- Email: [email protected]
Magnetic Resonance Imaging for Perianal Fistula
Magnetic Resonance Imaging for Perianal Fistula Damian J.M. Tolan Perianal fistulas and other inflammatory diseases of the anus and perianal soft tissues are a cause of substantial morbidity, and are a major part of the practice of any colorectal surgeon. Magnetic resonance imaging (MRI) has a key role in the assessment of patients for the extent of fistulizing Crohn disease, complications related to fistulas, and to assist in confirming the diagnosis or proposing an alternative. Technique is critical and in particular, the selection of sequences for diagnosis and characterization of abnormalities with the main choices being between standard anatomical sequences (T1 or T2), assessing for edema (FS T2 or STIR), assessing abnormal contrast enhancement (FS T1), and assessing for abnormal diffusion or a combination of these. Guidance on MRI sequence selection, classification of fistulas, the current guidance on the role of MRI in assessing patients, and advice on how to provide useful structured reports, as well as how to detect complications of perianal sepsis are included. Semin Ultrasound CT MRI 37:313-322 C 2016 Elsevier Inc. All rights reserved.
Introduction erianal fistulas and other inflammatory diseases of the anus and perianal soft tissues are a cause of substantial morbidity, in particular related to incontinence and consequently poor quality of life as a result of either the disease itself or the surgical interventions. These diseases form a substantial part of the practice of any colorectal surgeon with an increasing range of surgical procedures being performed for treatment. As with many other areas of surgical practice, radiology has developed a key role in the assessment of patients for the extent of fistulizing Crohn disease, complications related to fistulas, and to assist in confirming the diagnosis or proposing an alternative. This article assesses the role of magnetic resonance imaging (MRI), provides guidance on MRI sequence selection, and advices how to provide useful reports for clinical colleagues.
P
How Do Perianal Fistulas Arise? There are 2 main reasons for a perianal fistula to form. The first is related to the anal glands. These lie in the mid-to-lower anal canal at the dentate line, where the columnar epithelium of St James’s University Hospital, Leeds Teaching Hospitals NHS Trust, England, UK. Address reprint requests to Damian J.M.Tolan, St James’s University Hospital, Leeds Teaching Hospitals NHS Trust, Leeds, England, UK. E-mail: [email protected]
http://dx.doi.org/10.1053/j.sult.2016.04.004 0887-2171/& 2016 Elsevier Inc. All rights reserved.
lower gastrointestinal tract meets the squamous epithelium of the lower anus. When anal glands become occluded, an infection of the contents may develop leading to the formation of a small abscess. As this abscess grows, it typically passes toward the anal verge where it may discharge either spontaneously or following surgical incision and drainage if the patient presents to hospital with an acute perianal fistula. In either situation, an abnormal fistulous communication may form in between 15% and 40% of cases between the anal canal (the internal opening) and the perineum (external opening). Alternatively, rather than being formed from occlusion or infection of glands, a fistula may form as a result of deep ulcerating diseases of the anorectum. The commonest cause in most Western Countries relates to Crohn disease, whereas in developing countries, such as in the Indian subcontinent, tuberculosis (TB) is much more common. Other less frequent causes include fistulation from adenocarcinoma or squamous cell carcinoma of the anorectum or less commonly as a complication of high-dose pelvic radiotherapy, whereas other infections are recognized to cause perianal fistulation including actinomycosis and human immunodeficiency virus (HIV) infection.
Classifying Fistula-In-Ano The most frequently used classification system used in surgical practice is the Parks Classification.1 This describes the anatomical location of the fistula. However, to understand the 313
D.J.M. Tolan
314 classification of fistulas using this and all other systems we must first evaluate the relevant anatomy of the perianal region and pelvic floor.
