if required) 1 cm from the lower pancreatic surface, ensuring adequate lymph node dissection, using either an energy source device or clips. • Division of the right side of the greater omentum, the right gastrocolonic ligament, and the transverse mesocolon using an energy source device. • Medial to lateral taking down of the hepatic flexure and full mobilization of the right colon after widely opening the right gutter. • Endocorporeal or extracorporeal, stapled division of the ileal and colonic edges of the resection. In case of endocorporeal division and/or anastomosis, the stapling device was inserted through the 12-mm diameter port of the GelPoint. Alternatively, we could perform the anastomosis outside the abdominal cavity after deflating the pneumoperitoneum and retrieving the specimen. In either setting, a side-to-side, stapled, ileocolonic anastomosis was created. The enterotomy was closed with a slowly absorbable running suture. Staple lines were routinely oversewn for hemostatic purposes. The abdomen was then reinsufflated and checked for bleeding. No mesenterical closure was performed.
In the SIRC group, the aponeurosis was closed using a nonabsorbable running suture. The umbilicus was reimplanted and the skin closed with undyed, absorbable suture. Finally, we performed a periumbilical long-acting anesthesia blockade as inspired from the transversus abdominis plane block technique. With or without ultrasound guidance, 20 mL of 7.5% ropivacaine was distributed in six different sites, located each 3 cm away from the umbilicus, inside the rectus muscle just in front of the posterior sheath.
In the SLRC group, multilayer, running, nonabsorbable sutures were used for closure. The skin was closed with undyed, absorbable running suture. All skin incisions were injected with 20 mL of 7.5% ropivacaine.
All patients followed a perioperative, fast-track protocol, including no nasogastric tube, no abdominal drainage, restricted fluid intake, and accelerated resumption of oral intake on day 0 or 1. Patient's controlled opioid analgesia was maintained for only 24 hours. Nonopioid, oral painkillers were then proposed if necessary.
A member of the surgical team (surgeon or fellow) examined each patient at 1 month after surgery. Then, the clinical record was thoroughly completed for early postoperative complications (i.e., 30-day morbidity). Patients were systematically examined at 3-month intervals for the first 3 years after surgery, then every 6 months for the subsequent 2 years.
The local ethics committee of each center approved the study.
Data analysis was performed using the IBM SPSS statistics standard v20.0 software. Categorical data are expressed as percentages and quantitative data are presented as mean ± SD with median and range. Data were analyzed by using the Fisher F exact test for categorical values and the Student t-test for continuous variables. Data analysis was based on the initial dispatching of patients between groups, except for the length of incisions. For this parameter, patients were included in Group B when converted to SLRC or excluded when converted to laparotomy. Statistical significance was defined as P < .05.
Results
Between January 2010 and October 2013, 592 patients who had laparoscopic right colectomy for cancer were included in the study: 336 patients (56.8%) had SIRC (Group A) and 256 patients (43.2%) had SLRC (Group B). Mean age was 62.4 ± 21.2 years in Group A and 63.1 ± 18.7 years in Group B (P = .87). Male to female sex ratios were 1/1.1 in Group A and 1/1 in Group B (P = 1). Mean body mass index (BMI) was 27.0 ± 7.2 kg/m2 in Group A (range 16.5–43.9) and 26.2 ± 6.2 kg/m2 in Group B (range 17.1–42.4), (P = .57). The majority of patients in both groups had localized disease upon preoperative assessment without patent distant metastases or peritoneal carcinomatosis. The two groups were comparable for patients' classification according to ASA score (P = .62) with a majority of ASA 1 in both groups. A total of 16.7% of patients in Group A and 21.9% of patients in Group B had history of prior abdominal open surgery. A summary of the preoperative patients' characteristics is listed in Table 1.
P < .05 was considered statistically significant.
ASA, American Society of Anesthesiologists; BMI, body mass index; SD, standard deviation; SIRC, single-incision right colectomy; SLRC, standard laparoscopic right colectomy.
