Serologic and laboratory markers in prediction of the disease course in inflammatory bowel disease

Abstract

The search for biologic markers that can assess the natural history and perhaps predict the course of individual’s disease including response to treatments over time has become an important focus of inflammatory bowel disease research. The knowledge of an individual’s prognosis can help physicians and patients make important management decisions and aid communication on risk and benefits of disease and treatment.

Key Words: Inflammatory bowel disease; Ulcerative colitis; Crohn’s disease; Perinuclear anti-neutrophil antibody; Anti-Saccharomyces cerevisiae antibody; Pouchitis; Internal penetrating; Fibrostenosing

INTRODUCTION

The ideal serological and laboratory marker should be able to identify individuals at risk for the disease and should be disease specific. Moreover it should be able to detect disease activity and monitor the effect of treatment; and finally it should have a predictive value towards relapse or recurrence of the disease. Predicting disease course however has now been expanded beyond just disease recurrence, but perhaps more importantly to include predictors of disease complications including surgery. This topic highlight will focus on the prognostic role of serological and laboratory markers and the importance of phenotype stratification in inflammatory bowel disease (IBD) patients.

SEROLOGICAL IMMUNE MARKERS

The search for the underlying trigger of the abnormal intestinal inflammatory reaction characteristic of IBD has led to the discovery of antibodies present specifically in the blood of patients with Crohn’s disease and/or ulcerative colitis (UC)[1-6]. Perinuclear anti-neutrophil antibody (pANCA) is noted for its association with UC or a UC like phenotype[7-13]. Anti-Saccharomyces cerevisiae antibody (ASCA) is another important antibody marker that is present in the blood of individuals with IBD[7-9]. ASCA is part of the family of anti-glycan (carbohydrate) antibodies. Other anti-glycan antibodies, antibodies against laminaribioside (ALCA) and chitobioside (ACCA) have been studied in IBD[14]. In addition to the anti-glycan antibodies, 3 additional markers representative of microbial driven immune responses have been identified; antibodies to the Escherichia coli (E. coli) outer-membrane porin C (OmpC), the Pseudomonas fluorescens CD related protein [anti-CD related bacterial sequence (I2)] and the CBir1 flagellin[15,16]. As seen with the genetic and clinical heterogeneity of CD, studies have shown immune response (immune phenotype) heterogeneity exists among CD patients. Landers et al[15] analyzed immune response heterogeneity in 330 patients and demonstrated that CD patients could be clustered into 4 distinct groups depending on their immune response patterns to microbial or auto-antigens. One cluster was ASCA, a second was antibodies to OmpC and I2, the third pANCA and the fourth was low or no immune response to any tested antigens. Immune reactivity to CBir1 may further define CD phenotypes in that anti-CBir1 expression is present in 40%-44% of pANCA positive CD patients vs only 4% in pANCA positive UC patients. This difference may denote a unique etiopathogenic mechanism of disease that helps to further stratify patients based on immunogenetic phenotypes.

Vasiliauskas et al[17] introduced the notion of immune response stratification when he first reported that high ASCA levels were found to be associated with fibrostenosing (FS) and internal-penetrating (IP) disease as well as the need for small bowel surgery. Another cross sectional study demonstrated that patients who were ASCA IgA or IgG positive were 8.5 times and 5.5 times more likely to undergo early surgery (within 3 years of diagnosis) than ASCA IgA or IgG negative patients[18]. Mow et al[19] examined the association of multiple immune responses and disease phenotype. Reactivity to OmpC was independently associated with IP disease, while reactivity to anti-I2 was independently associated with FS disease and the need for surgery. Both the presence and magnitude of the immune response was associated with more aggressive disease behaviors. A similar study in a Scottish CD cohort reported that the cumulative reactivity to ASCA, I2 and OmpC was associated with small bowel complications[20]. Antibodies to CBir1 were examined in a later study and were found to be independently associated with small bowel disease, IP and FS disease[16]. Xue et al[21] demonstrated that reactivity to ASCA, OmpC and CBir1 was associated with early disease onset, FS and IP disease and the need for surgery. The presence and magnitude of the serological response to anti-glycan antibodies were found to be associated with complicated disease, disease duration, and need for surgery[22]. Subsequently, increasing amount and level of antibody responses toward gASCA, ALCA, ACCA, AMCA, and OmpC were found to be associated with more complicated disease behavior (P < 0.0001) and need for surgery in CD (P = 0.023) in an eastern European IBD cohort[23].