Anatomy for Fistula Classification The pelvic floor is a conical structure formed from the levator ani muscles (Fig. 1). As these pass medially and inferiorly from their attachment on the pelvic sidewall, they merge with the V-shaped or U-shaped puborectalis muscle sling. The puborectalis passes posterior to the anorectal junction and acts to pull the anorectal junction anteriorly toward the pubis to assist with continence. Inferiorly, the puborectalis blends with the continuous circular external anal sphincter forms, which passes inferiorly to the anal verge. Although anatomical studies have described 3 separate components of the external sphincter (deep, superficial, and subcutaneous), these are usually inseparable on imaging evaluation and act as a single functional unit. All of these pelvic floor muscles are striated, providing baseline tone in addition to voluntary squeeze or relaxation for the regulation of continence and defecation respectively. The internal sphincter lies medial to the external sphincter and is formed from smooth muscle, which is continuous with the muscularis propria of the rectum. This runs from the anorectal junction inferiorly to where it terminates as it interdigitates with the lowermost fibers of the external sphincter muscle. The relevance of the dentate line has already been described but this is not visible on radiological assessment. Instead, its position may be approximated to lie 2 cm superior to the anal verge, or alternatively caudal to the inferior border of the puborectalis muscle that is usually a position one-third to onehalf the distance of the anal canal from the anal verge. The anal
Figure 1 Normal anatomy of the pelvic floor and anal sphincters. The levator ani (lev ani) muscles blend with the puborectalis (pr), which in turn merges with the circular external sphincter (ext). The internal sphincter is continuous with the muscularis propria of the rectum, running the length of the anal canal. The intersphincteric space (white arrow) occupies the plane between the internal sphincter and the striated sphincter muscles. The dentate line, though not visible radiologically, lies just below puborectalis, approximately 2 cm from the anal verge (broken circular line).
glands arising at this location may penetrate into and beyond the internal anal sphincter muscle into a potential space called the intersphincteric plane. This is a critical potential space for perianal infection as such infections may track either inferiorly or superiorly along it. Alternatively, infection may pass from the intersphincteric plane through the external sphincter muscle into the fat of the ischioanal fossa, which lies below the levator ani muscles and lateral to the anal sphincters.
Parks Classification Sir Alan Parks from St Marks Hospital first produced a classification of perianal fistulas, based on the relationship of the fistula to the anal sphincter complex (Fig. 2).1 Superficial or submucous fistulas are rarely encountered in radiological practice but are the simplest to treat surgically. Intersphincteric fistulas arise in the anal canal before descending in the intersphincteric plane to reach the anal verge. A transsphincteric fistula may track for a variable distance in the intersphincteric plane before crossing the external sphincter muscle into the ischioanal fossa fat to reach the perineum or adjacent urogenital structures. Supralevator fistulas also originate in the intersphincteric plane but instead pass superiorly before they cross the levator ani muscles and descend through the ischioanal fossa to the perineum. The final group of fistulas are the extrasphincteric fistula; these usually arise from the rectum and entirely bypass the anal sphincter complex by crossing the levator ani muscle via the mesorectum before reaching the perineum through the ischioanal fossa.
Figure 2 Inter—intersphincteric fistula: confined to the intersphincteric plane. These may cross the lowermost fibers of the external sphincter (dotted arrow). Trans—transsphincteric fistula: may traverse the external sphincter high or low to pass some distance distally in the intersphincteric plane before crossing the external sphincter. Supra—supralevator fistula: sepsis tracks proximally in the intersphincteric plane before piercing the levator ani to descend through the ischioanal fossa. Extra—extrasphincteric fistula: pelvic pathology generates sepsis that tracks across the levator ani to the perineum.
MRI for perinatal fistula In addition to the main fistula, additional elements may extend from it that are commonly called extensions, secondary tracks, or branches. Although these extensions can be linear, they may also form a curved shape in either the intersphincteric plane, supralevator space, or within the ischioanal fossa fat, where they may be referred to as “horseshoe” extensions, which is a commonly used term in surgical practice.
St James’s University Classification This radiological classification system groups types of fistula based on outcomes from surgery and categorizes them from simple to progressively more complex using a combination of the fistula topography related to the sphincter complex and the presence or absence of extensions.2 Type 1 and type 2 refer to intersphincteric fistulas, the latter associated with additional secondary extensions or intersphincteric abscesses. Type 3 and type 4 are transsphincteric fistulas, the latter complicated by abscesses or secondary tracks in the ischioanal fossa. The final group type 5 includes supralevator and translevator fistulas (combining the suprasphincteric or extrasphincteric fistula classes from the Parks classification). As the class of fistula increases on this scale, the complexity of surgical intervention increases along as the risk of poorer outcome from surgery. Although this classification system has been adopted by some radiologists, and it has prognostic ability, most surgeons are unfamiliar with it which limits its applicability in day-today practice.
