Predictors of mortality following emergency open colectomy for ischemic colitis: a single-center experience

Background Ischemic colitis (IC) is a severe emergency in gastrointestinal surgery. The aim of the present study was to identify the predictors of postoperative mortality after emergent open colectomy for IC treatment. Additionally, we compared postoperative outcomes of patients undergoing emergent colectomy due to aortic surgery-related IC (AS-IC group) vs. other IC etiologies (Other-IC group). Methods We analyzed records of consecutive patients who underwent emergency open colectomy for IC between 2008 and 2019. Logistic regression analysis was performed to identify clinical and operative parameters associated with postoperative mortality. The AS-IC and Other-IC groups were compared for mortality, morbidity, ICU stay, hospital stay, and survival. Results During the study period, 94 patients (mean age, 67.4 ± 13.7 years) underwent emergent open colectomy for IC. In the majority of cases, IC involved the entire colon (53.2%) and vasopressor agents were required preoperatively (63.8%) and/or intraoperatively (78.8%). Thirty-four patients underwent surgery due to AS-IC, whereas 60 due to Other-IC causes. In the AS-IC group, 9 patients had undergone endovascular aortic repair and 25 open aortic surgery; 61.8% of patients needed aortic surgery for ruptured abdominal aortic aneurism (AAA). Overall, 66 patients (70.2%) died within 90 days from surgery. The AS-IC and Other-IC groups showed similar operative outcomes and postoperative complication rates. However, the duration of the ICU stay (19 days vs. 11 days; p = 0.003) and of the total hospital stay (22 days vs. 16 days; p = 0.016) was significantly longer for the AS-IC group than for the Other-IC group. The rate of intestinal continuity restoration at 1 year after surgery was higher for the Other-IC group than for the AS-IC group (58.8% vs. 22.2%; p = 0.05). In the multivariate model, preoperative increased lactate levels, a delay between signs/symptoms’ onset and surgery > 12 h, and the occurrence of postoperative acute kidney injury were statistically associated with postoperative mortality. Neither IC etiology (aortic surgery vs. other etiology) nor ruptured AAA was associated with postoperative mortality. Conclusion Emergency open colectomy for IC is associated with high postoperative mortality, which appears to be unrelated to the IC etiology. Preoperative lactate levels, > 12-h delay to surgery, and postoperative acute kidney injury are independent predictors of postoperative mortality.

The etiologies of IC are numerous, but they can be divided into two main categories: occlusive and nonocclusive diseases [4]. In both cases, IC is the result of blood supply insufficiency to the colon leading to various severities of ischemic lesions, from superficial mucosal lesions to full-thickness transmural necrosis [10]. The clinical presentation is related to the severity and extent of the ischemia lesions. The most frequent symptoms are abdominal pain, diarrhea, mild lower gastrointestinal bleeding, and elevated or persistent fever [4,11,12].
IC is considered as one of the most severe postoperative complications of aortic surgery, occurring in 2% of patients operated on in an elective setting and in 9% of patients operated on in emergency [13,14]. IC occurring after aortic surgery has been reported to increase mortality by 50% [15]. Several studies investigated the risk factors of IC after abdominal aneurysm (AAA) repair and compared the occurrence of IC after endovascular aortic repair (EVAR) vs. open aneurysm repair [16][17][18][19]. Older age, female gender, ruptured aneurysm, emergency surgery, prolonged surgery, aortic clamping duration, and previous cardiovascular history were associated with an increased risk of IC [5,15,17,20,21]. On the contrary, EVAR surgery was found to be protective towards IC compared to open aortic surgery probably due to lower ischemia-reperfusion injury and peripheral proinflammatory cytokine release [16,22,23].
Depending on the severity and IC grade, two treatment strategies can be discussed: a conservative approach or a surgical treatment [24,25]. IC patients who require surgical treatment are most of the time operated on in an emergency setting, which is associated with a higher risk of postoperative mortality than elective surgery [14,26,27].
The aim of the present study was to identify predictors of postoperative mortality in patients requiring emergency open colectomy for IC treatment. Additionally, we compared postoperative outcomes of patients undergoing emergency colectomy due to aortic surgery-related IC vs. other IC etiologies.

Methods
We scrutinized records of consecutive patients who underwent emergency open colectomy for IC at the Henri Mondor University Hospital of Creteil (France) between January 2008 and February 2019. The study protocol was approved by the Institutional Review Board (IRB: 00011558) and was conducted in conformity to the principles declared to the National Commission for Data Protection and Liberties (CNIL: 2210699) and in accordance with the ethical principles described in the Declaration of Helsinki.

