Predictors of successful fistula management

Predictors of successful fistula management

August 6, 2020

Issue 15

Clinical Question

What are the predictors of perianal fistula relapse in Crohn’s disease (CD)?

Editor’s Bottom Line

Continued treatment with a TNF antagonist is associated with reduced post-surgical recurrence of perianal fistulas in CD. Surgery is not a cure, and a combined medical/surgical approach is optimal.


Malian A, Rivière P, Bouchard D, et al. Pedictors [sic]of Perianal Fistula Relapse in Crohn’s Disease. Inflamm Bowel Dis. 2020;26(6):926–31;


Researchers in France set out to identify predictors of perianal fistula recurrence among patients treated successfully for this complication of CD. As the authors noted, up to 30% of patients treated for a perianal fistula experience recurrence within one year.

They retrospectively examined data from 137 consecutive CD patients treated surgically for a total of 157 perianal fistulae at their hospital between 2007 and 2015. All patients experienced initial remission within three months of surgery, defined by the absence of local pain or drainage. Participants were a median 35 years of age at the time of treatment and 57% were male. Roughly 76% received an anti-tumor necrosis factor (TNF) agent following surgery.

They found that during a median 43 months of follow-up (Range: 26– 64 months), 22% of patients experienced a fistula relapse—defined as subsequent perianal draining fistula or an abscess confirmed clinically or through pelvic MRI. The median time to recurrence among these individuals was 22 months (Range: 16–33 months). Approximately 97% of patients were recurrence-free one year after drainage, while 78.4% and 74.4% were recurrence-free at three and five years, respectively. Discontinuation of anti-TNF treatment increased the risk of relapse by 3.5 times (Odds Ratio [OR]: 3.49; P=0.04). Other significant risk factors included disease limited to the colon (OR: 6.25; P=0.01) and a stricturing disease phenotype (OR: 4.39; P=0.01). Discontinuing an immunosuppressant drug was associated with a lower likelihood of fistula relapse, a finding that the authors speculated likely reflected these patients’ positive response to treatment, rather than true causation. Fistula relapse rates did not differ by anti-TNF drug used.


Study Design: Retrospective cohort
Funding: Not disclosed
Allocation: Not applicable
Setting: Single-center database
Level of Evidence: 2b (Oxford Levels of Evidence)