Malignant Bowel Obstruction: A Retrospective Multicenter Cohort Study

Background: Malignant bowel obstruction (MBO) is a serious clinical entity that requires surgical intervention in almost 50% of cases. However, overall survival remains low even for operable cases. The aim of the present study was to investigate the correlation between patients’ characteristics, perioperative details, histopathological results and postoperative outcomes of patients who were operated on due to MBO. Methods: A retrospective search of patients who were operated on due to MBO in a university and a rural hospital was conducted. Patients’ characteristics, perioperative details, histopathological results and postoperative outcomes were reported. Univariable and multivariable analysis was performed. Results: Seventy patients were included with a mean age of 76.1 ± 10.6 years. The 30-day mortality rate was 18.6%, the Intensive Care Unit (ICU) admission rate was 17.1% and the mean length of stay (LOS) was 12.4 ± 5.7 days. Postoperative 30-day mortality was associated with increased age, known malignant recurrence, microscopically visible metastatic foci and defunctioning stoma creation. Colorectal malignancy type, sigmoid obstruction and primary anastomosis were correlated with decreased 30-day mortality. In addition, operation at the university hospital led to increased LOS, while stoma creation led to decreased LOS. Finally, ICU admission rates were increased for operations at university hospitals, at least one comorbidity, known malignant recurrence and longer preoperative waiting interval, whereas they were decreased for colorectal primary malignancy type. Conclusions: Surgery due to MBO leads to increased morbidity and mortality. Therefore, prospective studies are needed to highlight inter-patient differences regarding the best individualized therapeutic strategy.


Introduction
Colorectal cancer is the most common type of cancer of the gastrointestinal tract [1].Despite the large amount of available data, the increased awareness of screening and the improvement of treatment methods, the initial presentation of stage 4 colorectal cancer cases is not rare [2].A serious manifestation of advanced disease is malignant bowel obstruction (MBO), which is the most common indication for palliative surgical consultation [3].The definition of malignant bowel obstruction has not yet been completely clarified, but the most widely accepted criteria are (a) clinical evidence of bowel obstruction, physical examination, or radiographic examination, (b) obstruction distal to the Treitz ligament, (c) primary intra-abdominal incurable cancer, or (d) extra-abdominal primary cancer with peritoneal involvement [4].
Between 10 and 28% of patients with colorectal cancer and up to 50% of patients with ovarian cancer will develop malignant bowel obstruction at some point in their lives [5].
While most people with malignant bowel obstruction have distant metastases, 10% present an isolated metastatic lesion resulting in obstruction [6].Operable disease has a survival of 3-8 months, whereas inoperable cases present a survival of 4-5 weeks [7].Although ideal therapeutic management is still under controversy, surgical intervention could not be avoided in almost 50% of cases worldwide [8].
The aim of the present study was to investigate possible correlations between the preoperative characteristics of patients presenting with malignant bowel obstruction, as well as perioperative parameters and their histopathological reports with their postoperative outcomes, such as hospitalization interval, ICU admission and postoperative mortality.

Study Design and Participants
The present study included retrospectively enrolled patients who underwent emergent surgery due to malignant bowel obstruction at the ''Hippocration" General Hospital of Athens and the General Hospital of Sparta from 2017 until 2021.Inclusion criteria were age 18-90 years old, emergency admissions, urgent or emergent operations, bowel obstruction based on Computed Tomography (CT) with oral contrast, obstructing point at small or large intestine and preoperative or intraoperative diagnosis of gastrointestinal or other malignancy.Exclusion criteria were age below 18 or above 90 years old, inflammatory diseases leading to intestinal obstruction, bowel perforation and re-operations due to postoperative ileus.

Ethical Approvals
This study was approved by the Ethical Committees of ''Hippocration" General Hospital of Athens and General Hospital of Sparta.It was conducted in compliance with the Declaration of Helsinki guidelines about ethical principles for medical research involving human subjects.A written informed consent was obtained by all patients before participation in the study.The present manuscript has been prepared according to the Committee on Publication Ethics (COPE) guidelines.

Study Outcomes
A retrospective search of included patients' medical records was conducted after institutional approvals were obtained.Patient demographic data (age, gender, medical history, surgical history, type of hospital), patient information at the time of diagnosis (malignant disease status, location of obstruction, malignancy origin), operative details, intraoperative details and postoperative outcomes (ICU admission, length of stay, reoperation, 30-day mortality) were reported.In addition, histopathological analysis of specimens was conducted by two histopathologists independently and included information about the type of malignancy, its differentiation, the number of harvested lymph nodes and their infiltration rate, the surgical margins, Tumor Node Metastasis (TNM) staging, as well as perineural and perivascular infiltration.Postoperative mortality at 30 days was predefined as the main study outcome, whereas ICU admission and length of stay were considered as secondary outcomes.

