PISA-IIMust knowPerianal CD · Surgery
Short-term anti-TNF therapy with surgical closure versus anti-TNF therapy in the treatment of perianal fistulas in Crohn's disease (PISA-II): a patient preference randomised trial
Meima-van Praag EM, et al.
Lancet Gastroenterol Hepatol 2022 · PMID 35427495 (PubMed, opens in a new tab) · doi 10.1016/S2468-1253(22)00088-7 (full paper, opens in a new tab)
| Endpoint | Surgical closure | Anti-TNF alone | Difference |
|---|---|---|---|
| Radiological healing, MRI at 18 months (primary endpoint) | 32% (12/38) | 9% (5/56) | P=0.005 |
| Clinical closure, 18 months | 68% (26/38) | 52% (29/56) | P=0.076 |
| Reintervention, Lower is better | 13% (5/38) | 43% (24/56) | P=0.005 |
Recurrences after clinical closure occurred only in patients without MRI healing, which makes MRI healing the better target.
Source: Meima-van Praag 2022, abstract (intention to treat; randomised and patient-preference groups combined).
Structured abstract
BackgroundGuidelines recommended anti-TNF therapy for Crohn's perianal fistulas and suggested considering surgical closure when feasible, but the long-term outcomes of the two strategies had not been compared directly.
MethodsMulticentre patient-preference randomised trial (nine hospitals in the Netherlands, one in Italy); patients without a preference were randomised. 94 patients. Primary endpoint: Radiological healing on MRI at 18 months: a completely fibrotic tract or a MAGNIFI-CD score of 0.
ResultsPrimary result. 32% (12/38) with surgical closure vs 9% (5/56) with anti-TNF alone; P=0.005.
Take-homeFor a high perianal fistula with a single internal opening, surgical closure after seton and short anti-TNF gives more MRI healing and fewer reinterventions than anti-TNF alone, and MRI healing predicts durable closure.
| Patients | 94 |
|---|---|
| Population | Adults with Crohn's disease and an active high perianal fistula with a single internal opening, after seton insertion. 32 (34%) randomised and 62 (66%) chose their treatment. |
| Comparison | Surgical closure plus 4 months of anti-TNF vs anti-TNF for 1 year (both after seton placement) |
| Drugs | infliximab, adalimumab |
| Classes | Anti-TNF, Surgery |
| Published | Meima-van Praag EM, et al. Lancet Gastroenterol Hepatol 2022;7:617–26. |
Key points
- Clinical closure at 18 months: 68% (26/38) vs 52% (29/56); P=0.076.
- Reintervention: 13% (5/38) surgical closure vs 43% (24/56) anti-TNF; P=0.005.
- Recurrence after clinical closure occurred only in patients without radiological healing.
- Long-term follow-up (median 5.7 years): radiological healing 42% (15/36) vs 18% (10/55); P=0.014. No recurrences in patients with radiological healing vs 36% (27/76) recurrence with clinical closure only (PMID 37457118).
- In long-term follow-up, none of 14 patients with radiological healing who stopped anti-TNF had a recurrence, vs 41% (9/22) with clinical closure only.
Limitations
- Only a third of patients were randomised; most chose their treatment, so groups may differ.
- Small sample; only fistulas with a single internal opening and suitable for surgical closure.
- Long-term follow-up analysed observed data only.
Commentary
PISA-II changed counselling for perianal Crohn’s disease: patients whose fistula is suitable for surgical closure should be offered a surgical opinion about closure once sepsis is controlled with a seton and anti-TNF therapy. It also made radiological healing on MRI a meaningful treatment target, since recurrence happened only in patients whose tract had not healed on MRI.
The long-term follow-up found higher anti-TNF trough levels in patients with lasting clinical closure, in keeping with targeting higher levels in perianal disease.
Journal club
Design
Multicentre patient-preference randomised trial (nine hospitals in the Netherlands, one in Italy); patients without a preference were randomised. 94 patients.
Primary result. 32% (12/38) with surgical closure vs 9% (5/56) with anti-TNF alone; P=0.005.
Validity and applicability
- Only a third of patients were randomised; most chose their treatment, so groups may differ.
- Small sample; only fistulas with a single internal opening and suitable for surgical closure.
- Long-term follow-up analysed observed data only.
Where it sits now
For a high perianal fistula with a single internal opening, surgical closure after seton and short anti-TNF gives more MRI healing and fewer reinterventions than anti-TNF alone, and MRI healing predicts durable closure.
Questions for discussion
- Two-thirds of patients chose their treatment. How might preference-based allocation bias the comparison?
- Clinical closure and MRI healing diverged. Which would you use to decide when to stop anti-TNF therapy?
- Which patients with perianal Crohn's disease would you refer for surgical closure, and which would you not?