Incidence of and survival after surgical intervention for bowel obstruction in women with advanced ovarian cancer

Authors: Kolbrún Pálsdóttir, Sahar Salehi, Hemming Johansson, Nina Groes‐Kofoed, Henrik Falconer, Ulrika Joneborg

Abstract

Introduction

Women with advanced ovarian cancer commonly present with peritoneal disease both at primary diagnosis and relapse, with risk of subsequent bowel obstruction. The aims of this study were to assess the cumulative incidence of and survival after intervention for bowel obstruction in women with advanced ovarian cancer, to identify factors predictive of survival and the extent to which the intended outcome of the intervention was achieved.

Material and methods

Women diagnosed with advanced ovarian cancer stages III and IV in 2009–2011 and 2014–2016 in the Stockholm‐Gotland Region in Sweden were identified in the Swedish Quality Registry for Gynecologic Cancer. Through hospital records, types of intended and executed interventions for bowel obstruction were assessed, and as well as when in the course of oncologic treatment, the intervention was performed. Time from first intervention to death was analyzed with survival methodology and proportional hazard regression was used.

Results

Of 751 identified women, 108 had an intervention for bowel obstruction. Laparotomy was the most prevalent intervention and was used in 87% (94/108) of all women, with a success rate of 87% (82/94). An intervention for bowel obstruction was performed before or during first line treatment in 32% (35/108) with a cumulative incidence in the whole cohort of 14% (108/751, 95% confidence interval [CI] 11–16). Median survival after intervention for bowel obstruction was 4 months (95% CI 3–6). The hazard of death increased when the intervention was performed after completion of primary treatment (HR 4.46, 95% CI 1.61–12.29, P < 0.01), with a median survival of 3 months. In women subjected to radical surgery during primary treatment, the hazard of death after intervention for bowel obstruction decreased (hazard ratio [HR] 0.54, 95% CI 0.32–0.91, P = 0.02).

Conclusions

Women with advanced ovarian cancer undergoing intervention for bowel obstruction have a dismal prognosis, regardless of which line of oncologic treatment the intervention was performed. In the majority of women an intervention for bowel obstruction was performed in a relapse situation with an even worse survival. Our findings emphasize the importance of a holistic approach in the decision‐making before an intervention for bowel obstruction in women with advanced ovarian cancer.

Key messageIn a population‐based cohort of women with advanced ovarian cancer, the cumulative incidence of intervention for bowel obstruction was 14%. Median survival after intervention was only 4 months and the hazard of death increased after interventions performed after primary treatment.

INTRODUCTION

A large proportion of women with epithelial ovarian cancer have disseminated disease with widespread peritoneal carcinomatosis at time of diagnosis. Complete cytoreductive surgery in combination with adjuvant chemotherapy constitutes first line treatment. ^1^ , ^2^ The vast majority of women with seemingly successful primary treatment, in complete remission, will experience relapse of disease within 2 years. ^3^ , ^4^ The abdominal cavity is the most prevalent location of disease, with subsequent risk of obstruction at different levels of the gastrointestinal tract due to peritoneal carcinomatosis. ^5^ Recurrent epithelial ovarian cancer is non‐curable and treatment aims to prolong life with maintained optimal quality of life. ^6^ , ^7^ There is limited observational data describing the incidence, management and palliative considerations in women with ovarian cancer affected by bowel obstruction despite a suggested incidence of 17%–28%. ^8^ , ^9^ , ^10^ Surgical interventions to ameliorate symptoms of bowel obstruction may include laparotomy (enteric bypass with anastomosis, ostomy, adhesiolysis, gastrostomy), endoscopic percutaneous gastrostomy or bowel stent. ^5^ , ^11^ There is a lack of knowledge about which women are more likely to benefit from a surgical intervention for bowel obstruction as opposed to best supportive care. ^12^ , ^13^ Moreover, there are no previous studies to suggest outcomes after an intervention for bowel obstruction pertaining to when in the course of oncologic treatment, intervention of the bowel obstruction was performed.

The aim of our study was to describe the spectrum of bowel obstruction in women with advanced epithelial ovarian cancer irrespective of indication or intent by assessing the incidence of and survival after an intervention for bowel obstruction. Moreover, we aimed to examine how successful the performed intervention was, and to identify clinical factors that might predict a favorable survival outcome.

