- Top of page
- MATERIALS AND METHODS
Background: Noninvasive, sensitive, and specific tools for early identification of chronic inflammatory bowel disease (IBD) are needed for clinical practice. The aim was to identify new noninvasive test combinations for characterization of IBD in children and adolescents by comparing serological responses to microbial antigens and fecal calprotectin, a new promising marker for intestinal inflammation.
Methods: Our study included 73 children who underwent endoscopies because of suspicion of IBD. Their sera were tested for antibodies to the Pseudomonas fluorescens-associated sequence I2, a Bacteroides caccae TonB-linked outer membrane protein, OmpW, and anti-Saccharomyces cerevisiae (ASCA). Simultaneously, samples for fecal calprotectin measurements were obtained from 55 subjects.
Results: IBD was diagnosed in 60 patients (Crohn's disease [CD] in 18 patients, ulcerative colitis [UC] in 36, and indeterminate colitis [IC] in 6). Thirteen children had a non-IBD disease. Fecal calprotectin levels were elevated (≥100 μg/g) more frequently in IBD patients (89%, 39/44) compared to non-IBD cases (9%, 1/11, P < 0.001). ASCA antibodies in sera were detected in 67% (12/18) of patients with CD, in 14% (5/36) of the children with UC, and in 50% (3/6) of patients with IC. Seroreactivity for I2 was observed in 42% of the IBD patients, this frequency being higher than in non-IBD cases (7.7% seropositive; P = 0.025). Serum anti-I2 IgA levels (median absorbances) were higher in those with IBD compared to those without gut inflammation (P = 0.039). The combination of the measurements of fecal calprotectin and serological responses to microbial antigens (ASCA, I2, and OmpW) identified 100% of CD patients (sensitivity 100%, specificity 36%, positive predictive value [PPV] 66%, negative predictive value [NPV] 100%) and 89% of UC patients (sensitivity 89%, specificity 36%, PPV 77%, NPV 57%).
Conclusions: Increased levels of serological responses to microbial antigens (ASCA, I2, and OmpW) and fecal calprotectin are evident in both CD and UC patients. The combination of these markers provides valuable, noninvasive tools for the diagnosis of IBD.
Crohn's disease (CD) and ulcerative colitis (UC) are chronic inflammatory bowel diseases (IBD) characterized as a group of immune-mediated disorders of the intestine. To increase diagnostic accuracy, noninvasive, sensitive, and specific tools for early identification of IBD are needed. Fecal calprotectin, a cytoplasmatic protein released by activated neutrophilic polymorphonuclear cells and/or monocytes-macrophages, is a marker of intraluminal intestinal inflammation.1, 2 This protein has been demonstrated to be useful in detecting active IBD and clinical relapse, and in monitoring the response to treatment both in adults and children.1–7
Serological markers such as anti-Saccharomyces cerevisiae antibodies (ASCA) and perinuclear anti-neutrophilic antibodies (pANCAs) have proven to be valuable in the diagnosis and differentiation of CD and UC.8–16 These markers may also predict the development of IBD years before the disease is clinically diagnosed.17 Familial clustering of ASCA and pANCA has also been demonstrated, suggesting that these antibodies might be familial disease markers.18, 19 The variety of serologic markers for IBD has expanded and an increasing amount of experimental data are nowadays available on new antibodies directed against various microbial antigens such as the Pseudomonas fluorescens-associated sequence I2, a Bacteroides caccae TonB-linked outer membrane protein, OmpW, and E. coli outer membrane porin C, OmpC.20–26 The appearance of these antibodies reflects a loss of tolerance to different intestinal bacteria. Varying responses to selected microbial and autoantigens have been described in subsets of CD patients and also in UC patients.23, 26, 27
The aim of the present study was to examine the association of fecal calprotectin with serological markers in children and adolescents with IBD. Furthermore, we wanted to identify new possible noninvasive test combinations for detecting IBD in children and adolescents.
