Acute mesenteric ischemia (AMI) is a potentially life-threatening condition because of its diagnostic difficulty, operative challenges, and comorbidities a patient may have. The aim of this study was to identify factors associated with adverse outcomes in patients with AMI.
The hospital records and clinical data of all patients with AMI were reviewed for a recent 4-year period. Clinical outcomes and factors influencing mortality were analyzed.
Included in the study were 104 patients (46 females and 58 males) with an overall mean age of 66 ± 13.4 years. The cause of AMI was arterial pathology in 74 (71%) patients, venous thrombosis in 15 (14%) patients, and nonocclusive ischemia in 12 (12%) patients. Abdominal pain was the most common presenting symptom (97% of patients). The 30-day mortality rate was 66%. Univariate analysis showed that mortality was associated with renal insufficiency (p = 0.004), an age greater than 70 years (p = 0.02), the presence of comorbidities (p = 0.001), a leukocyte count greater than 18,000/mL (p = 0.04), and small bowel necrosis of more than 100 cm (p < 0.0001). Logistic regression analysis showed that independent predictors of mortality were small bowel necrosis of more than 100 cm (p = 0.002) and a serum creatinine level greater than 2 mg/dL (p = 0.04).
The length of the necrosis and renal insufficiency are the primary factors that result in a poor outcome in AMI patients. Prompt diagnostic evaluation and early therapeutic interventions may help to prevent the development of these fatal predictors.
intestinal necrosis;mesenteric ischemia;renal failure
Acute mesenteric ischemia (AMI) is an uncommon surgical emergency and is often diagnosed at a late stage. It is associated with a mortality rate of 60–80%, which has remained unchanged for the past few decades. These poor outcomes are primarily because of the difficulty of early diagnosis and the rapid nature of the ischemic tissue deterioration.1
Acute mesenteric ischemia can be classified primarily as ischemia of thrombotic origin or ischemia of nonthrombotic origin. Nonocclusive mesenteric ischemia results from low-flow states, whereas thrombotic conditions include arterial embolism, arterial thrombosis, and mesenteric venous thrombosis.2 Early diagnosis, resection of the infarcted bowel, restoration of blood flow, second-look laparotomy, and supportive intensive care are the basis of adequate management; however, the existing literature for the most part focuses on a relatively small number of patients.3
Despite the dismal prognosis, accurate presurgical and perioperative assessment of the risk of death in patients with AMI is poorly defined. Knowledge of preoperative and perioperative risk factors and estimating mortality would aid physicians, patients, and patients' families in decision-making and treatment planning. For this reason, researchers have sought to identify factors that may indicate an ominous prognosis and justify aggressive initial intervention for this condition.3; 4 ; 5 In the literature, there are unfortunately scant data evaluating these factors. The objective of this study was to determine clinical variables that could have an impact on the perioperative mortality of AMI.
The hospital charts and clinical records of all patients who underwent surgical intervention for AMI at Erciyes University Medical Faculty (Kayseri, Turkey) were evaluated with regard to demographic information, presenting symptoms, comorbid medical conditions, history and physical examination findings, laboratory and radiographic examinations, surgical procedure, hospital stay, and treatment outcome. The records of the radiologic examinations were abstracted and findings suggestive of bowel ischemia such as bowel wall thickening, ileus, pneumatosis intestinalis, or an occlusion in the superior mesenteric artery (SMA) or superior mesenteric vein (SMV) were classified as positive radiological findings. The time delay from the onset of the symptoms to surgery was also evaluated, and was categorized as less than 1 day or more than 1 day. Based on clinical, radiological, surgical, and histopathological findings, patients were diagnosed as having a mesenteric embolus, arterial or venous thrombosis, nonocclusive mesenteric ischemia, or an undetermined etiology. Operative findings with regard to small bowel necrosis were first divided into two groups: (1) patients with small bowel necrosis and (2) patients without small bowel necrosis. They were further divided into four groups: (1) patients without any necrosis; (2) patients with total necrosis; (3) patients with small bowel necrosis of more than 100 cm; and (4) patients with necrosis of less than 100 cm. For statistical analysis, patients with total necrosis were added to patients with small bowel necrosis more of than 100 cm. Patients with total intestinal necrosis underwent exploration alone because it was believed to be incompatible with survival.6 ; 7 The decision to perform a second-look operation was based on clinical findings of marginally viable bowel during the initial operation. The factors associated with survival were analyzed. Mortality was defined as mortality 30 days after the surgical procedure.
For statistical analysis, the associations between mortality rate and risk factors were assessed univariately with a two-sample t test, the Wilcoxon rank sum test, and the χ2 test or Fishers exact test. A multivariate logistic regression analysis model using JMP software (SAS, Cary, NC) was constructed with factors that were significantly associated with survival rate, based on univariate analysis results. Data are expressed as the mean ± the standard error of the mean. The test results were considered significant at p < 0.05.
