There is a lot of confusion amongst surgeons about the exact anatomy, pathways of spread, and extent of these fistulas[2]. Not uncommonly, a high transsphincteric fistula is labelled as an extrasphincteric fistula and a suprasphincteric fistula is reported as a high transsphincteric fistula[2].

In order to understand the pathways of extension of these fistulas, it is important to understand the anatomy of the anal canal. There are three muscle layers in the sphincter-complex in the anorectum which are downward extensions of the muscle layers of the gut. There are potential spaces between these muscle layers in which the fistula and pus can spread[9] (Figures 1, 2, 4, 7, 8 and 10).

The three muscle layers in the sphincter-complex[9] are (1) Inner circular muscle layer of gut continues downwards in the anal sphincter complex as IAS; (2) Outer longitudinal muscle layer of gut continues downwards in the anal sphincter complex as CLM; and (3) Puborectalis component of levator ani muscle continues downwards in the anal sphincter complex as EAS (Figure 1). Conventionally, it was assumed that inside the sphincter-complex, there is only one space where the fistula extends and this space between the IAS and EAS was labelled as the ‘intersphincteric space’. However, with the availability of higher resolution magnetic resonance imaging (MRI) and increasing experience, it has been demonstrated that there are three potential spaces associated with in these muscle layers through which the fistula can extend[9].

The three potential spaces in the anal sphincter-complex are (1) Inner intersphincteric space (inner space): Between IAS and CLM; (2) Middle intersphincteric space (middle space): Between CLM and EAS (conventional intersphincteric space); and (3) Outer-sphincteric space (outer space): Between the lateral muscle fibers of EAS and its covering fascia[9] (Figure 1).

The anal fistula, usually initiating at crypt glands at the level of the dentate line, can extend superiorly (cephalad) in any of these three spaces.

The fistulas traversing superiorly in (1) Inner intersphincteric space (inner space) between IAS and CLM inside the rectal wall to become a high intrarectal fistula as these fistulas are present between the muscle layers of the rectal wall[8]. At times, these are erroneously labelled as submucous fistula[8] (Figures 8 and 9); (2) Middle intersphincteric space (middle space) between CLM and EAS become a supralevator fistula (if it continues into the supralevator space) (Figures 2 and 3) or a suprasphincteric fistula (if pus extending superiorly in the middle space reaches up to the top of the EAS and then traverses through the junction between the EAS and puborectalis to enter the ischiorectal fossa[15]) (Figures 4-6); and (3) Outer-sphincteric space (outer space) between the EAS and its lateral fascia becomes a RIFIL fistula[9] (Figures 10-12).

Thus, the definition of these five fistulas is: (1) Supralevator (Figures 2 and 3): These fistulas traverse in the middle intersphincteric space to enter the supralevator space. The fistula may open into the rectum through an additional high rectal opening in the supralevator space (Figures 3 and 6); (2) Suprasphincteric (Figures 4-6): These fistulas extend superiorly in the middle intersphincteric space to reach up to the top of the EAS and then traverses through the junction between the EAS and puborectalis to enter the ischiorectal fossa; (3) Extrasphincteric (Figure 7): These fistulas traverse through the ischiorectal fossa and enter the supralevator space penetrating the levator muscle. They generally do not traverse through the sphincter-complex (EAS and IAS), which is why they are labelled as ‘extra-sphincteric’ fistulas; (4) High intra-rectal (Figures 8 and 9): These fistulas traverse cephalad between IAS and CLM in the rectal wall and is thus present between the muscle layers of the rectal wall; and (5) RIFIL (Figures 10-12): These fistulas traverse the EAS but do not enter the ischiorectal fossa. They extend cephalad between the EAS and its lateral fascia to continue on the undersurface of puborectalis and levator ani muscle[9].

Incidence: In recently published large cohorts, the prevalence of these high-5 fistulas has been highlighted[9,16]. The incidence of these fistulas in a cohort of 419 consecutively operated patients over a 2-year period were RIFIL–10% (42/419), supralevator-9.5% (40/419), suprasphincteric-5.5% (23/419) and extrasphincteric-0%[9,16].

