Biomarkers of connective tissue turnover in inflammatory bowel disease - diagnostic and predictive capacity

Chronic inflammatory disease (CID) including inflammatory bowel disease (IBD) constitute idiopathic, prevalent and potentially debilitating diseases. Albeit heterogeneity, they express similarities in symptomatology e.g., fluctuating disease pattern and overlapping co-morbidity. Biological medications targeting the proinflammatory tumor necrosis factor (TNF) are predominantly prescribed. Currently, a significant part of patients (up to 60%) has no or insufficient clinical treatment response, pinpointing the lack of adequate management. With a lifelong perspective and an overall increasing disease burden, CID inevitably constitutes a significant socio-economic burden. If focusing on present clinical challenges in IBD management, early diagnosis and timely and adequate therapeutic treatment stand out. Early diagnosis may pose a clinical challenge, due to the absence of a golden standard, and relative dependency on endoscopic procedures. Delayed diagnosis and suboptimal treatment may negatively affect disease progression (irreversible organ damage), timely therapeutic management and ultimately quality of life for patients and relatives.

A hallmark of chronic inflammation is excessive extracellular matrix (ECM) remodeling and disturbed connective tissue turnover. Biomarkers of the ECM mirroring inflammation, connective tissue turnover, and fibrosis/fibrogenesis have previously demonstrated diagnostic and predictive utility. Therefore, it was plausible that these biomarkers of ECM reflecting pathological processes had potential as clinical tools for aiding IBD diagnosis, prognosis, and prediction of treatment response. This thesis was based upon the planning, coordination and conduction of the observational NORDTREAT cohort study (consisting of a cross-sectional study and a longitudinal study) partnering in a Nordic collaboration with prospective enrolment of patients in Denmark, Sweden, Norway and Iceland. A protocol article highlighting details of study design, methodology, statistics, primary and secondary outcomes was published, as part of this thesis. The NORDTREAT cross-sectional study and the BELIEVE cohort study constituted the foundation of the serological investigations of ECM biomarkers in the two clinical challenges within IBD management presented above.

The NORDTREAT cross-sectional study enrolled 241 adult patients referred on suspicion of IBD (symptomatic and treatment naïve i.e., clinical relevant setting). Patients had blood samples taken at baseline, and were subsequently diagnosed with either Crohn’s disease (CD), ulcerative colitis (UC), IBDunclassified (IBD-U) or symptomatic controls. ELISA was conducted to assess levels of biomarkers of tissue fibrosis: CTX-III, PRO-C4, PRO-C5, and PRO-C16, immune cell activity (VICM), and mucosal damage (C3M, C4M, C6M), respectively. The primary aim of the NORDTREAT cross-sectional study was to explore the utility of a panel of ECM biomarkers on diagnosis of IBD (CD, UC, IBD-U) versus symptomatic controls at first clinical visit in patients referred to a gastroenterology department with suspected IBD. The logistic prediction model achieving the best distinction of patients with IBD versus symptomatic controls included CTXIII and C6M (AUC 0.64, 95% CI 0.56 - 0.70). Subsequent post-hoc analysis of the discriminative utility of the model to separate patients with colonic CD from patients diagnosed with UC included the biomarkers VICM, C4M, PRO-C5, and PRO-C16. This predictive model provided an AUC of 0.80 (95% CI 0.69 - 0.98).

The BELIEVE cohort study (finalized in 2020) prospectively enrolled 233 patients with CID including CD, UC, rheumatoid arthritis (RA), axial spondyloarthritis (axSpA), psoriatic arthritis (PsA), and psoriasis (PsO). Adult patients initiating or switching (after anti-TNF failure) biological therapy were eligible for participation. Blood samples were collected at baseline and 14-16 weeks after initiating biological therapy, and levels of eighteen ECM biomarkers were measured by ELISA. The primary aim of the BELIEVE study was to explore the clinical utility of ECM biomarkers as predictors of biological treatment response in a population of patients with CID. Disease-specific biological treatment response was achieved in 54% of the study population with CID (ranging from 44% to 65% in the subgroups of CID). The utility of the single ECM biomarkers for prediction of disease specific treatment response was addressed by measuring the AUC (values ranging from 0.47 to 0.59). An explorative principal component analysis was conducted and revealed a clustering of the ECM biomarkers in five groups/components. However, the logistic regression analysis did not demonstrate any significant association between the five components retrieved and clinical treatment response.

In conclusion, the NORDTREAT study documented, that even though exploring previously applied and thoroughly investigated ECM biomarkers, neither of the single candidate biomarkers nor profile of biomarkers had sufficient diagnostic utility (predefined lower limit of AUC of 0.80). However, post hoc analysis revealed a prediction model successfully discriminating patients with colonic CD from patients with UC. The BELIEVE study documented, that despite of multiple common features in patients with CID and application of an intensively assessed ECM biomarker panel, we did not succeed retrieving any single ECM biomarker nor biomarker profile predictive of biological treatment response.

With future prospects of sharply increasing prevalence of CIDs the search for valid clinical tools enabling early diagnosis, prognosis of disease course and reliable prediction of treatment response is as relevant as ever. To mitigate the inevitable consequences of chronic inflammation, increase the quality of live for patients, and minimize scarce health care resources (e.g., spent on hospitalizations, endoscopic procedures and excessive medication), cost-efficient and minimally invasive clinical measures to distinguish similar but essentially different diseases, and enabling prompt and efficient therapeutic intervention is of immense importance.