MATERIALS AND METHODS
- Top of page
- MATERIALS AND METHODS
Serum and fecal samples were collected in 73 children and adolescents examined for suspicion of IBD (CD, IC, UC) at the Hospital for Children and Adolescents in Helsinki, Finland, from May 2005 to November 2006. At the time of primary diagnosis samples of 64 patients with IBD suspicion were available for analyses (9 cases excluded, see Table 3). All 73 subjects underwent upper and lower endoscopies and their sera were collected for further analysis. The diagnosis of IBD was based on histopathological criteria.28 Subjects were grouped for the final analysis into IBD patients (n = 60), including patients with CD (n = 18), UC (n = 36), and IC (n = 6), and non-IBD control subjects (n = 13). The presence and degree of inflammation were determined in the upper gastrointestinal biopsies using the modified Sydney system.29
Measurement of Fecal Calprotectin
Fecal calprotectin was measured by enzyme immunoassay in fecal samples available for analysis in 55 patients (median age 12.8 years, range 5.8–19.9). Calprotectin levels were determined from feces as previously described and fecal calprotectin level higher than 100 μg/g was considered elevated.30
Sera for the determination of anti-I2 and anti-OmpW, IgA were drawn at the time of endoscopy from all 73 children and adolescents and stored at −20°C until testing. The majority of the samples (64/73) were collected at the time of primary diagnosis. In our laboratory, E. coli XL-1-blue and E. coli BL-21 (Stratagene, La Jolla, CA) strains were used for all cloning and recombinant expression experiments. I2-GST and OmpW were produced by using previously reported antigen purification techniques.20, 22 Sera were analyzed by IgA enzyme-linked immunosorbent assays (ELISA) for the determination of the Pseudomonas fluorescens-associated sequence I2 and the Bacteroides caccae TonB-linked outer membrane protein OmpW as previously described.26 An enzyme immunoassay kit (QUANTA Lite ASCA, INOVA Diagnostics, San Diego, CA) was used to determine ASCA of both IgG and IgA isotypes. Quantitative results in arbitrary enzyme immunoassay units were obtained from standard curves defined by the manufacturer, and the results were statistically handled as qualitative. Equivocal/borderline results were interpreted as negative. Results exceeding 25 U for IgG and/or IgA ASCA were regarded as positive.
Numerical analysis of laboratory measurements was made by Mann–Whitney U-test. Other parameters between study groups were compared by Pearson's χ2 and Fisher exact tests. Statistical calculations were carried out with SPSS (v. 11.0, Chicago IL). Analyses for specificity, sensitivity, and positive (PPV) and negative predictive value (NPV) were performed.
The study protocol was approved by the Ethics Committee of Helsinki University Hospital, and informed consent was obtained from the parents and from the children when appropriate.
- Top of page
- MATERIALS AND METHODS
IBD was detected in 60 patients (CD in 18 patients, UC in 36, and IC in 6), and 13 children presented a non-IBD disease including celiac disease (n = 1), allergy (n = 1), lymphatic adenopathy (n = 1), extrahepatic portal obstruction (n = 1), juvenile polyp (n = 1), hematochezia (n = 1), hemorrhagia intestinalis NAS (n = 1), operated intestinal invagination (n = 1), lumbosacralic meningomyelocele (n = 1), and normal (n = 4). The gender and age of study population are shown in Table 1.
Table 1. Gender and Age (Median, Range) in the Study Population
|Age median (range)||13.8 yrs (2.7–19.9)||15.5 yrs (9.0–19.6)||13.7 yrs (2.7–19.9)||7.2 yrs (6.1–16.3)||8.1 yrs (5.8–14.9)|
At the time of diagnosis fecal calprotectin levels were increased (≥100 μg/g) in 89% (39/44) of IBD patients. This frequency of positivity was significantly higher compared to non-IBD patients (9%, 1/11, P < 0.001). Frequencies of fecal calprotectin positivity in all IBD subgroups (CD, UC, and IC) are shown in Table 3. Calprotectin levels in feces were significantly elevated in patients with IBD when compared to patients who were excluded from the disease group (P < 0.001). (Fig. 1). The sensitivity, specificity, PPV, and NPV are presented in Table 2.