MATERIAL AND METHODS

This was an observational and descriptive cohort study. Karolinska University Hospital is the only tertiary referral center in the region of Stockholm‐Gotland in Sweden. Since 2004, all women with suspected epithelial ovarian cancer in the Stockholm‐Gotland region (2.4 million inhabitants), irrespective of which primary treatment the individual patient is eligible for, are referred to this hospital. Public cancer care is available to all residents in Sweden and no privately funded cancer care, by insurance or else, is available.

Patients

The patients included in this study were from the Stockholm Ovarian Cancer Project (STOOVCA), a registry‐based observational cohort study. ^14^ Women with International Federation of Gynecology and Obstetrics (FIGO) stages III and IV, epithelial ovarian/fallopian tube/peritoneal cancer and cancer in the abdomen of unknown origin (epithelial ovarian cancer without a biopsy specifically from the adnexae), diagnosed in 2009–2011 and 2014–2016, reported to the Swedish Quality Registry of Gynecologic Cancer – string ovary and controlled against the National Cancer Registry with 100% coverage, constituted the study population. Details on the registries, topographic and morphologic ICD codes, validation of data and how the final dataset was established are presented in a previous publication. ^14^ The two cohorts were compiled into one for this study and potential differences regarding primary surgical effort and outcome were accounted for in the results.

Additional variables for the present study were retrieved by review of hospital charts and the in‐hospital operation‐planning database Orbit. All women in the STOOVCA database were cross‐linked with our institutional operation‐planning database and computerized hospital records and matched for any of the following International Classification of Diseases (ICD) K56 (bowel obstruction), R10 (abdominal pain), JAH (explorative laparotomy), JFC (bowel anastomosis), JFF (bowel stoma), JFK (adhesiolysis in abdomen), JDB (gastrostomy). In women identified with any of the aforementioned ICD codes, a thorough review of hospital charts was performed to select only the women in whom an intervention for bowel obstruction was performed. The survival status of each patient was updated October 30, 2020, and controlled against the Swedish Population Registry.

Outcome variables

Survival was reported from the date of first intervention for bowel obstruction to death.

Successful intervention for bowel obstruction was defined as accomplishment of the intended outcome of the intervention.

Independent variables

The predictor variable investigated was intervention for bowel obstruction, defined as either an attempt or performance of laparotomy, percutaneous endoscopic gastrostomy or endoscopic stenting of bowel.

Confounding variables were predefined and chosen based on known clinical association with survival; age (<65 vs ≥65), disease stage (FIGO IV vs III), treatment at primary diagnosis (surgical vs non‐surgical), complete resection there was surgical treatment (yes vs no), when in the course of the oncologic treatment the intervention was performed (before primary treatment, during primary treatment or after primary treatment was completed), interval between diagnosis and intervention (<1 year, 1–2 years, >2 years), time of day when the intervention was performed (daytime 00–16:00 hours or on‐call 00–08:00 hours) and whether a gynecologic oncologist was involved in the decision on intervention (yes vs no).

Statistical analyses

Descriptive statistics are presented as numbers, proportions, medians, and minimum to maximum where appropriate. Survival time was calculated from the date of intervention for bowel obstruction to the date of death or last date of follow‐up. Median follow‐up was estimated using the reversed Kaplan–Meier method. Cumulative incidence functions, taking the competing risk of death into account, were used to estimate the probability of intervention for bowel obstruction over time. Overall survival was estimated and graphically illustrated using the Kaplan–Meier method. Proportional hazards regression was used to estimate the univariate and multivariate effect of clinical factors on time to death. Results from these models are presented as hazard ratios (HR) together with 95% confidence intervals (CI). The P‐values reported from these models refer to Wald's test. The significance level was set to 5% and all reported P‐values are two‐sided. All statistical analyses were performed using the statistical software STATA version 16.

Ethics statement

The study was regarded as quality assurance and deemed exempt from review by the Regional Ethics Committee in Stockholm (reference number 2016/1233‐31/4) on August 3, 2016.

