Abstract Background Because of very limited comparative data, there is still doubt whether complex perianal fistula (cPAF), in isolated form (I-cPAF) and in association with Crohn’s disease (CD-cPAF), are different phenotypes of CD and how different is their treatment responses. Methods We retrospectively compared patients with I-cPAF and CD-cPAF in terms of clinical characteristics and clinical and radiologic remission rates. Among 72 patients who presented with symptoms of cPAF, 29 had I-cPAF, while the remaining patients either had concomitant luminal Crohn’s disease at baseline or developed it during follow-up. I-cPAF was confirmed with no gastrointestinal involvement in cross sectional imaging, gastroscopy and colonoscopy both initially and finally in follow-up. Clinical remission (CR) was defined as the absence of fistula drainage with gentle compression for ³ 3 months, and radiologic remission (RR) as complete fibrotic closure of fistula tracts on MRI. Results Data from 72 patients were analysed, of whom 40.3% (n:29) had I-cPAF. The median follow-up for the entire cohort was 95.5 (98) months and was comparable between I-cPAF and CD-cPAF (p:0.167; 82 (97) vs 100 (96.5) months). Table 1a summarizes demographic, clinical, and laboratory features of the overall cohort and of I-cPAF versus CD-cPAF. In the CD-cPAF group, ileal and ileocolonic disease locations were present in 34.1% and 41.5% of patients, respectively, with rectal involvement in 34.1%; penetrating and stricturing phenotypes were observed in 20.5% and 10.3%. In both groups, nearly half of patients had more than one fistula tract, whereas trans-sphincteric fistulas were more frequent in I-cPAF (67.9% vs 43.2%). Rates of abscess, drainage, and seton placement were similar (Table 1a). Initial albumin and haemoglobin levels were lower and CRP higher in CD-cPAF than in I-cPAF. Infliximab (IFX) was the most commonly used anti-TNF, with overall IFX treatment persistence of 80.3%, without group differences (p:0.94). CR occurred in approximately 70% of patients and was similar between I-cPAF and CD-cPAF (p:0.86). However RR was significantly less frequent in I-cPAF (Table 1b, Figure 1) with comparable MRI interval patterns in both groups. The CR–RR interval was comparable between groups (p:0.559), with a median of 28 (32.5) months in the entire cohort. Conclusion Despite its lower inflammatory burden and more challenging trans-sphincteric PAF profile with a lengthy radiological healing, largely shared fistula characteristics and clinical remission rates between I-cPAF and CD-cPAF, strongly support that I-cPAF is a different phenotype of CD. Conflict of interest: Bickes, Emre: No conflict of interest Mr. Bakkaloglu, Oguz Kagan: No conflict of interest Bayar, Melikşah: No conflict of interest Akpınar, Atilla: No conflict of interest Eskazan, Tugçe: No conflict of interest Sirolu, Sabri: No conflict of interest Simsek, Osman: No conflict of interest Tutar, Onur: No conflict of interest Erzin, Yusuf Ziya: No conflict of interest Hatemi, Ali Ibrahim: No conflict of interest Çelik, Aykut Ferhat: No conflict of interest
Bickes et al. (Thu,) studied this question.