A longitudinal study of fecal calprotectin and the development of inflammatory bowel disease in ankylosing spondylitis
In this longitudinal 5-year study of a well-characterized cohort of patients with AS, we followed the intraindividual variation in fecal calprotectin in relation to disease activity and medication, and we determined the incidence and predictors for IBD. Fecal calprotectin was elevated (50 mg/kg) in approximately two-thirds of the patients at both the baseline and 5-year follow-up visits. The intraindividual variation was moderate; in 80% of the patients, fecal calprotectin changed 200 mg/kg between the two measurement time points. Smoking and elevated fecal calprotectin at baseline were the strongest predictors for high fecal calprotectin at 5-year follow-up. Of the 204 patients included at baseline, 1.5% were diagnosed with CD and 0.5% with lymphocytic colitis at 5-year follow-up. High fecal calprotectin and presence of mucus in diarrhea at baseline were the main predictors of development of CD in this cohort.
Our results indicate that gut inflammation in AS is associated with higher disease activity in rheumatic disease. Increased fecal calprotectin at baseline was associated with higher ASDASCRP, BASDAI, BASFI, CRP, and ESR at 5-year follow-up. Early prospective studies showed a link between active gut inflammation and persistence of joint and spine inflammation as well as more chronic radiographic changes in the sacroiliac joint and spine [18, 27]. Researchers in two cross-sectional studies reported higher fecal calprotectin in AS than in healthy control subjects [28, 29]. One of the studies showed a positive correlation between fecal calprotectin and BASDAI, BASFI, ESR, and CRP [29]. In studies with the Belgian Inflammatory Arthritis and spoNdylitis cohorT (GIANT) of early SpA, the presence of histological gut inflammation was associated with higher BASDAI and higher BASMI and was also linked to a greater degree of bone marrow edema in the sacroiliac joints visualized by magnetic resonance imaging [13, 30]. The longitudinal results of our study support they hypothesis that inflammation in the gut and the musculoskeletal system may be linked. If gut inflammation is the cause or the consequence of locomotor disease is unknown, however, and fecal calprotectin is only a surrogate measure for gut inflammation.
In a recent report based on the GIANT cohort, fecal calprotectin, serum calprotectin, and CRP were also significantly higher in patients with microscopic gut inflammation [31]. We found that serum calprotectin at baseline was positively correlated with fecal calprotectin, both at baseline and at 5-year follow-up, but serum calprotectin was not a significant predictor for the development of CD.
In the present study, the use of anti-TNF antibodies was associated with lower fecal calprotectin, whereas use of TNF receptor fusion proteins was associated with higher levels of fecal calprotectin, at 5-year follow-up. Infliximab and adalimumab, but not etanercept, are efficacious in the treatment of CD and UC [32]. Similarly, in earlier studies, more flares and onset of IBD were observed in patients with AS treated with etanercept than in patients treated with infliximab [33, 34].
The 5-year incidence of CD in the present study was 1.5%, which corresponds to an annual incidence of 294 new cases per 100,000 person-years. This can be compared with the highest reported annual incidence of CD in the general population of 12.7 per 100,000 person years in Europe and 20.2 per 100,000 person-years in North America [35]. Mielants et al. reported an incidence of IBD of 5% in a cohort of 217 patients with SpA or AS during a median follow-up of 5.5 years [19]. Similarly to our study, CD was more common than UC in their study. In contrast to our study, ileocolonoscopy was performed on all patients at inclusion, and 49 patients underwent a second endoscopy. Because patients with fecal calprotectin 200 mg/kg were not examined with endoscopy in our study, cases with IBD despite lower fecal calprotectin may have been missed. Moreover, the earlier studies were performed in the “prebiologic era,” which may have affected the results. Furthermore, a few patients with persistently elevated fecal calprotectin declined ileocolonoscopy.
In our study, elevated fecal calprotectin and presence of mucus in diarrhea at baseline were the only significant predictors of the development of CD. The low number of new cases of CD (n?=?3) may have affected the robustness of the results and the possibility of finding other significant predictors.
The high use of NSAIDs among patients with AS is a problem when studying gut inflammation. NSAID-induced enteropathy is difficult to distinguish from AS-associated gut inflammation and is associated with increased fecal calprotectin [23]. In the present study, NSAID use was stopped for 3 weeks before retesting fecal calprotectin and before ileocolonoscopies, which resulted in a drop of a median of 116 mg/kg in fecal calprotectin. We found that 3-week cessation of NSAID use before testing fecal calprotectin was a reasonable course of action and acceptable for most patients.
The strengths of the present study are its prospective design with two measurement time points, the relatively large and well-characterized cohort, and cessation of NSAIDs before ileocolonoscopy at follow-up. The study’s weaknesses are that ileocolonoscopy was not performed in all patients, which may have resulted in a lower incidence of IBD. The ileocolonoscopies were also performed at different thresholds of fecal calprotectin at baseline and 5-year follow-up (Fig. 1).