Introduction

Nonocclusive mesenteric ischaemia (NOMI) is an acute mesenteric circulatory disorder that is not caused by an organic occlusion of blood vessels, in contrast to a mesenteric arterial occlusion, which is induced by a blockage of the blood flow by emboli and thrombi [1]. It may involve all abdominal parenchymas and viscera and the whole gastrointestinal tract (from the oesophagus to the rectum), insomuch that the involvement of the entire colon should be considered a distinctive tract in diagnosing this condition in respect to the occlusive forms of ischaemia. Ende et al. [2] described what are considered to be the first reported cases of NOMI in three patients with heart failure-associated low output in 1958. Many more cases have been described since then and today NOMI is reported to be responsible for approximately 20–30 % of all intestinal ischaemic events, although its incidence has increased due to an ageing society, the high number of cardiothoracic and major abdominal surgeries, and an increase in the dialysis population [1, 3].

Intestinal vasospasm due to persistent low perfusion is the principal cause of the ischaemic disorder in NOMI and therefore the principal aim of current therapy is to reduce the spasm and improve perfusion of the mesenteric artery using vasodilators. Conversely, surgery should be limited to the excision of irreversibly necrotised intestine, in particular when reperfusion, a very frequent event in NOMI, appears to be ineffective [1, 4, 5]. However, because of the uncertain clinical status and the fact that many NOMI patients are misdiagnosed, the prognosis of NOMI is extremely poor and its mortality rate has not changed over time [6, 7]. That being so, the role of imaging in diagnosing NOMI is crucial: firstly for the early identification of this critical condition, to plan the correct therapeutic approach, which is based on a different treatment from those of the occlusive forms of acute mesenteric ischaemia (AMI), then to recognise the presence of reperfusion, characterise its effectiveness, and guide the surgical excision of irreversibly necrotised intestinal tract [3, 8]. This review aims to increase awareness regarding the pathogenesis and the clinical, laboratory and imaging features of NOMI, and in particular to describe the CT findings of NOMI with and without a reperfusion event and to correlate these with pathological features.

Pathogenesis

Regarding pathogenesis, in NOMI bowel ischaemia and infarction can occur because of a reduction in the mesenteric blood supply without vascular occlusion due to an intestinal vasospasm for a persistent low perfusion. It usually develops during an episode of cardiogenic shock, septicaemia, dehydration or a state of hypoperfusion in which excessive sympathetic activity results in secondary vasoconstriction of the mesenteric arteries (superior and inferior), which is often superimposed on a pre-existing atherosclerotic plaque. NOMI also frequently occurs postoperatively in patients with hypotension following heart or major abdominal surgery. However, pharmacological factors (administration or abuse of digitalis, ergotamine or other vasoconstrictive agents) can also represent common causes of NOMI [1–3, 9, 10]. The ischaemic injury in NOMI may range from reversible superficial damage, localised to the watershed areas, to a more severe form that extends to the entire bowel, depending also if reperfusion occurs [11]. The clinicopathologic diagnostic criteria include no occlusion of the mesenteric artery or vein in the area of the bowel necrosis, the presence of ischaemic and necrotic spots and segments that are distributed over a wide area also in a nonconsecutive manner and histopathological findings including haemorrhagic and necrotic changes, as well as small veins without fibrin plugs [12]. The discontinuous segmental necrosis is a marked distinguishing feature of NOMI from occlusive mesenteric ischaemia/infarction, in which the mesentery and the intestine are involved and eventually necrotised from the site of the thrombus, forming a sphenoidal necrotic area in the region served by the artery or vein (continuous necrosis) [12].

