Article Text

Download PDFPDF

Comparative study of three different managements after colorectal anastomosis in ovarian cancer: conservative management, diverting ileostomy, and ghost ileostomy
  1. Victor Lago1,
  2. Amalia Sanchez-Migallón1,
  3. Blas Flor2,
  4. Pablo Padilla-Iserte1,
  5. Luis Matute1,
  6. Álvaro García-Granero2,
  7. Marcos Bustamante3 and
  8. Santiago Domingo1
  1. 1 Department of Gynecologic Oncology, University Hospital La Fe, Valencia, Spain
  2. 2 Department of Surgery, Coloproctology Unit, University Hospital La Fe, Valencia, Spain
  3. 3 Endoscopy Unit, University Hospital La Fe, Valencia, Spain
  1. Correspondence to Dr Victor Lago, Gynecologic Oncology Department, La Fe University and Polytechnic Hospital, Avinguda de Fernando Abril Martorell, 106, 46026. Tower F; 3rd Floor. Valencia,Spain, Spain; victor.lago.leal{at}hotmail.com

Abstract

Objective Anastomotic leak remains the main concern after colorectal anastomosis in ovarian cancer. Our objective was to compare the use of three different management approaches after colorectal resection and anastomosis in patients with ovarian cancer.

Methods Between January 2010 and June 2018, a total of 133 patients with International Federation of Gynecology and Obstetrics (FIGO) stage II–IV ovarian cancer who underwent colorectal resection and anastomosis were included. According to the approach followed after colorectal anastomosis and during the post-operative period, patients were stratified into three groups: conservative management and observation, diverting ileostomy, or ghost ileostomy technique. Univariate analyses were performed for quantitative variables by applying Student’s t test or Mann-Whitney U test and for qualitative variables by using the χ2 test (or Fisher’s test according to the sample size).

Results A total of 145 patients underwent colorectal resection during cytoreduction for FIGO stage II–IV ovarian cancer. Twelve patients were excluded because a colostomy was required. Thus, 133 patients were included in the final analysis. Modified posterior pelvic exenteration was performed in 121 (91%) patients and recto-sigmoid resection in 12 (9%) patients with relapse. The approach after anastomosis was wait-and-see in 72 patients (54.1%), diverting ileostomy in 19 patients (14.4%), and ghost ileostomy in 42 patients (31.5%). There were no differences in diagnosis, age, body mass index, ECOG (Eastern Cooperative Oncology Group), histology, tumor grade, FIGO stage, or type of surgery between the groups. No differences were found regarding the anastomosis leak related factors or the rate of anastomotic leak between the three groups (5.6% vs 5.3% vs 4.8%; p=0.98). Two patients died because of the anastomotic leak in the wait-and-see group, and none died in the diverting ileostomy or ghost ileostomy group. In the diverting ileostomy group, a higher number of patients had complications compared with the ghost ileostomy group (78.9% vs 7.1%; p<0.01). Four patients (21.1%) developed dehydration due to high output stoma (>1500 mL) causing electrolyte imbalance in the diverting ileostomy group, and one patient (2.4%) in the ghost ileostomy group (p=0.03). The stoma reversal rate was 73.7% for the diverting ileostomy group and 100% for the ghost ileostomy group.

Conclusions There were no differences found in the rate of anastomotic leak among the three groups of patients. The use of ghost ileostomy avoids the drawbacks of diverting ileostomy and seems to have advantages over routine diverting ileostomy and wait-and-see approaches for ovarian cancer patients undergoing colorectal anastomosis. Rates of stoma reversal are lower after diverting ileostomy when compared with ghost ileostomy.

  • anastomotic leak
  • ovarian cancer
  • complications
  • ghost ileostomy

Statistics from Altmetric.com

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

HIGHLIGHTS

  • There was higher number of complications after diverting ileostomy compared with ghost ileostomy.

  • Ghost ileostomy may be easily converted to diverting ileostomy in the setting of an anastomotic leak.

