Introduction

Laparoscopic right hemicolectomy (LRHC) has gained general acceptance and is increasingly performed to treat benign and malignant colorectal disease, with oncological outcomes comparable to open surgery [1]. LRHC also improves postoperative recovery, with less postoperative pain, shorter hospital stay and decreased postoperative morbidity [1,2,3]. Nevertheless, standardization of the surgical technique for LRHC, in particular regarding the intracorporeal or extracorporeal ileocolic anastomosis, is lacking. The performance of intracorporeal or extracorporeal anastomoses in LRHC can involve broad variations of the procedure, ranging from a hand-assisted procedure, with hand-assisted dissection and mobilization of the specimen, to totally laparoscopic surgery. The first meta-analyses published, which included heterogeneous studies, reported no evidence regarding the benefits of intracorporeal (IC) versus extracorporeal (EC) anastomosis after LRHC in terms of postoperative morbidity [4]. However, more recent meta-analyses, reported a benefit for IC anastomosis concerning postoperative morbidity [5,6,7]. The purpose of the present retrospective, non-randomized, multicentric study was to compare the short-term postoperative results obtained in two homogeneous groups of patients undergoing LRHC in tertiary centers using either IC or EC anastomoses.

Materials and methods

Patients

The study population was retrieved from the CLIMHET Study Group database which included patients LRHC between January 2005 and December 2015 in five tertiary University Hospital centers in France (Clermont-Ferrand University Hospital, Strasbourg University Hospital-IRCAD, Henri-Mondor Hospital in Créteil, European Hospital Georges Pompidou in Paris, and Tours University Hospital). This study was conducted in accordance with the ethical principles determined by the Declaration of Helsinki. According to French law no formal ethics approval was required for this study.

For the present analyses, all patients who had had elective LRHC for colon cancer, inflammatory bowel disease or polyps located in the ascending colon (from the ileocecal valve to the right colonic flexure) were included. Patients with synchronous colon cancers, metastatic disease, locally advanced cancer requiring multivisceral resections or who had repeat surgery for a tumor recurrence were excluded. Conversion was defined as completion of the right colectomy procedure through either an enlarged incision or an abdominal incision of ≥ 6 cm, and when conversion was necessary, the patients were also excluded. Two groups were compared: laparoscopic right colectomy with IC or EC anastomosis.

The following variables were analyzed: age, sex, body mass index (BMI), obesity (BMI > 30 kg/m2), American Society of Anesthesiologists (ASA) score, smoking, diabetes mellitus, arteriopathy, coronaropathy, previous laparotomy or laparoscopy, American Joint Committee on Cancer (AJCC) staging, lesion location (e.g., caecum, ascending colon, right colonic flexure (= hepatic flexure), ileocaecal valve), operating time, anastomotic characteristics (e.g., manual or mechanical anastomosis), perioperative antibiotics, perioperative transfusions, and perioperative intravenous fluids.

Endpoints

Postoperative morbidity and mortality were defined as events occurring during the hospital stay or within 90 days after resection. Short-term outcomes included postoperative morbidity according to the Clavien–Dindo classification, 30-day postoperative severe morbidity (Clavien–Dindo ≥ 3), time to first flatus, postoperative ileus, length of hospital stay, anastomotic leak, anastomotic bleeding, wound infection and postoperative mortality [8].

Medical complications included cardiac, vascular, urinary and pulmonary events. All complications were assessed by a clinician and prospectively registered in the databases at discharge or during the first outpatient visit. Surgical complications included ileus, wound infections, anastomotic leakage and evisceration. Postoperative outcomes included the following: reintervention rate, time to resume a regular diet, time to first mobilization, time to flatus, time to the first stools, intensive care unit (ICU) stay, overall length of hospital stay, and the number of lymph nodes harvested.

Surgical techniques

Two standardized surgical techniques were performed and compared: IC and EC anastomosis. During the study period, the choice of technique was made according to the surgeon’s preference. The learning curve for either of these two surgical techniques was considered to be achieved after 30 procedures [9]. Under both approaches, the Veress needle technique was used to create pneumoperitoneum, and four ports in total were generally employed. The placement of the ports was the same for both approaches: two 12-mm ports were inserted in the left abdominal wall along the pararectal line (symmetrically, one below for a 30° camera and the other above the umbilicus) and a 10-mm port in median suprapubic position. A 5-mm assistant port was placed in the right hypochondrium (Fig. 1).

Fig. 1
figure 1

Port positioning during laparoscopic right colectomy

Full size image

The first step consisted of exposing the vascular pedicles of the ileocolic trunk and right colic trunk; the right trunk of the medium colic pedicle was then ligated intracorporeally. The second step consisted of complete mobilization of the right colon and of the final ileal loop, performed in a medial-to-lateral manner. The Cattell-Braasch manoeuver was then performed; this consists in lowering the hepatic flexure after the incision of the colo-hepatic peritoneum and thus exposing the duodenum [10]. After the specimen was completely mobilized, the right colon and final ileal loop were transected using a linear stapler.

