Table 3 Studies providing information on treatment patterns (n = 12 studies)

Corte, 2015 [9]

Not reported by fistula type

160 procedures in 34 patients with CD-related RVF

Specific number for each procedure not reported by fistula type

Not reported by fistula type

Pre-operative: Investigators routinely propose anti-TNF therapy prior to surgery in patients with CD

Post-operative:

Not reported

Pre-operative:

Not reported

Post-operative:

Not reported

El-Gazzaz, 2010 [14]

Overall (n = 65)

Seton: 32.3%^a (21/65)

Stoma: 60.0%^a (39/65)

Healed group (n = 30)

Seton: 40.0%^a (12/30)

Stoma: 66.7%^a (20/30)

Unhealed group (n = 35)

Seton: 25.7%^a (9/35)

Stoma: 54.3%^a (19/35)

Overall

Mucosal advancement flaps: 72.3% (47/65)

Episioproctotomy: 12.3% (8/65)

Proctectomy and pull-through procedure with coloanal anastomosis: 10.8% (7/65)

Fibrin glue: 3.1% (2/65)

Fistula plug placement: 1.5% (1/65)

27.7%^a (18/65) of patients received > 3 repairs

Median (range) number of repairs:

Healed group: 2 (1–5)

Unhealed group: 2 (1–8)

p = 0.5

Median (IQR) interval from last repair to current, months:

Healed group: 7.6 (4.1–11.1)

Unhealed group: 9.7 (4.9–41.5)

p = 0.1

Median (IQR) interval from seton to current repair, months:

Healed group: 7.3 (5–8.9)

Unhealed group: 4.2 (3.6–8.2)

p = 0.5

Median (IQR) interval from stoma to current repair, months:

Healed group: 5.7 (0.6–7.8)

Unhealed group: 8 (0.9–22.9)

p = 0.1

Not reported

Pre-operative:

Immunomodulator use (infliximab, adalimumab, 6-mercaptopurine, and azathioprine within the 3 months prior to surgery): 40.0%^a (26/65)

Steroids: 30.8%^a (20/65)

Post-operative:

Not reported

Pre-operative:

Not reported

Post-operative:

Not reported

Gaertner, 2011 [15]

1. Previously received medical therapy: 94%^a (48/51)

2. Among patients who received surgery only (n = 25):

- Previous RVF surgical repairs (median): 3

- Previous bowel resection for CD: 56% (14^a/25)

- Previous anorectal surgery for CD: 40% (10^a/25)

3. Among patients who received surgery + infliximab (n = 26):

- Previous RVF surgical repairs (median): 2

- Previous bowel resection for CD: 42% (11^a/26)

- Previous anorectal surgery for CD: 50% (13^a/26)

4. Pre-operative fecal diversion: 19.6% (10/51)

- 7 had undergone ileostomy

- 3 had undergone colostomy

1. Total (n = 51 patients, 65 procedures):

54%^a (35/65) seton drainage

12%^a (8/65) advancement flap

12%^a (8/65) fibrin glue injection

9%^a (6/65) transperineal repair

6%^a (4/65) collagen plug placement

6%^a (4/65) bulbocavernosus (Martius) flap

2. In the surgery only group (n = 25 patients, 30 procedures):

60% (18/30) seton drainage

7%^a (2/30) advancement flap

20% (6/30) fibrin glue injection

13%^a (4/30) transperineal repair

0% (0/30) collagen plug placement

0% (0/30 bulbocavernosus (Martius) flap

3. In the surgery + infliximab group (n = 26 patients, 35 procedures):

49% (17/35) seton drainage

17% (6/35) advancement flap

6% (2/35) fibrin glue injection

6% (2/35) transperineal repair

11% (4/35) collagen plug placement

11% (4/35) bulbocavernosus (Martius) flap

Note: a patient might have received > 1 surgery

Of the 9 patients who did not heal:

- Seton insertion: 33%^a (3/9)

- No seton insertion: 67%^a (6/9)

27% (14/51) patients eventually required proctectomy (n = 9 treated by surgery alone and n = 5 treated by surgery + infliximab)

Pre-operative:

1. In the surgery + infliximab group (n = 26):

100% received preoperative infliximab of 5 mg/kg at 0, 2, and 6 weeks (mean of 3.6 [range, 3– 6] infusions) and were also taking 6-mercaptopurine or azathioprine