  • There was no difference in the interval between surgery and the beginning of chemotherapy comparing diverting versus ghost ileostomy

Introduction

In ovarian cancer, modified posterior pelvic exenteration is a commonly performed procedure during cytoreductive surgery and it is often followed by an end-to-end colorectal anastomosis.1 2 Anastomotic leak remains the main concern after colorectal anastomosis, with a reported incidence of 1.26–9% in ovarian cancer patients.3–11 Anastomotic leak is associated with an increase in morbidity and mortality, days of hospitalization, as well as chemotherapy delays.12 To date, there are no conclusive data outlining the risk factors for anastomotic leak in ovarian cancer.3 6 12 13 In a recent study, age at surgery, serum albumin level, one or more additional small bowel resections, hand-sewn anastomosis, and distance of the anastomosis from the anal verge have been found to be independently associated with anastomotic leakage in a multivariate analysis.14 In colorectal cancer, these factors have been widely studied.15–17 In cytoreductive ovarian cancer surgery, as there is no definitive recommendation regarding which patients would benefit from diverting ileostomy,18 the choice is often left to surgeon preference. Nevertheless, this decision may be based on an estimated risk of anastomotic leak tailored in each case.14

There are three possible approaches after the colonic anastomosis is performed: conservative management described as wait-and-see during the post-operative period, a diverting ileostomy, or ghost ileostomy. Diverting ileostomy has been studied as a protective approach to reduce the consequences of anastomotic leak. However, diverting ileostomy is not without potential complications, and it may be associated with significant morbidity, such as dehydration, malnutrition, renal failure, prolapse, and stricture. The reversal of diverting ileostomy is also associated with possible complications such as small bowel anastomotic leak, wound infection, or incisional hernia.18–20 Diverting ileostomy may also impact self-image and has been associated with serious psychological effects and decreased quality of life.21 Ghost ileostomy has been previously described as an alternative to diverting ileostomy in ovarian cancer.22 Ghost ileostomy can be easily converted into a diverting ileostomy in cases of suspicion of anastomotic leak, and it minimizes the impact of fecal peritonitis in cases of leak or offers the advantage of avoiding a stoma in patients without anastomotic leak. An additional advantage is that the diverting ileostomy can be performed without re-laparotomy and it only requires removal of the loop for reversal in patients without anastomotic leak, while diverting ileostomy requires a second surgery for reversal. Lastly, a ghost ileostomy also avoids the impact on self-image and quality of life associated with a diverting ileostomy. Therefore, ghost ileostomy has been suggested as an alternative to diverting ileostomy in colorectal cancer treatment.23–26 Moreover, ghost ileostomy seems to be a reproducible, safe and low-cost technique.27 There is a lack of information in the gynecologic oncology literature pertaining particularly to these three different approaches in ovarian cancer tumor reductive surgery. The aim of this study is to evaluate patient outcomes after undergoing these three approaches following bowel resection and anastomosis in the setting of ovarian cancer surgery.

Methods

After the approval of the local ethics committee was obtained, a retrospective review of all patients who underwent cytoreduction surgery due to ovarian cancer between January 2010 and June 2018 was performed. Patients with International Federation of Gynecology and Obstetrics (FIGO) stage II–IV ovarian cancer who underwent colorectal resection during cytoreduction for ovarian cancer were assessed for eligibility. Those who underwent recto-sigmoid or rectal resection and colorectal anastomosis because of tumor involvement at the pouch of Douglas were included in the analysis. All surgeries were performed by a multi-disciplinary team, including a gynecologic oncologist and a general surgeon available for intra-operative consultation. Before surgery, bowel preparation and antibiotic prophylaxis were performed in each case. A circular end-to-end stapled colorectal anastomosis was performed in all cases. The integrity of the anastomosis was assessed in each patient intra-operatively. Patients were classified according to three different approaches after colorectal anastomosis was performed.

Group Stratification

Wait-and-see

After colorectal anastomosis, a drain was placed in the pelvis followed by abdominal wall closure without stoma creation. Diagnosis of anastomotic leak was performed based on symptoms (fever, drain output, wound or vaginal discharge) and blood test results (C-reactive protein, procalcitonin, and/or leucocytosis) associated with any defect in the integrity of the colorectal anastomosis (CT scan, contrast radiology imaging).