As these fistulas are deep and high, they are extremely difficult to diagnose on clinical examination. It usually requires advanced imaging modalities, MRI, or transrectal ultrasound (TRUS), to diagnose these fistulas[17-21]. Both MRI and 3D-TRUS are very effective in delineating various parameters of these fistulas, including the internal opening, primary tracts, and secondary extensions[17-20,22], though MRI has a slight edge over TRUS. It is really impressive that in the pre-MRI era, doyens in the field like Parks et al[8] could describe these deep fistulas with a reasonably high level of accuracy.

MRI helps to clearly differentiate between supralevator, suprasphincteric, high intrarectal, and RIFIL fistulas[23]. Rather it would not be wrong to say that MRI is mandatory to properly diagnose these fistulas. Therefore, whenever a fistula has recurred several times, or a tract or side branch of the main fistula is seen extending superiorly during surgery, an MRI should be done to rule out these fistulas[23].

Increasing experience with MRI in fistulas has highlighted that extrasphincteric fistulas do not occur or are extremely rare[24]. It stands to reason that it is almost impossible for a fistula in the ischiorectal fossa to penetrate through the strong levator plate when it can extend with ease in the surrounding fat of the ischiorectal fossa[24]. The only likely possibility for an extrasphincteric fistula to occur is iatrogenic when an artery forceps in a high transsphincteric fistula is pushed through the levator muscle. Even this seems improbable as it would require a lot of force by the operating surgeon. Moreover, due to increased awareness amongst surgeons of iatrogenic injuries, better understanding of regional anatomy, and easy availability of advanced radiological modalities, these iatrogenic blunders seem exceedingly unlikely. Due to this, extrasphincteric fistulas are not seen (or are extremely rare) these days[24]. It is quite possible that the fistulas diagnosed as extrasphincteric fistulas in the pre-MRI era were perhaps supralevator or RIFIL fistulas. Understandably, in the absence of MRI, it is extremely difficult to differentiate between an extrasphincteric fistula and other deep fistulas. So, a diagnosis of extrasphincteric fistula should be made only after due deliberation and detailed analysis of MRI[24].

The main challenge in these five fistulas is that they involve almost the complete EAS (Figures 2, 4, 7, 8 and 10). The EAS plays a major role in maintaining continence and risk of damage to the EAS is very high in these fistulas. Because this risk of incontinence is so high, these fistulas are extra concerning[4,5,25]. The most commonly performed procedure, fistulotomy, is contraindicated in these fistulas[25]. There is little published data and there are no standard guidelines to manage these fistulas. Therefore, the fear of managing these fistulas amongst surgeons in easy to understand.

One of the features that make these fistulas more challenging is the presence of an ASRO along with a primary internal opening at the dentate line[6]. In a large cohort, ASRO was present in 2.8% (23/836) anal fistulas, but in supralevator fistulas, the incidence was 16.6% (23/138)[6]. At times, the presence of ASRO is detected accidentally during surgery (intraoperatively) when a colored solution injected into the external opening comes out through the primary internal opening at the dentate line as well as through the ASRO[6]. In many cases, a granulation polyp or a papilla can be seen at the site of the ASRO. However, ASRO can be diagnosed preoperatively with the help of MRI in almost every case[6]. MRI is an excellent tool to diagnose ASRO with a high level of accuracy[6].

The management of ASRO seems challenging. Closure of an opening high up in the rectum which, at times, is even difficult to reach through the transanal route is not easy. However, there is a sliver of relief in these otherwise complex fistulas. A recent study has demonstrated that even if nothing is done to ASRO, there is no impact on ultimate fistula healing[6].

There is an explanation for this surprising finding. Anal fistulas usually initiate in crypt glands located at the dentate line. The fistula then can extend in different directions from here. Some fistulas or abscesses extend superiorly and enter the supralevator space[6]. In a small subset (16%) of these supralevator abscesses, the pressure generated by the supralevator abscess is so high that the abscess ruptures into the rectum, thus creating an ASRO. In such situations, the flow of infection (and bacteria) in ASRO is from the supralevator space into the rectum[6]. This continues in the same manner because, unlike in the anal canal where intraluminal pressure rises during defecation, pressure in the mid-rectum is quite low (as the rectum is primarily a storage organ). Therefore, the bacteria usually enter the fistula tract from the primary internal opening at the dentate line, bacterial multiplication leads to suppuration, and then the pus egresses through the external opening and the ASRO (if present)[6]. Hence, ultimate healing of the fistula depends primarily on successful closure of the primary internal opening at the dentate line. Once the primary internal opening at the dentate line heals, the fistula heals irrespective of whether the ASRO is closed or open[6]. Therefore, it is a reassuring fact that ASRO left untreated does not affect the ultimate fistula outcome[6].