Figure 1. Fecal calprotectin levels in children and adolescents according to diagnosis into inflammatory bowel disease (IBD) subgroups (Crohn's disease, CD; ulcerative colitis, UC; indeterminate colitis, IC) and non-IBD group. Horizontal line denotes the cutoff value for positivity (see text). P-values compare absorbances between the IBD subgroups and non-IBD group using the Mann–Whitney U-test.
Download figure to PowerPoint
Table 2. Sensitivity, Specificity, Positive, and Negative Predictive Values of Serum ASCA, I2, and OmpW and Fecal Calprotectin Tests
|Sensitivity||33% (67%, 14%)||42% (44%, 42%)||32% (28%, 33%)||89% (100%, 82%)|
|Specificity||77% (77%, 77%)||92% (92%, 92%)||77% (77%, 77%)||90% (90%, 91%)|
|PPV||87% (80%, 63%)||96% (89%, 94%)||86% (63%, 80%)||97% (93%, 96%)|
|NPV||20% (63%, 25%)||26% (55%, 36%)||20% (44%, 29%)||67% (100%, 67%)|
In our laboratory the mean optical density in the IgA-class I2-GST antibody ELISA test for all 60 children with untreated IBD was 0.72 (95% confidence interval [CI] 0.34–1.1), and the respective value for all non-IBD patients was 0.30 (95% CI 0.01–0.59). The mean optical density in the IgA-class OmpW antibody test with untreated IBD patients was 0.70 (95% CI 0.33–1.08) and in children with non-IBD disease 0.36 (95% CI 0.16–0.56). The cutoff levels were set as previously described.26 Sensitivity, specificity, PPV, and NPV of I2 and OmpW tests are shown in Table 2. The frequencies of seropositivity toward different antigens in children with IBD and children with non-IBD disease are presented in Table 3.
Table 3. Frequency of Positive Seroreactivity to Anti-Saccharomyces cerevisiae (ASCA), I2 and OmpW and Frequency of Fecal Calprotectin Positivity Among 60 Patients with IBD and 13 Non-IBD Disease Patients
| ||ASCA IgA and/or IgG||I2 IgA||OmpW IgA||Fecal Calprotectin|
|All IBD||20/60 (33.3%)||25/60 (41.7%)||19/60 (31.7%)||39/44 (88.6%)|
|N= 51a||18/51 (35.3%)||22/51 (44.0%)||17/51 (33.3%)||32/37 (86.5%)|
|CD||12/18 (66.7%)||8/18 (44.4%)||5/18 (27.8%)||14/15 (93.3%)|
| ||10/16 (62.5%)||8/16 (50%)||5/16 (31.3%)||12/13 (92.3%)|
|UC||5/36 (13.9%)||15/36 (41.7%)||12/36 (33.3%)||21/25 (84%)|
| ||5/30 (16.7%)||12/30 (40%)||10/30 (37%)||17/21 (81%)|
|IC||3/6 (50.0%)||2/6 (33.3%)||2/6 (33.3%)||4/4 (100%)|
| ||3/5 (60.0%)||2/5 (40.0%)||2/5(40.0%)||3/3 (100%)|
|Non-IBD||3/13 (23.1%)||1/13 (7.7%)||3/13 (23.1%)||1/11 (9.1%)|
Anti-I2 levels in the sera were significantly higher in IBD patients when compared to those without IBD (P = 0.039; median of absorbances, Mann–Whitney U-test) (Fig. 2A). At the time of diagnosis children and adolescents with CD and UC expressed higher serum anti-I2 IgA levels in comparison with those with non-IBD disease (P = 0.032 and P = 0.042, respectively) (Fig. 3). In contrast, serum anti-OmpW levels were not markedly elevated in IBD patients (Fig. 2B). In children and adolescents with CD and UC, anti-OmpW IgA levels did not differ significantly from patients with non-IBD disease in this material (P = 0.207, P = 0.217, respectively).
Figure 2. Serological responses to the Pseudomonas fluorescens-associated sequence I2 (A) and Bacteroides caccae TonB-linked outer membrane protein OmpW (B) in children and adolescents according to their diagnosis into inflammatory bowel disease (IBD) and non-IBD group. Horizontal lines denote the cutoff values for seropositivity (see text). P-values compare absorbances between the IBD and non-IBD groups using the Mann–Whitney U-test.