Diverting ileostomy

After colorectal anastomosis a drain was placed in the pelvis and, before the abdominal wall was closed, the stoma was created by a small circular skin incision at the previously marked point by the stoma therapist. After the abdominal incision was closed, the bowel was opened and fixed following the loop ileostomy technique.28

Ghost ileostomy

After colorectal anastomosis, a drain was placed in the pelvis. A portion of terminal ileum 20 cm distant from the ileocecal valve was identified. A small orifice was dissected in the mesenteric border to pass a vessel-loop. The afferent and efferent portions of the terminal ileum were marked with a long and short stitch, respectively. The loop was placed like a percutaneous surgical drainage at the point of the theoretical stoma (the point was previously marked the day before surgery by the stoma therapist) (Figure 1)29.

Figure 1

The ghost ileostomy procedure. (1) A portion of terminal ileum (20 cm) distant from ileocecal valve is identified. (2) A small orifice is dissected in the mesenteric border to pass a vessel loop. (3) The loop is placed like a percutaneous surgical drainage at the point of the theoretical stoma. (4) In case of clinical suspicion of leakage, the ghost ileostomy is converted into a diverting ileostomy (DI). (5) In case of leakage findings at the rectoscopy, a circular incision is performed and with gentle pulling movements the bowel loop externalized. (6) A diverting ileostomy is performed without re-laparotomy.

As previously reported,29 the diagnosis of anastomotic leak was based on a sequential post-operative rectoscopy with low pressure performed between the third and seventh post-operative day to check on the anastomosis status. If an anastomotic leak was found during the rectoscopy, with or without confirmation in a CT scan, the ghost ileostomy was converted into a diverting ileostomy. Otherwise, if no leak was found at rectoscopy, then oral intake (solids) was initiated, and the vessel-loop was simply removed approximately 2 weeks after discharge on an outpatient basis. In cases of clinical suspicion of leakage, the ghost ileostomy was converted to a diverting ileostomy. The same anastomotic leak diagnosis criteria used for wait-and-see and diverting ileostomy were also applied, but in association with post-operative rectoscopy.

The decision between the three approaches was based on surgeon choice. The following demographic and surgical variables were recorded: presentation of disease (first diagnosis or relapse), age, body mass index (BMI), performance status (ECOG (Eastern Cooperative Oncology Group)), adjuvant treatment, histologic data, FIGO stage, type of surgery, operative risk assessed according to the American Society of Anesthesiology (ASA) score, and extended surgical procedures. We also collected information on history of diabetes, smoking status, steroid use, previous pelvic radiotherapy, low serum albumin level (<3.5 g/dL), low colorectal anastomosis (<5 cm from anal verge), additional bowel resection, operative time, intra-operative transfusion, estimated blood loss, and incidence of anastomotic leak. For patients who underwent diverting ileostomy or ghost ileostomy, the following variables were recorded: length of hospitalization, interval to chemotherapy, diverting ileostomy reversal, and time to reversal. All complications directly related to the diverting ileostomy or ghost ileostomy (>30 days) were recorded: dehydration, acute renal failure, stoma stricture, gastrointestinal bleeding, paraestomal hernia, subacute bowel obstruction, complication rate, readmission rate, and reason for re-admission.

Data were described as n (%) for qualitative variables and mean±SD for quantitative variables. Univariate analyses were performed for quantitative variables by applying Student’s t test or Mann-Whitney U test and for qualitative variables by using the χ2 test (or Fisher’s test if the sample size was small).

