Exploring I-FABP, endothelin-1 and L-lactate as biomarkers of acute intestinal necrosis: a case-control study

Abstract Objective Acute intestinal necrosis (AIN) is a disease with devastating high mortality. AIN due to obstructed arterial blood flow has a blurred clinical presentation. Timely diagnosis is paramount, and a blood-based biomarker is warranted to increase patient survival. We aimed to assess intestinal fatty acid binding protein (I-FABP) and endothelin-1 as diagnostic biomarkers for AIN. To our knowledge, this is the first study exploring endothelin-1 in AIN patients from a general surgical population. Design We conducted a single-centre nested case–control study comparing acutely admitted AIN patients to age- and sex-matched non-AIN patients during 2015–2016. I-FABP and endothelin-1 were analysed using an enzyme-linked immunosorbent assay. L-lactate levels were also measured in all patients. Cut-offs were estimated using receiver operator characteristic curves, and the diagnostic performance was estimated using the area under the receiver operator characteristic curve (AUC). Results We identified 43 AIN patients and included 225 matched control patients. Median levels of I-FABP, endothelin-1 and L-lactate were 3550 (IQR: 1746–9235) pg/ml, 3.91 (IQR: 3.33–5.19) pg/ml and 0.92 (IQR: 0.74–1.45) mM in AIN patients and 1731 (IQR: 1124–2848) pg/ml, 2.94 (IQR: 2.32–3.82) pg/ml and 0.85 (IQR: 0.64-1.21) mM in control patients, respectively. The diagnostic performances of endothelin-1 and of I-FABP + endothelin-1 combined were moderate. Endothelin-1 alone revealed an AUC of 0.74 (0.67; 0.82). The sensitivity and specificity of endothelin-1 were 0.81 and 0.64, respectively. Conclusion I-FABP and endothelin-1 are promising biomarkers for AIN, with moderate diagnostic performance compared with the commonly used biomarker L-lactate. Preregistration ClinicalTrials.gov: NCT05665946.


Introduction
Acute intestinal necrosis (AIN) is a life-threatening disease caused by obstructed blood supply to the intestinal segment [1].This obstruction is either an intravascular obstruction, vascular AIN, extravascular compression, nonvascular AIN, or collapsed blood circulation, nonocclusive mesenteric ischemia (NOMI).NOMI was not further investigated in this study.The total lack of blood supply causes intestinal barrier disruption followed by leakage of endotoxins and intestinal bacteria followed by sepsis, multiorgan failure, and ultimately death [2].The diagnosis of extravascular compression due to a hernia or a volvulus is easily performed clinically or in abdominal computed tomography (CT) scans.Furthermore, vascular venous AIN is rare and chronic in nature and is easily diagnosed on abdominal CT.In contrast, vascular arterial AIN patients, referred to as 'vascular AIN' , suffer from delayed treatment due to the prolonged diagnostic process because of an unspecific clinical presentation.This might explain the devastating mortality in vascular AIN despite decades of research in blood-based biomarkers and the continuous development of increasingly sophisticated scanning modalities.A significant portion of vascular AIN patients appear without abdominal peritoneal signs and show an unspecific elevation in standard blood-based parameters shortly after the event [3].The imaging and radiological evaluation is compromised if the clinical suspicion of AIN is not raised in the initial diagnostic workup [4].Altogether, the violation of the prenecrotic timeframe of six hours before irreversible transmural necrosis might explain the high mortality in AIN patients [5].Due to the need for specific biomarkers, standard blood-based parameters, particularly L-lactate, have been widely used in the diagnostic workup of AIN [6].
L-Lactate, one of the two enantiomers of lactate, is a hydroxycarboxylic acid that is the end product of anaerobic glycolysis [6].Excess serum L-lactate is also observed in nonhypoxic conditions, e.g., during liver or kidney function impairment [7].L-lactate research contradicts the widespread use of L-lactate in the diagnostic process of AIN [6][7][8][9][10][11], including in a meta-analysis by Cudnik et al. [12] as well as a number of clinical guidelines [5,[13][14][15].Acosta et al. [8] and Van der Voort et al. [16] reported sensitivities of 52% and 78%, respectively, and Van der Voort et al. reported a specificity as low as 48%.
The focus of this study was to explore the diagnostic performance of two novel promising biomarkers, I-FABP and endothelin-1, related to the commonly used metabolic marker L-lactate.
The small epithelial molecular I-FABP is present in the cytosol of mature enterocytes throughout the small bowel and colon.I-FABP is involved in fatty acid transportation.The concentration in plasma is low in nonischaemic states but rapidly rises under inflammatory, ischaemic or necrotic conditions in the gastrointestinal tract [9].Elevation of serum concentration has been potentially seen within the first 60 min of obstructed blood supply and has been shown to fall after resection of a necrotic intestinal segment [17,18].The correlation between I-FABP concentration in plasma and the propagation of intestinal ischaemia/necrosis is described in Schellekens et al. [17].
Endothelin-1 is a small 21-amino acid peptide produced in endothelial cells with constricting function on blood vessels in the whole body and has been extensively researched with a focus on the kidney, brain, lungs and heart.Furthermore, endothelin-1 has been shown to be increased by hypoxemia in humans [19].In addition, endothelin-1 has been proposed to play a role in cell survival in some endothelial cells [20,21].Finally, endothelin-1 has been shown to be elevated before and after surgery in NOMI patients compared to controls [22], but to our knowledge, this study is the first investigation of endothelin-1 in vascular AIN and nonvascular AIN patients in a general surgical population.
The aim of this study was to explore the performance of two novel promising biomarkers, I-FABP and endothelin-1, compared to the commonly used metabolic marker L-lactate for diagnosing AIN.