RESULTS

The median follow‐up of all women was 75 months (minimum 47, maximum 141). Of the 751 women identified with advanced epithelial ovarian cancer and assessed for eligibility, 341 were coded with an ICD diagnosis potentially conformable with bowel obstruction or intervention for bowel obstruction and were enrolled in the study (Figure 1). After review of hospital records, 155 of these women were identified as hospitalized with suspected bowel obstruction, 47 of whom were assessed as having a bowel obstruction without any intervention and were therefore excluded, leaving 108 women for inclusion in the final analysis (Figure 1). The characteristics of the included women, treatment and interventions are presented in Table 1. At the time of primary diagnosis, the majority of women had FIGO stage III disease (68.5%, n = 74) and were treated with cytoreductive surgery (87.0%, n = 94); in the majority, complete macroscopic resection was not achieved (59.6%, n = 56). The median age at time of intervention for bowel obstruction was 66 years (minimum 44, maximum 93). In most women, the intervention was performed after completion of primary treatment (67.6%, n = 73). Laparotomy was the most commonly performed intervention (87%, n = 94), followed by an equal distribution of endoscopic bowel stent and percutaneous gastrostomy (11.1%, n = 12). The majority of interventions were performed during office hours (74.5%, n = 79) and by general surgeons (79.6%, 86) (Table 1). The success rates using laparotomy, endoscopic bowel stent or percutaneous gastrostomy were 87%, 58% and 66%, respectively. The median length of hospital stay was 14 days, and the majority of patients (70%, 76/108) were able to return home.

The probability of experiencing a first intervention for bowel obstruction before death within 5 years of diagnosis was 14% (95% CI 11–16) (Figure 2). The median survival after intervention for bowel obstruction was 4 months (95% CI 3–6) (see Figure 3).

The effect of each chosen variable on survival after an intervention is presented in Figure S1. The unadjusted survival was improved if complete macroscopic resection (0 mm residual disease) was achieved in the primary surgery (P = 0.04), if the intervention was performed prior to completion of primary treatment (P < 0.001), if the intervention was performed <2 years after primary diagnosis (P = 0.01), if a gynecologic oncologist was included in the discussion leading to an intervention or performed the intervention (P = 0.02) (Figure S1). In the adjusted analysis, an intervention for bowel obstruction performed after completion of primary treatment increased the hazard of death (HR 4.46, 95% CI 1.61–12.29, P < 0.01) with a median survival of 3 months; if complete resection was achieved at time of surgery during primary treatment, the hazard of death was reduced (HR 0.54, 95% CI 0.32–0.91, P = 0.02) (Table 2).

DISCUSSION

Our study suggests a very poor prognosis in women with advanced epithelial ovarian cancer in whom a surgical intervention for bowel obstruction was performed, irrespective of indication or intent. An increased hazard of death was evident if the surgical intervention was performed after completion of primary treatment, in contrast to women in whom complete macroscopic resection was achieved at primary debulking surgery.

The hallmark of dissemination in ovarian cancer is exfoliation to the peritoneal cavity with extensive shedding to the peritoneum. For this reason, cytoreductive surgery in the primary situation often includes resection(s) of the gastrointestinal tract. ^15^ , ^16^ Similarly, in the relapse situation the peritoneum is the most prevalent site of tumor manifestation with risk of bowel obstruction. The risk of hospitalization for malignant bowel obstruction after an ovarian cancer diagnosis has been reported as 17.6% in a large cohort from the Surveillance, Epidemiology and End Results (SEER) database. ^9^ The proportion of women subjected to a surgical intervention was 4.3 % (n = 373), which is much lower than the 14% in our study. In contrast to the aforementioned study, we selected only women in whom an intervention for bowel obstruction was performed and investigated bowel obstruction and not only malignant obstruction during the course of the disease from diagnosis to death. The reason to limit selection of women to those in whom an intervention was performed, was to ensure accurate estimates and assess outcomes in this particular group. We were not able to capture the true prevalence of women with bowel obstruction for two main reasons pertaining to our women hospitalized at other institutions are not accounted for and, in our setting, preterminal women receive end of life care in their homes without need of hospitalization.

