P0570 - Biologic Therapy in Patients With Perianal Crohn’s Disease and Association With Long-Term Rates of Surgical and Clinical Outcomes

John Gubatan, MD¹, Spencer Frost, MD², Steven Levitte, MD, PhD², Kian Keyashian, MD^2
1Stanford University, Stanford, CA; ²Stanford University School of Medicine, Stanford, CA

Introduction: Perianal Crohn’s disease (pCD) represents an aggressive phenotype with limited studies on long-term outcomes. We evaluated long-term outcomes of these patients on biologics at the time of perianal disease diagnosis and the association of type of biologic use after first anti-TNF failure .

Methods: We performed a retrospective analysis of patients with pCD at a tertiary medical center through December 2020. We used Kaplan-Meier curves to estimate rates and multivariate logistic regression to identify predictors of outcomes.

Results: We included 311 patients with pCD (31 patients with ulcers, 72 with fissures, 236 with abscesses, and 262 with fistulas) of which 168 patients were on biologics (138 anti-TNF, 14 vedolizumab, 16 ustekinumab) at time of perianal disease diagnosis. At 5-year follow-up, biologic use (anti-TNF, vedolizumab, or ustekinumab) at time of perianal disease diagnosis was not associated with overall rates of any surgical intervention compared to no biologics. Anti-TNF use at time of diagnosis was associated with decreased rates of perianal abscess recurrence at 5 years (HR 0.48, 95% CI 0.32-0.74), whereas ustekinumab use was associated with increased rates of perianal fistula closure (HR 3.58, 95% CI 1.04- 12.35) and decreased rates of perianal abscess recurrence (HR 0.20, 95% CI 0.07- 0.56). Among patients who failed anti-TNF, switching to another anti-TNF was associated with decreased rates of colectomy (HR 0.20, 95% CI 0.04-0.90) and permanent diversion (HR 0.16, 95% CI 0.03-0.94) compared to ustekinumab, whereas vedolizumab use was associated with decreased rates of perianal fistula closure (HR 0.22, 95% CI 0.05-0.96) compared to ustekinumab. In multivariate analysis, baseline BMI (OR 1.04, 95% CI 1.01-1.08, P=0.037), presence of perianal abscess (OR 10.37, 95% CI 4.45-24.14, P< 0.0001), perianal fistula (OR 5.50, 95% CI 2.47-12.25, P< 0.0001), and antibiotic use (OR 2.57, 95% CI 1.31-5.07, P=0.006) were independently associated with increased odds of any surgical intervention at 5 years. Predictors of colectomy and permanent diversion at 5 years among patients with pCD included colonic disease (Montreal L2) and anal stenosis.

Discussion: Anti-TNF and ustekinumab use at time of perianal disease diagnosis is associated with improved long-term outcomes. Among pCD patients who fail anti-TNF, switching to another anti-TNF may be more favorable to ustekinumab, whereas switching to ustekinumab may be more favorable to vedolizumab with regards to long-term outcomes.


Figure: A) Anti-TNF use (N= 138) at time of perianal disease diagnosis was associated with decreased rates of perianal abscess recurrence at 5 years (HR 0.48, 95% CI 0.32-0.74, P < 0.001) compared to no biologic use (N=143). Ustekinumab use (N=16) was associated with increased rates of perianal fistula closure (HR 3.58, 95% CI 1.04- 12.35, P < 0.05) and decreased rates of perianal abscess recurrence (HR 0.20, 95% CI 0.07- 0.56, P < 0.01) at 5 years compared to no biologic use (N=143).
B) Anti-TNF use after first Anti-TNF failure (N=70) was associated with decreased rates of colectomy (HR 0.20, 95% CI 0.04-0.90, P < 0.05) and permanent diversion (HR 0.16, 95% CI 0.03-0.94, P < 0.05) compared to Ustekinumab use (N=17). Vedolizumab use after first anti-TNF failure (N=8) was associated with decreased rates of perianal fistula closure (HR 0.22, 95% CI 0.05-0.96, P < 0.05) compared to Ustekinumab use (N=17).

Table 1. Baseline Clinical Characteristics of Patients with Perianal Crohn's Disease
Clinical Variables	Perianal Crohn's Disease Patients, N= 311
Age
CD Diagnosis, years (mean ± SD)	25.9 (± 13.8)
Perianal Disease Diagnosis, years (mean ± SD)	32.5 (± 15.6)
Timing of Perianal Disease Diagnosis
Diagnosed Before CD Diagnosis, no, (%)	5 (1.6)
Diagnosed During CD Diagnosis, no, (%)	104 (33.4)
Diagnosed After CD Diagnosis, no, (%)	202 (65.0)
Sex
Male, no. (%)	193 (62.1)
Female, no. (%)	118 (37.9)
Ethnicity
White, no. (%)	158 (50.8)
Hispanic, no. (%)	25 (8.0)
Black, no. (%)	17 (5.5)
Asian, no. (%)	51(16.4)
Native American, no. (%)	2(0.6)
Other, no. (%)	58(18.7)
Montreal Classification Location at CD Diagnosis
L1 , no. (%)	19 (6.1)
L2, no. (%)	132 (42.4)
L3, no. (%)	158(50.8)
L4, no. (%)	2 (0.64)
Montreal Classification Behavior at CD Diagnosis
B1, no. (%)	13 (4.2)
B2, no. (%)	14 (4.5)
B3, no. (%)	284 (91.3)
Perianal Crohns Disease Manifestation
Ulcer, no. (%)	31 (10.0)
Fissure, no. (%)	72 (23.2)
Abscess, no. (%)	236 (76.1)
Fistula, no. (%)	262 (84.2)
Mode of Perianal CD Diagnosis
Physical Exam, no. (%)	249 (80.1)
Pelvic MRI, no. (%)	154 (49.5)
Endoscopy/EUS, no. (%)	4 (1.29)
Exam Under Anesthesia (EUA), no. (%)	205 (65.9)
BMI, kg/m2 (mean ± SD)	22.3 (± 9.6)
Smoker, no. (%)	10 (3.2)
Alcohol Use, no. (%)	111 (35.9)
Medications at time of Perianal CD Diagnosis
Steroids, no. (%)	78 (25.1)
5-ASA, no. (%)	95 (30.9)
6MP/Azathioprine, no. (%)	68 (22.2)
Methotrexate, no. (%)	25 (8.1)
Infliximab, no. (%)	73 (23.4)
Adalimumab, no. (%)	57 (18.6)
Golimumab, no. (%)	2 (0.7)
Certolizumab, no. (%)	6 (2.0)
Ustekinumab, no. (%)	16 (5.2)
Vedolizumab, no. (%)	14 (4.6)
NSAID use, no. (%)	18 (6.1)

Disclosures:
John Gubatan indicated no relevant financial relationships.
Spencer Frost indicated no relevant financial relationships.
Steven Levitte indicated no relevant financial relationships.
Kian Keyashian indicated no relevant financial relationships.


John Gubatan, MD¹, Spencer Frost, MD², Steven Levitte, MD, PhD², Kian Keyashian, MD². P0570 - Biologic Therapy in Patients With Perianal Crohn’s Disease and Association With Long-Term Rates of Surgical and Clinical Outcomes, ACG 2021 Annual Scientific Meeting Abstracts. Las Vegas, Nevada: American College of Gastroenterology.