Study population
To be included in the present analyses, IC patients should have been treated with an emergency open colectomy; IC diagnosis should have been confirmed preoperatively by computed tomography scan and colonoscopy and then by histology on the surgical specimen. At the contrastenhancement computed tomography scan, IC was identified as large bowel wall thinning ("paper-thin-wall"), or any large bowel wall thickening and/or abnormal wall enhancement. Porto-mesenteric venous gas, pneumatosis, free intraperitoneal air, and mesenteric arterial or venous thrombus may also be present [5,13,28,29]. Colonoscopy findings were classified according to Favier's classification [30].
According to the IC etiology, patients were divided into 2 groups: those with aorta surgery-related IC (AS-IC group) and those with IC of other etiologies (Other-IC group). IC in the AS-IC group should have occurred during the aorta surgery hospitalization. We collected baseline characteristics including demographics, comorbidities, IC etiologies and location, and the need of preoperative vasopressor agents (norepinephrine/epinephrine). We estimated the Ischemic Colitis Mortality Risk (ICMR) score as described by Reissfelder et al. [31]. Operative and postoperative variables were also retrieved and compared between the groups.

Emergency open colectomy procedure
The extent of the colonic resection depended on the severity of IC, based on preoperative and intraoperative findings, varying from segmental to total colectomies. After a midline laparotomy, the peritoneal cavity was explored. All necrotic, thickened, or edematous colonic segments were resected. The colon was transected in a macroscopically healthy area identified by the operating surgeon. For a left colon lesion, a Hartmann resection or a left colectomy, including the resection of the rectosigmoid junction, was performed. For IC located on the right colon, a right colectomy with resection of the right flexure and a double-barreled ileo-colostomy were performed. For ischemic lesions involving the right colon up to the splenic flexure, a subtotal colectomy was performed. For pancolic ischemia, a total colectomy with terminal ileostomy was performed. After surgery, all patients were admitted to the intensive care unit (ICU).

Study outcomes
The main study outcome was postoperative mortality, defined as in-hospital mortality or within 90 days from colonic resection. Secondary outcomes included intraoperative variables, postoperative morbidity, and ICU stay and hospital stay duration. The hospital stay duration was counted starting from the day of the emergency colectomy for IC. Acute kidney injury was defined according to the KDIGO classification [32]. Myocardial infarction was defined as isolated troponin elevation, electrocardiogram changes, and/or clinical evidence of heart attack. Respiratory complications were defined as postoperative occurrence of pneumonia or need for prolonged intubation (> 48 h). Severity of postoperative complications was categorized by using the Clavien-Dindo Classification [33]. Survivors were followed up; during the first year, restoration of intestinal continuity was performed if suitable.

Statistical analysis
Statistics were performed with tee SPSS software (Statistical Package for Social Science, IBM SPSS Statistics, version 23 for Macintosh; IBM Corp., Armonk, NY, USA). Continuous variables are expressed as mean (SD) or median (range), as appropriate. Categorical variables are expressed as frequency (n, %). Student t test, Mann-Whitney U test, Fisher exact test, or chi-square test was used for group comparisons, as appropriate. Logistic regression analysis was performed to identify clinical and operative parameters associated with postoperative mortality including the model variables that reached a p value < 0.2 at the univariate analysis. Odds ratios (OR) and adjusted OR (AOR) are presented with 95% confidence interval (CI). A p value ≤ 0.05 was considered as statistically significant.

Results
During the study period, 94 patients underwent emergency open colectomy for IC. The study population had a mean age of 67.4 ± 13.7 years, with 65.9% being women. In the majority of cases, IC involved the entire colon (53.2%), requiring a total colectomy with terminal ileostomy. The time interval from signs/symptoms onset to surgery was > 12 h in 75.5% of patients. The majority of patients needed vasopressors preoperatively (63.8%) and/or intraoperatively (78.8%) ( Table 1).
Operative and postoperative outcomes for the entire study populations are summarized in Table 2. Overall, 66 patients (70.2%) died within 90 days from surgery.

