Impact of timing and duration of thiopurine treatment on first perianal surgery in Crohn's disease: UK population-based study (1995-2009).

Abstract

The impact of thiopurine (TP) use on perianal surgery is uncertain. Our aim was to determine trends in perianal surgery and the impact of timing and duration of TPs on the risk of first perianal surgery. We identified a population-based cohort of incident cases of Crohn's disease between 1995 and 2009. We used Kaplan-Meier analysis to determine trends in TP usage and first perianal surgery by era of diagnosis: era 1 (1995-2002) and era 2 (2003-2009). We quantified the impact of duration and timing of TPs on the risk of perianal surgery using a Cox regression model. We identified a cohort of 5235 incident cases of Crohn's disease. The 5-year cumulative probability of first perianal surgery decreased from 2.7% to 1.7% between era 1 and era 2, respectively (P = 0.03). TP use for greater than 18 months was associated with a 40% risk reduction for first perianal surgery (hazard ratio: 0.60, 95% confidence interval: 0.39-0.95) and 49% if TPs were used for 2 years or more (hazard ratio: 0.51, 95% confidence interval: 0.32-0.99). There was no demonstrable additional benefit from early TP use within the first year after diagnosis (hazard ratio: 0.85, 95% confidence interval: 0.52-1.40, P = 0.53). Over the past 15 years, TP use has increased by 50%, whereas perianal surgery rates have decreased by 37% among UK population with Crohn's disease. Sustained use for 18 months was associated with a reduced risk of perianal surgery by almost a half in the first 5 years after diagnosis.