Download figure to PowerPoint
Figure 3. Serological responses to the Pseudomonas fluorescens-associated sequence I2 in 51 children and adolescents with inflammatory bowel disease (IBD) at the time of diagnosis. Patients are grouped according to their final diagnosis into Crohn's disease (CD), ulcerative colitis (UC), indeterminate colitis (IC), and non-IBD patients. Horizontal lines denote the cutoff value for seropositivity (see text). P-values compare absorbances between the IBD and non-IBD groups using the Mann–Whitney U-test.
Download figure to PowerPoint
ASCA antibodies were detected in 67% (12/18) of patients with CD. Serum titers of IgA and IgG ASCA were significantly elevated in children and adolescents with CD as compared with the patients with non-IBD disease (P = 0.001 and P = 0.007, respectively). Three out of 6 CD patients negative for ASCA showed seropositivity to I2 and/or OmpW. The combination of ASCA, OmpW, and I2 tests was able to detect 83% of all CD patients and 61% of all UC patients and 46% of patients with non-IBD disease. Among UC patients 14% (5/36) of the children and adolescents and 50% (3/6) with IC were ASCA-positive (Table 3). Serum IgA and IgG ASCA titers were not significantly higher in UC patients than in patients with non-IBD disease, however (P = 0.226 and P = 0.461, respectively). IgA ASCA titers were significantly increased in IC patients compared with patients excluded for IBD (P = 0.041). Fifty-six percent (20/36) of UC patients were positive for I2 and /or OmpW antibodies. Fourteen UC cases were negative for all serological tests (ASCA, I2, OmpW) used in this study. Samples of 7 out of those 14 UC patients were available for fecal calprotectin measurement. Fifty-seven percent (4/7) of those turned out to have positive (>100 μg/g) levels. Four UC patients were positive only for OmpW antibodies in serological testing (Fig. 4). Altogether, the combination of the measurements identified 100% (15/15) of CD patients and 88% (22/25) of UC patients who were available for all 4 tests.
Figure 4. Relationship between serum antibodies to Saccharomyces cerevisiae, the Pseudomonas fluorescens-associated sequence I2, and to a Bacteroides caccae TonB-linked outer membrane protein, OmpW by presence versus absence in 51 inflammatory bowel disease (IBD) patients at the time of primary diagnosis. The percentage of patients positive for any of each marker and of 2 or all 3 markers is presented.
Download figure to PowerPoint
When analyzing the whole IBD group, 66% of patients were seropositive toward at least 1 of the microbial antigens (ASCA, I2, OmpW) at the time of primary diagnosis (Fig. 4). The best 3 test combinations concerning relationship between serum antibodies to microbial antigens and fecal calprotectin in IBD patients by the presence versus absence were examined. Samples of 37 IBD patients were available for testing of all 4 tests (ASCA, I2, OmpW, and fecal calprotectin) at the time of primary diagnosis. By testing serology (ASCA, I2, OmpW), 76% of patients were positive at least in 1 of the tests. A fecal calprotectin test alone was positive in 87% of IBD patients at the time of diagnosis. When selecting fecal calprotectin and any 2 out of 3 serological tests, 89%–92% of IBD patients showed positivity at least in 1 of the tests. The combination of these 4 noninvasive tests was highly sensitive at the time of diagnosis for IBD (sensitivity 92%, specificity 36%, PPV 83%, and NPV 57%) and also for subgroups; CD (sensitivity 100%, specificity 36%, PPV 63%, and NPV 100%), UC (sensitivity 87%, specificity 36%, PPV 74%, and NPV 57%), and IC (sensitivity 100%, specificity 36%, PPV 22%, and NPV 100%).
