Evaluation of the diagnostic performance of estimated fecal calprotectin and serum intelectin-1 and C-reactive protein solo or in combination for differentiation between patients with query ulcerative colitis and irritable bowel syndrome

Evaluation of the ability of estimated levels of fecal calprotectin (FCP), serum intelectin-1 (ITLN1), and C-reactive protein (CRP) to differentiate between patients with ulcerative colitis (UC) and irritable bowel syndrome (IBS). Three-hundred forty-two patients were evaluated clinically for diagnostic criteria of UC and IBD and underwent colonoscopic examination and grading according to Mayo endoscopic scores (MES). Colorectal biopsies were taken for microscopic examination. Fecal and blood samples were obtained for ELISA estimation of levels of the studied variate. Patients were grouped according to microscopic examination of the obtained biopsies as UC and IBD groups. Study outcome is the ability of the laboratory variate for prediction of the microscopic diagnosis. In UC patients, FCP and serum CRP levels were notably elevated compared to controls and IBS patients. Conversely, UC patients exhibited significantly reduced serum ITLN1 levels in comparison to controls and IBS patients with insignificantly lower levels in samples of IBS patients. Statistical analyses defined high FCP and low serum ITLN1 as the significant predictors for UC diagnosis with high specificity for FCP level > 150 µg/ml and high sensitivity for serum ITLN1 ≤ 30 and ≤ 18 ng/ml to predict colonoscopic and microscopic UC diagnosis, respectively. The combination of high FCP and low serum ITLN1 could accurately predict the colonoscopic and microscopic findings of UC and can differentiate UC from IBS and may spare the need for colonoscopy and biopsy especially for IBS patients.


Introduction
Gastrointestinal (GI) immune homeostasis is the result of collaborative functions of the intestinal epithelium, the immune system and gut nerve supply, and specifically for the colon, gut microbiome, which dynamically regulates the local immune function [1].The disrupted intestinal epithelial barrier and leakage of the gut microbiome are the key factors in different pathophysiological conditions, especially inflammatory bowel disease (IBD), irritable bowel syndrome (IBS), and chronic liver diseases [2].
Inflammatory bowel diseases (IBD), including Crohn's disease, ulcerative colitis (UC), and indeterminate colitis, are distinguished by their idiopathic and chronic inflammation of the digestive tract.Unfortunately, they are incurable and demonstrate a worldwide increasing incidence and prevalence [3].On contrary, IBS is a functional GI disorder that is characterized by gastrointestinal dysmotility and visceral hyperalgesia [4] and is highly prevalent conditions with bothersome abdominal symptoms in the absence of structural abnormalities [5].
Functional GI disorders are recently attributed to gut-brain interaction that starts with mucosal infiltration with immune cells, which release and/or induce the release of nociceptive mediators by intestinal cells [5].These mediators can activate the sensitized neurons leading to visceral hypersensitivity [6].The interaction between immune activation and an impaired barrier function of the gut is most likely bidirectional, and the altered microbiota, psychological stress, and food staffs are components of the pathophysiology [7].
Distinguishing patients with IBS and IBD is still dependent on colonoscopic evaluation; however, recently, the development of IBS diarrhea secondary to minimal inflammation was reported in patients with quiescent CD which increased the dilemma [8].

Objectives
In trial to find a noninvasive differentiating modality between cases with query UC or IBS, this study targets to evaluate the diagnostic performance of estimated levels of fecal calprotectin (FCP), serum intelectin-1 (ITLN1), and C-reactive protein (CRP) solely or in combination for such purpose.

Design
It is a prospective double-blinded comparative study.

Setting
It is the Departments of Internal Medicine and Medical Biochemistry, Faculty of Medicine, Benha University, in conjunction with multiple private GIT centers.

Ethical considerations
The study was started in June 2020 after obtaining the preliminary approval of the Faculty of Medicine, Local Ethical Committee, and after completion of case collection in Dec 2022; the final approval was obtained and registered by no.RC: 2.2.23.

Blindness
One author was responsible for case collection and obtaining clinical data, and another was responsible for colonoscopic examination and biopsy taking, and each author was blinded about the data collected by the other.Fecal samples were obtained and sent for estimation of FCP and blood samples for estimation of serum CRP and ITLN1 levels.The obtained biopsies were preserved in formalin and sent for pathological examination at the Department of Pathology, Faculty of Medicine.At the end of case collection, the clinical, colonoscopic, laboratory, and microscopic data were interpreted.

Patients
All patients attending the outpatient clinic with manifestations of either IBD or IBS were evaluated for exclusion and inclusion criteria.Twenty healthy controls free of inclusion and exclusion criteria were enrolled to give fecal specimens and serum samples as control group for laboratory investigations.