The high-5 fistulas are more complex and refractory to treatment as compared to other fistulas. The incidence of associated complicating pathology like Crohn’s disease and tuberculosis is higher in these fistulas[24,26,27]. Therefore, it is important that these diseases should be ruled out whenever any of these high fistulas are diagnosed or suspected.

In regions where Crohn’s disease is common, a colonoscopy should be done in all cases of high fistulas[28]. On the other hand, in regions where tuberculosis is common, polymerase chain-reaction (PCR) should be done in samples from the fistula (pus or fistula tract wall or tract lining)[24,29]. In a large series published from a TB-endemic region (India), tuberculosis was detected in 10% (133/1336) of tested samples (fistula tract lining or pus)[29]. The detection rate of tuberculosis was 12.5% (129/1034) by PCR, 1.5% (3/197) by histopathology, and 0.9% (1/105) by GeneXpert tests[29]. Therefore, PCR is significantly more sensitive than histopathology or GeneXpert to detect tuberculosis[29]. However, in spite of high sensitivity, it is not uncommon that tuberculosis is missed and goes undetected in the first sample. Therefore, in regions where tuberculosis is endemic, it is recommended that repeated samples should be sent for testing, especially in more complex refractory fistulas[29].

MRI or TRUS are preferred modalities to evaluate Hi-5 fistulas. MRI is better than TRUS to evaluate high secondary extensions. Level of Evidence-2B. Grade of recommendation-B.

MRI and TRUS are the diagnostic modalities of choice to evaluate complex anal fistulas[17,19-21]. MRI and 3D-TRUS have comparable efficacy in outlining the internal opening and primary tracts; however, MRI is significantly more effective than 3D-TRUS in detecting deep secondary extensions[35]. The sensitivity of 3D-TRUS to detect deep secondary extensions was 73.9% while MRI had sensitivity of 97% (P = 0.041)[35]. Therefore, MRI is the preferred modality to diagnose high fistulas especially the high-5 fistulas[10,36].

Due to the challenging factors discussed above, the management of high-5 fistulas is quite difficult and far from satisfactory. Apart from the associated complex parameters, the additional problem is minimal experience or published data on the management of these fistulas. Due to this, the quality of data is of a low evidence level.

As discussed earlier, due to complete involvement of EAS, the most commonly performed procedure, fistulotomy, is contraindicated in these fistulas[25] unless it is coupled with primary sphincter repair. The newer sphincter-sparing procedures, like video-assisted anal fistula treatment[24], fistula laser closure[24], over-the scope clip (OTSC)[37-39], anal fistula plug (AFP)[40], stem cells[28] etc, have dismal success rates in complex fistulas. In fact, there are hardly any studies that have analyzed the success rates of these newer procedures in high-5 fistulas. Therefore, the sphincter-preserving procedures, such as LIFT, FPR, advancement flaps, and TROPIS, seem more suitable procedures for high-5 fistulas.

TROPIS provides a ray of hope, as it has been shown to have satisfactory results in these fistulas[2,3,16,24,41]. In the TROPIS procedure, the intersphincteric space is laid open in the anal canal and the resultant wound is allowed to heal by secondary intention[16,24]. The transsphincteric tracts (tracts lateral to the EAS in the ischiorectal fossa) are thoroughly curetted and cleaned[3,41]. Thus, the tracts on both sides of the EAS are managed separately (tracts inner to the EAS are laid open into the anal canal and tracts outside the EAS in the ischiorectal fossa are curetted and cleaned) and the EAS is not cut or damaged at all[16,24]. Due to this, it has been shown that there is no deterioration in continence after the TROPIS procedure[16,24].

The reason for the higher success rate of the TROPIS procedure could be that, unlike other procedures, TROPIS adequately tackles the sepsis in the intersphincteric space[3,41]. It is now understood that the pus/sepsis in the intersphincteric part of the fistula tract is like an abscess in a closed space[24]. Therefore, this sepsis is best managed in the manner an abscess anywhere else in the body is managed-deroofing the abscess cavity and allowing it to heal by secondary intention[24].

A supralevator abscesses can be drained into the rectum with a moderate success rate[7,42-44]. The level of evidence for this is 2B with a grade of recommendation of B. For example, Garcia-Granero et al[7] drained the supralevator abscess into the rectum in four patients and all of them healed.