Clinical aspects and laboratory findings

NOMI occurs more frequently in elderly patients who have associated comorbidities such as systemic atherosclerotic disease, congestive heart failure, cardiac arrhythmias, coronary artery disease, or hypovolemia and hypotension due to various causes [1–3]. Clinically, NOMI can be present with abdominal pain, nausea, vomiting, gastrointestinal haemorrhage and ileus symptoms, but the characteristic early symptoms and laboratory test results are unclear [13–15]. In particular, the detection of subjective symptoms is difficult in cases that develop and advance after surgery because the patient’s clinical condition and the effects of general and epidural anaesthesia may conceal the symptoms and disease may develop and advance unnoticed [3, 16]. Clinical symptoms may be accompanied by leukocytosis and acidosis, as evidenced by an altered pH and elevated serum lactate [13]. Different plasma biomarkers such as d-lactate, intestinal fatty acid binding protein (I-FABP), alpha-glutathione S-transferase and d-dimer have been tested in humans. The most promising plasma markers according to a recent review were I-FABP, alpha-glutathione S-tranferase and d-lactate [17–20]. I-FABP and alpha-glutathione S-transferase are both located in the mucosa of the small intestine, and d-lactate originates from bacteria in the intestinal lumen as a normal product of bacterial fermentation. As intestinal damage during AMI starts from the intestinal mucosa, all these markers seem to have the potential to be used as diagnostic tests in the early phases of intestinal ischaemia. In particular, in contrast to l-lactate, the d-isomer of lactate (d-lactate) is metabolised very slowly by the liver, and elevated levels may be an indicator of bacterial translocation, intestinal mucosal injury and, although not exclusively, AMI [17]. However, none of the proposed plasma-derived tests for AMI has, as yet, entered routine clinical practice because the performance of the currently available serological markers is suboptimal for routine clinical use. Furthermore, because the diagnosis of NOMI is clinically difficult to be suspected, the routine blood biomarkers are not always tested before imaging and often the tested biomarkers, like white blood cell (WBC) count or lactate dehydrogenase (LDH), may provide a normal result in particular in hospitalised patients, who often arrive for a CT examination during the early phase of NOMI. As a result, the clinical lesson about blood biomarkers should be that when a patient suffers from vague abdominal symptoms soon after a major cardiovascular or abdominal operation, even if the blood markers, such as an elevation of lactic acid, LDH, and/or CK are absent, NOMI should be considered as one of the presumed causes [3, 16].

Diagnostic imaging

Because early diagnosis is difficult, today NOMI represents the most lethal form of AMI with a mortality rate ranging from 70 to 90 % [6]. Recognition of a NOMI from an occlusive infarction or from other causes of acute abdomen is paramount as the condition requires different treatments [14, 15]. Indeed, medical treatment, and in recent years perfusion therapy using vasoactive substances (continuous administration of vasodilators such as papaverine, PEG1, and nitroglycerine into the mesenteric artery) has gained an important role in the successful treatment of NOMI, confining only the advanced forms with necrosis of the involved tract to surgery [4, 5]. In this context, imaging plays a pivotal role in the differential diagnosis between occlusive and nonocclusive forms of ischaemia [4, 5, 8, 21–24]. Despite the technical evolution of noninvasive methods, such as multi-detector CT (MDCT), many authors claim that angiography should still be considered as the gold standard diagnostic technique which reliably establishes an early diagnosis of NOMI by demonstrating the narrowing of the branches of the superior mesenteric artery as described by Boley et al. (the so-called “string-of-sausages” sign: alternate dilation and stenosis of the superior mesenteric arterial branches), spasm of the intestinal marginal artery, and poor contrast enhancement of veins into the muscular layer, as the many features of vasospasm associated with NOMI [25–28]. However, NOMI often occurs in patients with poor or unstable systemic conditions, and an angiography may not be possible due to its complexity and invasiveness; however, today MDCT can provide high-quality images that are comparable to an angiogram with less invasiveness and less time [29, 30]. Furthermore, as recently stated in the ACR appropriateness criteria on imaging of mesenteric ischaemia, published by Oliva et al. [31], angiography allows diagnosis and treatment but MDCT with intravenous contrast media presents a higher rating for various reasons. Firstly, MDCT represents a fast noninvasive study that also evaluates other causes of abdominal pain and secondly it is a fast and noninvasive diagnostic tool for evaluating both vascular CT findings, like the narrowing of the superior mesenteric artery (SMA) with poor demonstration of their branches, and nonvascular CT findings, such as abnormalities in the bowel wall, mesentery and peritoneal cavity and not only restricted to the mesenteric vessels [31]. At the time of writing, there are only a few articles relating to the effectiveness of MDCT for the diagnosis of NOMI and some with only a small case series (four patients) [4, 5, 29, 30]. Moreover, these authors underline the effectiveness of MDCT for the diagnosis of NOMI by assessing the morphology and diameter of the SMA on multi-planar reconstructed images, comparing it to an angiography examination. Nevertheless, the great advantage of MDCT is the possibility to also show the bowel wall abnormalities that often guide the treatment of the patient (surgical or nonsurgical). MDCT is also the best diagnostic modality to investigate possible reperfusion events, which are more frequent in NOMI than in the occlusive type of AMI; as such, also the appearances of NOMI at imaging could be different depending on these reasons and on the timing of the CT examination [23, 24].