Exclusion criteria
Patients who were previously diagnosed as IBD or IBS and were under treatment or not, patients who had recurrent ulceration after surgical treatment to any levels, and patients who had malignancy elsewhere in the body, autoimmune disorders, maintained on immunosuppressant, or refused to participate in the study were excluded from the study.

Inclusion criteria
Patients free of exclusion criteria and presented by symptoms suggestive of IBD or IBS were enrolled in the study.

Evaluation tools
A) Evaluation tools for UC diseases 1. UC diagnosis was relied on the presence of at least three of the following items: abdominal pain, diarrhea, hematochezia, and/or pus in stools, the presence of rectal with or without colonic mucosal ulcerations on colonoscopic examination, microscopic findings consistent with UC in the obtained colorectal biopsy, and the absence of manifestations and colonoscopic diagnostic findings of Crohn's disease.To simplify the evaluation, each item was scored by 0 or 1 for a collective score of 0-8 [9].and MES-3 if there is spontaneous bleeding and/or large ulcerations, so indicates severe disease activity [15].

Fecal sampling and processing
Fresh fecal samples (about 30 ml) were obtained and stored at − 20 °C.The stored feces samples were thaw at room temperature; about 100 mg of feces was taken by the inoculation loop, placed into a screw-cap tube, and weighted to obtained net feces weight.Then, the prediluted extraction buffer was added to tube contents in 1:50 Wt.:vol., the tube was closed and shake vigorously for 30 s, and contents were homogenized for about 25 min on a shaker.One milliliter of the homogenate was put in a tube, centrifuged for 20 min at 10,000 g, and 0.5 ml of the clear extract supernatant was pipetted and put in Eppendorf tube and stored at − 80 °C until being assayed.

Blood sampling and processing
At time of obtaining the fecal sample, blood samples were obtained from the antecubital vein, allowed to clot, and centrifuged at 1500 rpm for 15 min, and serum was collected and put in dry Eppendorf tube at − 80 °C until being assayed.

Patients' grouping
Considering the gold standard for UC diagnosis is the microscopic detection of mucosal changes consistent with UC, patients with positive microscopic examination for UC were categorized as UC group and patients with negative findings consistent with UC diagnosis as IBS group.

Study outcomes
1.Primary outcome: Is the predictability of the estimated biomarkers for the results of the microscopic examination of the obtained mucosal biopsies 2. Secondary outcomes -The diagnostic performance of the studied biomarkers solely or in combination for blinded differentiation between patients with UC and IBS -The relation of estimated biomarkers to disease activity scores

Statistical analysis
Results were analyzed using one-way ANOVA and chisquare test (Χ 2 test), and correlation analysis was performed using Pearson's correlation analysis.Regression analysis of correlated variate was conducted by the use of the stepwise method.Receiver operating characteristic (ROC) curve analysis was performed to compare the area under curve (AUC) for each variate versus the area under the standard line (AUC = 0.5), and performance of each variate was judged according to the AUC as sensitive (AUC < 0.5) or specific (AUC > 0.5).Paired analysis of difference of AUC was performed for each two variates to determine the best predictor.Kaplan-Meyer analysis was used to determine the cutoff point for the studied variate that might predict the cumulative hazard for UC diagnosis.Performance characters of lab variate were evaluated versus MES and microscopic diagnosis of UC using the determined cutoff points.IBM ® SPSS ® Statistics (Version 22, 2015; Armonk, USA) which was for Windows statistical package was the applied system.Significance was considered at the cutoff point for P-value at < 0.05.

