The different treatments for AL after colorectal cancer surgery resulted in comparable 5-year OS and DFS, and distant and local recurrence rates did not differ significantly between surgical and conservative treatment. Multivariable analysis showed no association between treatment and oncological outcomes. Regarding short-term outcomes, postoperative hospital stay was shorter in the conservative group compared with the surgical group.
Previous studies indicated a rate of AL of 4%–29 %, [11–14] and several risk factors for AL after colorectal cancer surgery have been discussed [20–22], as well as methods to reduce the rate, prevent AL, and the related risk factors [23–26].
Loop ileostomy diversion has been performed for patients undergoing rectal cancer surgery (up to 5 cm from the anus), intersphincteric resection, chemoradiotherapy, and total mesorectal excision, in our institution. In the current study, 9 (18%) patients with diversion developed AL in in the conservative group, while 3 (12%) patients with diversion developed AL in the surgical group. There were no significant differences between the groups regarding the loop ileostomy diversion rate. Kulu et al. [1] and Crippa et al. [27] reported that diversion did not protect against AL, although there might be a protective effect of diversion in patients with more severe grades of AL.
Some studies suggested that laparoscopic surgery is associated with increased anastomotic failure rates [28], while a Cochrane review found no difference in AL rates between laparoscopic and open surgery [29]. We found no difference in AL grade in our study between laparoscopic and open surgery. To our knowledge, no studies have evaluated whether the surgical approach is associated with AL grade.
AL was diagnosed postoperatively earlier in the surgical group vs the conservative group, in our study, and a previous study showed that early AL may lead to technical failure of the anastomosis [22]. However, no significant differences were found with Cox regression analysis, and for whether early vs late AL is associated with oncological outcomes.
To our knowledge, studies have not evaluated surgical and oncological outcomes once AL occurs. AL is reported to be significantly associated with increased rates of local recurrence [1–10], although large retrospective studies reported a survival advantage of chemotherapy [30, 31]. Our institution routinely prescribes chemotherapy for patients with stage II and III colorectal cancer, as a high-risk group [32, 33].
Kulu et al. [1] performed a propensity score analysis for rectal cancer, and concluded that grade B and C leaks were not associated with oncological outcomes. We consider three possibilities to explain this finding: First, excluding nine patients with grade A AL decreased statistical power and led to a loss of statistical significance. Second, the prolonged chronic inflammation associated with conservatively-treated AL led to immunosuppression and a persistence of tumor cells [34, 35]. Third, chronic inflammation prevented early adjuvant treatment, which is known to be beneficial regarding oncological outcomes [36, 37].
Some previous studies reported that microperforation could occur before or during surgery, which is associated with increased local recurrence rates [38, 39]. Vital malignant cells have been detected intraluminally and on staple and suture lines during surgery. Furthermore, in vitro and experimental animal studies have shown growth of these cells and their ability to metastasize [40–44].
Crippa et al. [27] stated that AL negatively impacted the adjuvant chemotherapy rate. In our study, 8 patients (32%) in the surgical group received adjuvant chemotherapy, and 17 patients (34%) in the conservative group received adjuvant chemotherapy. However, these differences were not statistically significant, and both rates were low. Surgical treatment for AL was associated with longer postoperative hospital stay, need for intensive care, and elevated serum inflammatory markers. The time from surgery to receiving chemotherapy did not differ significantly between the groups. Additionally, hospital stay was shorter in the conservative group vs the surgical group, and patients receiving surgical treatment required subsequent stoma closure.
Noh et al. [45] stated that a delay in initiating adjuvant treatment for patients with AL could explain their poorer DFS. den Dulk et al. [46] reported no difference in AL rates between patients undergoing neoadjuvant chemoradiotherapy and those who do not. In our study, the AL rate in patients who underwent neoadjuvant chemoradiotherapy was similar to those who did not receive preoperative radiation.
Our study suggested that the type of AL treatment was not associated with oncological outcomes. However, recurrence location was significantly different between the two groups; liver metastasis was more common in the surgical group. Additionally, more patients had left-sided colon cancer in the surgical group, and a previous study reported that left-sided colon cancer was more likely to result in liver metastasis than right-sided colon cancer [47]. Our results were consistent with these previous results.
There were several limitations in our study. First, this was a retrospective and single-institution study. Second, surgeries were performed by several surgeons, and the use of nonstandardized techniques may have affected the incidence of AL and oncological outcomes. Finally, the sample size was small, and the observation term was short.