CT examination

MDCT imaging protocol and suggestion for imaging evaluation

Oral and rectal administration of contrast material is not recommended (or useful) for accurate CT examination and assessment of acute bowel ischaemia, as it does not add any additional information for the final diagnosis, but also because of the severe clinical condition of the patient and the need to keep the investigation time to a minimum. CT images should be obtained from the dome of the liver to the perineum to cover the entire course of the intestine. The examination should be performed with a spiral technique in the cranio-caudal direction (from the base of the lungs to the pelvic brim) and with the patient in the supine position. With MDCT scanners, the following technical parameters could be used: effective slice thickness of 3.75 mm for plain acquisition and 1–2.5 mm for contrast-enhanced CT (both late arterial, 45–50 s, and portal venous phase, 70–80 s), beam pitch of about 1, reconstruction interval of about half or less than half of the effective slice thickness to improve the multi-planar 2D reconstruction and the CT angiography analysis; tube voltage of 120–140 kVp and reference mAs of 250–500 mAs according to the patient’s body mass index. Automatic tube current modulation was used to minimise radiation exposure, even though this is not clinically important considering the median age of the patients. A standard reconstruction algorithm should also be used. Patients were instructed not to breathe during helical imaging to avoid motion artefacts, when they were compliant. Acquisition of both unenhanced and contrast-enhanced CT scans is always necessary. The main roles of unenhanced CT are to identify a possible high attenuation of the bowel wall (reperfusion condition or intramural haemorrhage) or to obtain a baseline attenuation measurement of the bowel wall for the assessment of its enhancement. It also could be useful for monitoring the patient after medical treatment in the case of reperfusion, evaluating the reduction or increase of free intraperitoneal fluid or of the dilation of the bowel lumen, and for detecting signs of bowel wall necrosis like the presence of intestinal pneumatosis [32]. On the other hand, the main roles of contrast-enhanced CT are: to make a differential diagnosis with occlusive ischaemia (by identifying thrombi in the mesenteric arteries and veins) or other causes of acute abdomen (including ischaemic changes in bowel obstruction), to evaluate the degree of bowel wall contrast enhancement (absent in the case of necrosis), and the presence of hypoperfusion phenomena of other organs. Contrast-enhanced scans should be performed in the late arterial phase (start delay 45–50 s) and in the portal venous phase (start delay 70–80 s) with an intravenous injection of 2 ml/kg of nonionic contrast material followed by 40 ml of saline solution using an automated power injector (3–4 ml/s flow rate), with an 18-gauge needle in the antecubital vein. It is always recommended to visualise the CT images with different window and level settings. In fact, a window set on the soft tissue values (width, W: 300–350; level, L 40–50) could demonstrate alterations of the bowel wall, abdominal organs, mesentery and vascular structures; the window setting used to visualise the lung parenchyma (W: 450–100; L: 100–0) will aid the recognition of extraluminal gas [21]. In cases of NOMI, 2D multi-planar reconstruction (MPR) and maximum intensity projection (MIP) are useful for evaluating the morphologic appearance and diameter of the superior mesenteric artery, by searching for the same angiographic criteria addressed by Siegelman et al. [27] for the diagnosis of mesenteric vasospasm: (1) narrowing at the origins of the major branches with involvement of segments of the superior mesenteric artery; (2) irregularities at the origins of the major branches of the superior mesenteric artery (beading sign or string-of-sausage sign); and (3) spasm of the arcades of mesenteric artery, and impaired filling of intramural vessels.