IC anastomosis

The IC anastomosis was performed using a stapler or with 4-0 absorbable sutures and the mesenteric breach was closed. The specimen was then completely mobilized and freed before extraction. The wound was then protected before extracting the specimen.

EC anastomosis

During EC construction of the anastomosis, after the protection of the abdominal wall, the specimen was extracted via an abdominal incision with subsequent extracorporeal bowel transsection.

The anastomosis was then performed either with a stapler or manual suture.

In the case of mechanical anastomosis, the enterotomy used to introduce the stapler was closed with sutures. An abdominal drain was not routinely inserted.

Perioperative management

No preoperative bowel preparation was performed. The patients were admitted in the afternoon of the day before the procedure. Deep venous prophylaxis with low-molecular-weight heparin (4000 IU) was started the evening before surgery and pursued until the postoperative day (POD) 30. Antibiotic prophylaxis with 2 g cefazolin and 500 mg metronidazole was administered before the induction of anesthesia. A nasogastric tube was always removed at the end if no longer required. The urinary catheter was generally removed on POD 1. No spinal anesthesia was associated with general anesthesia. Fluids were infused throughout the procedure at a rate of 10 ml/kg/h. All the centers were experts in enhanced recovery after surgery (ERAS).

Statistical analysis

For descriptive statistics, continuous variables are presented as medians (ranges) and categorical variables as numbers and percentages. Binary, multimodal and linear regression analyses were performed to control the effects of covariates on operative and postoperative outcomes; multivariate analyses included those variables which reached a significant p value under univariate analysis. In addition, a multivariate analysis of demographic, clinical and pathological variables was used to identify independent factors associated with the incidence of postoperative complications. A p value < 0.05 was considered to be significant.

Results

Demographic data

A total of 637 patients had LRHC between January 2005 and December 2015, 40 of whom were converted to open surgery and therefore excluded from this study. Therefore, 597 patients were included (Fig. 2).

Fig. 2
figure 2

Flow chart of the study population

Full size image

There were 306 males and 291 females with a mean age of 69 years (range: 21–96 years). Their mean BMI was 25.4 kg/m2 (range: 15–45 kg/m2).

The indications for surgery were neoplasm in 451 cases (75.54%), including 1 patient with a neuroendocrine tumor, polyps in 93 cases (15.58%) cases, and inflammatory bowel disease in 53 cases (8.88%). Among the 451 right colon cancers, 43 were stage 0, 109 were stage I, 178 were stage II, and 120 were stage III according to the TNM classification [11].

Among patients having LRHC, lesions were located in the caecum in 244 (41%), 156 (26%) in the ascending colon in 156 (26%), 71 (12%) at the hepatic flexure in 71 (12%), 36 (6%) in the right transverse colon in 36 (6%), and 90 (15%) at the ileocæcal valve. Only three patients received neoadjuvant chemotherapy.

Both types of anastomosis were performed in each expert center, following the gradual introduction of the IC anastomosis technique since 2005. An IC anastomosis was performed in 150 patients (25.13%) and an EC anastomosis in 447 patients (74.87%). There were more males (p = 0.005) and a higher BMI (p = 0.003) in the IC group, but otherwise the patients in the two groups were demographically comparable (Table 1).

Table 1 Preoperative patient characteristics in the IC and EC groups
Full size table

Intraoperative data

During IC anastomoses, a mechanical (88.67% versus 46.7%; p < 0.001), side-to-side (IC = 99.3%, EC = 88.6%; p < 0.001), isoperistaltic (IC 87.2%, EC = 45.8%; p < 0.001) anastomosis was generally performed.

During EC anastomoses, vascular ligation was mostly performed intracorporeally (87%). The rate of manual and mechanical techniques was comparable and a side-to-side anastomosis was mostly performed (88.6%).

The duration of the surgical intervention was longer in the EC group with a median of 150 min in the IC anastomosis group and 195 min in the EC anastomosis group (p < 0.001). Transfusion rates were similar in the two groups (p = 0.25). These results are shown in Table 2.

Table 2 Intraoperative outcomes in IC and EC groups
Full size table

Postoperative outcomes

Postoperative outcomes are summarized in Table 3. A significant difference was found regarding the incidence of medical complications, with a higher incidence of cardiac, vascular and pulmonary complications in the EC anastomosis group (13% versus 20%; p = 0.049). This difference remained significant after adjusting for field characteristics (age, sex, BMI, smoking, diabetes, arteriopathy, coronaropathy or history of previous surgery), under multivariate analysis (p = 0.009; Fig. 3). A shorter period elapsed before the resumption of a regular diet in the IC group (1 versus 3 days; p = 0.001). In addition, a shorter hospital stay (median of 7 versus 8 days; p = 0.003) was observed in the IC group. This difference remained significant under multivariate analysis (p = 0.029; Fig. 4). Surgical complications rates were similar between the two groups (p = 0.76). The most common complication in both groups was prolonged ileus with 23 cases (15.3%) in the IC anastomosis group and 89 cases (19.9%) in the EC anastomosis group (p = 0.83). There were no differences between the groups in terms of time for mobilization (p = 0.93), reintervention rate (8 versus 24; p = 1) and 90-day mortality (p = 0.47).

