External Validation of a Microscopic Colitis Clinical Scoring System in Patients With Chronic Watery Diarrhoea

Overview

Chronic watery diarrhoea is a very common problem in the population and most of these patients will be referred for colonoscopy. If no macroscopic findings are observed during colonoscopy to justify the diarrhoea, serial colonic biopsies will be taken to rule out Microscopic Colitis (MC). However, it has been estimated that only 10-15% of these patients will be diagnosed with MC after colonoscopy. Therefore, about 80% of the biopsies collected and analysed will not be useful to establish a diagnosis, considerably increasing costs. To predict the risk of developing MC, a new promising clinical scoring system has been recently developed. This score will be useful in the diagnostic work-up of chronic watery diarrhoea to prioritize colonoscopy with stepwise colonic biopsies in patients with a positive highly specific score for MC. In cases with a negative score, colonoscopy plus biopsies should be performed only if other diagnostic tests are negative. The aim of this current study is to externally validate the new scoring system to predict MC in patients with chronic watery diarrhoea.

Two diagnostic scoring systems have been proposed to predict the risk of developing MC: 1) The Kane score (sensitivity (SN) 96%; specificity (SP) 46%) combines eight risk factors, including female sex, age over 50 years, weight loss, absence of abdominal pain, and use of proton pump inhibitors (PPIs) and non-steroidal anti-inflammatory drugs (NSAIDs), presence of nocturnal diarrhoea and diarrhoea duration of less than 6 months. 2) The Cotter score (SN 93%; SP 49%) includes age ≥55 years, duration of diarrhoea ≤6 months, ≥5 bowel movements per day, body mass index \<30 kg/m2, current smoking, and current use of selective serotonin reuptake inhibitors/serotonin-norepinephrine reuptake inhibitors and NSAIDs. However, both scoring systems have been derived from retrospective studies and watery chronic diarrhoea, the hallmark clinical symptom of MC, was not well defined. Besides MC risk factors, faecal markers could also be useful for MC screening. Most studies have found that 60-75% of the patients with active collagenous colitis, a subtype of MC, have elevated faecal calprotectin levels. Moreover, a recent study showed that faecal calprotectin concentrations \>100 μg/g (AUC, 0.73) showed a 67% sensitivity and 75% specificity to predict MC. To further assess the value of the Kane and Cotter scoring systems and to derive a new score including faecal calprotectin aiming to increase specificity, the investigators performed a prospective observational two-centres study. Chronic watery diarrhoea was defined as ≥2 watery stools (Bristol 6-7) per day, of frequent occurrence (≥3 times per week), of at least 1 month duration. A registry of demographic and clinical characteristics (VAS of 0-100 of abdominal pain and abdominal distension, presence of nocturnal diarrhoea, urgency and faecal incontinence, smoking, body mass index, weight loss, use of drugs) was performed. The study included 118 patients with chronic watery diarrhoea, from which 41 were diagnosed with MC (21 lymphocytic colitis, 16 collagenous colitis and 4 paucicellular colitis) and obtained lower SN and SP values than those published of the Cotter (SN 78%; SP 57%) and Kane (SN 78%; SP 38%) scores. The AUC of both scores was 0.71 and 0.66, respectively. The multivariate analysis identified 5 variables associated with MC: \>5 stools/day (OR 12.5), duration of diarrhoea ≤8 months (OR 5), regular use of low-dose acetylsalicylic acid (ASA) (OR 4), BMI ≤26 Kg/m2 (OR 4.1) and faecal calprotectin \>500 µg/g (OR 5.5). A new score was developed using the variables mentioned above with an AUC of 0.86 (p\<0.001 vs. Kane and Cotter scores). A score \>10 had a sensitivity of 61.5% and a specificity of 92%. A score \>17 gave a specificity of 100% with a sensitivity of 36%. The score was internally validated using bootstrapping techniques. Although promising, the new scoring system must be externally validated before generalizing its use in clinical practice.