Results

A total of 145 patients who underwent colorectal resection during cytoreduction for FIGO stage II–IV ovarian cancer were evaluated. In 12 patients a colostomy was performed, as an anastomosis could not be achieved due to the extent of the large bowel resection or the overall condition of the patient. Finally, 133 patients were included in the analysis. Modified posterior pelvic exenteration was performed in 121 patients (91%) and recto-sigmoid resection in 12 patients with recurrent disease (9%). The approach after anastomosis was wait-and-see in 72 patients (54.1%), diverting ileostomy in 19 patients (14.4%), and ghost ileostomy in 42 patients (31.5%). Regarding the baseline characteristics (Table 1), there were no differences in diagnosis, age, body mass index, ECOG, histologic subtype, tumor grade, FIGO stage or type of surgery. Patients in the diverting ileostomy group had a higher ASA score than the other two groups (p=0.02). In patients with relapsed ovarian cancer involving the rectum, a colorectal resection was not previously performed in the primary surgery. A laparotomy was performed in all cases. No differences were noted in history of diabetes, smoking status, steroid use, pelvic radiotherapy, albumin level (<3.5 g/dL), distance (<5 cm) from anastomosis to anal verge, additional bowel resection, or operative time. Additional procedures were performed to achieve an optimal cytoreduction in the wait-and-see, diverting ileostomy, and ghost ileostomy groups as follows: omentectomy in 90%, 84.2%, and 88.1% (p=0.79); pelvic and para-aortic lymphadenectomy in 55.6%, 47.4%, and 69% (p=0.21); diaphragm stripping in 43.1%, 57.9%, and 59.5% (p=0.18); and splenectomy in 18.1%, 21.1%, and 38.1% (p=0.05), respectively. A higher rate of intra-operative transfusion (61.1% vs 78.9% vs 42.9%; p=0.02) and a higher estimated blood loss (446±224 vs 683±413 vs 557±318 mL; p=0.006) were found in the diverting ileostomy group. The rate of anastomotic leak was 5.6% after wait-and-see (4/72), 5.3% after diverting ileostomy (1/19), and 4.8% after ghost ileostomy (2/42); no differences were found between the three groups (p=0.98) (Table 2).

Table 1

Patients characteristics

Table 2

Anastomosis leak related factors

In two cases of the ghost ileostomy group, a loop ileostomy was created due to anastomotic leak. All patients with anastomotic leak in the wait-and-see group required intensive care unit (ICU) admission, but no patients in the diverting ileostomy or ghost ileostomy groups required ICU admission. A laparotomy was needed in all patients of the wait-and-see group for stoma formation (75% end colostomy), but no laparotomy was required for stoma formation in the ghost ileostomy group. Anastomotic leak resulted in the death of two patients (50%) in the wait-and-see group, while none died in the diverting ileostomy or ghost ileostomy groups (Table 3).

Table 3

Anastomotic leak

The median hospital stay was 11±8.7 vs 10±4.2 days for diverting ileostomy and ghost ileostomy, respectively (p=0.23) (table 4). There was no difference in the interval between surgery and the beginning of chemotherapy (42±11.3 vs 46±13.6 days; diverting ileostomy vs ghost ileostomy; p=0.32). The stoma reversal rate was 73.7% for the diverting ileostomy group and 100% for the ghost ileostomy group. There was no difference in time to reversal between the two groups (324±161 vs 356±days; p=0.39). The reasons for no reversal in the diverting ileostomy group were: disease progression in two patients, worsening general condition in two patients, and patient refusal in one patient.

Table 4

Stoma outcomes

A higher number of patients in the diverting ileostomy group developed complications (>30 days) compared with the ghost ileostomy group (78.9% vs 7.1%; p<0.01). Dehydration, due to a high output stoma (>1500 mL) causing an electrolyte imbalance, developed in four patients (21.1%) in the diverting ileostomy group and in one patient (2.4%) in the ghost ileostomy group (p=0.03). Acute renal failure developed in two patients (10.5%) in the diverting ileostomy group, but in no patients in the ghost ileostomy group (p=0.09). A higher incidence of subacute obstruction was found in the diverting ileostomy group (21.1% vs 2.4%; p=0.03). No differences were found regarding stomal stricture (p=0.09) or parastomal hernia (p=0.23). Most of the complications in the ghost ileostomy group occurred in the two patients who underwent loop ileostomy creation, except for one patient who developed subacute obstruction that resolved by removing the vessel-loop. The hospital readmission rate associated with stomal complications was 47.4% and 7.1% for diverting ileostomy and ghost ileostomy, respectively. When comparing diverting versus ghost ileostomy, bowel obstruction was the most common reason for re-admission (21.1%, 4/19 vs 4.8%, 2/42), followed by dehydration (21.1%, 4/19 vs 2.4%, 1/42) and acute renal failure (10.5%, 2/19 vs 0%, 0/42).