How Does MRI Make a Difference? MRI vs Other Modalities MRI offers superior contrast resolution to computed tomography in the assessment of perianal fistula, thereby limiting the role of computed tomography to the exclusion of pelvic organ pathology or abnormalities of the large or small bowel. Endoanal ultrasound is able to assess perianal fistula but because of the depth of penetration it is difficult to track the distribution of more complex features like deep extensions. However, it is better able to visualizing the internal opening of the fistula than MRI. Patients can find endoanal ultrasound assessments painful in the setting of active sepsis and in particular abscess formation. Contrast studies with fistulography no longer have a contributory role in the assessment of these patients.
Clinical Decision-Making With MRI MRI is better able to detect complex features than standard clinical evaluation.3 This added value has been shown to reduce the relapse rate in patients with recurrent perianal fistula after initial failed surgery by up to 75%.4 Surgical
315 guidelines have since been produced with recommendations for practice with the Association of Coloproctologists of Great Britain and Ireland advising MRI as the modality of choice for any patients with recurrent fistula or where clinical assessment indicates that this is a complex fistula at presentation.5 Guidance from the American Society of Colon and Rectal Surgeons does not give any specific recommendations on the value of imaging in the assessment of patients.6 MRI is also the method of choice for the assessment of patients with active perianal Crohn disease. The European Crohn’s and Colitis Organisation guidelines recommend MRI assessment at presentation in all patients, except for those with a simple fistula judged by a surgical expert, either with acute perianal sepsis or those presenting with a more indolent course to classify the fistula.7 The commonest cause for treatment failure is untreated or undetected fistula extensions and abscesses, and as MRI is the most accurate method of assessing perianal Crohn disease, surpassing examination under anesthesia, its role in initial assessment and in treatment planning is essential.8 MRI is then recommended in patients in remission at 12 months or in those patients with persistent disease despite surgery and medical therapy.9 In a follow-up of patients with Crohn disease on medical therapy, radiological healing lags behind the clinical response to treatment, and patients who are withdrawn from maintenance therapy can relapse in the same fistula distribution as seen on prior MRI scans where healing has been demonstrated in the past.10 A scoring system of signs of disease activity and assessment of severity has been developed by Van Assche et al,11 using a combination of an assessment of the extent of the fistula (the complexity of the fistula, the relationship to the sphincters, and whether there is supralevator extension) and the activity of the fistula (judged by the T2 signal intensity of the fistula, whether abscesses or cavities are present and the presence or absence of rectal wall involvement). Although these signs of disease activity are useful, a formal scoring of activity is not considered practical in routine clinical practice.8
Hardware Requirements for Imaging Endoanal coils are not required for assessment of perianal fistula. Multichannel phased array body coils obtain highquality imaging and allow a wider field of view to be obtained. They are also less uncomfortable for a patient than endoanal coils, who may have active perianal inflammation. Magnetic field strength is not an essential factor and 1.5-T MRI is more than adequate for diagnosis. In general, a slice thickness acquisition between 3-5 mm is required to allow detection of fistulas and extensions and the relationship of the fistula to the anal sphincter mechanism.
Choosing the Right Plane for Imaging Scanning with a “true” anatomical plane produces images acquired obliquely through the sphincter complex, which can
D.J.M. Tolan
316
A
B
C
D
Figure 3 High transsphincteric fistula (white arrow). (A) oblique coronal, (B) axial FS T2, (C) oblique coronal, and (D) axial postcontrast FS T1. Oblique coronal image shows a long track anterior to the external sphincter. Oblique axial image more clearly shows the fistula lying anterior to the external sphincter muscle. Note the dark dot in the center of the fistula denoting a seton suture.
