Can the systematic use of non-invasive biomarkers in patients with lower gastrointestional sysmtpoms help rationalise the demand for limited endoscopy resources?

Abstract

Introduction: Gastrointestinal complaints are very common in the general population. Demand for endoscopy continues to outstrip capacity, despite significant annual increase in the number of endoscopies performed nationwide, leading to expanding waiting lists, and potential delay in significant diagnoses. Hypothesis: We hypothesised that the demand on endoscopy can be optimised using investigation algorithms for common lower gastrointestinal complaints through the systematic use of a combination of non-invasive serum and faecal biomarkers. Methods: One retrospective and two prospective studies were carried out focusing on Colonoscopy outcomes in patients with lower GI symptoms and asymptomatic surveillance patients, to explore the above hypothesis looking at CRP, Faecal Calprotectin, symptom based diagnostic questionnaire and Faecal CologuardTM. We also performed a patient survey to examine potential factors influencing patient compliance with faecal testing. Results: Both retrospective and prospective studies of symptomatic patients showed mucosal inflammation was significantly associated with diarrhoea (p=0.008). Most Colonoscopies showed no evidence of inflammation, especially in patients with nondiarrhoeal symptoms (79-93%). Raised FCal (>50 ug/g) was strongly associated with inflammation (OR = 10.59), with a stronger relationship between inflammation and FCal >100 ug/g (OR=15.66). ROC curve suggested the diagnostic model of age, gender, F Cal and the questions relating to frequent loose/ mushy stool was best at predicting inflammation. FCal-CRP score was found to be very strongly associated with mucosal inflammation. The incidence of adenomas was low in young patients (8.7% ADR) compared to 24.7% in older symptomatic patients and 35.7% in surveillance patients. Faecal CologuardTM performed poorly in our study with NPV and PPV of 77.5% and 40%, with a sensitivity and specificity of 30.8% and 83.8%, respectively. Conclusion: We recommend the use of integrated investigation pathways that include a validated clinical symptom questionnaire, non-invasive serum and faecal inflammatory biomarkers in the initial assessment of patients referred with non-?red flag? lower gastrointestinal symptoms prior to referral for Colonoscopy. This allows accurate triage of patients who require Colonoscopy and those who can be reassured without the need for it, or who may have an alternative organic cause for their symptoms. A potential diagnostic algorithm extrapolated from our data, for young patient with diarrhoea is; exclude Coeliac disease and hyperthyroidism, check FCal-CRP score, and if all are negative, do not refer for Colonoscopy, and consider management in primary care. Colonoscopy may be avoided in young patients with non-diarrhoeal symptoms (Constipation, abdominal pain and bloating). For older patients with diarrhoea the same model can be applied, with the addition of a validated faecal marker for adenomatous polyps and CRC. Patient compliance with faecal testing remains a major limiting factor to the successful implementation of such pathways.