Short- and Long-term Outcomes of 2-Step Stapled Intracorporeal Versus Extracorporeal Anastomosis in Laparoscopic Colectomy for Colon Cancer

Discussion

Since it was first reported, laparoscopic colectomy has rapidly become widely used. The noninferiority of laparoscopic colectomy to open surgery was established in large-scale clinical studies including the COST study (4-6), COLOR study (7, 8), and CLASICC study (9, 10). Several studies have also found that because laparoscopic surgery is less invasive than open surgery, postoperative recovery is faster (7, 11). Laparoscopic surgery for colon cancer is more difficult than open surgery but remarkable recent advances in endoscopic surgical instruments and improved surgical techniques mean that it has now become virtually standardized. However, even under these circumstances, it may prove difficult to deliver the bowel outside the body via a small incision during laparoscopic surgery for colon cancer, making it difficult to complete the procedure properly. In colon cancer surgery, many hospitals usually carry out lymph node dissection and dissection and mobilization of the bowel laparoscopically before delivering the bowel outside the body to transect it and create the anastomosis. However, in obese patients with thick abdominal walls or mesenteric fat tissue, it may be difficult to deliver the bowel outside the body via a small incision, and this may sometimes damage the mesentery, causing unnecessary haemorrhage. If the operations involved in anastomosis formation must be conducted with the bowel insufficiently delivered outside the body, this can hinder the creation of an accurate, reliable anastomosis, increasing the risk of subsequent anastomotic leakage. In patients with cancer of the transverse, descending, or sigmoid colon, additional mobilization of the left colonic flexure may also be necessary simply to deliver the bowel outside the body. Not only does IA in laparoscopic colectomy reduce the extent of bowel mobilization and avoid the risk of damaging the bowel or mesentery during delivery of the bowel outside the body, but adequate laparoscopic anastomosis operations may also reduce the occurrence of anastomotic leakage.

Recently, the number of reports of the performance of IA in hospitals where EA had been the first choice in laparoscopic colectomy has gradually been increasing, although this involves only a proportion of hospitals. IA for laparoscopic colectomy was first described in 2003 (12), but although over a decade has elapsed since that report, comparisons of IA and EA have yet to reach a conclusion as to which procedure is best. Some comparative studies of IA and EA have found that IA reduces operative time (13), others have found that it prolongs it (14, 15), and still others reported that the two are equivalent (16, 17). In the present study, operative time was longer in the IA group but this may have been because IA requires more coordination with the assistant than does EA, and a somewhat higher level of surgical skill is required than for EA. Laparoscopic surgery entails a learning curve (18), and, in this study, it was found that it was possible to complete the procedure in an operative time almost the same as that for EA after 27 cases. Even after the 27th patient, however, the operative time remained unstable, and since this study included only 44 cases, the possibility must be taken into account that the stable phase may not have been achieved. Associated intraoperative blood loss was significantly lower for IA than for EA. Previous studies have also reported lower volumes of haemorrhage (19-21). In EA, haemorrhage may occur due to damage to the mesentery when the bowel is delivered outside the body, and unnecessary haemorrhage may also occur if the mesentery must be divided in a position in which it is not directly visible, but in IA, the use of magnified imaging and of devices with improved haemostatic function may contribute to reducing haemorrhage from the mesentery.

Several studies have reported that incision length is shorter for IA (13, 17, 20, 21). In the present study, it was also significantly shorter. In IA, the resected bowel is straightened to remove it from the body, enabling the use of a smaller incision for tumours of the same diameter. By reducing the length of the incision, IA provides superior cosmetic results to those of EA. Attempts have also been made to combine IA with specimen extraction through a natural orifice, extracting specimens via the vagina or rectum to minimize destruction of the abdominal wall and achieve less invasiveness and better cosmetic results in a concept different from that of reduced port surgery (22, 23).

Some comparative studies of postoperative recovery time have found that it is equivalent for IA and EA (13, 20), while others have found that recovery is faster after IA (15-17, 21). In the present study, bowel movements resumed significantly earlier after IA, and the first passage of gas also tended to be earlier. Because IA does not require the bowel to be delivered outside the body, there is no need to perform bowel mobilization that is not oncologically necessary but is carried out solely to enable bowel delivery for EA, and this reduces the extent of mobilization. It is also conceivable that the bowel may recover faster because it is neither touched directly by hand nor exposed to the outside air at any point in the procedure.