2. In the surgery only group (who did not receive infliximab) (n = 25):

23% (5/25^c) 5-ASA derivative

9% (2/25^c) azulfidine

64% (14/25^c) prednisone

9% (2/25^c) azathioprine

5% (1/25^c) methotrexate

50% (11/25^c) 6-mercaptopurine

Post-operative:

Not reported

Pre-operative:

In the operation only group (who did not receive infliximab) (n = 25), as reported by the authors:

73% (n = 16) metronidazole

50% (n = 11) ciprofloxacin

Peri-operative short-course antibiotics were given to 78.4%^a (40/51) patients; assumed prophylactic

Post-operative:

Not reported

Göttgens, 2017 [27]

Not reported by fistula type

Not reported by fistula type

Not reported by fistula type

Not reported with respect to surgery

Exposure to immunomodulator or anti-TNFα therapy at any time prior to diagnosis of RVF:

1991–1998: 70% (7/10)

1999–2011: 60% (3/5)

Pre-operative:

Not reported by fistula type

Post-operative:

Not reported by fistula type

Jarrar, 2011 [16]

All patients underwent initial seton drainage and then flap repair ≥ 6 weeks later

Other prior operations not reported by fistula type

Transanal endorectal advancement flap repair: 100%

Note: if the fistula track was long it was drained with a mushroom catheter that was removed 10 days later. If the track was short the external opening was opened widely

Crohn’s AVF:

n = 7 received 2nd flap

n = 3 received 3rd flap

n = 1 diverted

Cryptoglandular perianal:

n = 7 received 2nd flap

n = 6 received 3rd flap

n = 0 diverted

Cryptoglandular anovaginal:

n = 1 received 2nd flap

n = 2 received 3rd flap

n = 0 diverted

Pre-operative:

Not reported

Post-operative:

Not reported

Pre-operative:

Stated within 24 h prior to surgery and specified as prophylactic

Post-operative:

- Intravenous antibiotics continued post-surgery until discharge; unclear whether prophylactic

- Oral antibiotics prescribed for 1 week

Korsun, 2019^b [17]

Not reported by fistula type

Not reported by fistula type

GMT: 100%

Peri-operative^d:

RVFs (n = 22 patients, including a patient with pouch and RVF)

- Short-chain fatty acid: 4.5%^a (1/22)

- Enema: 4.5%^a (1/22)

- Azathioprine: 22.7%^a (5/22)

- Steroids: 22.7%^a (5/22)

- Colifoam: 4.5%^a (1/22)

- Mercaptopurine: 9.1%^a (2/22)

- Adalimumab: 9.1%^a (2/22)

- MTX: 4.5%^a (1/22)

- Sulfasalazine: 9.1%^a (2/22)

- None: 45.5%^a (10/22)

AVFs (n = 2 patients)

- Azathioprine, steroids, mesalazine foam: 50%^a (1/2)

- None: 50%^a (1/2)

Medication before GMT:

RVFs (n = 22 patients, including a patient with pouch and RVF)

- Steroids: 9.1%^a (2/22)

- Azathioprine: 13.6%^a (3/22)

- Adalimumab: 9.1%^a (2/22)

- Mercaptopurine: 9.1%^a (2/22)

- Infliximab: 4.5%^a (1/22)

- Sulfasalazine: 4.5%^a (1/22)

- Mesalazine: 4.5%^a (1/22)

- None: 4.5%^a (1/22)

- Unknown: 59.1%^a (13/22)

AVFs (n = 2 patients)

- Unknown: 100.0%^a (2/2)

Medication after GMT:

RVFs (n = 22 patients, including a patient with pouch and RVF)

- Mesalazine foam: 4.5%^a (1/22)

- Steroids: 27.3%^a (6/22)

- Azathioprine: 27.3%^a (6/22)

- Adalimumab: 9.1%^a (2/22)

- Sulfasalazine suppository/

sulfasalazine: 9.1%^a (2/22)

- MTX: 4.5%^a (1/22)

- Golimumab: 4.5%^a (1/22)

- Mercaptopurine: 4.5%^a (1/22)

- Unknown: 27.3%^a (6/22)

- None: 22.7%^a (5/22)

AVFs (n = 2 patients)

- None: 50.0%^a (1/2)

- Azathioprine: 50.0%^a (1/2)

 ≤ 1 medication per patient

Pre-operative:

100% received antibiotic (cefuroxime und metronidazole) 24 h prior to surgery—specified as prophylactic

Post-operative:

Not reported

Manne, 2016^b [26]

Past RVF surgery:

- Cases: 50% (8/16)

- Controls: 43% (20/47)

Proportion of patients who underwent mucosal flap procedure:

- Cases: 88%

- Controls: 12%

Proportion of patients who underwent seton:

- Cases: 13%

- Controls: 77%

Note: numbers for calculation not reported

Not reported

Pre-operative steroid use (timing prior to surgery unclear)

- Cases: 25% (4/16)

- Controls: 21% (10/47)

Azathioprine/6-mercaptopurine use

- Cases: 6% (1/16)

- Controls: 15% (7/47)

Biologic use:

- Cases: 44% (7/16)

- Controls: 62% (29/47)

MTX use:

- Cases: 6% (1/16)

- Controls: 6% (3/47)

Pre-operative:

Not reported

Post-operative:

Not reported

Narang, 2016 [19]

1. Had a seton before undergoing an attempted definitive surgical procedure:

- Yes: 43.4% (43/99)

- No: 56.6% (56/99)

2. Had a diverting stoma at the time of surgical repair:

- Yes: 36.3% (36/99)

- No: 63.6% (63/99)

Transrectal approach with endorectal advancement flap: 59.5%^c (59/99)

Transvaginal repair: 14.1%^c (14/99)

Muscle interposition: 14.1%^c (14/99)

Martius or groin flaps: 9.6%^c (9/99)

GMT: 5.3%^c (5/99)

Episioproctotomy: 6.4%^c (6/99)

Overlapping sphincteroplasty: 3.2%^c (3/99)

Fibrin glue placement: 2.1%^c (2/99)

Biological plug insertion: 1.1%^c (1/99)

Note: reported calculations could not be replicated

Not reported

At baseline:

Steroids: 57.6%^a (57/99)

Infliximab: 48.5%^a (48/99)

Adalimumab: 20.2%^a (20/99)

Azathioprine: 4.0%^a (4/99)

6-mercaptopurine: 4.0%^a (4/99)

Follow-up:

Not reported

Pre-operative:

Not reported

Post-operative:

Not reported

Oakley, 2015 [20]

Not reported by fistula type

Patients with Crohn’s RVF:

20%^a (4/20) patients received initial expectant therapy

80%^a (16/20) patients received initial surgery

Not reported

Pre-operative:

Not reported by fistula type

Post-operative:

Not reported by fistula type

Pre-operative:

Not reported by fistula type

Post-operative:

Not reported by fistula type

Pinto, 2010 [22]

Not reported by fistula type

In the 45 patients with CD, 80 procedures were performed:

- Endorectal advancement flap: 47.5% (38/80)

- GMT: 7.5% (6/80)

- Transvaginal repair: 3.8% (3/80)

- Transperineal repair: 3.8% (3/80)

- Others: 37.5% (30/80)

Not reported

Pre-operative:

Not reported by fistula type

Post-operative:

Not reported

Pre-operative:

Not reported

Post-operative:

Not reported

Sapci, 2019 [23]

1. Previous surgery to close fistula: 57.9%^a (11/19)

2. History of ≥ 2 surgeries to close fistula: 52.6% (10^a/19)

Transanal advancement flap: 42.1% (8/19)

Transanal repair with tissue interposition (Martius or gracilis flap): 15.8% (3/19)

Episioproctotomy: 10.5% (2/19)

Fistulotomy: 10.5% (2/19)

Coloanal anastomosis: 10.5% (2/19)

Fistula plug: 10.5% (2/19)

Active smoker: 42.1% (8/19)

Not reported

Pre-operative:

Not reported

Post-operative:

Not reported

Pre-operative:

Not reported

Post-operative:

Not reported

Schloericke, 2017 [24]

Recurrent cases included, but exact numbers and previous treatments are unclear

Patients with CD received resective surgical treatment only:

- Low anterior resection: n = 6

- Subtotal colectomy: n = 3 (all patients indicated for this surgery based on presence of toxic megacolon)

- Proctectomy: n = 1

- Pelvic exenteration: n = 1

Note: total number of patients with CD = 15, but only 11 surgeries reported

Proctectomy was performed in 1 case of recurrent fistulas in CD that led to severe sepsis

Not reported

Not reported