Setting and study population
In this case-control study, all patients acutely referred to The Department of Gastrointestinal Surgery, Aalborg University Hospital, Denmark, between 6 January 2015 and 10 June 2016 were included.Additionally, in-hospital patients suspected to suffer from AIN were included from 10 June 2016, to 24 March 2019, referred to as 'high-risk patients' .AIN diagnosis was secured during surgery.Control patients were selected retrospectively by random, age and sex-matched among all acute non-AIN patients included in the study.Thus, inclusion criteria was every acutely referred patient, and in the high-risk patient group in-hospital patients suspected to suffer from AIN. Exclusion criteria was patients below 18 years.

Definitions
AIN was verified during surgery and defined as nonviable necrotic bowel demanding resection.Acute intestinal ischaemia is defined as visible pale bowel with an ischaemic appearance with a total recovery of viability intraoperatively.Vascular AIN is defined as the absence of a physical reason explaining the pathology eventually with diminished pulsation in the mesentery.Nonvascular AIN is defined as AIN with obvious reasons for compromised blood circulation, such as adhesions, herniation, or strangulation.Patients 'suspected to suffer from AIN' were in-hospital patients with abdominal pain requiring surgery and emerging or manifesting single or multiple organ failure.

Data collection
Findings during surgery, comorbidity, medicine use, tobacco use, previous surgery, and clinical findings were registered at the time of inclusion.

Blood sample collection and pre-analytical handling
Blood samples were collected after verbal and written consent just after admission in the non-high-risk patients or at the time of the clinical evaluation for surgery in the high-risk patients in accordance with the approval by The North Denmark Region Committee on Health Research Ethics.Blood samples for L-lactate analysis were collected in 3-ml Vacuette® FC-mix (FC) tubes (Cat # 454513; Greiner bio One International GmbH), and blood samples for endothelin-1 and I-FABP were collected in 9-ml serum tubes with clot activator (Cat # 455092; Greiner bio One International GmbH).Tubes were centrifuged at 2200 g at 20 °C for 15 min before storage at −80 °C.Since these research samples were handled in an acute routine laboratory setting, we designed a study to test the impact of worst-case scenarios regarding pre-analytical handling compared with the optimal workflow.Details of the stability test setup are described in Supplementary Materials 1.