The mean operative time was 129.0 ± 31.8 minutes (range 65–245) in Group A and 168.1 ± 48.3 minutes in Group B (range 70–290), (P < .001). Mean operative blood loss was 84.9 ± 69.0 mL in Group A and 98.3 ± 47.2 mL in Group B, (P = .32). No patient was transfused during the procedure. Additional procedures were performed in 40 patients in Group A (11.9%) and in 24 patients in Group B (9.4%), (P = .12). In Group A, additional procedures comprised 16 cholecytectomies, 8 hysterectomies, 8 bilateral oophorectomies, 8 partial gastrectomies, and 8 liver resections. In Group B, additional procedures comprised 12 left liver lobectomies, 8 cholecytectomies, 8 small bowel resections, and 4 atypical liver resections. In Group A, 16 patients had additional heated intraperitoneal chemotherapy for synchronous peritoneal metastasis. Conversion occurred in 32 patients in Group A (9.5%) and in 8 patients in Group B (3.1%), (P = .1). Tumor size or invasion of adjacent organs did not require conversion in any patient in this series. The causes for conversion were poor exposure with or without adhesions in 32 patients and bleeding in 8 patients. The bleeding incidents, all occurring in Group A patients, were adequately controlled with the addition of two conventional ports. Operative data are summarized in Table 2.
P < .05 was considered statistically significant.
If combined with other procedures, only time of colon resection is measured.
All skin incisions have been added including those of trocars and specimen retrieval site.
The mean doses of total intravenous morphine delivered postoperatively by the patient's controlled analgesia pump were 5.4 ± 2.7 mg in Group A and 6.9 ± 3.2 mg in Group B (P = .09). The median length of hospital stay was 4.95 days (range 3–14) in Group A and 5.5 days in Group B (range 3–12), (P = .28). The overall 30-day operative morbidity rates were 21.4% in Group A and 25% in Group B, the majority of these being of grade 1 in the Clavien Dindo classification.
No anastomotic leak or delayed intraperitoneal bleeding was noticed in Group A. Sixteen patients had anastomotic bleeding in Group A, including one with small bowel obstruction due to paralytic ileus. An additional patient had liver hematoma secondary to atypical resection for metastasis. Later on, the patient had right lower lobe pneumonia and deep venous catheter infection. No reoperation occurred in Group A.
In Group B, 8 patients had anastomotic leak (3.1%). One patient had a hematoma in the mesocolon next to the anastomosis. He was reoperated on day 5 for a peritoneal lavage, dismantling of the anastomosis, and a loop ileocolostomy. Sixteen patients had anastomotic bleeding, of whom 8 were transfused. Two patients, both with uncontrolled type 2 diabetes mellitus, had suprapubic, surgical site infection treated conservatively.
Globally, the two groups were comparable in terms of morbidity (P = .64) with no reported mortality. Postoperative parameters and morbidity–mortality profiles are analyzed and given in Table 3.
P < .05 was considered statistically significant.
At 30-day follow-up, the mean length of the skin incision(s) was significantly shorter in patients in Group A than in those in Group B (2.99 ± 0.63 cm versus 4.94 ± 0.65 cm, P < .001). Three hundred twenty patients (95.2%) in Group A and 200 patients (78.1%) in Group B were satisfied with the cosmetic quality of their incision(s) (P = .03). No delayed morbidity was found. The mean follow-up period is 33 months (range 13–49). To this day, we found no patient with clinically patent incisional hernia in either group.
Pathology report analysis revealed no positive margins situation in either group. The mean proximal margin as calculated from the ileocecal valve was 6.9 ± 4.9 cm in Group A and 7.2 ± 4.5 cm in Group B(P = .83). Mean distal margin was 6.7 ± 3.0 cm in Group A and 7.3 ± 2.4 cm in Group B(P = .36). The median number of harvested lymph nodes was 16.5 ± 6.3 in Group A and 17.7 ± 5.2 in Group B (P = .38). An overall analysis of histological parameters is shown in Table 4.
P < .05 was considered statistically significant.
Measured as the distance between the ileocecal valve and the proximal limit of the tumor, considered equal to zero whenever the valve is invaded.
Discussion
In this series, the short-term results of patients who had SIRC compare favorably with those who had SLRC. Both techniques were found to be safe with no mortality. No major intraoperative complications such as hollow viscus or visceral injury or uncontrollable bleeding occurred. Relative advantages of SIRC included better abdominal wall preservation, less postoperative morbidity, shorter hospital stay, and enhanced cosmesis. Moreover, the oncological quality of the resection was preserved with negative margins and comparable number of harvested lymph nodes. No incisional hernia was found in any patient during the study period.