CT findings

Mesenteric vessels

No sign of SMA, inferior mesenteric artery (IMA) or superior and inferior mesenteric vein (SMV/IMV) thrombosis or occlusion were found at contrast-enhanced CT examination. In the early phase of NOMI, CT appearances regarding the arterial findings are superimposable to those of angiography. In particular, the original axial images and MPR images show irregular narrowing of the SMA and IMA, spasm of the arcades of mesenteric arteries, and impaired filling of intramural vessels. In our personal case series (25 NOMI cases), the mean value of the SMA diameter of NOMI cases was statistically smaller than that of controls [4.6 ± 1.4 mm (range 2.2–7.4 mm) versus 7.3 ± 1.0 mm (range 5.40–9.20 mm)], and this fact was confirmed both in the study of Woodhams (four NOMI cases) and in the study of Nakamura (11 NOMI cases) [3, 29, 30] (Fig. 1). However, the calibre of the SMA and IMA could be regular during and after the reperfusion event.

Fig. 1
figure 1

Vessel and mesentery computed tomography (CT) findings in the early phase of nonocclusive mesenteric ischaemia (NOMI): contrast-enhanced coronal image of abdomen in a 76-year-old man hospitalised for heart failure and with abdominal pain: a narrowing of the superior mesenteric artery (SMA) with poor demonstration of the branches of the arteries is shown (a); a reduction in calibre and number of mesenteric vessels is also shown in axial CT image (b)

Full size image

Mesentery

The most common feature described regarding the mesentery in AMI is mesenteric fat stranding. Mesenteric fat stranding and ascites appear with transudation of fluid in the mesentery or peritoneal cavity caused by an elevation of mesenteric venous pressure, which is commonly seen in strangulating bowel obstruction and veno-occlusive bowel ischaemia [23, 24, 33]. It is also frequently seen in ischaemic colitis because of a superinfection of the ischaemic colonic segments [34, 35]. Therefore, mesenteric fat stranding in these conditions can appear without bowel infarction and has limited value in estimating the severity of bowel ischaemia. On the contrary, in patients with NOMI, as in arterial occlusive mesenteric ischaemia, mesenteric fat stranding appears only when reperfusion occurs or in advanced conditions of ischaemia with transmural infarction, when hypoperfusion results in increased vascular permeability that leads to extravascular leakage of plasma, red blood cells (RBCs), or both into the bowel wall, mesentery, and peritoneal cavity. In this sense, mesenteric fat stranding in NOMI can be helpful in estimating the severity of bowel ischaemia [23, 24, 36] whereas in the early phase of NOMI, the mesentery appears pale because of the reduction in calibre and number of mesenteric vessels (Fig. 1). In our case series, a strong correlation between the reperfusion event and mesenteric fat stranding (p = 0.026) was demonstrated. Other CT findings of mesentery are mesenteric fluid and mesenteric pneumatosis. The mesenteric fluid can be present in NOMI after reperfusion, and its increase has negative prognostic significance, as does mesenteric pneumatosis that represents a sign of late stage ischaemia (Fig. 2).

Fig. 2
figure 2

Mesentery and bowel wall CT findings of NOMI with an ineffective reperfusion event: Unenhanced CT coronal (a) and axial (b, c) images of the abdomen in a 49-year-old man with a history of surgical intervention for a type A aortic dissection, hospitalised for heart failure during a wedding party: both mesenteric fat stranding and free peritoneal fluid due to a reperfusion event are shown in (a); thickened loops with high attenuation of the wall is also shown in the axial CT image (b). A CT examination (c) performed 6 days later because of clinical worsening demonstrated increased wall thickening and its high attenuation (intramural haemorrhage), and increased fluid in the peritoneal cavity (c) due to an ineffective reperfusion event

