Introduction Analysis of faecal calprotectin (FC) is a non-invasive method for differentiating IBS from IBD. As a screening test, a normal FC has been employed to support a diagnosis of IBS, thereby avoiding invasive endoscopy in the patient cohort. We sought to investigate the diagnostic yield on subsequent colonoscopy in patients with mildly elevated FC and lower GI symptoms.
Methods Between November 2009 and November 2010, all patients with a FC value of 50–100 μg/g of stool (normal <50) were identified from our FC database. All of the stool samples were prepared and analysed according to the manufacturer's instructions (Bühlmann calprotectin ELISA kit). Patients were excluded from analysis if they were outwith the age range 16–50, had previous faecal calprotectin levels >100 ug/g stool, were known cases of IBD, had a history of NSAID intake, positive stool cultures, or any “alarm” GI symptoms. A similar cohort was identified with age & sex matched controls with FC values <50 μg/g. All the patients who did not go on to have a complete colonoscopy were removed from further analysis. Patients' records were analysed electronically using the NHS Great Glasgow & Clyde Clinical Portal.
Results 216 patients were identified with a FC of 50–100 μg/g. After exclusion criteria, 158 patients remained. Of these 82 underwent complete colonoscopy (mean age 36.7, M:F 1:2.2) which was abnormal in only six cases (three cases of a single adenoma<10 mm, one diverticulosis, one helminth infection & one non-specific acute inflammation). 280 patients were identified with a FC<50 μg/g. After exclusion criteria, 176 patients remained. Of these 65 underwent complete colonoscopy (mean age 36.6, M:F 1:2.3) which was abnormal in only eight cases (six cases of non-specific acute inflammation, one adenoma <10 mm & one diverticulosis). The colonoscopy outcome data, as expected, demonstrated that the pathology rate was very low in both groups. There was no difference in the rate of pathology detection between to two groups (p=0.3) and an FC<100 μg/g has an NPV of 88% to exclude any pathology or 100% for significant pathology (IBD, advanced adenoma or colonic carcinoma).
Conclusion In our population, the diagnostic yield of colonoscopy in patients below the age of fifty with new lower GI symptoms and a mildly elevated FC is very low. If our data can be replicated in a prospective manner, we suggest that invasive colonoscopy can be safely avoided in this cohort and interval FC analysis may be more appropriate.
Competing interests None declared.