After first being described in 2008,22,23 SILS colectomy has emerged as a viable method of minimally invasive surgical treatment of benign and malignant colorectal disease. In the past, for our first cases of SILS colectomy, we highly selected our patients as recommended by many reported series, focusing on safety and feasibility. We avoided situations that recognized to increase the difficulty of dissection and/or lengthen the operative time and/or predispose to conversion such as obesity17,24,25 with BMI >35 kg/m2, history of previous surgery, or locally advanced tumors.17,24,26–32 With increase in experience, indications for SIRC have been broadened as in Group A where 38% of the patients were more than 65 years old, 19% had a history of prior open surgery, and 17% had a BMI superior to 35 kg/m2, with a maximum of 44 kg/m2. Moreover, the diameter of the tumor or its local invasiveness (i.e., T4 tumors) did not deny our patients' access to SIRC. For these reasons, we consider this series as a representative specimen of the real-life practice of minimally invasive surgery for right-sided colon cancer. We believe that SIRC has surpassed the theoretical hurdles of early experience and can be now proposed as a first-line approach.
Identification of the learning curve is critical for surgeons wishing to perform single-incision laparoscopy. The mishaps of inexperience could generate dangerous complications to the patient and dissuade the surgeon from pursuing skill developments in a nowadays promising field of surgical practice. In a recently published series of consecutive cases, 33 the authors found that for a surgeon trained in advanced laparoscopic techniques, the learning curve is optimized after 40 procedures. Each surgeon performed more than 50 single incisions and NOTES-inspired colorectal resections before starting to enroll patients in the database used for this study.
Patients' safety remains a prerequisite if a newer technique (i.e., SIRC) was to challenge a more conventional one (i.e., SLRC). This is first achieved by the standardization of the technique. The key steps of right colectomy as well as the fundamental principles of oncological resection remained unchanged. The only changes in SIRC as compared with SLRC concern the used instruments. We have adopted the Gelpoint device as a single port platform for our colorectal resections. In our opinion, the Gelpoint allows adequate movements, enabling dissection in different abdominal quadrants. It also guaranties a tight protection of the abdominal wall and permits a scrupulous control of the length of skin incision. In addition, we prefer to use straight instruments like most authors34,35 and confirm that specially designed, curved, S-shaped, and flexible instruments could add complexity.36,37
When the surgeon holds the camera in one hand and the operative instrument in the other, eyes and hands depend of the same brain. This considerably reduces inefficacious time-outs during the procedure and accelerates the execution of orders coming from the brain. The role of the assistant is to hold tissues and retract organs, providing a “neo-triangulation” while reducing the conflict of instruments in the narrowed available space. Single-incision surgery is inspired from NOTES, which has endoscopic origins. In contrast, laparoscopy has been inspired by open surgery. Thus, the dogma of triangulation as the cornerstone of both open and laparoscopic surgeries is challenged with the more recent NOTES and single-incision techniques.
A technique can also be considered as secure when it allows a fallback solution at any time, should the patient's safety be compromised or the expected result be suboptimal. An initial SIRC technique does not preclude the addition of supplementary trocar(s) whenever deemed necessary by the surgeon just as laparoscopy permits conversion to laparotomy. In one patient of Group A, an intraoperative bleeding has been effectively controlled after conversion to conventional laparoscopy. The addition of extra trocar(s) does not compromise safety but offers technical solutions in delicate situations. Conversion should not be perceived as a failure of the approach but rather a practical alternative to provide a better outcome. With increase in experience, surgeons would naturally accomplish the procedures in SIRC more frequently.
No anastomotic leak occurred in Group A patients. The overall postoperative morbidity rate was comparable in both groups (21.4% in Group A and 25% in Group B). Although on the upper limit, these rates fit within the range of randomized laparoscopic trials,9,11,12 as well as single-incision laparoscopic case series.24,25,28 Differences in definitions, methods of recognition, and exhaustiveness in reports may explain such differences. It is noteworthy to state that severe complications (Clavien grade III and IV) occurred in 16 patients in each group, corresponding to 4.8% in Group A and 6.3% in Group B. This confirms other reports conclusions that in colorectal minimally invasive surgery most complications are minor.30,32,38 In fact, no patient required reoperation or admission to the intensive care unit for perioperative complication in Group A.