Blood sample analysis
L-lactate was analysed on a Cobas® 8000 (Roche, Mannheim, Germany) in a routine laboratory.L-lactate controls (Roche, Mannheim, Germany) were run at two levels in quadruplicate for 5 days.The day-day coefficient of variation (CV) was 0.9% at both levels (1.71 and 3.77 mmol/l).Endothelin-1 was analysed using the Quantikine® ELISA endothelin-1 immune-assay (R&D, Abingdon, UK).Serum samples were run in duplicate essentially as described by the manufacturer.Assay controls in two levels (R&D, Abingdon, UK) and a serum pool were included in each run.In our lab, we found a mean CV on duplicates of 3.3%, interassay CV of 1.5-3.7%(levels 9.1 pg/ml and 3.8 pg/ml, low and medium controls), and 7.5% on the serum pool (level 2.3 pg/ml).I-FABP was analysed using the Quantikine® ELISA FABP2/I-FABP (R&D, Abingdon, UK).Serum samples were run in duplicate essentially as described by the manufacturer.Assay controls at three levels (R&D, Abingdon, UK) and a serum pool were included in each run.In our lab, we found a mean CV on duplicates of 3.0% and interassay CVs of 9.7% (high control, level 472 pg/ml), 8% (medium assay control, level 224 pg/ml), 8.4% (low assay control, level 73 pg/ml), and 9.7% on the serum pool (level 1608 pg/ml).Stability tests were performed in serum from 7 (L-lactate) and 10 (endothelin-1 and I-FABP) healthy controls.Stability tests demonstrated a significant positive bias of a median 10.7% for L-lactate tested under worst-case scenarios for pre-analytical handling.CVs of 12.4% for endothelin-1 (level 1.4 pg/ml) and 12.5% for I-FABP (level 1400 pg/ml) were found.Both are comparable to the serum pool CV and considered acceptable.

Statistics
Since detailed information concerning the mean and SD of the investigated variables are largely unknown in study populations comparative to ours, power calculation was unachievable and we decided to include age and gender-matched control patients in a ratio of 1:5.
Continuous variables are reported as the mean and standard deviation (SD) for normally distributed variables.For skewed data, the median and interquartile range are reported.Categorical variables are reported as numbers and percentages.Comparisons between AIN patients and controls are reported by differences in means or proportions with 95% confidence intervals (CIs).Confidence intervals and p values were calculated by generalised linear models with identity links and robust variance estimation.
The diagnostic performance of the biomarkers was assessed by receiver operator characteristic (ROC) curves and reported as the area under the curve (AUC) with 95% CI.Cut-off values were calculated by the youden index, and the corresponding sensitivity and specificity are reported.For the combined performance of I-FABP and endothelin-1, leave-oneout cross-validation was used for AUC to account for an overly optimistic estimate of diagnostic performance.The cut-off value per se is not interpretable and hence has not been tabulated.

Results
We included 43 patients with verified AIN, 27 of whom were from the high-risk patient group.We identified 2914 non-AIN patients and included 227 controls, with 9 controls from the high-risk group.One high-risk and one non-high-risk control were excluded due to missing preoperative blood sample collection.The patient flow is outlined in Figure 1.The characteristics of AIN patients and controls are displayed in Table 1.
Thromboembolic risk factors, such as smoking, arteriosclerotic diseases, and heart disease, were equally frequent in AIN patients and controls.Additionally, anticoagulation treatment in AIN subgroups was most frequent in vascular AIN patients (78%) but was also prevalent in nonvascular AIN (69%) and nonvascular ischaemia (43%).Perioperative data for AIN subgroups are displayed in Table 2.
Nonvascular necrosis primarily affects the small intestine, and vascular necrosis affects the colon in every second patient.
The final diagnoses of all patients are displayed in Supplementary Table 1.
No notable elevation in sensitivity or AUC was displayed in the high-risk patient subgroup.In non-high-risk patients, the sensitivity of I-FABP + endothelin-1 combined approached 0.9, but the optimism-corrected AUC was unchanged.Finally, endothelin-1 sensitivity reached 0.9 with an AUC approaching 0.8 in only vascular AIN patients compared to controls.
Analyte stability tested according to Supplementary Table 2 shows no significant difference between the gold standard and worst-case scenarios (Supplementary Materials 1).