Statistical Analysis
A power analysis was performed considering a power level of 80% and a significance level (alpha) of 0.05.The minimum sample size was 52 patients.
The Kolmogorov-Smirnov test was used to check the normality of distributions among quantitative variables.Mean values and standard deviations (SD) were used for normally distributed outcomes, while medians and interquartile ranges were used for abnormally distributed outcomes.Absolute (N) and relative (%) frequencies were used to describe qualitative variables.Comparisons of proportions were performed with Pearson's χ 2 test or Fisher's exact test.Comparisons of quantitative variables between the two groups were conducted with the Student's t-test or the non-parametric Mann-Whitney U test.Correlations between two quantitative variables were performed by the Pearson or Spearman correlation coefficient.
Stepwise linear regression analysis was conducted to investigate independent factors associated with length of stay, retrieving dependence coefficients (β) and their standard errors (SE).Independent factors related to ICU admission and postoperative 30-day mortality were investigated by logistic regression analyses that were performed with the stepwise inclusion/exclusion procedure and odds ratios along with their 95% confidence intervals (95% CI) were calculated.Significance levels were two-sided and the statistical significance level was set at p = 0.05.The statistical program SPSS 22.0 was utilized for the analysis.

Discussion
Malignant bowel obstruction (MBO) seems to remain an important healthcare issue, despite the increased screening programs for healthy individuals and meticulous followup protocols for already diagnosed cancer patients [9].The present retrospective study which included patients who were operated on due to malignant intestinal obstruction demonstrated a significant correlation between increased 30-day mortality and increased age, already known recurrence during oncologic follow-up, already known metastatic lesion at the site of obstruction, macroscopically visible metastatic foci intraoperatively and defunctioning stoma creation.On the other hand, lower 30-day mortality rates were observed in patients with colorectal cancer compared to other malignancy types, in patients who presented an obstruction point at the sigmoid colon and in patients who underwent a primary anastomosis.In addition, an operation at a university hospital was associated with an increased length of stay, while defunctioning stoma creation was associated with a decreased length of stay.Finally, increased ICU admission rate was correlated with operation at a university hospital, presence of at least patient comorbidity, already known recurrence during oncologic follow-up, already known metastatic lesion at the site of obstruction and longer preoperative waiting interval, whereas colorectal malignancy was associated with lower ICU admission rate.
Length of hospital stay due to malignant bowel obstruction has been described as about 7 days for conservative treatment, while it significantly increases (mean 11 days, range 6-18 days) for surgical management [10].In our study, the length of stay was 12.4 ± 5.7 days, which is in concordance with international data.Nevertheless, the length of stay was significantly different between operations at a university (13.6 ± 5.1 days) and a rural hospital (11.4 ± 6.1, p = 0.042).A multidisciplinary preoperative approach examining all treatment options, including non-operative ones, such as stenting or medical treatment, which are available in a university hospital could be a reason [11].In addition, preoperative resuscitation based on ICU protocols that would be available in university hospitals could lead to a longer hospital stay [12].Nevertheless, longer preoperative waiting interval was correlated to increased ICU admission rates (5.1 ± 3.5 days vs. 2.4 ± 2.7 days, p = 0.002) indicating the severity of such cases and the meticulous preoperative preparation that is necessary.
Intra-abdominal "oncologic load", which is expressed by an already known malignancy recurrence during oncologic follow-up or the presence of macroscopically visible metastatic foci intraoperatively, seems to significantly affect the mortality and morbidity of patients undergoing surgery due to malignant bowel obstruction.Our study indicated that already known recurrences, especially at the site of obstruction, and visible metastatic lesions during operation led to increased 30-day mortality and ICU admission rate.However, a 30-day mortality of 15.7% and an ICU admission rate of 16.3% is reported worldwide [13], which is in concordance with our results (18.6% and 17.1%, respectively).On the other hand, colorectal type as the primary malignancy seemed to be associated with decreased 30-day mortality and ICU admission rates compared to other malignancies, like gastric or pancreatic cancer.Nevertheless, a malignant intestinal obstruction with or without contemporary distant metastases, results in a 5-year survival of 10% and 40.9%, respectively [14].
Making the decision for surgical intervention in a patient with malignant intestinal obstruction is a multifactor procedure, during which a surgeon should examine the availability of alternative means of treatment, the nutritional condition of the patient, the preoperative clinical staging including features of excess "oncologic load" (ascites, extended omental infiltrations) and the possibility of urgent complications such as volvulus, ischemia and perforation [15].However, when the decision for surgical intervention is made, principles of surgical oncology should be followed when the patients' condition allows it and the operation could have a curative intent potential.The total number of lymph nodes in the specimen has been associated with overall survival and a number of at least 12 lymph nodes are necessary in order to achieve a precise N-status classification [16].In addition, a distance of at least 5-7 cm on either side of the tumor is recommended.Infiltrated surgical margins entail an increased risk of local recurrence, occurrence of late metastases, and reduced overall and disease-free survival [17].In the present study, the negative resection margin rate was 89.8% and the mean number of harvested lymph nodes was 18.6 ± 9.7, which provided an accurate N-stage histopathological assessment (N0 44.4%, N1 24.4%, N2 26.7% and the mean number of infiltrated lymph nodes 2.9 ± 5.3).
The medical condition of patients suffering from malignant intestinal obstruction is usually burdened due to the emergent situation of bowel obstruction and the advanced stage of their malignancy [18].Consequently, the decision for surgical intervention should be made meticulously, based on the operability of the disease.A curative intent surgical operation should be considered in patients with operable intra-abdominal disease, whereas a quick bypass of the obstruction or a defunctioning stoma creation should be followed in extended-malignancy cases [19].Nevertheless, our study indicated that a stoma creation was associated with increased 30-day mortality (p < 0.001), indicating severely ill patients that undergo a stoma creation.Therefore, non-surgical treatment options should be considered in such cases.The only randomized trial which is available by Krouse et al. demonstrated no difference in terms of good days out of hospital in patients with small bowel malignant obstruction who underwent surgical therapy compared to patients who received non-surgical management both for randomized and patient-choice groups [20].Endoscopic stent placement shows a high rate of symptom relief (64-100%), but there is a risk for perforation (0-15%), stent migration (0-40%) and re-occlusion (0-33%) [21].Medical options include anti-secretory agents (somatostatin analog, steroids, scopolamine), pain medications (morphine) and antiemetic therapy (haloperidol, prochlorperazine) target against symptoms with controversial outcomes [22].Under these circumstances, large prospectively designed clinical trials should be conducted in order to outline the potential benefits of surgical treatment and identify the patient groups that would benefit more.
The present retrospective study focuses on the surgical management and postoperative outcomes of patients with malignant bowel obstruction (MBO) due to several primary cancer types.To the best of our knowledge, this is the first study that includes patients with MBO due to several primary cancers and investigates their impact on postoperative outcomes.In addition, the predefined meticulous scientific design and the variety of investigated parameters were two advantages of the present study.On the other hand, its retrospective nature is an important disadvantage of this study.Another limitation could be its small sample size (n = 70).However, power analysis demonstrated that the minimum sample size was 52 patients.Therefore, the risk for type-I and type-II errors has been diminished.