Comparison of AS-IC vs. Other-IC patients
The clinical characteristics of patients in the AS-IC vs. Other-IC group are presented in Table 3. Significant between-group differences were found only for the prevalence of hypertension, the location of ischemia and consequently the extent of colonic resection, and the use of open abdomen. Comparisons between the AS-IC group and Other-IC group for postoperative outcomes are presented in Table 4. After adjustment on the demographic and clinical variables that were significantly different between the AS-IC and Other-IC groups, the two groups showed similar rates of postoperative multiorgan failure, acute kidney injury, respiratory failure, and postoperative mortality (Table 4). However, the ICU stay (19 days vs. 11 days; p = 0.003) and the total hospital stay durations (22 days vs. 16 days; p = 0.016) were significantly longer for the AS-IC group than for the Other-IC group. The rate of restoration of the intestinal continuity at 1 year was higher in the Other-IC group than in the AS-IC group (58.8% vs. 22.2%; p = 0.05) ( Table 4).

Predictors of postoperative mortality
At the univariate analysis, lactate level > 2.5 mmol/L, time from signs/symptoms' onset to surgery > 12 h, need of preoperative and/or intraoperative adrenergic vasopressors, sepsis, and blood loss > 500 mL were found to be significantly associated with postoperative mortality (Table 5). In the multivariate model, preoperative lactate levels, time from signs/symptoms onset to surgery, and postoperative acute kidney injury were statistically associated with the occurrence of postoperative mortality. The presence of AAA rupture and the IC etiology (aortic surgery vs. other etiology) were not significantly associated with postoperative mortality (Table 5).