Discussion
AIN mortality is related to the restoration of blood flow to the intestinal segment within the first six hours [5].The clinical suspicion of vascular AIN is often absent due to unspecific clinical and biochemical presentation [3].An early and reliable biomarker is crucial in the initial diagnostic process, entailing the initiation of lifesaving treatment.In this study, we examined the diagnostic performance of a widely used blood-based parameter, L-lactate, as well as the two proposed AIN biomarkers endothelin-1 and I-FABP in AIN patients compared to non-AIN controls.L-Lactate showed insufficient diagnostic performance with low AUC and specificity.However, the diagnostic performance of endothelin-1 and I-FABP was promising.
To reduce AIN related mortality, the diagnostic process should focus on vascular AIN patients [23].The blurred clinical and biochemical appearance compromises suspicion of AIN in the first important hours thereby delaying lifesaving treatment resulting in high mortality rate.A reliable AIN biomarker might rise suspicion in the early phase, potentially before irreversible intestinal damage or clinical deroute emerges.Endothelin-1 and I-FABP might have such potential and could be candidates to be included in a diagnostic algorithm in supplement to other blood-based parameters as well as clinical and radiological findings.
Our study shows promising diagnostic performance of endothelin-1 and I-FABP.The AUC for both tests in all AIN patients was moderate, the sensitivity for endothelin-1 was good, and the specificity for I-FABP was moderate.The diagnostic performance in the vascular AIN patients shows even better AUC for endothelin-1 with high sensitivity.This might have the potential to help initiate the lifesaving initial suspicion and treatment in this important subgroup of AIN patients.
In addition, high-risk patients where AIN is suspected mimics a frequently used research design in the literature with the inclusion of only suspected cases compared to controls [16,24,25].Our subgroup analysis in high-risk patients reveals almost unchanged results.Altogether, high-risk patients are suspected to suffer from AIN, and the operation decision is already taken in the clinical evaluation, making this blood testing too late and irrelevant.
Overall, our findings correspond well to the literature, although the combination of endothelin-1 + I-FABP has not been researched before.
Our study supports the major portion of the literature rejecting L-lactate as an AIN biomarker, including several papers and a few newer reviews [6][7][8][9]26,27] as well as the ESTES guideline [5].Elevated I-FABP in AIN has been shown in animal [28][29][30] and human studies.A review from 2018 supports the use of I-FABP as an AIN biomarker [26], and two newer metanalyses [31,32] and a recent multicenter study [33] reported good diagnostic performance in humans.In fact, Treskes et al. reported very high pooled specificity (>0.90), and Sun et al. reported an AUC of 0.86.Nevertheless, in a new case-control study from France, unsatisfactory diagnostic performance for I-FABP was found [24].They reported a very low AUC and sensitivity but high specificity based on 50 AIN patients compared to 79 non-AIN controls.Among AIN patients, only vascular AIN was included, and only 12 patients had a part of their intestine resected, reflecting nonnecrotic intestinal manifestations.This might explain the difference in diagnostic performance compared to our study, where most AIN patients had intestinal necrosis demanding intestinal resection.Finally, Bourcier et al. reported an AUC of 0.83, sensitivity of 0.70 and specificity of 0.85 diagnosing AIN in a recent multicentre study of critically ill NOMI patients [33].The AIN diagnosis was confirmed endoscopically and by CT scans, and not only by surgery.Additionally, they reported non-surgical treatment of AIN in roughly 1/3 of the patients.However, since AIN-pathology (critically ill NOMI patients), the definition of AIN (both irreversible and reversible AIN as well as both mucosal and transmural necrosis) and diagnostic modalities (CT scan/endoscopy/surgery) were different from our study, impeding appropriateness of comparison of study results.
Endothelin-1 and its role in AIN have been scarcely researched.In a review by Derikx et al. [9], they found that only one study from 2005 to 2016 reported on endothelin-1 and AIN.This single study reported 78 patients with NOMI who had undergone cardiac surgery [22].The AUC in our study corresponds well with their results.However, their reported specificity was very high (94%), whereas the sensitivity was lower than ours (51%), which might be explained by difficulties with generalizing from patients with NOMI to non-NOMI AIN patients.