Conclusions
Malignant intestinal obstruction is a serious clinical entity that usually needs surgical intervention, which is accompanied by high rates of postoperative mortality and morbidity.Therefore, a meticulous preoperative assessment considering the risks of intervention, the ideal preoperative waiting time and the operative strategy is indicated.Factors such as age, comorbidities, "oncologic load" and primary malignancy type, which affect postoperative course should also be taken into account.In advanced staged cases a multidisciplinary approach considering non-surgical treatment options should be followed.However, large prospective studies should be designed in order to highlight the benefits of different treatment options for different groups of patients.

Supplementary Materials:
The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/jcm13010263/s1,Table S1: The comorbidities of patients that underwent surgery due to malignant bowel obstruction; Table S2: The surgical history of patients that underwent surgery due to malignant bowel obstruction; Table S3: Histopathological characteristics of specimens after malignant bowel obstruction; Table S4: The primary malignancy type of tumors causing malignant bowel obstruction.

Table 1 .
Characteristics of included patients who underwent surgical operation due to malignant bowel obstruction.

Table 2 .
The postoperative outcomes of patients that underwent surgery due to malignant bowel obstruction.Length of stay (days), mean (SD) 12.4 (5.7) ICU, Intensive Care Unit.

Table 3 .
Univariable analysis of patients' characteristics, perioperative details and histopathological assessment with 30-day mortality after malignant bowel obstruction surgery.
The significant p-values are in bold.T, tumor; M, metastasis.

Table 4 .
Univariable analysis of patients' characteristics, perioperative details and histopathological assessment with length of hospital stay after malignant bowel obstruction surgery.

Table 5 .
Univariable analysis of patients' characteristics, perioperative details and histopathological assessment with ICU admission after malignant bowel obstruction surgery.
The significant p-values are in bold.ICU, Intensive Care Unit; T, tumor; M, metastasis.

Table 6 .
Multivariable analysis of patients' characteristics, perioperative details and histopathological assessment with postoperative outcomes after malignant bowel obstruction surgery.
The significant p-values are in bold.