A period of starvation after gastrointestinal surgery during which an intestinal anastomosis has been formed has been a common practice. The rationale for starvation was to prevent postoperative nausea and vomiting and to protect the anastomosis, allowing it time to heal before being stressed by food [11]. ERAS programs incorporate evidence-based practices to minimize perioperative stress, intestinal dysfunction, iatrogenic infections, and postoperative pain and to promote early mobilization and recovery. Most data supporting ERAS programs have come from colorectal surgery. The key components for ERAS in colorectal surgery are as follows: a thorough preoperative patient history, thoracic epidural anesthesia during open (but not laparoscopic) colonic surgery, the avoidance of fluid overload and hypovolemia, no use of a nasogastric tube, and EOF and mobilization [12]. In the field of gynecologic surgery, some studies have demonstrated that increased ERAS guideline compliance is associated with a decrease in the length of the hospital stay across all patients and a lower risk of complications [14, 15]. In a prospective controlled trial of a COF group (n = 72) versus an EOF group (n = 71) following major abdominal gynecologic surgery, the average length of stay for the COF group was 5.8 days, while that for the EOF group was 4.7 days (P = 0.006) [19]. In this study, 83 patients (58%) had ovarian cancer and 19 patients (13%) received upper abdominal surgery. The proportion of women who underwent rectosigmoid resection was not mentioned. In another prospective controlled trial comparing COF (n = 22) and EOF (n = 18) groups of gynecologic oncology patients undergoing intestinal resection, the average length of stay for the COF group was 9.1 days, while that for the EOF group was 6.9 days (P = 0.022) [14]. In this study, 30 patients (75%) underwent rectosigmoid resection and 15 patients (38%) received upper abdominal surgery. In these two prospective controlled trials, patients in the EOF group were initially given a clear liquid diet on POD 1. If well tolerated, they were then given a regular diet. In reviews of ERAS programs for major abdominal gynecologic surgery, EOF appeared to be safe without increased gastrointestinal morbidities or other postoperative complications. The benefits of this approach include a faster recovery of bowel function, lower rates of infectious complications, a shorter hospital stay, and higher patient satisfaction [20–23].
The current study also demonstrated that EOF was feasible and safe for ovarian cancer patients undergoing rectosigmoid resection with anastomosis as a part of cytoreductive surgery. A long period of postoperative starvation seems to be unnecessary, even after high-complexity surgery. Consequently, EOF was associated with a shorter postoperative hospital stay. The lengths of the hospital stay were longer for both groups in the current series than those reported in previous studies. In our study, forty-seven patients (23%) underwent multiple bowel resections and 117 patients (58%) underwent upper abdominal surgery. Therefore, our operating time was much longer than the operating times in previously reported prospective controlled trials, which averaged about 3 hours. The shortness of operative times might partly explain the intestinal function recovery immediately after cytoreductive surgery. Patients in the EOF group were given only liquids on POD 1 and a semisolid diet on POD 2. Our impression from the current study is that a diet of either clear liquids or semisolid food is not tolerable for most ovarian cancer patients undergoing high-complexity surgery, including both rectosigmoid resection and upper abdominal procedures, on POD 1. A proportion of advanced ovarian cancer patients can tolerate a diet on POD 1. We are planning a trial to investigate the safety and clinical usefulness of providing a postoperative diet on POD 1 for patients undergoing cytoreductive surgery with intermediate SCS. The medical insurance system in Japan fully covered the costs of their hospital stay regardless of the length. This system may influence the relatively longer hospital stays that are typical for patients in Japan. However, a comparison of the length of the hospital stay for patients undergoing rectosigmoid resection with anastomosis between the COF and EOF groups still showed a significant difference even after accounting for potential biases.
Some studies have shown that diverting stomas have the potential to decrease the frequency of anastomotic leakage in ovarian cancer patients undergoing rectosigmoid resection. Houvenaeghel et al. reported that 59 (20%) of 305 patients who underwent rectosigmoid resection with anastomosis during up-front surgery, interval debulking surgery, or secondary debulking surgery received a diverting stoma, and the overall anastomotic leakage rate was 8% in 9 French cancer centers [24]. Another study analyzed retrospective data from 331 patients with stage II–IV ovarian cancer who underwent colon resection during up-front surgery. Forty-four patients (13%) received a diverting ileostomy, and the overall anastomotic leakage rate was 6% [25]. In the current series, the incidence of anastomotic leakage was relatively low (1%) in both COF and EOF groups. However, it should be noted that the patient selection was limited. A diverting stoma was created for 8% of the patients who underwent rectosigmoid resection with anastomosis. The proportion of the patients who had a diverting stoma in the COF group was higher than that in the EOF group. Since the middle of this study period, we have routinely utilized a TDT after rectosigmoid resection with anastomosis for patients with ovarian cancer [17]. TDT placement can decrease the need for a diverting stoma after rectosigmoid resection with anastomosis. In both the COF and EDF groups, all patients with a diverting ileostomy started diet intake on POD 2 or 3.
Some limitations must be considered when interpreting the data from this study. This was a retrospective, not a randomized, study. However, as the safety and benefits of EOF are apparent, planning a prospective randomized study to evaluate the impact of EOF on the postoperative outcomes, including postoperative hospital stay, following rectosigmoid resection in comparison with that of COF is ethically impossible. Despite this limitation, the use of the same surgical techniques by the same gynecologic oncologists at a single institution, the consecutiveness of the patients, and the consistency of the surgical decisions should reduce the possibility of major biases.