In the present study, we assessed the clinical utility of anti-CdtB and anti-vinculin antibodies to diagnose superimposed IBS-D in IBD subjects, and to diagnose IBS-D in Brazil. Circulating levels of anti-CdtB and anti-vinculin antibodies did not differ between groups when compared to healthy controls (18).
IBS prevalence in IBD is three times greater than that in the general population (19). Recently, various biomarkers have been developed with the aim of making IBS a diagnosis of inclusion. Considering that the anti-CdtB and anti-vinculin antibodies are directed against microbial epitopes, we hypothesized that the accuracy of these biomarkers could be influenced by the high prevalence of intestinal pathogens in developing countries. As far as we know, this is the first study that analyzed an antibody-based approach in IBD-IBS subjects and the first one to compare antibodies levels with IBD endoscopic activity.
The results of the present study diverge from those observed by Pimentel et al (13). They analyzed anti-CdtB and anti-vinculin antibodies in a cohort of 2,681 patients from 180 hospitals in the United States. The primary endpoint was to assess the accuracy of these biomarkers in a group of individuals diagnosed with IBS-D according to the ROME IV criteria, and to compare them to IBD subjects, subjects with celiac disease, and healthy controls (13). More recently, the same authors developed a second-generation test that incorporated epitope stabilization for CdtB and vinculin (17). They retrospectively analyzed samples from 100 patients with IBS-D and 31 patients with IBD, and found the sensitivity and specificity of anti-CdtB and anti-vinculin were 43% and 52.2%, and 93.5% and 90.9%, respectively (17). In the present study, we used second generation tests. However, our groups were divided between active IBD, IBD-IBS, IBS-D and controls. The present study was single-centered and included subjects with a severe disease phenotype from a quaternary hospital. Therefore, the majority of IBD subjects were treated with immunosuppressive drugs or biologics. We can argue whether the levels of both antibodies were lower due to the effect of immunosuppressants. Supporting this argument, it is known that these agents can reduce the development of anti-drug antibodies and, theoretically, other antibodies (20).
The second difference was the IBS duration between the two studies. In our study, both the active and quiescent IBD groups had a long-standing disease. Klem et al published a meta-analysis that included 45 prospective studies of PI-IBS (21). The authors calculated an incidence of 10.1% at 3 months and 14.5% at more than 1 year after infectious enteritis. The risk of PI-IBS was 4.2 times higher in subjects who had intestinal infection compared to controls (21). We can further speculate as to whether anti-CdtB and anti-vinculin antibodies would perform differently if a previous gastrointestinal infection was identified prior to subject inclusion.
Scallan Walter et al showed the burden of Campylobacter infection in the development PI-IBS (22). They calculated a 1-year incidence rate of IBS following Campylobacter jejuni infection of 16.7 and 3.9 per 1,000 among cases and non-cases, respectively, with an unadjusted risk ratio of 4.3 (95% CI: 3.0–6.2) (22). However, in Brazil, data on the prevalence of PI-IBS and its most frequent pathogens are limited.
C. jejuni infection is one of the most common causes of bacterial gastroenteritis in the world, with different epidemiological profiles between developed and underdeveloped countries. In developed countries, prevalence is mainly related to outbreaks, but is endemic, asymptomatic, and rarely seasonal in low-income countries. Veras et al. analyzed the virulence genes of C. jejuni and inflammatory biomarkers in a case-control study with 340 children in Brazil (23). Fecal DNA samples were extracted and exposure to C. jejuni was evaluated using a polymerase chain reaction assay. The prevalence of C. jejuni in this population was 9.7% and the most expressed genes were CadF, IamA, CheW and SodB. The CdtB gene was found to be expressed in 15.2% of the population. In addition, the authors observed that malnourished children had higher rates of infection with strains of C. jejuni lacking the CdtB gene (23). This might explain the lower titers of anti-CdtB in our study. In Brazil, 85% of Campylobacter strains did not express the CdtB antigen and this may have negatively impacted our results (23).
Antibodies raised against CdtB are thought to cross-react with vinculin due to molecular mimicry (24). An experimental study by Narcisi et al demonstrated the presence of vinculin in the cytoskeleton of Giardia duodenalis (25). Giardiasis affects 280 million individuals a year and its estimated prevalence in Brazil is around 30% (26). Therefore, the positivity of anti-vinculin amongst the four groups could be explained by the high prevalence of giardiasis and could possibly represent cross-reactivity.
Schmulson et al analyzed anti-CdtB and anti-vinculin antibodies in Mexican subjects with chronic diarrhea. Thirty patients with IBS-D, IBS-M, functional diarrhea, microscopic colitis, and tropical sprue were included. Antibody positivity was 55% in IBS subjects, similar to that in the original study by Pimentel et al (58.6%) (27).
Talley et al evaluated the diagnostic utility of anti-CdtB and anti-vinculin antibodies among individuals with IBS, functional dyspepsia, and healthy controls (28). The authors found no statistically significant difference in anti-CdtB and anti-vinculin levels between the groups. However, anti-CdtB results were numerically higher in IBS patients compared to controls (2.36 vs 2.14, p = 0.06). IBS diagnosis was based on clinical assessment, rather than the ROME IV criteria, and IBD subjects were not included. Furthermore, Talley et al assessed serum levels of the antibodies rather than plasma, which differed from our work as well as that of Pimentel et al.
Our study has some limitations. IBS-D subjects were not classified as PI-IBS. They were included if they had 3 or more bowel movements a day and met ROME IV criteria. According to the ROME IV, the diagnosis of PI-IBS requires the onset of symptoms following a resolution of an acute gastroenterocolitis, defined by positive stool culture in a symptomatic patient or by the presence of two of the symptoms, such as fever, vomiting or diarrhea (29). Since the anti-CdtB antibody results from the pathophysiology of PI-IBS, negative results should be analyzed with caution. Furthermore, the antibodies were not assessed at the time of the IBS-D diagnosis. Some authors have demonstrated that IBS symptoms, such as abdominal pain and diarrhea, begin in the first 6 months following infectious enteritis (30, 31). Clinical improvement occurs in 25% of patients in the first year and in 50% in 6 to 8 years.
The lack of statistical significance of the present study does not weaken the antibody-based approach to diagnose IBS-D. The antibodies were developed in light of the current knowledge of IBS pathophysiology: altered microbiota composition, dysmotility, increased intestinal permeability, immune dysregulation, and visceral hypersensitivity (24, 32, 33). However, in the IBD setting, they have not contributed to the diagnosis of superimposed IBS-D.
Our results cannot be generalized to the wider population. This study included complex subjects from a quaternary hospital, with long-term conditions treated with immunosuppressive drugs. IBS-D is a prevalent disease and non-invasive diagnosis reduces costs to the health system, avoids futile tests and guides proper treatment, as previously demonstrated in the literature (34). Future studies are needed for the external validation of anti-CdtB and anti-vinculin antibodies and to establish ideal cutoff values in different populations.