render accurate assessment of these features impossible, and this should be avoided. Instead, by following the same conventions as used for imaging the rectum and gynecological structures for assessing malignancy, it is important to perform imaging in planes that are orthogonal to the anal canal itself. As the anal canal passes craniocaudally, it moves posteriorly at an angle of approximately 451. It is important to scan using this plane to allow evaluation of the height of the internal opening relative to the anal canal and also the relationship of the fistula to the sphincter muscles. It is a matter of personal preference which scan planes are acquired. Although the axial and coronal planes are essential, it is the author’s personal experience that the sagittal plane is less useful, except for patients where there is clinical suspicion of fistula-to-urogenital structures or involvement of the natal cleft. This can usually be deduced before the scan at the time the examination is protocolled. It is also important to evaluate the whole of the perineum and to include an assessment of the lower pelvis or mesorectum as these are potential sites of fistula extension. This is particularly important in patients with complex fistulas, for example, from Crohn disease.
Sequence Selection for Interpretation Different authors have used a range of sequences for imaging evaluation of perianal fistula. These may be broadly divided into the following categories:
“Standard” Pelvic Imaging Sequences (T1 and T2 Spin Echo) Some authors advocate standard T1 or T2 (or a combination of both) for assessment of fistula.12,13 These provide good demonstration of sphincter anatomy. Fistula tracks appear dark on T1-weighted sequences (similar to adjacent muscle) and so this reminds of the dark fistula being visible on the bright background of fat in the soft tissues. A healed fibrotic fistula track would look indistinguishable from an active fistula on this sequence alone. Similarly, a highsignal fistula on T2-weighted sequences may have similar signal to adjacent intersphincteric or ischioanal fossa fat that may limit conspicuity.
MRI for perinatal fistula
Sequences Assessing Edema (STIR or T2 With Fat Saturation T2) An advantage of these sequences is an increase in conspicuity of inflammatory changes related to an active fistula on the scan. The fistula may have a dark wall, whereas the fistula contents would appear bright from fluid or inflammatory granulation tissue (Fig. 3A and B).13-15 Inflammatory changes may extend beyond the fistula itself, particularly in the context of a contained abscess. Typically, the anatomy of the anal canal is clearly visible enough to allow identification of the internal and external sphincter muscles. The internal sphincter smooth muscle is intermediate-to-high signal compared with the lower-signal striated external sphincter muscle and puborectalis in a normal subject. Fibrosis appears dark, as sometimes seen at the site of a healed fistula.13
Sequences Assessing Contrast Enhancement (T1 With or Without Fat Saturation Post Gadolinium) Post–gadolinium T1-weighted MRI assess abnormal enhancement related to inflammation (Fig. 3 C and D). This is made more conspicuous where fat saturation is applied and can be derived from 2-dimensional or 3-dimensional sequences. Although contrast injection adds cost and time, it allows differentiation of abscess or fluid within a fistula track from granulation tissue, as an abscess demonstrates rim enhancement with central nonenhancement (from the fluid within it) (Fig. 4).16 This can be particularly important in surgical planning for complex cases or where patients are being treated with powerful immunosuppressants, such as anti–TNF agents, where it may assist in deciding to pause treatment and offer antibiotic therapy or surgery. In the author’s and others’ experience, this can also make small fistulas more conspicuous and increase diagnostic confidence.13 Healed fibrotic fistula tracks conversely either demonstrate no enhancement or only low-level enhancement similar to adjacent muscle.
Diffusion-Weighted Imaging Although some authors propose a potential routine role for diffusion-weighted imaging,17 and in particular for the diagnosis of abscess complicating fistula-in-ano that might obviate the requirement for contrast administration, there are downsides to diffusion-weighted imaging assessment of the perianal region. In particular, there are difficulties in interpretation resulting from artifacts close to air–soft-tissue interfaces, which can degrade image quality.18
Sequence Combinations for MRI In arriving at a standard technique for fistula assessment, a combination of different imaging sequences is typically preferred. Although a combination of T2 and postcontrast 3-dimensional gradient-echo T1-weighted sequences with fat saturation has been proposed as optimal for the accurate assessment of fistula,13 the author’s own experience prefers a combination of a fat-saturated T2-weighted sequence, to look
317 for edema, in combination with postcontrast fat-saturated T1-weighted sequences to look for abnormal enhancement.
