Splanchnic hypoperfusion has an important impact on the treatment of critically ill patients in intensive care units, either as transient mesenteric ischemia triggering a gut-derived systemic inflammatory response syndrome (SIRS) or as mesenteric ischemic necrosis [
Intestinal fatty acid-binding protein (I-FABP) is a valuable marker for epithelial damage of the intestine and, therefore, for mesenteric ischemia. This small cytosolic protein with a molecular weight of approximately 15 kDa is expressed only in mature enterocytes. Due to its low molecular weight and its distribution at the tips of the intestinal villi, it is rapidly released into the circulation and excreted in urine in case of mucosal damage [
A number of clinical studies demonstrated the high sensitivity and specificity of I-FABP in detecting mesenteric ischemia in patients presenting with acute abdomen [
We conducted a prospective observational trial including critically ill patients in the ICU to address the following questions: Are I-FABP values different in patients with and without mesenteric ischemia at the time of initial clinical suspicion? Does the interval between the event which triggered ischemia and the I-FABP test influence the diagnostic value?
In the multidisciplinary ICU of Jena University Hospital, all patients suspected of having bowel ischemia were included immediately when tested positive for at least one of the following inclusion criteria:
Clinical signs (prolonged ileus after surgery and increasing need for vasopressors) Hyperlactatemia (>3.5 mmol/l or no drop below 5 mmol/l within 12 h after cardiac surgery) Signs of bowel ischemia found in endoscopy (ischemic mucosal or transmural necrosis) or computed tomography (pneumatosis intestinalis, occlusion of the superior mesenteric artery, NOMI).
These criteria were compiled from our own clinical experience and have not been restricted to proof of bowel ischemia by means of CT or endoscopy since clinical suspicion often raises due to unspecific signs like delayed bowel function, distension of the abdomen, or hyperlactatemia.
Patients in whom bowel ischemia was confirmed during laparotomy were assigned to group 1 (ischemia). Patients surviving more than 7 days after inclusion into the study without any proof of bowel ischemia were assigned to group 2 (no ischemia). Exclusion criteria were
A urine specimen was taken at the time of study inclusion (specimen “U0”). Blood was collected from the routine blood samples at the following time points: day of inclusion (“d0”), day before inclusion (“d-1”), and the following three days (“d1,” “d2,” and “d3”). No extra blood samples were taken for study purposes. Study samples were aliquoted and stored at −80°C until analysis.
If informed consent could not be obtained within 96 h after study inclusion, all aliquots of the respective patient were discarded.
Laboratory analyses (serum
All continuous variables were expressed as median with interquartile range (IQR), since Kolmogorov–Smirnov test showed non-normal distribution of all I-FABP values. The groups were compared by the Mann–Whitney
During a 9-month period, 2072 patients were treated on ICU. 82 patients met the inclusion criteria. Of 32 patients, informed consent could not be obtained within 96 hours after study inclusion. Six patients were excluded because less than three specimens were available since they died before day 1. One patient without confirmation of mesenteric ischemia died on day 5 and, therefore, could not be assigned to any of the groups. Data of 43 patients were analysed, 21 of them being assigned to group 1 and 22 to group 2. Patients’ characteristics and the initial reason for admission into the ICU are listed in Table
Characteristics of 43 patients at d0 and reason for admission on ICU.
Group 1 [ |
Group 2 [ |
|
|
---|---|---|---|
|
21 | 22 | — |
Male | 11 (52.4) | 14 (63.6) | 0.455 |
Age [years] | 67 | 65 | 0.770 |
Mean BMI | 28.37 | 28.07 | 0.369 |
Mean Apache 2 score | 22.4 | 20.6 | 0.214 |
On dialysis | 10 (47.6) | 5 (22.7) | 0.087 |
30 d mortality |
|
5 (22.7) |
|
Cardiac surgery | 8 (38.1) |
|
|
Including CPB | 4 (19) | 8 (36.4) | 0.206 |
Cardiogenic shock | 3 | 1 | — |
Embolism of SMA | 3 | 0 | — |
Abdominal surgery | 2 | 1 | — |
Vascular surgery | 2 | 1 | — |
Urological surgery | 2 | 1 | — |
Multiple trauma | 1 | 2 | — |
Brain surgery | 0 | 1 | — |
CPB: cardiopulmonary bypass; SMA: superior mesenteric artery; BMI: body mass index.