An important concern regarding single incision and NOTES-inspired colorectal procedures is the oncological quality of the resection. Large-scale prospective studies demonstrated that oncologic outcomes for conventional laparoscopy were similar to those for open surgery in colon cancer patients.9–12,14,26 In our series, all longitudinal and circumferential specimen margins were negative for cancer cells invasion and the number of retrieved lymph nodes was comparable in both groups (16.5 ± 6.3 versus 17.7 ± 5.2, P = .38). This finding was supported by the meta-analysis conducted by Zhou et al., stating that SILS was equivalent to conventional laparoscopic surgery in terms of margin-free resections and specimen lengths. 39 The two techniques were also equivalent for lymph node retrieval. 39 However, long-term outcomes of colorectal SILS for malignant diseases cannot presently be assessed given the lack of long-term follow-up in the available data. 38
In this study, the conversion rates to standard laparoscopy or open surgery were 2.4% and 3.1%. Despite excessive conversion rates (as high as 40%) in early experience, 40 our results perfectly fit within the range of most recent published works of which are two collective reviews.41,42 The estimated conversion rates range from 3.7% to 8.0% and from 1.6% to 2% for common laparoscopy and open surgery, respectively. Moreover, the rate of conversion to open surgery from SIRC (2.4%) is comparable to that from SLRC (3.1%). This finding was already observed in three meta-analyses on colorectal procedures, confirming that SILS does not increase conversion rate to open surgery when compared to standard laparoscopy.39,41,43 Acquired experience is of paramount importance in reducing the rate of conversion as no single conversion occurred in the last 30 SIRCs reported in this series.
The mean operative time of an SIRC is significantly shorter than of an SLRC (129.0 ± 31.8 minutes versus 168.1 ± 48.3 minutes, P < .001). It is worthy to note that anastomoses were mostly done extracorporeally except for 28 patients (16.7%) having SLRC. Endocorporeal anastomosis has been proved to significantly prolong the operative time in recent prospective series.44,45 This may explain the longer operative time found in Group B than in Group A. Besides that, it is now confirmed that accumulating experience considerably decreases operative time.35,46,47 The 50th case is considered the time point when SILS becomes fully efficient. 32 Three recent meta-analyses39,41,43 confirm that duration of procedure is not different between SLRC and SIRC.
We did not identify any clinically patent incisional hernia during the study period. Such low rates after colorectal resection are reported in most studies, averaging around 0%–1%.28,48 This may also be explained by our care in closing the incision as if it was a minilaparotomy, using nonabsorbable running sutures. Then, the umbiliculus was meticulously reimplanted and the skin closed cosmetically. The estimated incidence of incisional hernia after laparoscopic colorectal surgery is around 6% at more than 10 years follow-up. 49 High BMI and surgical site infection have been shown to be independent factors related to the occurrence of an incisional hernia. 49 In our experience with SILS and NOTES-related surgery, the overall incidence of incisional hernias ∼2% at 2 years, almost exclusively in obese patients, with or without ongoing weight loss (nonpublished data). Today's literature data have failed to prove that parietal complications are more frequent in the transumbilical SILS approach than in conventional laparosocopy.50–52
According to our results, SIRC offers a better pain profile than SLRC as shown by the significantly lowered intravenous morphine use. The reported results on postoperative pain after SILS remain controversial. Disparity of results may be secondary to inherent bias in retrospective comparative studies, lack of standardization of techniques, pain assessment, and management protocols.
SIRC is expected to be cosmetically superior to SLRC because reduced number and/or length of incision(s) logically improves the results. A significant difference in patients' satisfaction could not be drawn from this study. Ideally, an objective cosmetic score would have been more reliable to measure patient's satisfaction of his/her body image. The incision in the SLRC group was commonly performed either in the suprapubic area or in the right lower quadrant. Since anastomosis was mainly done, we estimate that the location of the incision led to a larger diameter (i.e., possibly a less cosmetic result) than in the SIRC group. As it happened previously, lessons from the laparoscopic experience enabled surgeons to optimize the practice of open surgery. Consequently, we presume that conclusions from the single port experience would help us optimize the multiport standard laparoscopic practice.
This study has several limitations to ascertain the clinical significance of our conclusions. The retrospective nature of this study implies inherent selection bias that may underpower the results. However, our study confirms previous conclusions that SIRC provides results at least similar to those of SLRC in the setting of cancer, while offering potential for new benefits. However, presently available data remain limited to retrospective, nonrandomized studies. At this level of knowledge, only large prospective multicenter studies could validate potential benefits of this minimally invasive colorectal procedure.
Footnotes
Disclosure Statement
No competing financial interests exist.
