Occurrence and aetiology of gastrointestinal perforation in patients with vasculitis.

Objective. This study aimed to characterise the presenting features and outcomes of patients with vasculitis and gastrointestinal perforation. Methods. Using a retrospective cohort design, this study included 20 cases with verified vasculitis and gastrointestinal perforation at Mayo Clinic, Rochester, USA, between 1998 and 2017. Results. Four of the twenty cases experienced vasculitis-induced perforation. Cases with perforations due to vasculitic involvement had more small bowel involvement, longer duration of abdominal pain prior to perforation (41 days vs. 0 days, p=0.005), and a higher proportion of active tobacco use (75% vs. 7%, p=0.01) compared to the cases with non-vasculitis perforation. A majority (88%) of the non-vasculitis perforations were associated with glucocorticoid use. The median cumulative glucocorticoid dose prior to perforation in patients with additional, non-vasculitic risk factors for perforation was 4,320 mg prednisone and was 22,170 mg for those without additional risk factors. Mortality rates for the whole cohort were higher than the general population (standardised mortality ratio: 2.19, 95% confidence interval 1.05-4.02). The cases with vasculitis-induced perforation tended to have increased number of surgeries and length of stay compared to the non-vasculitis cases; however, those differences failed to reach statistical significance. Conclusion. Small bowel location and longer abdominal pain duration may help distinguish vasculitis-induced bowel perforation from other aetiologies. Overall mortality in patients with vasculitis and bowel perforation is increased, highlighting the importance of prompt diagnosis and management.