- Top of page
- MATERIALS AND METHODS
There is a need, especially in children, for noninvasive, sensitive, and specific tools to improve the early diagnosis of IBD. Fecal calprotectin is remarkably stable and easy to detect in stool. Calprotectin levels have been shown to correlate well with the results of invasive measures of colonic and small bowel inflammation in pediatric IBD.3 We confirm the results of previous reports by showing that the frequency of positive (>100 μg/g) fecal calprotectin levels is significantly increased in children and adolescents with IBD compared with those with no intestinal inflammation.5–7, 30–32 The highest levels of calprotectin were found in patients with CD, which may refer to different histopathology and pathogenesis of CD and UC.31
Serological markers are promising in the diagnosis of IBD and in the differentiation of CD and UC. ASCA is thought to be a specific marker of CD, as several studies have provided evidence of seroreactivity to ASCA both in adult and pediatric CD patients.7–13, 27, 28, 33, 34 Our study agrees with the previous ones, as ASCA positivity was found in 65% of patients with CD. Interestingly, ASCA positivity was also observed in a small proportion of UC patients (14%) and in 3 children with non-IBD disease. These patients, especially those with non-IBD disease, are of major interest for follow-up because it has been suggested that ASCA is an early marker preceding the onset of clinical disease.17
Recently, the variety of serologic markers for IBD has expanded and antibodies directed to various bacterial antigens are under continuous investigation.21–26, 34, 35 High prevalence of antibodies to Pseudomonas fluorescens-associated sequence I2 in the sera has been pointed out in patients with CD.23, 26, 36 In the present study the frequency of anti-I2 antibodies was significantly increased in children and adolescents with IBD. Concordant with our previous findings, positive seroreactivity to a Bacteroides caccae TonB-linked outer membrane protein OmpW was frequent in children with CD and with UC (61%, 42%, respectively).26 The prevalence of anti-OmpW antibodies alone was not significantly elevated in IBD subgroups. However, a possibility is that serum OmpW-antibodies could be useful to identify those few cases that are negative for all other serological tests.
To support this idea, a CD patient subset has been reported in which the IgA response to the E. coli outer membrane porin C (OmpC) was suggested to have clinical utility in diagnosing IBD, specifically in ASCA-negative patients.23–25, 36 We have previously pointed out that most ASCA-negative children with CD and pANCA-negative patients with UC express positive seroreactivity to OmpW and/or I2.26 In the present study, 50% of ASCA-negative CD patients had elevated levels of serum anti-I2 and/or anti-OmpW antibodies. In addition, unlike adult IBD patients (in whom bacterial seromarkers are uncommon), 53% of pediatric UC patients were positive for serum anti-I2 and/or anti-OmpW antibodies. Recently, reactivity to bacterial flagellins (CBir) has also been found to be characteristic for CD patients.36–39 Thus, inclusion of these and other emerging bacterial seromarkers of IBD may further improve the specificity and sensitivity of early diagnosis of both pediatric UC and CD.
The key conclusion of this study is that the combination of fecal calprotectin levels with bacterial seromarkers increases the sensitivity of noninvasive identification of children and adolescents with IBD. Fecal calprotectin measurement together with serological tests (ASCA, I2, OmpW) showed high sensitivity in detecting IBD patients (92%), both CD and UC (100%, 87%, respectively) in this patient cohort. The fecal calprotectin test showed the best sensitivity and specificity for IBD than any other test alone.
As shown here, marked serological responses to microbial antigens are evident among children and adolescents with IBD, suggesting that immune responses to commensal enteric bacteria have an important role in the induction of mucosal damage in the pathogenesis of chronic bowel inflammation. In clinical practice, noninvasive, sensitive, and specific tools for early detection of IBD are needed and serological specificities together with fecal calprotectin measurement provide valuable help in the diagnosis of the disease. Noninvasive testing may decrease unnecessary invasive procedures and further facilitate clinical decision-making when the workup of IBD in children is initially uncertain. Early identification of the subgroups of IBD patients is important also because it has been shown that seroreactivity to microbial components in CD is associated with disease severity, progression, and the frequency of complications.40, 41 Additionally, it has also been suggested that a proportion of IBD patients may benefit from a relatively aggressive therapeutic approach and this could even decrease the necessity for surgery in the future.42 The positivity of any of the present tests (ASCA, I2, OmpW, fecal calprotectin) should encourage investigators to initiate further examinations aimed at a final diagnosis and in addressing treatment.