The present study corroborates previous reports of median overall survival in women with malignant bowel obstruction only with a diagnosis of advanced epithelial ovarian cancer. ^13^ , ^17^ However, our study investigates the whole spectrum of interventions for bowel obstruction, not only those in women with disease relapse. Surprisingly, a third of included women had an intervention for bowel obstruction before or during primary treatment (32.4%). This group of women certainly represents non‐palliative patients in whom the intervention was performed as part of primary cytoreductive surgery when the patient presented with bowel obstruction, or to lyse adhesions during adjuvant chemotherapy, which is why a longer survival would have been expected. Our results indicate that an intervention for bowel obstruction may be a proxy for a generally poor prognosis in women with advanced epithelial ovarian cancer. Nevertheless, the majority of women in our study had completed primary treatment and had already received one or multiple courses of chemotherapy due to relapse or progression of disease at the time of the intervention. These women represent a palliative group with a malignant bowel obstruction and a subsequent 446% increased hazard of death and short survival.

In women subjected to complete macroscopic resection at the time of primary diagnosis, the hazard of death after an intervention for bowel obstruction decreased. However, it is noteworthy that among women subjected to an intervention for bowel obstruction, complete macroscopic resection at the primary situation was only 40.4%. This is a further indication that bowel obstruction requiring intervention is a symptom of an aggressive natural course of the disease in a group of patients with expected poorer prognosis. This may therefore be an additional variable to take into consideration when deciding on treatment of malignant bowel obstruction in women with advanced epithelial ovarian cancer.

Despite relatively high success rates for all interventions in our study, the survival was very poor. Moreover, the most commonly performed intervention was laparotomy. This choice may not be appropriate, at least not in the relapse situation, considering the poor survival and much increased hazard of death. The main purpose of management of malignant bowel obstruction is to maintain quality of life with effective symptom control. This could be achieved by non‐surgical methods, such as nasogastric tubes, intravenous hydration and palliative pharmacologic therapies. ^18^ Comparisons of treatment modalities are difficult due to lack of prospective trials. Although some reports have shown a survival benefit after surgical management, a systematic review comparing palliative surgery and medical management for malignant bowel obstruction was unable to reach definite conclusions. ^19^ , ^20^ , ^21^ The decision to perform surgical interventions in palliative circumstances is not easily made. The indication for surgery must be carefully weighed and requires detailed knowledge of the ovarian cancer diagnosis, treatment and the individual patient. Palliative intervention for bowel obstruction is often performed in an emergency setting and is associated with high postoperative morbidity and mortality. The length of hospitalization is considerable and re‐obstruction is common. ^22^ , ^23^ Corroborating our results, previous studies have suggested an increased survival after intervention for bowel obstruction if a gynecologic oncologist was involved in the treatment decision. ^24^ In patients with a short life expectancy, end of life care at home or in a hospice, with focus on symptom control and pain relief, could be the best option. Therefore, we propose that a discussion with a gynecologic oncologist is required before making a decision about surgical intervention, to optimize the care of women with malignant bowel obstruction.

Our study is limited by its retrospective and observational design, and unknown confounders are not accounted for. Moreover, the estimates might have been more precise with a larger sample size. In addition, as we investigated survival as an outcome measure in a predominantly palliative cohort, we cannot conclude the extent to which the intervention provided women provided palliation.

Strengths include its population‐based and centralized setting, which allowed the inclusion of all women in the population, with complete follow‐up on treatment in the primary and recurrent situation and on outcome, including details on clinical variables.

CONCLUSION

One‐third of women with advanced epithelial ovarian cancer who underwent intervention for any type of bowel obstruction were subjected to the intervention before or during first line treatment. However, in the majority of women, the intervention was performed in the relapse situation. Irrespective of when during oncologic treatment the intervention for bowel obstruction was performed, the prognosis was dismal. Management of women with advanced epithelial ovarian cancer with bowel obstruction remains a challenge and for this reason the decision on surgical intervention should be carefully considered.

AUTHOR CONTRIBUTIONS

KP, SS and UJ contributed to conceptualisation, data collection and manuscript writing. HJ was responsible for the statistical analysis. NGK and HF participated in the conceptualisation and manuscript writing. All the authors approved the final version of the manuscript for submission.

CONFLICT OF INTEREST STATEMENT

The authors have stated explicitly that there are no conflicts of interest in connection with this article.

Supporting information