Discussion

In this study, ghost ileostomy was found to be a safe procedure that avoids the morbidity related to diverting ileostomy. Ghost ileostomy presents also the advantages of diverting ileostomy when it is compared with conservative management after colorectal anastomosis in ovarian cancer. To date, neither diverting ileostomy nor ghost ileostomy have been associated with a reduction in the incidence of anastomotic leak in ovarian cancer, but both could decrease its severity.3–11 The importance of ghost ileostomy is based on the advantages of the diverting ileostomy when it is converted to an active stoma without the potential diverting ileostomy drawbacks in patients without anastomotic leak.22 The use of diverting ileostomy has been traditionally considered in order to diminish the incidence of anastomotic leak after colorectal anastomosis. Nevertheless, we found no differences regarding the three different approaches (5.6% vs 5.3% vs 4.8%; p=0.98). Ideally, a multivariate analysis should be performed to elucidate the specific weight of stoma formation in the incidence of anastomotic leak in ovarian cancer. Unfortunately, the anastomotic leak incidence in the entire sample was 5.3%, representing a low incidence that limits the performance of a multivariate analysis of risk factors. However, no difference in the risk factors that could explain the anastomotic leak rate between the three approaches was found in the univariate analysis .

The use of diverting ileostomy in ovarian cancer remains a controversial issue because of the limited number of published articles. It has been reported that the use of diverting ileostomy may reduce the rate of anastomotic leak18; however, this has not been confirmed by other investigators.12 Ghost ileostomy has been reported as a reliable alternative to diverting ileostomy in ovarian cancer22 as it may decrease the number of patients in which diverting ileostomy is performed and, therefore, its consequences and the rate of associated morbidity.22 In a prospective observational study, patients who underwent rectal resection and diverting ileostomy were compared with patients in whom ghost ileostomy was performed.25 In the ghost ileostomy group, stoma related morbidity and length of hospital stay were lower and quality of life was better.30 In our institution, rectoscopy represents a new approach for early leak detection. We believe that the performance of a scheduled rectoscopy and serum parameters allows for early diagnosis of anastomotic leak in a subclinical stage compared with waiting for leak symptoms, potentially leading to a reduction of fecal peritonitis and its consequences.27 It is important to highlight that with the use of ghost ileostomy, up to 95.2% of stomas were avoided without increasing the morbidity or mortality in patients who developed an anastomotic leak. Ghost ileostomy seems to have several advantages compared with diverting ileostomy. From our point of view, its use is a preferable choice in case of low leak rate as it occurs in ovarian cancer. Notwithstanding, selective diverting ileostomy must co-exist with ghost ileostomy being performed in cases with increased risk factors. However, the indication of diverting ileostomy in ovarian cancer should be delineated in further studies.

A higher rate of stoma-related complications (>30 days) was found in the diverting ileostomy group when compared with the ghost ileostomy group (78.9% vs 7.1%; p<0.01). Dehydration, subacute obstruction, and peristoma skin irritation were associated with the use of diverting ileostomy. A unique set of complications has been associated with diverting ileostomy such as: prolapse, stricture, ileus, bowel obstruction, malnutrition, and electrolyte disturbances due to high output ileostomy.31 In gynecologic malignancies, 30- and 60-day re-admission rates have been reported as high as 34% and 16–21%, respectively, with dehydration being the most frequent complication. These complications were in patients in which diverting ileostomy was performed after emergency surgery (bowel obstruction, bowel perforation, treatment of anastomotic leak, or others), and only 18.9% of patients had a first diagnosis of ovarian cancer.19 The time interval from surgery to chemotherapy correlates with the incidence of recurrence and progression of disease and it may be influenced by these complications. In our study, no differences were found between diverting ileostomy and ghost ileostomy as it pertains to interval time to initiation of chemotherapy.