Discussion
The present single-center study showed that IC requiring an emergency open colectomy is associated with a high postoperative mortality, which can be predicted by high levels of preoperative lactates, delayed interval from signs/symptoms onset to surgery (> 12 h), and the occurrence of postoperative acute kidney injury. On the contrary, the specific etiology of IC does not impact on the outcome.
Prediction of postoperative outcomes and risk assessment can be of particular importance facing a pathology that requires an optimized emergency management. In Other etiologies 60 (63. 8) this perspective, some authors proposed different predictive scores aiming at guiding the clinical treatment choice and anticipating postoperative complications after colonic resection for IC [5,34,35].
In 2010, Chung et al. [35] proposed a prognostic model able to accurately identify severe IC requiring surgery or life-threatening based on 3 preoperative variables: tachycardia, shock within 24 h after admission, and endoscopic evidence of ulceration. The probability of severe IC was 1% if none of these 3 variables was present and was 74% if the 3 components were all present [35]. This score is of interest to assess the patient's risk once IC is diagnosed and prior to surgery. By providing a method to assess a patient's prognosis, it guides clinicians in the choice of the best management strategy and postoperative protocol, which must take into account the individual patient's risk.
In 2011, Reissfelder et al. [31] developed the ICMR score, a risk score to predict perioperative mortality in patients undergoing surgery for IC. In 2013, Castleberry et al. [34] assessed the validity of the ICMR score in their 10-year cohort composed of 115 patients who underwent colectomy for acute IC. They reported an adequate discriminatory accuracy in predicting in-hospital mortality with an area under the receiver operating characteristic curve of 0.75. When analyzed separately, the individual components of the ICMR score that performed better were the preoperative lactate levels > 2.5 mmol/L, the occurrence of postoperative kidney injury requiring hemodialysis, and the use of intraoperative vasopressor. On the contrary, the non-occlusive etiology of IC and the need of total/subtotal colectomy were not significantly associated with in-hospital mortality [34].
In accordance with the previous literature (Table 6), the present study confirms the predictive value of the preoperative lactate levels and the occurrence of postoperative acute kidney injury, whereas the need of intraoperatory adrenergic vasopressors and the ICRM score showed only a statistical trend. We also observed that the delay to surgery can significantly impact on the risk of in-hospital mortality, being those operated on after > 12 h from signs/symptom onset drastically more at risk than those receiving an early diagnosis and a prompt surgical treatment.
The definition of immediate and delayed surgery varies among the studies [8,20,31,36,37]. In some studies, as in the present one, immediate surgery was defined as the surgical treatment performed within 12 h from admission or surgical consultation, whereas delayed surgery as the intervention performed more than 12 h after admission owing to severe or worsening clinical conditions, or no response to conservative therapy [8]. As known, the clinical presentation of IC can be vague and unspecific, and the frequent presence of pre-existing medical conditions may hamper a timely diagnosis [4,26,38]. Severe cases at presentation have more chance to be rapidly taken in the operating room, but they are also associated with poorer outcomes due to the severity and extent of the ischemic lesion [4,5,8,15,26,31,36,39]. When comparing mortality rates after emergency colectomy for IC, it is thus important to take into account the type of procedure performed (e.g., segmental resection, total colectomy), the surgical setting (e.g., emergency or delayed surgery), and the patient's status (e.g., comorbidities, hemodynamic stability). Some authors advocated for an aggressive approach with extended colonic resections (e.g., left colectomy or subtotal colectomy) and only rarely limited to a segment (e.g., right colectomy) [3,10,13]. Indeed, acute IC located in the right colon should be considered as an extended colonic hypoperfusion with global hypo-perfusion in the mesenteric superior artery territory. However, an aggressive management is not supported by all surgeons. Paterno et al. [8] suggested that in case of limited IC extension and no hemodynamic instability, a segmental colon resection is  [15,16,20,26,40], with some differences related to the type of aortic surgery [14,22,40]. Indeed, a recent systematic review and metaanalysis demonstrated that IC occurs more frequently after open repair (2.1-3.6%) than in EVAR (0.5-1%) in the elective setting [22]. However, once IC occurred and an emergency colectomy is required, the mortality appears to be independent of the type of aortic surgery performed. The present study confirms these findings and suggests that there is no evidence of an impact on mortality of the specific IC etiology, namely the aortic surgery-related cause or other causes, including occlusive and non-occlusive diseases.
In the present study, survivors of open colectomy underwent a restoration of the intestinal continuity in almost 50% of cases, with no related postoperative mortality. Interestingly, the rate of intestinal continuity restoration within 1 year from surgery was higher for patients in the Other-IC group than in the AS-IC one. This difference may be explained by several factors, such as     the severity of comorbid diseases in AS-IC patients. Indeed, these patients showed most of the time cardiovascular comorbidities, which have detrimental effects on disease evolution and may be related to worse postoperative outcomes. In the study by Castleberry et al. [34], 24% of IC survivors underwent elective end-ostomy reversal with a mortality rate of 18%. Mariani et al. [41] described good outcomes after the restoration of intestinal continuity following colectomy for IC. This was possible in 40% of survivors, and the median delay between colectomy and the intestinal continuity restoration was 7.9 months (range 0.2-35 months). There was no postoperative mortality, but 13% developed surgical complications (e.g., lower digestive tract bleeding, wound   [41]. The authors did not find any clinical or biological discriminant factor that may help in identifying which patients should undergo reversal surgery. However, they supported that the restorati0on of intestinal continuity should be discussed in multidisciplinary meeting for every surviving IC patient, and although challenging, it should be considered to improve quality of life in selected, motivated patients [41]. In the same vein, further studies may explore the possibility to manage IC by emergency laparoscopy to obtain additional advantages in the postoperative period related to the minimally invasive approach [42,43]. However, this may be conceivable only in highly selected patients, depending on the IC severity and etiology (nonrelated to aorta surgery). The present study has limitations that must be considered when interpreting the results. It is a retrospective analysis of an 11-year cohort of IC patients requiring emergency surgery. Patients were treated in a single institution over a long period of time; although the health database was prospectively maintained, reporting and selection bias cannot be ruled out. Furthermore, the external validity of the findings is limited. However, the present study investigated a relatively large sample of IC patients all treated by open colectomy in an emergency setting. Studies in the literature are rather heterogeneous in methods and target populations, making direct comparisons difficult. Efforts should be made in future studies, both prospective and retrospective, to standardize outcome measures and preoperative evaluation, in order to facilitate study comparisons as well as a comprehensive appraisal of the body of literature.

Conclusions
Emergency open colectomy for IC is associated with high postoperative mortality. The levels of preoperative lactates, a delay to surgery > 12 h, and the occurrence of postoperative acute kidney injury are found to be independent predictors of postoperative mortality. The specific etiology of IC, namely aortic surgery complications or other causes, appears to have no impact on postoperative mortality. These results support the key role of prompt diagnosis and surgical intervention in the management of severe IC. Moreover, the present data indicate that survivors in the Other-IC group are more likely to undergo intestinal continuity restoration within 1 year from IC surgery compared to AS-IC patients, which may have a considerable impact on IC survivors' quality of life.  IC ischemic colitis, AAA aortic abdominal repair, AFib/RVR Atrial fibrillation with rapid ventricular response, EVAR endovascular aortic repair, GI gastrointestinal, IMA inferior mesenteric artery, nrAAA non-ruptured AAA, rAAA ruptured AAA, RCI right colonic ischemia, SMA superior mesenteric artery