More recent pediatric cohort studies suggest that these markers are present in patients before a complication occurs and thus predictive of disease progression from uncomplicated to complicated disease state. Desir et al[24] demonstrated that baseline ASCA reactivity was associated with a more relapsing course in a pediatric CD cohort (OR = 2.9; 95% CI: 1.33-6.35). A multi-center study examined the association of ASCA, anti-I2, anti-OmpC, and anti-CBir1 reactivity with disease course in close to 800 pediatric CD patients[25]. The frequency of development of disease complications (IP and or FS) increased in parallel with reactivity to increasing numbers of antigens. The odds ratio for the development of IP disease was 5.0 and 9.5 for children with reactivity to 2 and 3 antigens, respectively. Furthermore, survival analysis demonstrated that reactivity to at least one microbial antigen was associated with the development of surgery faster as compared to patients negative for all markers suggesting that these markers may predict more aggressive disease behaviors. Amre et al[26] also studied a cohort of pediatric CD patients who had sera drawn at diagnosis and were studied for the subsequent development of complications of their disease. Survival analysis revealed that the time to first complication was more rapid for ASCA positive patients than those who were ASCA negative. Moreover the relative risk (RR) of a recurrent complication (RR = 3.68) and needs for an additional surgery (RR = 1.95) was significantly higher in ASCA positive patients. As compared to the positive association between ASCA, antibodies to OmpC, I2 and CBir1 and disease complication, pANCA has been shown to be associated with a more benign, UC-like disease course and negatively associated with small bowel complicating disease[17,27]. High pre-colectomy levels of pANCA (> 100 EU/mL) have been prospectively shown to be associated with the development chronic pouchitis all IBD patients undergoing IPAA[28]. More recently, the same group reported that anti-CBir1 may accelerate the development of chronic pouchitis in the face of high pANCA levels[29]. Melmed demonstrated that ASCA positivity and a family history of Crohn’s disease was most predictive of CD of the pouch after IPAA[30]. This information may not change the need for colectomy but the surgical procedure chosen and the post operative management would certainly be impacted by this prognostic information.

FECAL CALPROTECTIN AND C-REACTIVE PROTEIN

Other biomarkers such as fecal calprotectin and C-reactive protein (CRP) have been examined as non invasive tools to predict disease course but with a focus on relapse or disease recurrence. Calprotectin is an abundant neutrophil protein found in both plasma and stool that is markedly elevated in infectious and inflammatory conditions including IBD. Calprotectin constitutes approximately 5% of the total protein and 60% of the cytosolic protein in human neutrophils[31]. Plasma calprotectin has been shown to increase 5 to 40 fold in infectious and inflammatory conditions[32] and fecal calprotectin concentration is approximately six times that of normal plasma[33]. The clinical course of most patients with IBD is typically marked by periods of remission with intermittent relapses. Because the timing of these relapses can be unpredictable, clinical symptoms more so than invasive tests are used to monitor disease activity. Most IBD relapses, however, come to medical attention after the inflammatory response has become well-established and clinical symptoms are worsening. Given that fecal calprotectin has been shown to correlate well with endoscopic and histologic intestinal inflammation, its use as a predictor of relapse in patients with quiescent IBD has been examined. Tibble et al[34] studied 43 patients with CD and 37 patients with UC in clinical remission for 1-4 mo as determined by clinical disease activity indices and Harvey Bradshaw index. These patients had fecal calprotectin measurements at baseline and were then followed for 1 year. Approximately half of the patients in each group relapsed during the 12 mo follow up period. The sensitivity and specificity of calprotectin for predicting relapse within the year in all patients with IBD were 90% and 83%, respectively by using a cut off of 50 mg/L. This cut off was associated with a 13-fold increased risk of relapse. These values were similar when patients with CD or UC were examined separately. Costa et al[35] found nearly identical sensitivity and specificity of calprotectin for predicting relapse in UC, but their specificity of calprotectin in CD was 43%, markedly lower than the 83% reported by Tibble et al[34] noted above. This large discrepancy may be explained by the fact that their study used a different assay and a lower cut-off of fecal calprotectin (150 μg/g, corresponding to 30 mg/L). A more recent study prospectively evaluated 97 patients with UC and 65 with CD in clinical remission for at least 6 mo[36]. The cutoff level was set at 130 mg/kg of feces. Patients were followed up for 1 year after the test or until relapse. Fecal calprotectin was positive in 44 UC patients and 26 of them relapsed within a year, while 11 of 53 UC patients with a negative calprotectin relapsed within the year. Thirty CD patients had a positive calprotectin and 13 of them relapsed within a year, as did 7 of the 35 with a negative test result. A significant correlation emerged between a positive calprotectin and the probability of relapse in UC patients (P = 0.000). In CD patients, only cases of colonic CD showed a significant correlation between a positive calprotectin test and the probability of relapse, such that 6 colonic CD patients were positive for the calprotectin test and 4 relapsed (P = 0.02). A subsequent multicenter study demonstrated that fecal calprotectin concentrations in patients who suffered a relapse were higher than in nonrelapsing patients P < 0.001)[37]. Relapse risk was higher in patients having high (> 150 μg/g) calprotectin concentrations (30% vs 7.8%, P < 0.001). The area under the receiver operating characteristic curve to predict relapse using calprotectin determination was 0.73 (0.69 for UC and 0.77 for CD). Better results were obtained when only colonic CD disease or only relapses during the first 3 mo were considered (100% sensitivity).