Results
During the study duration, 373 patients presented by manifestations suggestive of UC or IBS; 13 patients were excluded because 9 patients were under treatment for UC, and 4 patients had recurrent ulceration after previous surgical resection, and 360 patients were enrolled in the study.Eighteen patients refused to undergo colonoscopic examination and were excluded from the study, and data of 342 patients were analyzed.According to the microscopic examination of the obtained mucosal biopsies, 51 biopsies showed changes consistent with diagnosis of UC (UC group), while 291 biopsies were free of these changes (IBS group) as shown in Fig. 1.Patients of UC group were significantly older than patients who had IBS, while other demographic data showed nonsignificant differences between patients of both groups.Recurrent abdominal pain was the main complaint of 330 patients (96.5%) including all IBS patients and 39 (76.5%) of UC patients with significantly higher incidence among IBS patients.One-hundred and five patients complained of diarrhea with significantly higher frequency among UC patients.Among IBS patients, constipation was the main bowel habit compliant of 42 patients, while 49 patients had bowel habit alternating between diarrhea and constipation, and 137 could not classify their bowel habits.Hematochezia and pus in stool were reported by 51 and 59 patients, respectively, with significantly higher incidence among UC patients (Table 1).
Mean FCP levels estimated in samples of UC patients were significantly (P < 0.001) higher than levels estimated in samples of controls and IBS patients with significantly (P = 0.0015) higher levels in samples of IBS patients than control samples.Twenty-eight patients had FCP level ≥ 150 µg/ml, 152 patients had levels ranging between > 50 and < 150 µg/ml, and 162 patients had levels ≤ 50 µg/ml with significantly (P < 0.001) higher frequency of patients who had FCP ≥ 150 µg/ml among UC patients than IBS patients.Mean serum CRP levels estimated in samples of UC and IBS patients were significantly (P < 0.001) higher than control levels with significantly (P < 0.001) higher levels in samples of UC than samples of IBS patients.In relation to 5 mg/L as cutoff point for serum CRP, 55 patients had serum CRP levels < 5, while 287 patients had serum CRP level ≥ 5 mg/L with nonsignificant (P = 0.893) difference between UC and IBS patients.Mean serum ITLN1 levels estimated in samples of UC patients were significantly (P < 0.001) lower than levels estimated in control and IBS patients' samples with nonsignificantly (P = 0.366) lower levels in samples of IBS patients than in control samples (Table 3).
Regarding the remaining 291 patients, according to Rome IV criteria, in addition to recurrent abdominal pain, 177 patients (60.8%) had change in stool frequency, and 152 patients (52.2%) had change in stool form.Collectively, there were 38 patients (13.1%) fulfilling > 2 of Rome IV criteria, while 253 patients (86.9%) fulfilled two  Regression analysis defined high FCP and low serum ITLN1 as the significant predictors for colonoscopic diagnosis for UC.Regarding prediction of the microscopic findings suggestive of UC, regression analysis defined old age, high FCP, and low serum ITLN1 as the significant predictors while excluded other correlated variables (Table 6).Furthermore, estimated serum ITLN1 levels showed negative significant correlation with BMI (r = − 0.122, P = 0.025), FCP (r = − 0.132, P = 0.012), and serum CRP (r = − 0.265, P < 0.001).
Evaluation of the diagnostic performance characteristics of the studied lab variate for colonoscopic and microscopic diagnosis of UC showed high specificity for FCP level > 150 µg/ml and high sensitivity for serum ITLN1 ≤ 30 and ≤ 18 ng/ml to predict colonoscopic and microscopic UC diagnosis, respectively.Both high FCP and low serum ITLN1 showed accuracy for diagnosis of 76% and > 90%, respectively, for colonoscopic and microscopic diagnosis of UC (Table 7).