From January 2008 to December 2011, 104 patients underwent surgical interventions for AMI at Erciyes University (Kayseri, Turkey). There were 58 men and 46 women (mean age, 66 ± 13.4 years; age range, 32–83 years). Sixty-nine patients died as a result of AMI during the perioperative period. Of the 69 deceased patients, 23 patients had total small bowel necrosis and no surgical procedure was performed. They died during the first 3–96 hours. In addition, 25 patients had widespread intestinal necrosis and only three of these patients survived. The cause of AMI was arterial pathology in 74 (71%) patients, venous thrombosis in 15 (14%) patients, and nonocclusive mesenteric ischemia in 12 (12%) patients. In three patients, we were unable to determine the exact cause. They were observed to have shock with worsening acidosis and were taken immediately to the operating room for exploration. Acute abdominal pain was the most frequent symptom [97 (93%) patients], followed by nausea. Most patients (73 patients) had acute abdominal pain for a median time of more than 24 hours. Plain abdominal radiography was performed in 96 (92%) patients, abdominal computed tomography (CT) scan in 87 (84%) patients, mesenteric Doppler ultrasound in 5 (5%) patients, and mesenteric angiography in 2 (2%) patients. The time delay from the onset of symptoms to surgery was less than 1 day in 27 (26%) patients and more than 1 day in 77 (74%) patients. All patients underwent an operation. Twenty-six patients underwent a second-look operation; among these patients, 21 patients had additional bowel necrosis that necessitated an additional intestinal resection in 19 of them, but no further resection was necessary in the remaining two patients because of total bowel necrosis. A vascular mesenteric procedure (i.e., mesenteric embolectomy in all patients) was performed in only eight (8%) patients. Univariate analysis revealed that mortality was significantly associated with an age of greater than 70 years, a leukocyte count greater than 18000/dL, a creatinine level greater than 2 mg/dL, the presence of at least one comorbidity, and small bowel necrosis of more than 100 cm (Table 1). On multivariate analysis, small bowel necrosis of more than 100 cm and a creatinine level greater than 2 mg/dL were independent predictors of survival (Table 2).
|Number of patients||Mortality, n (%)||p|
|<70 y||52||29 (56)||0.02|
|≥70 y||52||40 (77)|
|+ for ischemia||73||48 (66)||0.3|
|− for ischemia||23||14 (61)|
|<2 mg/dL||63||35 (56)||0.004|
|≥2 mg/dL||41||34 (83)|
|Length of necrosis|
|<100 cm||37||17 (46)||<0.0001|
|≥100 cm||36||26 (72)|
|<1 d||27||17 (63)||0.3|
|≥1 d||77||52 (68)|
|Age ≥ 70 y||0.66||1.02|
|Creatinine level ≥ 2 mg/dL||0.04||2.4|
|Length of necrosis ≥ 100 cm||0.002||5.6|
|Leukocyte count ≥ 18,000/dL||0.086||1.08|
Acute mesenteric ischemia is an urgent life-threatening condition with high complication rates and poor overall survival, regardless of the therapeutic approach. Acute mesenteric ischemia is a prevalent pathology in the elderly population, and it is becoming more frequent as the mean age of the population increases.8 In this study, which focused on the clinical outcome of patients who underwent surgical procedures for AMI, mesenteric arterial thrombosis was the most common cause of AMI. The 30-day postoperative mortality of the patients was 66.3%. The most frequent cause of death was multiorgan failure, which resulted from a combination of bowel ischemia and related sepsis.
A significant portion of our patients in their anamnestic data had previous nonspecific initial symptoms of intestinal angina, although most patients had peritonitis with findings of a gangrenous intestine during the abdominal exploration that necessitated bowel resection or that rendered life impossible with total necrosis (93%). Early recognition of AMI is crucial before the development of irreversible bowel necrosis; however, nonspecific initial symptomatology unfortunately makes diagnosis at an early stage challenging.9 We believe that patients who have previously experienced ischemic abdominal pain generally wait with the hope that it will resolve itself and only go to the emergency department if they are becoming worse. Thus, it is nearly impossible to determine the exact time of the onset of symptoms in these patients. Therefore, in contrast to the medical literature, which suggests early intervention as life-saving, our analysis showed that the time delay to the surgical procedure was not statistically significant, when based on survival.
In the medical literature, age is associated with poor prognosis in these patients.10 ; 11 In our series, elderly patients aged 70 years or older were associated with a greater mortality rate, compared to patients under 70 years of age (p = 0.02). However, in multivariate analysis, this difference was not significant (p = 0.66), and therefore age alone should not be considered a contraindication for surgery.
Acidosis has been associated with perioperative mortality in many reports.3; 12 ; 13 We found similar results; acidosis was a significant predictor of mortality in univariate analysis (p = 0.03). However, we did not analyze it in the logistic regression model because of two limitations: (1) it was interrelated with almost every factor such as renal failure, length of necrosis, comorbidity, symptom duration; and (2) in only the most affected patients of our series was an arterial blood sample obtained.