Numerous studies have reported that postoperative complication rates are equivalent (24-28). In the present study, the results of IA were also equivalent to those of EA. The rates of suture failure, a complication that is one focus of concern in gastrointestinal surgery, were also equivalent. Different studies have reported different types of anastomosis created in IA but side-to-side anastomosis is basically the most common. A randomized, clinical trial found no difference between isoperistaltic and anti-peristaltic anastomosis formation (29). In the present study, isoperistaltic side-to-side anastomoses were created because they provide physiological bowel motility, require a smaller extent of bowel mobilization, are a simple procedure, have a large anastomotic orifice with little risk of stenosis, and have a low risk of anastomotic leakage. The insertion opening may be closed by hand suturing or mechanical suturing, but practice varies between hospitals, and since there has been no comparative study of the two techniques, it is unclear at this point which is better. We use mechanical suture closure because of the simplicity and reliability of the procedure. Technical issues with side-to-side anastomoses created entirely with linear staplers mean that if the staplers are inserted via the same 12-mm port, the axis of the linear stapler used to create the side-to-side anastomosis is different from that of the linear stapler used to close the insertion opening, making the procedure somewhat complicated. In the right colon, the procedure is ileocolostomy, and the good mobility of the small intestine means that problems are unlikely, but in the left colon, this difficulty becomes more pronounced the closer the anastomosis site is to the promontory angle. This problem can be resolved by inserting the stapler via the 12-mm port in the umbilicus during side-to-side anastomosis creation and through the 12-mm port in the left lower abdomen when closing the insertion opening.

Problems with IA include (i) bacterial contamination of the peritoneum as a result of faecal leakage, and (ii) the possibility of cancer cell dissemination. IA has been widely performed in laparoscopic gastric cancer surgery for some time, but in colorectal disease, any intestinal fluid that leaks contains stool, increasing the risk of the occurrence of SSIs such as peritoneal abscess; for this reason, IA has not become generally used despite its technical feasibility. In the present study, ascites bacterial cultures from peritoneal lavage after anastomosis completion were positive in a significantly higher proportion of patients who underwent IA. The effect was evident in postoperative biological reactions, with maximum temperature and blood test results indicating an inflammatory reaction on postoperative day 1 both higher in patients who underwent IA, although the incidence of SSIs was similar. This is because the bacterial species constituting the human normal flora cannot be judged clinically even if the isolated bacteria are identified, there are many bacterial species that are difficult to isolate and identify, obligatory anaerobic bacteria rarely cause infection alone, and it seems to be affected by the fact that there are many cases of multiple bacterial infections with aerobic bacteria and other anaerobic bacteria.

We had initially carried out the procedure after solely mechanical pretreatment of the bowel. Because leakage of watery mushy stool had also been noted during EA before the introduction of IA, and we had also seen postoperative intraperitoneal abscess formation due to leakage of watery mushy stool into the peritoneal cavity in patients who underwent IA when it was first introduced, we considered that mechanical pretreatment of the bowel made the stool more watery, increasing the risk of leakage during anastomosis creation and the extent of bacterial contamination. The surgical site infection guidelines issued jointly by the American College of Surgeons and the Surgical Infection Society (30) recommend the combined use of mechanical pretreatment and oral antibiotics for elective colorectal surgery. However, since mechanical pretreatment of the bowel is not recommended for the Enhanced Recovery After Surgery program, which aims to promote postoperative recovery, we replaced mechanical pretreatment with chemical pretreatment using oral antibiotics. The results of the present study clearly demonstrated that the use of this pretreatment may significantly reduce the rate of positive bacterial cultures in IA. This suggests that although the surgical procedure of IA entails a high risk of intraperitoneal contamination, carrying out preoperative pretreatment by chemical rather than mechanical means may limit such contamination to the same level as in EA.

No cancer cells were detected in ascites cytology after peritoneal lavage in either group. Many patients with gastric cancer with positive peritoneal lavage cytology go on to develop peritoneal recurrence, and this is regarded as a poor prognostic factor (31), but the clinical significance of peritoneal lavage cytology in colorectal cancer has yet to be established (32, 33). Because the sensitivity of peritoneal lavage cytology for detection is not very high (34), the argument for oncological safety is somewhat difficult to make on the basis of cytology alone. However, comparisons of recurrence rates have found that they are equivalent, and the type of recurrence is not peritoneal recurrence due to cancer cell dissemination or recurrence at the anastomosis site. In the present study, thus, no oncological problems were identified.

This study was limited by its nature as a non-randomized, single-centre study. Despite the equivalence of the IA and EA groups in terms of patient characteristics, surgery-related factors, and tumour-related factors, it is possible that bias may still have occurred. The results of the comparisons between the two groups must therefore be interpreted with caution. Although the present study included a large number of patients overall, the number in each group may still have been too small to permit proper analysis. Despite all these limitations, however, the present study clearly identified a number of advantages of IA, as well as its problems and methods for dealing with them. However, further randomized trials are required to compare IA and EA properly.