Discussion
Decreased serum levels of intelectin-1 (ITLN1) could differentiate patients with UC from healthy controls and IBS patients due to the significantly lower levels in samples of UC patients than controls and IBS samples and might differentiate between IBS patients and healthy controls because of the nonsignificant difference between samples of IBS patients and controls.
These findings supported the earlier studies detected significantly lower serum ITLN1 in patients with active Crohn's disease (CD) than in patients in remission and patients with IBS and healthy controls [20].Another study also found serum ITLN1 is significantly decreased in CD and UC patients than healthy controls and with active than inactive diseases [21].These results showed an inverse relation between the presence and severity of inflammation and serum levels of ITLN1; in support of this assumption, the applied correlation analysis showed negative significant correlation between serum ITLN1 levels and Mayo's endoscopic score (MES), which is recently documented as the most common endoscopic index recommended in guidelines and widely used in clinical trials and practice [22] and with the microscopic diagnosis of UC.Intelectin-1 (omentin-1) is one of the anti-inflammatory adipocytokines including chemerin, vaspin, omentin, and visfatin [23], and recently, the relation between various adipocytokines and IBD presence and severity was reported, wherein Sochal et al. [24] reported a relation between serum chemerin and IBD severity of inflammation and Saadoun et al. [25] documented the use of estimated serum visfatin levels to detect UC activity and as predictor for disease extension.
In support of the relation between serum ITLN1 and inflammation, Kukla et al. [26] detected significantly decreased serum levels of ITLN1 and chemerin in samples of COVID-19 patients compared to healthy volunteers, and Gültekin et al. [27] found serum ITLN1 levels were significantly higher in non-sepsis than sepsis and septic shock patients and concluded that ITLN1 may have a role in development of inflammatory and metabolic complications in intensive care patients and is associated with poor outcomes and mortality.
In trial to explore the mechanisms underlying the relation between serum ITLN1 and UC disease, using animal model of induced UC inflammation, Ma et al. [28] detected decreased expression levels of ITLN1 in colonic tissues and found overexpression of ITLN1 inhibited endoplasmic reticulum stress-related proteins, colonic damage, inflammation, barrier damage, and cell apoptosis.Clinically, Nasir et al. [29] found an inverse significant relation between serum ITLN1 levels and levels of cholesterol and consumption of monounsaturated fatty acid and total fat intakes and established a relation between obesity and low serum ITLN1 levels.In support of this, the current study detected a negative relation between serum ITLN1 and BMI of the studied patients.Also, Li et al. [30] detected a similar negative relation between total body fat mass and serum ITLN1 levels.
On contrary, estimated serum CRP levels were significantly higher in UC and IBS patients than controls' levels, despite the significantly higher levels in UC patients than IBS patients.Similarly, Xu et al. [31]  reported increased serum CRP levels with decreased serum total leucocytic count, and Sakemi et al. [32] detected significantly higher levels of inflammatory markers including CRP in UC and CD patients than in controls with significant difference between UC and CD.Further, the current study showed a negative significant relation between serum levels of CRP and ITLN1; such relation assured the antiinflammatory action of ITLN1.The action of ITLN1, which was consumed or its synthesis, was suppressed by the inflammatory cascade.Similarly, inverse relations were detected between serum CRP and lymphocyte:monocyte ratio [31] and interleukin-22, which have an essential role in mucosal repair in IBD [32].Estimated levels of FCP were significantly higher in patients than in controls and in UC than IBS patients with significantly higher frequency of patients who had FCP level > 150 µg/ml among UC than IBS patients.Further, there was positive relation between FCP and endoscopic and microscopic UC diagnostic findings and serum CRP, while the relation was negative with serum ITLN1 levels.These findings go in hand with recent studies that applied FCP for different purposes to diagnose or monitor UC patients [33,34].
Statistical analyses found high FCP and serum ITLN1 levels were highly predictive for the UC disease, but with different performance where high FCP was specific and low serum ITLN1 was sensitive biomarker for the presence of UC, and were correlated with disease activity as judged by colonoscopic and microscopic findings for UC.Further, estimated levels of FCP and serum ITLN1 might predict the colonoscopic and microscopic findings with accuracy rate of about 76% and 90%, respectively.These complimentary roles of both markers allow using this combination for differentiation between UC and IBS and sparing colonoscopy with its financial and psychological impacts.
In line with this assumption for using complimentary investigation to replace colonoscopy goes in hand with Goodsall et al. [35] who found the composite use of FCP and intestinal ultrasonography to reduce the need for colonoscopy in routine care for UC patients.Also, Singh et al. [36] suggested the applicability a biomarker combination based on FCP and symptombased monitoring strategy over a symptom-based monitoring strategy and to reduce endoscopic use for assessment of disease activity to guide treatment decisions.Further, Magalhaes et al. [37] in a systemic review found blood neutrophil-expressed biomarkers as adjuvants to basic diagnostics for IBD can help to modify treatment decision-making.

Conclusion
The combined estimation of serum ITLN1 and FCP could be used as an array for screening patients presented by manifestations suggestive of UC.Elevated FCP and decreased serum ITLN1 could accurately predict the colonoscopic and microscopic diagnostic findings of UC and thus can differentiate UC from IBS. Serum ITLN1 near normal levels suggest IBS especially with FCP < 50 or in the gray zone of > 50-< 150 µg/ml.The suggested combination may spare the need for colonoscopy and biopsy taking for UC diagnosis.

Limitation
Estimation of serum ITLN1 in UC patients in remission was a limitation of the current study.

Recommendations
Wider-scale comparative studies were mandatory to establish the suggested cutoff points for ITLN1 as predictor for MES grade and microscopic findings.Also, comparative studies for serum ITLN1 levels in UC patients who were in remission versus IBS patients and controls are required to assure its distinguishing ability between various colorectal disorders.

Fig. 2 Fig. 3
Fig. 2 Kaplan-Meyer curve for estimated serum ITLN1 as predictor for the cumulative hazard of UC endoscopic diagnosis of activity

Table 1
Demographic data and clinical presentations of the enrolled UC and IBS patients

Table 2
Colonoscopic findings of the enrolled patients

Table 3
Estimated laboratory parameters estimated in samples of the enrolled patients compared to samples of control subjects

Table 4
Diagnostic scorings of UC

Table 5
Diagnostic scoring of IBS

Table 6
Correlation and regression analyses for patients' data and laboratory findings as predictors for colonoscopic and microscopic diagnosis of UC

Table 7
Diagnostic performance characters of the estimated levels of the studied lab variate for UC diagnosis