When Is a Fistula Not a Fistula? The Differential Diagnosis Although fistula-in-ano is a clinical diagnosis, other perianal conditions can mimic the symptoms and imaging appearances. Pilonidal sinus results from infection of ingrowing hairs, most commonly in the subcutaneous tissues of the natal cleft. These can produce a variety of appearances ranging from focal inflammation in the subcutaneous fat to complex branching sinuses. Sinuses may extend superiorly to the posterior aspect of the sacrum and inferiorly they may reach the outer aspect of the external anal sphincter. However, the sphincter mechanism itself is only rarely affected.19 Sparing of the sphincter complex is typical and allows differentiation from true fistula-in-ano (Fig. 5). Hidradenitis suppurativa is an apocrine gland disorder producing axillary and genitofemoral abscesses. Abscesses, sinuses, and fistulas in chronic cases may extend posteriorly to the perirectal area.20,21 Although it is less common, isolated anorectal involvement is recognised, with perianal fistula and abscess formation.21 In such cases, Crohn disease is a recognized association and these 2 conditions may coexist.20 Atypical infection can produce perianal fistula and is recognized in HIV, TB, and actinomycosis.22-24 The diagnosis is usually indicated either by a failure of surgical therapy, prior known infection, or a high prevalence in the local population (in the case of TB). Patients with HIV can develop perianal symptoms unrelated to fistula-in-ano as 2 or more of fissures, tumors, condylomas, and abscesses may be present in up to 66% of cases.22 Immunocompromised patients, for example, neutropenic patients having chemotherapy for solid-organ malignancy or patients with hematological malignancy, are a final group for consideration. These are a difficult group of patients to assess clinically and surgeons are understandably reluctant to perform examination under anesthesia without an intention to intervene to drain an abscess. MRI typically demonstrates generalized and quite-extensive perianal edema, but without abscess formation. Only a minority have an abscess requiring surgical intervention.25
Recommendations for Structured Reporting The purpose of the MRI is to determine confirmation of the presence of abnormality and to provide a surgical road map to guide treatment. In particular, the report needs to accurately map the distribution of the fistula and the location of any complex features, as these can be difficult to detect at examination under anesthesia and particularly in more complex cases.26 To achieve this, the following points should be included in reports to make them clinically useful.
D.J.M. Tolan
318
A
B
Figure 4 Recurrent intersphincteric fistula due to clinically occult left supralevator extension (arrow), a strength of MRI assessment. (A) Note the uniform high signal on FS T2 and (B) high-signal rim with central nonenhancement from pus on postcontrast FS T1.
on the scan, such as rectal inflammation may suggest underlying Crohn disease as a cause for fistula formation.
Classification of the Abnormality Fistula-In-Ano or Alternative Diagnosis? Sometimes the radiologist will be the first to suggest an alternative diagnosis such as pilonidal sinus or hidradenitis, for example, if there are inflammatory changes in the groins, or only if they are localized to the natal cleft. Additional features
A
Single Fistula or Multiple Fistulas? If a single fistula is present, then the structured report is usually straightforward. If multiple fistulas are present, then these
B
Figure 5 Extensive pilonidal sinus (black arrow). FS T2 sagittal (left) and axial (right). Note the typical sparing of the external sphincter (thick arrow) to differentiate from fistula-in-ano.
MRI for perinatal fistula
319
should be described in sequence and in a logical order (for example, clockwise around the circumference of the anal canal according to the internal opening).
What Fistula Type? The Parks classification of the fistula should be provided depending on the relationship of the fistula track to the sphincter complex, as described earlier.1 In some cases, a suture called a “Seton” may have been already placed within the fistula following examination under anesthesia. This would appear as a continuous low-signal loop within the fistula track (Fig. 6). This is sometimes best appreciated on coronal sequences. Figure 7 Axial FS T2 midcanal. The internal opening of a fistula at 11o’clock seen as an oval high-signal area crossing the internal sphincter (black arrow head). The internal opening is not often visible in this way. Clock face position common surgical terminology and should be used routinely for reporting.