The reported re-admission rate associated with diverting ileostomy in patients with ovarian cancer ranges between 4.5–15%.12 18 In our study, diverting ileostomy was associated with a very high rate (47.4%) of stoma-related re-admission while patients who underwent ghost ileostomy had a lower rate of re-admission (7.1%). As previously reported, dehydration was the most common reason for re-admission (31.6%), followed by bowel obstruction (21.1%). The timing of diverting ileostomy closure may impact the time to the start of adjuvant chemotherapy.32 In colorectal cancer, the sooner the stoma closure is undertaken, the lower the rate of associated morbidities.33 In the setting of ovarian cancer, diverting ileostomy is not likely to be associated with delays in the initiation of chemotherapy,12 but may impact adherence to the chemotherapy schedule.34 Similarly, having a temporary stoma may impact patient quality of life.21 35 This is particularly relevant, given the fact that many patients with temporary ostomies will never have such an ostomy reversed. The stoma non-reversal rate in our study was 26.3%, higher than that reported in the literature (7.4–17.2%).12 18 34 The reasons for no reversal in the diverting ileostomy group were progression of disease, worsening of the patient's general condition, and patient refusal. The timing of stoma closure will continue to be a matter of debate, with some advocating stoma closure as soon as possible19 because of the related complications and delay in the initiation of chemotherapy.

To our knowledge this is the first study in which ghost ileostomy has been compared with other managements after colorectal anastomosis in ovarian cancer (wait-and-see or diverting ileostomy).

The limitations of our study include the small sample size, the retrospective nature of the analysis, and the fact that there were no strict criteria to determine one strategy versus another. The low incidence of anastomotic leak also limited our ability to perform a multivariate analysis in order to establish the role of the stoma in reducing the rate of leak. Tailoring the use of a stoma by using a predictive model of anastomotic leak risk may result in a decreased use of diverting ileostomy.14

In conclusion, passive management after colorectal anastomosis based solely on imaging, blood tests, and symptoms may lead to serious consequences. The use of ghost ileostomy avoids the drawbacks of diverting ileostomy and seems to have advantages over routine diverting ileostomy and wait-and-see approaches for ovarian cancer patients undergoing colorectal anastomosis.

Acknowledgments

The authors want to thank Ms Sara García Álvarez for her helpful review and support in medical English editing during the writing of this article.

References

  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
  13. 13.
  14. 14.
  15. 15.
  16. 16.
  17. 17.
  18. 18.
  19. 19.
  20. 20.
  21. 21.
  22. 22.
  23. 23.
  24. 24.
  25. 25.
  26. 26.
  27. 27.
  28. 28.
  29. 29.
  30. 30.
  31. 31.
  32. 32.
  33. 33.
  34. 34.
  35. 35.

Footnotes

  • Contributors All of the authors (VL, AS-M, BF, PP-I, LM, ÁG-G, MB and SD) of the present manuscript declare that there are no conflicts of interest and have actively participated in the work, providing input including: (1) substantial contributions to conception and study design, (2) drafting of the article, and (3) final approval of the version to be published.

  • Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.

  • Disclaimer The manuscript contains original material that has not been published previously and it is not under consideration by another journal. All of the authors of the present manuscript declare that there are no conflicts of interest and have actively participated in the study providing input including: (1) substantial contributions to conception and design, or acquisition of data, or analysis and interpretation of data; (2) drafting of the article, or provision of critical revision for important intellectual content; and (3) final approval of the version to be published.

  • Competing interests None declared.

  • Patient consent for publication Not required.

  • Provenance and peer review Not commissioned; externally peer reviewed.

  • Data availability statement Data are available upon reasonable request. All data relevant to the study are included in the article or uploaded as supplementary information.