Limitations and strength
First, diagnostic challenges are most pronounced in vascular AIN.To reduce mortality, it is crucial to find and treat unsuspected vascular AIN patients as quickly as possible.Only a minor fraction of our patients were in this group, compromising the generalizability of our results to this important group in the search for a screening blood test for patients referred to the hospital with abdominal pain.Nevertheless, increasing the number of verified unsuspected AIN would be extremely costly because every admitted patient must be included.Second, due to the low number of unsuspected AIN patients, we introduced the secondary inclusion period to increase the number of AIN patients in the cohort thereby introducing bias to the positive predictive value and negative predictive value.Nevertheless, the sensitivity and specificity were unaffected.Third, the duration of AIN was not taken into account statistically.It is plausible that biomarker concentration is related to the duration of AIN.Omitting this might weaken our conclusion.
One strength of the study is the very early inclusion and sample collection at the time of admission and before the time-consuming diagnostic workup, hence revealing the ability of biomarkers to pinpoint the AIN patient before a clinical de route.In other words, this means uncovering the AIN patient in the non-high-risk patient group before it turns into a high-risk patient.Second, to our knowledge, this is the first study that has explored the diagnostic performance of endothelin-1 in non-NOMI AIN patients, uncovering very promising results for this biomarker.Third, comprehensive pre-analytical demands in blood samples might be a challenge in the daily clinical setting.We show that I-FABP and endothelin-1 are very robust in eliminating this potential problem.Furthermore, the analyses are considered transferrable to automated analytical setups in hospital laboratories.

Conclusion
AIN is a life-threatening disease, and timely diagnosis of vascular AIN, in particular, is reflected in the high mortality rate.Early blood-based biomarkers are thought to be crucial in the diagnostic workup since early diagnosis of AIN fails due to unspecific results of standard blood tests and clinical findings.We found the promising diagnostic performance of endothelin-1 and I-FABP and the two combined.Additionally, our study supports that L-lactate does not deserve widespread use in diagnosing AIN.Further validation of the endothelin-1 combined with I-FABP setup for the diagnosis of AIN patients is warranted.

Figure 1 .
Figure 1.flowchart for ain patients and controls.
skewness 'Hours between blood sample and operation' is reported by median(iQr); all other continuous variables are reported in means(Sd); Sd: standard deviation.

Figure 2 .
Figure 2. Scatterplot of endothelin-1 vs. i-faBP.a logarithmic y-axis scale was used for i-faBP, and a linear scale was used for endothelin-1.ain patients are depicted in red.
intestinal necrosis; cut-off value in mM; auc: area under the receiver operating characteristic curve; ci: confidence interval.the sensitivity and specificity are relative to the cut-off corresponding to the youden index for l-lactate, i-faBP, endothelin-1 and the combination of i-faBP and endothelin-1.a optimism-corrected.

Figure 3 .
Figure 3. comparison of roc curves of the biomarkers and the combination of endothelin-1 and i-faBP in the prediction of ain in the population.
(.) 0.74 (0.55; 0.94) 0.70 0.80 ain: acute intestinal necrosis.cut-off value in mM.auc: area under the receiver operating characteristic curve.ci: confidence interval.the sensitivity and specificity are relative to the cut-off corresponding to the youden index for l-lactate, i-faBP, endothelin-1 and the combination of i-faBP + endothelin-1.a optimism-corrected.

Table 1 .
characteristics of 268 acute surgical patients.

Table 3 .
diagnostic performance for l-lactate, i-faBP and endothelin-1 in 43 ain patients and 225 controls.

Table 4 .
diagnostic performance for l-lactate, i-faBP and endothelin-1 in subgroups compared to 225 controls.