Describing the Path of the Fistula Internal Opening The internal opening position should be given both in the height within the anal canal and the “clock position,” which is the standard surgical terminology, using the anal clock face with 12-o’clock anteriorly, 3-o’clock to the left, 6-o’clock posteriorly, and 9-o’clock to the right (Fig. 7). The internal opening is not always visible as a clear defect. Most commonly, its position is adjacent to the epicenter of the inflammation in the intersphincteric plane and so the position can be “predicted” from this. It is important to provide this prediction, as the surgeons will rely on this to guide their initial exploration.
sphincter,as this may alter the therapeutic options for the surgeon. In particular, the surgeon may choose to avoid a sphincterotomy (cut the sphincter) if the fistula is deemed to be high. External Opening The external opening should be provided with an estimated clock position and distance from the anal verge. This is particularly important where more than 1 fistula is present.
Path of the Fistula The course of the fistula should be described from the internal opening. In particular, details need to be provided of the clock position that the fistula passes inferiorly either within the intersphincteric plane (for an intersphincteric fistula) or where it crosses the external sphincter (in a transsphincteric fistula). For a transsphincteric fistula, it is important to provide an estimate of the height that the fistula traverses the external
A
Complex Features Position of extensions. The position of extensions from the main fistula should now be provided if these are present. The location should be clearly stated, for example, whether it is within the intersphincteric plane or within the supralevator compartment or ischioanal fossa and the height relative to the
B
Figure 6 (A) Oblique coronal and (B) axial FS T2. Chronic high transsphincteric fistula containing a seton (thin white arrow). The fistula is surrounded by low-signal fibrosis (thick white arrow). Ongoing high-signal inflammation is present in surrounding ischioanal fat (thin black arrow).
D.J.M. Tolan
320
A
B
Figure 8 Perianal pain in Crohn disease. (A) FS T2 and (B) postcontrast FS T1-weighted image shows a 2.5-cm intersphincteric abscess with avid rim enhancement (thin white arrow), The external sphincter (thick white arrow) is thinned out between 7 and 10-o’clock position. Note the eccentric displacement of the anal canal (asterisk).
anal canal and the clock position that the abnormality is located. Where a horseshoe extension is present, the clock positions of the arc of the extension should be provided (for example, “between 2-7–o’clock in the intersphincteric plane in the mid canal”).
T2-weighted); postcontrast FS T1-weighted sequences allow more easy differentiation, as fluid will not enhance within the fistula. Abscesses will often also have a surrounding “halo” of edema in the surrounding tissues as the inflammation spreads beyond the wall of the abscess itself.
Undrained fluid and abscesses. Where a blind-ending extension contains fluid that has not been drained, the estimated diameter of this collection should be provided, in addition to the position, as this would allow the clinician to determine whether it requires specific intervention (Fig. 4). This can be particularly important in the setting of Crohn disease where patients are on biological therapy, where patients are potentially at risk of sepsis (Fig. 8), or in patients with complex fistula-in-ano where the presence of an undrained extension might increase the risk of failed surgery. Differentiating undrained fluid from granulation tissue can be very difficult without contrast agent administration, as both show high signal on edema sequences (STIR or FS
Bone edema and osteomyelitis. In patients with established complex fistulation, particularly with Crohn disease, fistulas may extend to involve the bony pelvis and particularly the coccyx and lower sacrum. In such cases, the addition of a sagittal plane is very useful to cover the whole of the sacrum to determine the extent of the abnormality. This would appear as localized high-signal bone edema on either STIR or FS T2-weighted sequences and may be associated with disruption of the cortex when osteomyelitis develops. The supervising clinician will need urgent notification of this finding, as it is usually clinically occult and may require cessation of biological therapy and long-term antibiotics in addition to surgical exploration to drain the fistula in question.
A
B
Figure 9 Recurrent urinary tract infection with history of radiotherapy for prostate cancer. (A) Sagittal FS T2 and (B) postcontrast FS T1 acquisition. Foley catheter (thin arrow) and erectile device (asterisk). Small fistula with defect visible only on the post–contrast-enhanced image (thick white arrow).
MRI for perinatal fistula
A
321
B
Figure 10 Axial T2. History of fistula-in-ano. Pain and discharge. (A) Dark fibrosis related to chronic intersphincteric fistulas (thin arrow). (B) Mucin filling a fistula track (thick arrow) and displacing the vagina (V) anteriorly from mucinous fistula adenocarcinoma.
