**3. Intracorporeal or extracorporeal anastomosis: differences in clinical outcome**

When an anastomosis has been performed, the main complication that surgeons try to avoid is anastomotic leakage (AL) which means that bowel content can move from bowel lumen into abdominal space. In EA, despite the entire operation is carried out laparoscopically, the anastomosis is comparable to that performed during open surgery. The IA has been proved safe by several study, showing no statistically significant difference in AL rate between IA and EA [7]. A recent international snapshot audit [8] has identified 3 surgeon-dependent variables significantly associated with AL: duration of surgery, surgical approach, and anastomotic technique. Regarding duration of surgery, operating time varied widely: Magistro et al. [9] reported a significant longer duration of surgery for IA. Although the IA technique is retained faster by some [10], most studies showed no significant difference. However, it has been shown that the learning curve plays a major role in reducing the operative time [11].

Laparoscopic approach decreases morbidity and mortality after colorectal resection [4, 12]. Similarly, a laparoscopic approach is associated also with a lower AL rate compared with an open approach [8]. Considering anastomotic technique, the last Cochrane review [13] concluded that stapled ileocolic anastomosis was associated with fewer leaks than handsewn anastomosis. Two large observational

studies [14, 15] showed that the stapled technique is an independent risk factor for ileocolic anastomotic leak. Future large, randomized controlled trials are needed to identify the best anastomotic technique. To the authors' knowledge in 2018 has been proposed a study protocol for a randomized controlled trial IA versus EA in which primary endpoint is to compare hospital stay and secondary endpoints are intraoperative and postoperative events included AL. The results os this study will be available in 2021, depending on the volume of patients.

Surgical site infection (SSI) is reported in several case series; a meta-analysis by Ricc et al. reported a reduced risk of wound infection in favor of IA. The higher incidence of infection at the extraction site incision in EA anastomosis may be due to wound contamination during exteriorization of the bowel ends and performing the anastomosis through the incision [16].

The length of incision is another factor that influence morbidity after laparoscopic surgery: patients who had an EA were more likely to develop incisional hernia due to the longer incision required for specimen extraction and anastomosis: in EA group the extraction site is about 2.2 cm longer than IA group. Beside its length, the location of the extraction site incision may favor the development of incisional hernia. This was most frequently observed in cases of midline incision in the EC group, as compared to the IC group, where a Pfannenstiel incision was preferred [17]. Moreover, shorter incision is associated to less postoperative pain which result in early recovery after surgery [18].

Gastrointestinal function, demonstrated by time to first flatus and time to bowel movement, resume sooner in IA group than EA.

The technical challenges of EA may explain the earlier recover observed in the IA group. Indeed, delayed recovery of GI function may arise from traction on the bowel ends and mesentery needed to allow complete mobilization of the transverse colon during EA [17]. A recent RCT supports this hypothesis [19] by showing a significantly less surgical stress response after IA. Interleukin-6 and C-reactive protein levels were indeed markedly lower in this group.

Another aspect of intracorporeal ileocolic anastomosis which deserve to be studied is the configuration between ileum and colon. The anastomosis can be carried out in isoperistaltic or antiperistaltic configuration. ISOVANTI randomized clinical trial, performed in 2017 and published in 2018, has compared iso- and antiperistaltic configuration in order to understand if there is any difference in postoperative outcome. The results show that no differences were found in conversion rate, total operative time, and global complication rates after applying Clavien-Dindo's classification. Regarding functional results, the antiperistaltic group showed better results than the isoperistaltic group with less time to first flatus, less time to first stool and shorter time to satisfactory oral intake with statistically significant differences in all cases. However, this fact did not reduce hospital stay and there was no difference between both groups [20].