Full size image

Bowel wall and calibre

Bowel wall CT findings in NOMI are similar to those observed in acute arterial ischaemia [36, 37]. In particular, in the early phase of NOMI or when reperfusion does not occur, the bowel wall appears paper-thin because there is no arterial flow or mural oedema, whereas bowel wall thickening (>5 mm) occurs only after a reperfusion event (Fig. 3). Another important CT finding is the degree of attenuation of the bowel wall. It should always be assessed on both unenhanced and contrast-enhanced CT. On unenhanced CT, it is possible to observe low attenuation of the bowel wall that indicates bowel wall oedema caused by a reperfusion event, or high attenuation of the wall that generally represents a condition of intramural haemorrhage or haemorrhagic infarction in the late phase of ineffective reperfusion. On contrast-enhanced CT, enhancement of the bowel wall can be absent or diminished in the ischaemic phase or during an ineffective reperfusion, whereas contrast hyperenhancement of the bowel wall is typical in the reperfusion phase (Fig. 4). The evaluation of attenuation differences (HU measurements) of the bowel wall between the unenhanced and contrast-enhanced CT in cases of reperfusion is interesting, in particular when there is a high attenuation of the bowel wall on unenhanced CT, because when the HU difference is low this could represent a sign of ineffective reperfusion (Fig. 5). From a pathological point of view, the bowel wall can become thickened in cases of reperfusion due to the effects of the reperfusion that are superimposed on the ischaemic damage by hypoperfusion, with oedema and thickening of the submucosa, marked vascular congestion, and granulation tissue [12, 36]. There are inflammatory cells with numerous neutrophils and nuclear detritus developing pseudomembranes. The remaining epithelium shows regenerative changes with denudation of villi, crypts lined by hyperchromatic cells that proliferate to replace the destroyed epithelium [12]. The superimposition of the effects of the reperfusion on the ischaemic damage, characterised by marked vascular congestion, could also explain the high attenuation of the bowel wall on unenhanced CT scans, which in the current literature is usually only associated with intramural haemorrhage and haemorrhagic infarction [23]. In our case series, a strong correlation between the reperfusion event and bowel wall thickening (p = 3.2, E-05) was demonstrated. There was also a strong correlation between the reperfusion events and both wall thickening and mesenteric fat stranding combined. Furthermore, the reperfusion event was significantly correlated with the absence of paper-thin wall (p = 0.049). The last finding that could be considered regarding the bowel in NOMI is its appearance in terms of dilation or contraction. A contraction of the bowel loops could be observed in the early phase of NOMI (spastic reflex ileus), whereas in the following phase the bowel lumen is often dilated because of an interruption of normal bowel peristalsis with only a gas-filled appearance (hypotonic reflex ileus), whereas in the late phase of NOMI both the fluid-filled appearance (paralytic ileus) and intestinal pneumatosis could be observed. Noteworthy, the bowel dilation in NOMI always involves the colon and this should be considered a distinctive finding in the differential diagnosis between NOMI and occlusive forms of AMI and also has prognostic significance because its increase represents a sign of late stage of ischaemia or an ineffective reperfusion event (Fig. 6) [38–40].

Fig. 3
figure 3

Bowel wall CT findings of NOMI with an effective reperfusion event: Contrast-enhanced CT axial (a, b) and coronal (c, d) images of the abdomen in a 74-year-old woman with a diagnosis of NOMI with reperfusion after a cardiogenic shock during major abdominal surgery: thickened loops (duodenum and jejunum) with mural oedema and mucosal hyperenhancement are shown at first CT examination (a, c); the CT examination performed after 5 days of medical therapy demonstrated a significant reduction of bowel wall mural oedema (b, d)

Full size image
Fig. 4
figure 4

Bowel wall CT findings of NOMI without an with reperfusion event: Contrast-enhanced CT axial (a) image of the abdomen in a 63-year-old man with a prolonged cardiac arrest during pacemaker implantation. Both a focal absence or reduction of contrast enhancement of the bowel wall and free peritoneal fluid were present (a); a surgical specimen demonstrated a discontinuous segmental necrosis (b): the patient survived the surgical procedure. Contrast-enhanced CT coronal (c) image of the abdomen in a 72-year-old man with acute abdominal pain during an acute heart failure. Both thickened loops with mural oedema and small bowel mucosal hyperenhancement are shown (c): a surgical exploration demonstrated signs of effective reperfusion (d): the patient received medical treatment