Imaging follow-up is usually required to see a reduction in edema in response to treatment. Anogenital fistulation. Depending on the length, distribution, and severity of fistulas, the vagina and base of penis can be involved by fistulation. This is not always very symptomatic or clinically suspected but obviously important for surgical management. In some cases, symptomatic patients may have defunctioning of the rectum, particularly where surgical intervention has failed. Fistulas may also be detected between the rectum and the bladder or vagina. These are often caused by other diseases, such as diverticular disease, Crohn disease, and as a delayed complication from radiotherapy (Fig. 9). Malignancy in chronic fistulas. Few patients will develop malignancy as a consequence of chronic fistula inflammation.27 This is usually where a fistula has been present for 10 or more years. Both squamous cell carcinoma and adenocarcinoma tumor types are recognized. Identification of these tumors can be very challenging. Signs include development of a new soft-tissue signal component when compared with previous assessments or alternatively, where mucinous adenocarcinoma develops, a lobulated pattern with a more complex internal architecture to the fistula than a usual more uniform fluid or granulation tissue– related content (Fig. 10). In such cases, biopsy is required to confirm the diagnosis and it would be directed by the imaging appearances.
Conclusion MRI is an essential tool for the assessment of patients with fistula-in-ano, and particularly in more complex disease. Radiologists must be aware of the anatomy of the anal canal,
the classification of fistulas and able to produce a succinct report detailing the course of the fistula to allow planning for surgical intervention. A thorough understanding of the implications of more complex features of fistulas is required, in particular for the management of patients with perianal fistulas from Crohn disease on biological therapy, as well as an ability to differentiate conditions mimicking fistula-in-ano, such as hidradenitis and pilonidal sinus.
References 1. Parks AG, Gordon PH, Hardcastle JD: A classification of fistula-in-ano. Br J Surg 63:1-12, 1976 2. Spencer JA, Chapple K, Wilson D, et al: Outcome after surgery for perianal fistula: Predictive value of MR imaging. Am J Roentgenol 171:403-406, 1998 3. Buchanan GN, Halligan S, Bartram CI, et al: Clinical examination, endosonography, and MR imaging in preoperative assessment of fistula in ano: Comparison with outcome-based reference standard. Radiology 233:674-681, 2004 4. Buchanan GN, Halligan S, Williams AB, et al: Magnetic resonance imaging for primary fistula in ano. Br J Surg 90:877-881, 2003 5. Williams JG, Farrands PA, Williams AB, et al: The treatment of anal fistula: ACPGBI position statement. Colorectal Dis 9:18-50, 2007 6. Steele SR, Kumar R, Feingold DL, et al: Practice parameters for the management of perianal abscess and fistula-in-ano. Dis Colon Rectum 54:1465-1474, 2011 7. Van Assche G, Dignass A, Reinisch W, et al: The second European evidence-based Consensus on the diagnosis and management of Crohn’s disease: Special situations. J Crohns Colitis 4:63-101, 2010 8. Panes J, Bouhnik Y, Reinisch W, et al: Imaging techniques for assessment of inflammatory bowel disease: Joint ECCO and ESGAR evidence-based consensus guidelines. J Crohns Colitis 7:556-585, 2013 9. Avaliable at: http://www.e-guide.ecco-ibd.eu/algorithm/treatment-fistula ting-perianal-disease. Accessed March 30, 2016. 10. Tozer P, Ng SC, Siddiqui MR, et al: Long-term MRI-guided combined anti-TNF-α and thiopurine therapy for Crohn’s perianal fistulas. Inflamm Bowel Dis 18:1825-1834, 2012 11. Van Assche G, Vanbeckevoort D, Bielen D, et al: Magnetic resonance imaging of the effects of infliximab on perianal fistulizing Crohn’s disease. Am J Gastroenterol 98:332-339, 2003
