Abstract
Background: Patients with axial spondyloarthritis (axSpA) frequently exhibit both small and large bowel inflammation. This finding has been documented in numerous studies using ileocolonoscopy and capsule endoscopy, where the prevalence of gut inflammation is estimated to range from 40% to 65%. Gut lesions in axSpA are also known to correspond with non-invasive markers of inflammation, including elevated C-reactive protein (CRP) and faecal calprotectin. However, symptomatic Inflammatory bowel disease (IBD) is observed in only ~8% of patients with axSpA, with the remainder described as having silent, asymptomatic or subclinical bowel inflammation. Considering this evidence, we hypothesized that gut inflammation was unlikely to be entirely asymptomatic in patients with axSpA. Previously, we validated a questionnaire for assessing bowel symptoms in axSpA, the Dudley Inflammatory Bowel Symptoms Questionnaire -axSpA modification (DISQ)[1], in a study comparing bowel symptoms in people with axSpA, Crohn's disease (CD), and healthy controls. Whilst those with axSpA reported milder symptoms than people with CD, they had a significantly higher prevalence of bowel symptoms compared with healthy controls.
Objectives: In this study we aimed to determine 1) the frequency and severity of bowel symptoms in a large cohort of people with axSpA; (2) the correlation between bowel symptoms and non-invasive indicators of gut inflammation, axSpA disease activity, and quality of life; and (3) the persistence of bowel symptoms assessed at yearly intervals.
Methods: A cohort of 370 people with axSpA who fulfilled the ASAS criteria, were recruited across six New Zealand centres. Participants completed the DISQ score (scored 0-60) at baseline, 6 months, and then annually for three years. Faecal samples for calprotectin analysis were available for 205 participants at baseline, with levels >50 μg/g considered abnormal. The following variables were measured at each visit: CRP, the Ankylosing Spondylitis Disease Activity Score (ASDAS-CRP), the Ankylosing Spondylitis Quality of Life score (ASQoL) Medication use was documented at each visit: Non-steroidal anti-inflammatories (NSAIDs), and anti-TNF therapies (TNFi's). DISQ scores and ASDAS-CRP were correlated with ASQoL scores using Pearson's correlations.
Results: At baseline, mean DISQ scores were 13.5 (SD 9.5), mean ASDAS-CRP scores were 2.74 (SD 1.31) and mean CRP 10.41 (SD 14.45). A total of 9.5% of participants were taking TNFi's alone, 61.9% NSAIDs alone, 13.8% both TNFi's and NSAIDs, and 14.9% were on no treatment. At baseline mean faecal calprotectin was 109.1 (SD 147.9) and showed a positive correlation with DISQ scores (r(1)=0.146 p=0.04). The DISQ correlated positively with ASQoL (r(1) = 0.285, p<0.001) and with ASDAS-CRP (r(1) = 0.199, p<0.001). Categorising participants according to previously defined clinically relevant subsets [1], 28% had a DISQ score between 11 and 19 (highly symptomatic), 27.84% >19 (scores comparable with patients who have active inflammatory bowel disease) and 32.16% <7.6 (normal to low). At baseline, ASQoL score correlated significantly with DISQ (one-way ANOVA p<0.001. n=322). Participants with DISQ ≤7.6 (32%) had significantly lower mean ASQoL scores than both participants with DISQ scores ≥11and ≤19 (28% of participants, Tukey's p<0.001) and with DISQ ≥19 (28% of participants, Tukey's p<0.001). Repeated measures showed that over time there was a significant improvement in ASDAS-CRP scores (one-way ANOVA, p<0.001), but no significant change in DISQ scores (one-way ANOVA, p=0.486). Tukey's post-hoc tests, showed DISQ scores were significantly lower across all time points in participants treated with TNFi's as monotherapy, compared with those on no medication (p=0.028). There was no difference between participants treated with NSAIDs alone and those on no medication (p=0.723) or between those on NSAIDs alone and those on TNFi monotherapy (p=0.06).
Conclusion: Clinically relevant bowel symptoms are highly prevalent in people with axSpA and correlate with elevated faecal calprotectin and axSpA disease activity measures at baseline. There is a strong association between diminished quality of life and adverse bowel symptoms. Whilst disease activity scores decreased over time, bowel symptoms remained static in the cohort overall. In contrast, bowel symptoms improved over time in participants treated with TNF-inhibitor monotherapy. These findings provide evidence that axSpA is associated with adverse bowel symptoms which correlate with non-invasive markers of gut inflammation and adversely impact quality of life.
Oral Presentation.