17 patients met one inclusion criterion (hyperlactatemia only:
Hyperlactatemia was the most frequent reason for study inclusion in group 2, whereas patients of group 1 frequently had signs of bowel ischemia in endoscopy or CT in addition to clinical suspicion. Only 52% of group 1 had elevated serum lactate levels compared to 96% in group 2.
All patients of group 1 underwent laparotomy. In 20 of them, a bowel resection was performed because of ischemia (partial or complete colon resection:
Urine I-FABP was measured in 36 patients. Seven patients had acute or chronic renal failure on the day of inclusion into the study. Median urine I-FABP was significantly different in both groups (1310 pg/ml in group 1 versus 227 pg/ml in group 2,
I-FABP of serum and urine samples.
Group |
|
Median |
IQR |
Minimum |
Maximum |
| |
---|---|---|---|---|---|---|---|
Urine | 1 | 16 | 1310 | 2391 | 47 | 23,631 |
|
2 | 20 | 227 | 1377 | 47 | 29,000 | ||
Subgroup |
1 | 6 | 2464 | 15,781 | 475 | 23,631 |
|
2 | 15 | 230 | 848 | 47 | 29,000 | ||
Serum day 0 | 1 | 21 | 213 | 930 | 47 | 29,000 | 0.460 |
2 | 22 | 109 | 179 | 47 | 420 | ||
Subgroup |
1 | 8 | 977 | 3150 | 47 | 29,000 |
|
2 | 14 | 82 | 338 | 47 | 420 | ||
Serum day 1 | 1 | 21 | 47 | 290 | 47 | 3294 | 0.100 |
2 | 22 | 268 | 484 | 47 | 3998 |
I-FABP at day 0. Comparison of urine and serum samples of both groups. The statistical significance was higher in samples taken 12 to 48 hours after the event that most likely triggered ischemia. Logarithmic scale.
Urine I-FABP at study inclusion (d0)
Urine I-FABP obtained 12–48 h after ischemia
Serum I-FABP at study inclusion (d0)
Serum I-FABP obtained 12–48 h after ischemia
In all 43 patients, serum samples of d0 were analysed. Median I-FABP was not significantly different in both groups (213 pg/ml in group 1 versus 109 pg/ml in group 2,
In 40 patients, it was possible to identify the event which most probably triggered mesenteric ischemia (major surgery
Most I-FABP tests performed within 48 hours after the ischemia-triggering event led to correct or false positive results, as shown in Figure
Accuracy of the I-FABP test and its dependence on the interval to the triggering ischemic event, based on all d0 serum and urine samples. Among tests performed later than 48 hours after the event, the number of false negative results increased.
Sensitivity, specificity, and AUC were calculated for I-FABP levels at day 0 with analyses of all 43 patients and of the subgroups having their study inclusion 12 to 48 hours after the triggering event.
Table
Key figures of the diagnostic value.
Urine | Subgroup urine | Serum |
Subgroup serum | |
---|---|---|---|---|
Sensitivity | 81.3% | 100% | 33.3 | 75% |
Specificity | 70% | 73.3% | 95.5 | 100% |
AUC | 0.694 | 0.856 | 0.565 | 0.853 |
Cutoff [pg/ml] | 402.2 | 402.2 | 410.3 | 410.3 |
Making the diagnosis of mesenteric ischemia is difficult in ICU patients. Previous studies showed promising results with I-FABP in patients with acute abdomen [
For the detection of bowel ischemia, urine I-FABP had the highest diagnostic value with a sensitivity and specificity of 81.3 and 70%, respectively. Serum samples of day 0 (study inclusion) showed no statistically significant difference of I-FABP levels in both groups. This may be partially explained by the fact that, according to the study protocol, serum measurements were performed from the daily routine samples which were not always taken at the time of study inclusion.
