麻豆传媒

Despite higher utilization of immunomodulators and biologics, perianal Crohn's disease (PCD), especially the severe phenotype, is associated with poor outcomes, a large population-based study from Israel found.

Data from the epidemiology group of the Israeli IBD Research Nucleus (epi-IIRN) cohort showed that perianal involvement is tied to higher rates of hospitalization, surgery, colon cancer, and steroid dependency. The risk of anorectal cancer was almost three times higher in patients with PCD compared with those without perianal involvement, reported Dan Turner, MD, PhD, of Shaare Zedek Medical Center at the Hebrew University in Jerusalem, and colleagues.

"PCD is a strong predictor of complicated disease course and requires early intensified treatment, especially in severe perianal disease," they wrote online in .

Turner and colleagues reviewed data on 12,904 Crohn's disease patients from an inception cohort dating from 2005 and providing 86,119 person-years of follow-up. Patients were followed until January 2019, with a median follow-up of 6.6 years.

PCD was strictly defined as fistulizing disease or abscess, although it can include fissures and skin tags. Fistulizing disease was diagnosed in 1,530 (12%) patients, and of these 574 (4%) had the severe phenotype.

Perianal disease can occur at any time, the authors explained, even before Crohn's is diagnosed, with the highest risk in patients with colonic disease involving the rectum.

Over time, the study found PCD prevalences of 7.9%, 9.4%, 10.3%, and 11.6% at 1, 3, 5, and 10 years out from Crohn's disease diagnosis, respectively. Perianal disease preceded the Crohn's diagnosis in 602 (39%) PCD patients, and 165 patients (11%) developed the condition despite previous biologic treatment.

Disease outcomes included inflammatory bowel disease (IBD)-related hospitalizations and surgeries (such as small or large bowel resection), perianal surgeries, the need for at least two biologics from different classes, growth failure in children, steroid dependency (any course duration >90 days), mortality, and post-diagnosis IBD-related malignancies such as non-Hodgkin's lymphoma, skin cancers, small and large bowel adenocarcinoma, anorectal cancer, cervical cancer, and cholangiocarcinoma.

At 5 years, patients with PCD were more likely to be hospitalized (64% vs 36% of non-PCD patients; P<0.001), undergo IBD-related surgeries (38% vs 9%; P<0.001), and develop anorectal cancer (4.2 vs 1.2 per 10,000 person-years; P=0.01).

Severe PCD was also associated with poorer outcomes compared with non-severe PCD, as indicated by hospitalizations (73% vs 61%, respectively; P=0.004) and surgeries (43% vs 35%; P=0.001).

As in previous studies, the Israeli analysis found a higher risk of perianal involvement in men (22.6 per 1,000 person-years) compared with women (12.8 per 1,000 person-years; OR 1.89, 95% CI 1.69-2.11, P<0.001). Risk was also higher in pediatric-onset cases (21.4 per 1,000 person-years) versus adult-onset cases (16.6 per 1,000 person-years; OR 1.28, 95% CI 1.11-1.46, P<0.001).

According to the authors, this is the first population-based study to compare management and outcomes in severe versus non-severe PCD and the first to include pediatric-onset patients.

and guidelines recommend early biologic treatment in the presence of fistulizing perianal disease. "However, similar to previous reports we found that despite intensive management, PCD was still associated with a higher complication rate," Turner and team wrote. "These results emphasize the importance of early optimization of treatment in patients with PCD."

Compared with the current study, previous analyses have found a higher prevalence of PCD, ranging from 20% to 47%, possibly stemming from a referral bias due to patients' treatment at academic centers rather than a population-based cohort. "Another explanation relates to our stringent definition of perianal disease, including only those with fistulizing disease or abscess," the authors noted.

Among the study's limitations was its reliance on administrative data in the electronic records, which lacked some valuable details, such as disease extent and smoking. In addition, one of the algorithms used to differentiate severe and non-severe PCD achieved only modest sensitivity and specificity. Furthermore, some patients with PCD may have been missed and therefore not included.

Comment