Search in the Abstract Database

Abstracts Search 2018

OP002 Assessment of disease activity patterns during the first 10 years after diagnosis in a population-based Crohn’s disease cohort shows a quiescent disease course for a substantial proportion of the population

D. Wintjens1,2*, F. Bergey3, E. Saccenti4, S. Jeuring1,2, M. Romberg-Camps5, L. Oostenbrug5, A. Masclee1,2, D. Jonkers1,2, V. Martins dos Santos3,4, M. Pierik1,2

1Maastricht University Medical Center, Department of Internal Medicine, Division of Gastroenterology and Hepatology, Maastricht, The Netherlands, 2Maastricht University Medical Center, NUTRIM School for Nutrition and Translational Research in Metabolism, Maastricht, The Netherlands, 3LifeGlimmer GmbH, Berlin, Germany, 4Wageningen University & Research, Laboratory of Systems and Synthetic Biology, Wageningen, The Netherlands, 5Zuyderland Medical Centre, Department of Gastroenterology and Hepatology, Sittard-Geleen/Heerlen, The Netherlands

Background

Representative studies concerning the long-term prognosis and disease course in Crohn’s disease (CD) primarily describe steroid exposure, need for surgery or hospitalisations, and disease progression as characteristics of an unfavourable outcome. Real-life data on long-term disease activity are lacking. We aimed to define clusters with different disease activity patterns in the population-based IBDSL cohort.

Methods

All CD patients from the IBDSL cohort with at least 10 years follow-up (>18 years, diagnosed between 1991 and 2004) were included. Data on demographics, disease phenotype, medication use, hospitalisations, and surgery were available. In addition, all endoscopy and imaging reports were scrutinised. Since diagnosis, active disease was defined for each yearly quarter by (i) active disease on endoscopy or imaging, (ii) hospitalisation, (iii) surgery, or (iv) treatment adjustment for increased symptoms. Subsequently, formula-based clusters were generated based on four previously published questionnaire-based disease activity patterns,1 completed with two additional clusters (Figure 1). Prediction models were created using discriminant analysis with PLS regression based on characteristics at baseline and 6 months after diagnosis.

Results

In total, 432 patients were included. During 10 years follow-up after diagnosis, patients experienced 4.2 (SD 3.8) quarters of active disease on average. The distribution of patients over different clusters is shown in Figure 1. Notably, 128 patients (29.6%) were classified as quiescent and, of these, 89.8% never received immunomodulators or biologics. Ileocolonic disease location (OR 0.45; 95% CI 0.21–0.91) and smoking at diagnosis (OR 0.44; 95% CI 0.26–0.70) were negatively associated with a quiescent disease course, while surgery at diagnosis (OR 3.02; 95% CI 1.39–6.64) was positively associated. Our best prediction model for a quiescent course had an area under the ROC curve of 0.72 (p < 0.001) at baseline and 0.75 (p < 0.001) at 6 months after diagnosis.

Figure. IFX serum concentrations productive vs. closed fistula.

Figure 1. Formula-based disease activity clusters based on the first 10 years after diagnosis.

Conclusion

This first population-based clustering analysis of disease activity patterns in CD shows that a substantial group of patients experience a quiescent disease course during the first 10 years and early clinical markers can only moderately predict this. This study underscores the importance of patient stratification at baseline to prevent under- and overtreatment. Further studies are warranted to confirm our findings and to identify predictive biochemical, molecular or genetic markers.

Reference

1.Solberg IC, Vatn MH, Høie O, et al. Clinical course in Crohn’s disease: results of a Norwegian population-based ten-year follow-up study. Clin Gastroenterol Hepatol, 2007;5:1430–8.