However, the exact point in time when the mesenteric ischemia occurs is often obscure in ICU patients since clinical signs of acute abdomen are frequently masked. I-FABP is abundant only at the tips of the villi of bowel mucosa and rapidly released into the circulation in case of severe mucosal ischemia [
The time interval from mesenteric hypoperfusion to I-FABP measurement therefore seems to significantly influence the diagnostic value of the test and may lead to false negative (interval too long) or false positive (interval too short) results. Therefore, one single I-FABP measurement in ICU patients at the time of clinical suspicion failed to reliably detect or exclude mesenteric ischemia in our study. It can be assumed that repeated I-FABP measurements in critical situations like major surgery or cardiogenic shock may better reflect severity of intestinal damage and its recovery. For this clinical applicability, availability of a rapid laboratory assay is needed.
van der Voort et al. evaluated I-FABP test performance in 44 ICU patients with suspected mesenteric ischemia [
Symptoms of mesenteric ischemia in ICU patients are often unspecific. Thus, we applied a wide range of inclusion criteria for this study leading to a balanced recruitment to both groups. The retrospective analysis shows that more than 80% of patients of group 1 had signs of ischemia in endoscopy or CT and a large part of group 2 patients was included because of hyperlactatemia. However, in 12 of 17 patients of group 1 who were positive in endoscopy or CT, clinical signs or hyperlactatemia lead to suspicion of mesenteric ischemia and subsequently to further diagnostic measures. In 5 patients, the diagnosis of mesenteric ischemia was based on findings in CT scans that were performed without prior suspicion of mesenteric hypoperfusion. For this reason, we do not think that the unequal distribution of inclusion criteria to both groups influenced the results.
Mesenteric ischemia was confirmed in all patients of group 1 by surgical and histological findings. The definition of the absence of mesenteric ischemia is difficult in general since a comprehensive examination of the entire gut is almost impossible in vivo. Thus, our definition of the negative control group was patient survival of at least 7 days without confirmation of mesenteric ischemia, presuming that a mesenteric ischemia would be fatal within this period. However, some patients of the control group may have had mesenteric hypoperfusion that was resolved and, furthermore, surviving a significant mesenteric ischemia for more than 7 days seems occasionally possible.
The diagnostic value of I-FABP was higher using urine samples as compared to using blood samples. However, critically ill patients in ICU frequently suffer from renal insufficiency impairing both serum I-FABP clearance and urine I-FABP levels. Due to the small heterogeneous sample size, we were not able to investigate the effect of renal function or dialysis on the test results.
Another factor that might influence I-FABP levels to a more or less relevant degree is surgery or manipulation of the bowel. In our study population, 12 patients in both groups had some form of abdominal surgery prior to inclusion to the study (e.g., bowel resection, liver or pancreas resection, and aortic or renal surgery). We did not exclude these patients, because this pilot study aimed to evaluate the performance of the I-FABP test in the real world population of ICU patients. Again, the sample size was too small to perform a subgroup analysis of patients with and without surgical bowel manipulation.
In ICU patients, one single I-FABP measurement at the time of clinical suspicion failed to reliably detect or exclude mesenteric ischemia. A high diagnostic value of I-FABP was only confirmed in the time interval between 12 and 48 hours after the onset of mesenteric ischemia. I-FABP may be used most appropriately in perioperative monitoring. Further studies should consider the influence of the interval between mesenteric ischemia and I-FABP measurement on the diagnostic accuracy of the test.
Area under the curve
Body mass index
Cardiopulmonary bypass
Cardiopulmonary resuscitation
Computed tomography
Enzyme-linked immunosorbent assay
Intensive care unit
Intestinal fatty acid-binding protein
Lactate dehydrogenase
Nonocclusive mesenteric ischemia
Septic inflammatory response syndrome
Superior mesenteric artery
Transcatheter aortic valve implantation.
The authors report no proprietary or commercial interest in any product mentioned or concept discussed in this article.
The authors declare that they have no conflicts of interest.
Stefan Ludewig, Rami Jarbouh, and Henning Mothes made substantial contributions to the conception and design of the study. Stefan Ludewig, Rami Jarbouh, Michael Ardelt, Henning Mothes, Falk Rauchfuß, and René Fahrner permanently participated in screening patients for and including them into the study. Stefan Ludewig, Falk Rauchfuß, and Utz Settmacher contributed to the statistical analyses. Stefan Ludewig, Falk Rauchfuß, René Fahrner, Michael Ardelt, Jürgen Zanow, and Utz Settmacher were involved in drafting and revising the manuscript. All authors gave the final approval to all parts of the manuscript.
The authors wish to thank Cora Richert for laboratory assistance and Karin Jandt for editorial services.