A notable association has been reported between survival and comorbidities, especially a previous history of cardiac disease.14 ; 15 More than one-half (69%) of our patients had at least one comorbidity. Hypertension was the most frequent comorbidity, followed by coronary artery disease (31 patients), and diabetes mellitus (21 patients). Twenty-five patients had three or more comorbidities. An association between hypertension, coronary artery disease, and diabetes mellitus was present in nine patients. These patients with three comorbidities postoperatively had severe difficulty in maintaining their blood pressure and glucose level within the normal range, and they died during hospitalization. Our findings were similar: the survival of patients with at least one comorbidity was 24%, whereas the survival of patients without any comorbidities was 56%. This difference was statistically significant in univariate analysis (p = 0.001).
Laboratory abnormalities are generally present in AMI patients.16 ; 17 In this study, elevated white blood cell count and serum creatinine levels were statistically significant factors of death in univariate analysis. Because the sensitivity and specificity of the leukocyte count in acute abdominal pathology are relatively low, we found it was not statistically significant in multivariate analysis. In the medical literature, a high serum creatinine level is reportedly an important determinant of death in AMI patients.4 ; 18 We obtained similar results in univariate and multivariate analyses. These analyses demonstrated that renal function impairment is a significant risk factor for perioperative mortality in these patients and highlight the importance of normal kidney function. Treatment strategies must not focus only on early surgical intervention. They should also focus on conserving renal function by proper fluid replacement and by avoiding contrast agents or drug toxicity. Incomplete resuscitation eventually leads to the development of volume depletion, metabolic acidosis, and renal failure.
Once the diagnosis of AMI is determined, a surgeon is faced with multiple therapeutic challenges. The goal of surgery is to assess intestinal viability, remove necrotic bowel, and minimize and prevent further bowel infarction.5 Because the clinical features of AMI originate from local and systemic responses, which are induced by the degree and length of impaired microcirculation, damage to microcirculation leads to irreversible intestinal necrosis locally or leads to systemic inflammatory response syndrome (SIRS) in the whole body or in remote organs systematically. The damaged intestinal microcirculation loses its resistance to bacteria and to water, which lead to endotoxemia or bacteremia. Bacterial translocation has an important role in the development of SIRS, adult respiratory distress syndrome, cardiac dysfunction, and eventually multiple organ failure.16 Exploratory laparotomy or laparoscopy is still the gold standard for determining bowel viability and surgery is the only way to remove necrotic bowel.8 Most of our patients had intestinal necrosis at the time of the initial operation, and this necessitated bowel resection. Embolectomy for patients with mesenteric embolism was performed through the root of the small bowel mesentery in 8% of patients. All eight patients who underwent embolectomy had intestinal necrosis with a time delay from the onset of symptoms to surgery being greater than 1 day. Four patients had small bowel necrosis of more than 100 cm and six patients had at least one comorbidity. Among these patients, five patients died and three patients survived (i.e., 62.5% mortality). All patients who survived had small bowel necrosis of less than 100 cm. With regard to the mortality rate, patients who underwent the surgical embolectomy procedure had a similar outcome as patients who did not undergo embolectomy, although this finding was too small for statistical comment. We did not perform any bypass procedure because all operations were performed in an emergency setting in patients in whom most had severe atherosclerotic vessels, multiple comorbidities, and massive necrosis. Preoperative noncontrast CT imaging that revealed disseminating atherosclerotic plaques in the major arteries was a good indicator for the lethality of an emergent bypass procedure because of the factors listed previously.
Second-look procedures allow bowel viability and further bowel resection to be reassessed, as needed.19 ; 20 Twenty-six patients with questionable bowel viability underwent a second-look procedure within 24 hours. The high percentage (81%) of additional bowel necrosis observed in the abdominal re-explorations warranted the need for a close follow up of these patients and supported the liberal use of second-look operations. In this study and surprisingly to us, the presence of bowel necrosis was far below statistical significance (p = 0.3). We instead identified the length of bowel necrosis as the most significant factor for perioperative mortality with an odds ratio of 5.6. Proximal vascular involvement (and consequently more expansive involvement) in patients with distal small bowel and colonic involvement, and the larger and denser variety of microbiologic flora of the terminal ileum and colon may be some factors that contribute to this result. Preoperatively detecting the exact length of intestinal necrosis may improve the prediction of outcomes in patients with AMI. Biphasic CT with mesenteric CT angiography offers the ability to exclude other causes of abdominal pain. It can also assess arterial narrowing or occlusion and show associated changes in bowel wall thickness, pneumatosis, and mucosal or bowel wall enhancement patterns that support the diagnosis of ischemia or necrosis with a high positive (100%) and high negative (96%) predictive values. 6 ; 21 However, the medical literature lacks data concerning its accuracy in detecting the exact length of necrosis before laparotomy.
Parameters associated with increased mortality included an age of greater than 70 years, the presence of comorbidities, renal insufficiency, elevated leukocyte count, and the length of intestinal necrosis. Further analysis of this data with logistic regression showed that renal insufficiency and the length of necrosis were independent predictors of mortality. Knowledge of the independent prognostic factors for mortality such as the development of renal failure and the involvement of a long segment of bowel could help improve the treatment algorithms of this challenging disease. The development of new diagnostic and therapeutic approaches supporting proper fluid and drug management to maintain renal perfusion and function, and efforts to preserve as much small bowel as possible may decrease the high mortality associated with AMI.