This not only leads to resolution of the acute symptoms, but also leads to fistula healing in many cases[7,44].

Fistulectomy with or without advancement flap may be done in selected cases with a moderate success rate[44]. The level of evidence is a -4, with the grade of recommendation a-D. Van Onkelen et al[44] performed advancement flap in three patients after draining the supralevator abscess. All patients healed but multiple surgeries were required in order to achieve fistula healing[44].

Supralevator fistula can be managed with laying open of the supralevator extension into the rectum through the transanal route (TROPIS) procedure with a satisfactory healing rate[6,16]. The level of evidence for this is 2B with a grade of recommendation of B. The success rate was 84.6% (11/13) in the initial series and 82.1% (92/112) on long-term follow-up (median 30 mo) in the same series[16].

LIFT and FPR can also be done in selected cases with good success rates (80%-91%) in high complex fistulas (the studies included supralevator fistulas)[45,46]. The level of evidence for this is 5 with a grade of recommendation of D. However, data for these two procedures specific to supralevator fistulas is not available.

Advancement flap, fistulotomy/fistulectomy with primary sphincter repair and stem cells have moderate success rates, with a success rate range from 70%-85%, though the sample size was quite small[47]. The level of evidence for this is 2B with a grade of recommendation of D. Perez et al[47] compared advancement flap vs fistulotomy with sphincter reconstruction in suprasphincteric fistulas and the healing rate was 80% (4/5) and 83.3% (5/6) respectively[47]. However, the risk of cutting the entire EAS and then repairing the same seems a difficult option to most surgeons. Also, in the rare eventuality of suture dehiscence, the risk of incontinence would be very high.

Fibrin glue has a poor success rate in suprasphincteric fistulas. The level of evidence for this is 2B with a grade of recommendation of C. Garcia-Olmo et al[48] compared the healing rate of fibrin glue vs adipose stem cells in suprasphincteric fistulas. The healing rate in the fibrin glue group was 8% (2/16) while it was 71% (10/14) in the stem cells group[48].

The TROPIS procedure has had promising results with a good success rate with minimal impact on continence The level of evidence for this is 4 with a grade of recommendation of C. Out of 14 suprasphincteric fistulas, 78.6% (11/14) patients were cured with no deterioration in continence levels after surgery on long-term follow-up (median 30 mo)[16].

LIFT can be done in selected cases but is technically difficult. The level of evidence for this is 2B with a grade of recommendation of C.

There are studies in which LFT has shown a good success rate (80%-91%) in high complex fistulas[45,46]. but data for LIFT exclusive to supralevator fistulas is not available. It is also expected that performing the LIFT procedure in suprasphincteric fistulas (ligating the intersphincteric tract high up in the intersphincteric plane) would be a technically demanding procedure.

Temporary colostomy with management of the primary pathology can be used for extrasphincteric fisutlas. The level of evidence for this is 5 with a grade of recommendation of D.

Extrasphincteric fistulas are extremely rare these days, which means there is little data available on the management of these fistulas. As most of these fistulas are caused by secondary pelvic pathology or iatrogenic factors, a diverting stoma along with management of the underlying cause would be the mainstay of treatment[8].

Intra-anal fistulotomy (TROPIS procedure) has a high success rate with minimal impact on continence. The level of evidence for this is 4 with a grade of recommendation of B. These fistulas are perhaps the easiest to treat. Simple intra-rectal fistulotomy (laying open the fistula tract into the anorectum) would cure these fistulas with a high success rate (> 90%)[8,16].

TROPIS has a moderate success rate with minimal impact on continence in RIFIL fistulas. The level of evidence for this is 2B with a grade of recommendation of B. RIFIL fistulas are challenging to manage because access to the RIFIL component might become difficult. TROPIS has shown a moderate success rate in RIFIL fistulas[9]. The healing rate in RIFIL fistulas by the TROPIS procedure was 69.4% (25/36) with a follow-up (median) of 12 mo[9]. There was no negative impact on continence following surgery.

LIFT and FPR are expected to have a moderate success rate; however, since RIFIL fistulas have been recently described[9] and till now, there are no studies of LIFT or FPR procedure in RIFIL fistulas. Conceptually it is expected that these procedures would also prove at least moderately effective for these fistulas. The level of evidence for this is 5 with a grade of recommendation of D.