Graham D Naismith1, Lyn A Smith2, Sarah J E Barry3, Joanna I Munro4, Susan Laird5, Karen Rankin6, Allan J Morris7, Jack W Winter8, Daniel R Gaya9. 1. Department of Gastroenterology, Glasgow Royal Infirmary, Glasgow, United Kingdom. Electronic address: gnaismith@doctors.org.uk. 2. Department of Gastroenterology, Glasgow Royal Infirmary, Glasgow, United Kingdom. Electronic address: lynsmith1@nhs.net. 3. Robertson Centre for Biostatistics, Institute of Health and Wellbeing, University of Glasgow, Glasgow, United Kingdom. Electronic address: sarah.barry@glasgow.ac.uk. 4. Department of Gastroenterology, Glasgow Royal Infirmary, Glasgow, United Kingdom. Electronic address: Joanna.munro@ggc.scot.nhs.uk. 5. Department of Gastroenterology, Glasgow Royal Infirmary, Glasgow, United Kingdom. Electronic address: susan.laird@ggc.scot.nhs.uk. 6. Department of Clinical Biochemistry, Glasgow Royal Infirmary, Glasgow, United Kingdom. Electronic address: karen.rankin@ggc.scot.nhs.uk. 7. Department of Gastroenterology, Glasgow Royal Infirmary, Glasgow, United Kingdom. Electronic address: john.morris@ggc.scot.nhs.uk. 8. Department of Gastroenterology, Glasgow Royal Infirmary, Glasgow, United Kingdom. Electronic address: jack.winter@ggc.scot.nhs.uk. 9. Department of Gastroenterology, Glasgow Royal Infirmary, Glasgow, United Kingdom. Electronic address: daniel.gaya@ggc.scot.nhs.uk.
Abstract
BACKGROUND: Faecal calprotectin (FC) is a non-invasive marker of gastrointestinal inflammation. AIM: To determine whether higher FC levels in individuals with quiescent Crohn's disease are associated with clinical relapse over the ensuing 12 months. METHODS: A single centre prospective study was undertaken in Crohn's disease patients in clinical remission. The receiver operating characteristic (ROC) curve for the primary endpoint of clinical relapse by 12 months, based on FC at baseline, was calculated. Kaplan-Meier curves of time to relapse were based on the resulting optimal FC cutoff for predicting relapse. RESULTS: Of 97 patients recruited, 92 were either followed up for 12 months without relapsing, or reached the primary endpoint within that period. Of these, 10 (11%) relapsed by 12 months. Median FC was lower for non-relapsers, 96 μg/g (IQR 39-237), than for relapsers, 414 μg/g (IQR 259-590), (p=0.005). The area under the ROC curve to predict relapse using FC was 77.4%. An optimal cutoff FC value of 240 μg/g to predict relapse had sensitivity of 80.0% and specificity of 74.4%. Negative predictive value was 96.8% and positive predictive value was 27.6%, FC ≥240 μg/g was associated with likelihood of relapse by 12-months 12.18 (95% CI 2.55-58.2) times higher than lower values (p=0.002). CONCLUSIONS: In this prospective dataset, FC is a useful tool to help identify quiescent Crohn's disease patients at a low risk of relapse over the ensuing 12 months. FC of 240 μg/g was the optimal cutoff in this cohort.
BACKGROUND: Faecal calprotectin (FC) is a non-invasive marker of gastrointestinal inflammation. AIM: To determine whether higher FC levels in individuals with quiescent Crohn's disease are associated with clinical relapse over the ensuing 12 months. METHODS: A single centre prospective study was undertaken in Crohn's diseasepatients in clinical remission. The receiver operating characteristic (ROC) curve for the primary endpoint of clinical relapse by 12 months, based on FC at baseline, was calculated. Kaplan-Meier curves of time to relapse were based on the resulting optimal FC cutoff for predicting relapse. RESULTS: Of 97 patients recruited, 92 were either followed up for 12 months without relapsing, or reached the primary endpoint within that period. Of these, 10 (11%) relapsed by 12 months. Median FC was lower for non-relapsers, 96 μg/g (IQR 39-237), than for relapsers, 414 μg/g (IQR 259-590), (p=0.005). The area under the ROC curve to predict relapse using FC was 77.4%. An optimal cutoff FC value of 240 μg/g to predict relapse had sensitivity of 80.0% and specificity of 74.4%. Negative predictive value was 96.8% and positive predictive value was 27.6%, FC ≥240 μg/g was associated with likelihood of relapse by 12-months 12.18 (95% CI 2.55-58.2) times higher than lower values (p=0.002). CONCLUSIONS: In this prospective dataset, FC is a useful tool to help identify quiescent Crohn's diseasepatients at a low risk of relapse over the ensuing 12 months. FC of 240 μg/g was the optimal cutoff in this cohort.