CRP is one of the most important acute phase proteins in humans. Under normal circumstances CRP is produced by hepatocytes in low quantities but following an inflammatory stimulus, hepatocytes rapidly increase production of CRP under the influence of interleukin (IL)-6, tumour necrosis factor α, and IL-1β. CRP has a short half life (19 h) compared with other acute phase proteins and will therefore rise early after the onset of inflammation and rapidly decrease after the stimulus is resolved. Although CRP is upregulated in most inflammatory diseases, including IBD, CD is associated with a strong CRP response and UC only a modest and sometimes absent CRP response[38,39]. Moreover not all IBD patients mount a CRP response, even in CD and this must be kept in mind when measuring inflammatory markers in individual patients. It is unclear if this is due to differences in cytokine levels such as IL-6 or due to mucosal vs transmural disease differences among UC and CD or is genetically driven[40-43]. CRP has been shown to be a good marker for predicting disease course and outcome in a number of diseases with cardiovascular disease and poor outcome after myocardial infarction as well as in multiple myeloma[44-47]. Boirivant et al[48] prospectively followed 101 CD outpatients with 50% having raised CRP which correlated well with disease activity. However, 1/3 of patients presented with active disease despite normal CRP and 1/3 had raised CRP but reported inactive disease. The likelihood of relapse after 2 years was higher in CD patients with an increased CRP compared with patients with normal CRP. Perhaps the best study to date on the predictability of CRP and other biomarkers was reported by the GETAID group who prospectively followed 71 CD patients with medically induced remission and measured laboratory markers (full blood count, CRP, ESR, α1 antitrypsin, orosomucoid) every 6 wk[49]. In total, 38 patients relapsed after a median follow up of 31 wk. In the end only two laboratory markers were predictive of relapse: CRP (20 mg/L) and ESR (15 mm). Patients with both markers positive had an 8 fold increased risk for relapse with a negative predictive value of 97%, suggesting that normal CRP and ESR has a very high probability of ruling out relapse in the preceding 6 wk. There are fewer studies examining the role of these markers in predicting UC recurrence or outcome. One study evaluated 49 severe hospitalized UC patients treated with steroids and/or cyclosporine. On day 3, a stool frequency of > 8/d or 3-8 stools/d plus an increased CRP (0.45 mg/L) predicted with 85% certainty the need for colectomy[50].

CONCLUSION

Research has fostered a novel approach to understanding the intricate relationship between genetic and clinical expression of disease. Serum immune markers hold the most promise in helping researchers better comprehend disease heterogeneity and natural history. Although our current gold standard diagnostic tests do not possess this capability, prospective research studies now suggest that IBD-specific genetic and antibody markers may serve as predictors of an individual’s disease course. Although promising at predicting disease recurrence, larger studies are needed with more frequent biomarker testing as this may allow for early detection of active, perhaps sub clinical inflammation which may alter treatment decisions. Their role as non-invasive predictors of post-operative recurrence is currently under investigation and will be of interest to help guide post-operative maintenance strategies.