Table 3 Postoperative outcomes in both IC and EC groups
Full size table
Fig. 3
figure 3

Forest-plot representing postoperative medical complications after multivariate analysis

Full size image
Fig. 4
figure 4

Forest-plot representing the length of hospital stay after multivariate analysis

Full size image

Discussion

Our results from five tertiary centers involving expert surgeons in the achievement of a specific procedure show a benefit of the IC approach in terms of medical complications, hospital stay and operative time.

The principal strength of this multicenter study was that all patient data were collected over a short period of time and from five different tertiary centers involving experts in either IC or EC anastomosis techniques. There were therefore no technical changes to practice during the study period and the results obtained were a good reflection of current practice. To our knowledge and to date, this is the largest series published on this subject.

The main weakness of this study was its non-randomized design. However, randomization would have been difficult to interpret because the surgeons might have performed a technique in which they were not specialized. Our results were only conclusive after the learning curve had been achieved in tertiary center experts in laparoscopy. Although both techniques were practiced in each center, surgeons only used the technique they preferred. Differences in experience between surgeons could, therefore, have constituted a bias by influencing postoperative outcomes as a function of the learning curve. This study was retrospective and some information such as readmission rate and long-term complications (hernias, recurrence) was not available hence a prospective study including this information is needed.

In addition, the lengths of hospital stay and postoperative ileus could have been reduced during this 10-year period as a result of improvements in care and adherence to the ERAS program.

Laparoscopic colectomy is increasingly being considered as the gold standard for both benign and malignant colonic lesions [12]. Several studies have produced definitive results on the superiority of a laparoscopic approach over traditional open colectomy in terms of a more rapid return of bowel function, smaller incisions with less postoperative pain and better aesthetics, fewer pulmonary complications, and shorter hospital stay [13,14,15,16,17].

Jamali et al. [18] revealed that laparoscopic right colectomy with EC anastomosis is considered to be technically more difficult than laparoscopic sigmoidectomy, and that this difficulty increases significantly when the anastomosis is performed IC. This may explain why only a few surgeons have performed IC anastomoses to date [19].

This study consisted of a multicenter observational analysis of data from five tertiary centers in France involving a large cohort of 597 patients. The technique most widely employed in France remains EC anastomosis, which explains why the groups were not homogeneous in terms of the numbers of patients. EC anastomosis is more common because of the technical difficulties connected with IC anastomosis and the need in most cases to perform laparoscopic hand-sewn sutures, even in the case of a mechanical anastomosis. In terms of patient characteristics, stages of cancers and intraoperative data, the two groups were homogeneous and therefore comparable.

Vascular exposure was mostly achieved intracorporeally in our study. This was understandable due to the problems inherent in exposing the base of the mesentery via a small incision, especially in the case of obesity. Therefore, according to our study, most surgeons believe that the mesentery and ileocolic vessels should be divided intracorporeally.

The hypothesis underlying the development of a totally LRHC technique with IC anastomosis was that it could achieve more rapid recovery of intestinal function because traction on the colon and mesocolon is reduced, thus enabling greater postoperative comfort for the patient [20]. Our results totally supported this hypothesis, as patients experienced a more rapid recovery and shorter hospital stays after IC anastomosis, with no differences in terms of surgical outcomes. These findings reinforce those of a previous study which included colectomies by straight laparoscopy for colon cancer, adding colectomies for benign pathologies, and notably inflammatory bowel disease in the current study [21]. Although patients had a significantly higher BMI in the IC group, the duration of the surgical procedure was significantly shorter. This suggests that this technique is safe and less time-consuming in obese patients. In addition, a Pfannenstiel incision to extract the specimen was significantly more frequent in the IC group than in the EC group (48% versus 2%; p < 0.001). This lower rate was associated with less pain, resulting in potentially improved pulmonary recovery. Moreover, the occurrence of incisional hernia is much lower with a Pfannenstiel incision than with midline (umbilical) wounds [22, 23].

Postoperative medical complications were significantly more frequent in the EC group, probably because EC procedures require a larger incision. However, the rate of more severe complications was no higher in this group, which supports the fact that both techniques are safe.

Conclusions

LRHC with IC anastomosis could become the reference technique once surgical teams have completed the learning curve and achieved operative times comparable to those of right colectomies with EC anastomoses. IC is associated with faster recovery, a shorter hospital stay, and less medical morbidity. Our results suggest that IC anastomosis in LRHC is superior to EC anastomosis. This study clearly provides the rationale for a randomized prospective clinical trial.