麻豆传媒

Surgical closure of perianal fistulas in Crohn's disease (CD), when combined with short-term tumor necrosis factor (TNF) inhibitors, improved radiologic healing and other outcomes compared to medication alone, the PISA II trial showed.

In an intention-to-treat analysis involving 94 CD patients, healing as assessed by MRI at 18 months was substantially more common for those who received TNF inhibitors for 4 months combined with surgical closure, compared with those on TNF inhibitors for 1 year (32% vs 9%, P=0.005), reported Christianne Buskens, MD, of the University of Amsterdam, and colleagues.

And fewer surgical group patients required surgical reintervention after an average follow-up of 8 months (13% vs 43%, with a median of one vs two interventions, P=0.005), the authors wrote in .

The surgical closure group also had significantly better quality of life after 18 months (median perianal disease activity index score [PDAI], 1 vs 4).

No significant difference was seen for clinical closure (68% vs 52%), although a post-hoc as-treated analysis found a significantly higher closure rate in the surgical closure group (71% vs 50%, respectively, P=0.016).

"As treatment of Crohn's perianal fistulas predominantly concerns quality of life, to date clinical closure was often considered more important than radiological healing," coauthor Elise Meima-van Praag, MD, a PhD fellow at Amsterdam University Medical Centre, told 麻豆传媒. "Our results demonstrate that this is not the case and that physicians should aim for radiological healing to help optimize patients' quality of life."

Crohn's perianal fistulas often lead to pain, discharge, and morbidity that significantly affects quality of life, Buskens's group noted.

In the prior , chronic seton drainage proved inferior to surgical closure after the use of short-term TNF inhibitors. Guidelines no longer recommend chronic seton drainage as a sole treatment option and instead with consideration of surgical closure for Crohn's perianal fistulas with surgically amenable disease.

"These results show that instead of recommending anti-TNF treatment and only considering surgical closure in patients with surgically amenable disease, guidelines should recommend a multidisciplinary approach of combining surgical closure with short-term anti-TNF treatment," the trialists concluded.

"A combined approach of medical and surgical evaluation and appropriate treatment by providers with expertise in Crohn's disease is very strongly advised," agreed Russell Cohen, MD, of the University of Chicago Medicine, who was not involved in this study. "Not all fistulae are created equal, some fistulae cross the very important rectal sphincter muscles, and fistula surgery may render the patient permanently incontinent of stool, which is a very undesirable result."

Since PISA I ended early due to futility with too few patients to assess closure rates, PISA II enrolled 94 patients from nine hospitals in The Netherlands and one in Italy between September 2013 and December 2019. The trial had a patient preference design, randomizing 32 with no preference for therapy and allowing the remaining 62 (66%) to stick with their preferred therapy.

"Although a randomised controlled trial is still generally considered the gold standard to provide unbiased data, this type of study was originally developed for trials comparing the effect of medication to placebo," the researchers noted. A prior systematic review had shown that many patients in such trials refuse to be randomized, but "not being randomly assigned to receive a treatment did not affect the primary outcome. Therefore, the patient preference trial design increases external validity without compromising internal validity compared with traditional randomised trials."

In their trial, 38 patients received surgical closure after 4 months of TNF inhibitors infliximab (Remicade) or adalimumab (Humira), and 56 received 1 year of TNF inhibitors. TNF inhibitors (dosed at 5 mg/kg) were combined with azathioprine at the gastroenterologist's discretion, with dose escalation. All, except five patients, initially received seton drainage, which was removed after 6 weeks for the TNF group or 8-12 weeks for the surgical group.

The trial enrolled adults with "an active, high (intersphincteric, trans-sphincteric, or suprasphincteric) perianal fistula with a single internal opening," including some who had participated in PISA I.

Patients had a median age of 33, and the majority were women (60%). About two-thirds had an ileal Montreal disease location. Some 36% to 40% had prior TNF inhibitor exposure.

Radiological healing was defined as a complete fibrotic tract on MRI or a MAGNIFI-CD (Magnetic Resonance Index for Fistula Imaging in Crohn's Disease) score of 0.

Of those who achieved clinical closure, a similar proportion had a recurrence (14% vs 16%), which only occurred among those who did not have radiologic healing.

TNF inhibitors led to adverse events -- hair loss, fatigue, headache, rash, among others -- in 11% of patients. Two serious adverse events of myocardial infarction and appendicitis occurred, along with one death due to tongue base carcinoma, but these were deemed unrelated to the intervention.

Eight patients crossed over to the surgical closure group. Five more crossed over to seek outside treatment options.

The authors acknowledged limitations to the data. The lower number of patients requiring surgical reintervention in the surgical closure group was predominantly due to the TNF group undergoing planned surgical closure from persistent fistula complaints.

Comment