Full size image
Fig. 5
figure 5

CT evaluation of attenuation differences (Hounsfield unit measurements) of the bowel wall between the unenhanced and contrast-enhanced CT in a case of ineffective reperfusion event: there is high attenuation of the bowel wall without a significant HU difference between the unenhanced (a) and contrast-enhanced CT (b), which could represent a sign of ineffective reperfusion; a surgical specimen of that loop showed both signs of necrosis and reperfusion (c, d): the patient survived the surgical procedure

Full size image
Fig. 6
figure 6

Consecutive abdominal X-ray examinations (a–c) of a 84-year-old man with a history of chronic heart failure and acute abdominal pain: a progressive bowel dilation involving also the entire colon is shown. CT examination (d) performed immediately after the last X-ray examination demonstrated a colonic pneumatosis: the patient died 1 h after the CT examination

Full size image

Free peritoneal fluid and other abdominal cavity findings

Free fluid is frequently observed in NOMI. In our case series, it was present in 78.57 % of CT examinations whereas free air in the peritoneal cavity was observed rarely and as a consequence of bowel wall perforation owing to an ineffective reperfusion event. The increase of peritoneal free fluid, like mesenteric fluid, has a negative prognostic meaning as it occurs in the late stage of ischaemia or during an ineffective reperfusion event. Other signs of peritoneal cavity are a reduction of contrast enhancement of abdominal parenchyma, most common in the spleen, kidneys and liver, even though any other abdominal organ may be involved (Figs. 7, 8).

Fig. 7
figure 7

CT findings of peritoneal cavity (1): axial contrast-enhanced CT image of abdomen in a 64-year-old woman waiting for heart transplantation and hospitalised for abdominal pain showed hypoperfusion phenomena involving the spleen (a) and both kidneys (b)

Full size image
Fig. 8
figure 8

CT findings of peritoneal cavity (2): axial (a) and sagittal (b) contrast-enhanced CT images of the abdomen in a 76-year-old woman with a cardiogenic shock during major abdominal surgery. Focal absence of contrast enhancement of the gallbladder wall is present; a surgical specimen demonstrated a discontinuous segmental necrosis (c, d): the patient survived the surgical procedure

Full size image

Conclusion

An accurate and early diagnosis is essential for the appropriate and successful treatment of patients with NOMI to improve their prognoses. However, amongst various conditions of mesenteric ischaemia, NOMI is the most difficult to diagnose at imaging because it is often a dynamic condition owing to the reperfusion event. Today CT represents an essential diagnostic tool in this critical condition thank to the possibility of demonstrating both vascular and nonvascular findings; however, for a correct diagnosis, a technically appropriate CT examination and an accurate interpretation of images are mandatory. The evaluation of the CT findings of mesentery, bowel wall, and peritoneal findings, in addition to the vessel features, suggests a possible diagnostic utility for discriminating NOMI with and without reperfusion as well as a possible prognostic value (Table 1). In particular, drawing a distinction between NOMI with and without reperfusion may have a great clinical importance because these conditions may have different therapeutic approaches. Therefore, the introduction of MDCT in the decision tree of NOMI treatment may bring the benefit of an early identification of this critical condition, so as to plan the correct therapeutic approach—which is based on different treatments from those required by the occlusive forms of AMI—to recognise the presence of reperfusion and to characterise its effectiveness and to guide the surgical excision of irreversibly necrotised intestinal tract. Finally, MDCT allows the differential diagnosis with other painful abdominal conditions even if the diagnosis of NOMI should always be considered one of the presumed causes in a typical clinical setting.

Table 1 Typical computed tomography (CT) findings of nonocclusive mesenteric ischaemia (NOMI) with and without reperfusion event
Full size table