322 12. Koelbel G, Schmeidl U, Majer MC, et al: Diagnosis of fistulae and sinus tracts in patients with Crohn’s disease: Value of MR imaging. Am J Roentgenol 152:999-1003, 1989 13. Torkzad MR, Ahlström H, Karlbom U: Comparison of different magnetic resonance imaging sequences for assessment of fistula-in-ano. World J Radiol 6:203-209, 2014 14. Barker PG, Lunniss PJ, Armstrong P, et al: Magnetic resonance imaging of fistula-in-ano: Technique, interpretation, and accuracy. Clin Radiol 49: 7-13, 1994 15. Haggett PJ, Moore NM, Shearman JD, et al: Pelvic and perianal complications of Crohn’s disease: Assessment using magnetic resonance imaging. Gut 36:407-410, 1995 16. Spencer JA, Ward J, Beckingham IJ, et al: Dynamic contrast-enhanced MR imaging of perianal fistulas. Am J Roentgenol 167:735-741, 1995 17. Hori M, Oto A, Orrin S, et al: Diffusion-weighted MRI: A new tool for the diagnosis of fistula in ano. J Magn Reson Imaging 30:1021-1026, 2009 18. Dohan A, Eveno C, Oprea R, et al: Diffusion-weighted MR imaging for the diagnosis of abscess complicating fistula-in-ano: Preliminary experience. Eur Radiol 24:2906-2915, 2014 19. Taylor SA, Halligan S, Bartram CI: Pilonidal sinus disease: MR imaging distinction from fistula in ano. Radiology 226:662-667, 2003
D.J.M. Tolan 20. Kelly AM, Cronin P: MRI features of hidradenitis suppurativa and review of the literature. Am J Roentgenol 185:1201-1204, 2005 21. Griffin N, Williams AB, Anderson S, et al: Hidradenitis suppurativa: MRI features in anogenital disease. Dis Colon Rectum 57:762-771, 2014 22. Barrett WL, Callahan TD, Orkin BA: Perianal manifestations of human immunodeficiency virus infection: Experience with 260 patients. Dis Colon Rectum 41:606-611, 1998 23. Sultan S, Azria F, Bauer P, et al: Anoperineal tuberculosis: Diagnostic and management considerations in seven cases. Dis Colon Rectum 45: 407-410, 2002 24. Coremans G, Margaritis V, Van Poppel HP, et al: Actinomycosis, a rare and unsuspected cause of anal fistulous abscess: Report of three cases and review of the literature. Dis Colon Rectum 48:575-581, 2005 25. Plumb AA, Halligan S, Bhatnagar G, et al: Perianal sepsis in hematologic malignancy: MR imaging appearances and distinction from cryptoglandular infection in immunocompetent patients. Radiology 276:147-155, 2015 26. Buchanan G, Halligan S, Williams A, et al: Effect of MRI on clinical outcome of recurrent fistula-in-ano. Lancet 360:1661-1662, 2002 27. Egan L, D’Inca R, Jess T, et al: Non-colorectal intestinal tract carcinomas in inflammatory bowel disease: Results of the 3rd ECCO Pathogenesis Scientific Workshop (II). J Crohns Colitis 8:19-30, 2014
Introduction erianal fistulas and other inflammatory diseases of the anus and perianal soft tissues are a cause of substantial morbidity, in particular related to incontinence and consequently poor quality of life as a result of either the disease itself or the surgical interventions. These diseases form a substantial part of the practice of any colorectal surgeon with an increasing range of surgical procedures being performed for treatment. As with many other areas of surgical practice, radiology has developed a key role in the assessment of patients for the extent of fistulizing Crohn disease, complications related to fistulas, and to assist in confirming the diagnosis or proposing an alternative. This article assesses the role of magnetic resonance imaging (MRI), provides guidance on MRI sequence selection, and advices how to provide useful reports for clinical colleagues.
P
How Do Perianal Fistulas Arise? There are 2 main reasons for a perianal fistula to form. The first is related to the anal glands. These lie in the mid-to-lower anal canal at the dentate line, where the columnar epithelium of St James’s University Hospital, Leeds Teaching Hospitals NHS Trust, England, UK. Address reprint requests to Damian J.M.Tolan, St James’s University Hospital, Leeds Teaching Hospitals NHS Trust, Leeds, England, UK. E-mail: [email protected]
http://dx.doi.org/10.1053/j.sult.2016.04.004 0887-2171/& 2016 Elsevier Inc. All rights reserved.
lower gastrointestinal tract meets the squamous epithelium of the lower anus. When anal glands become occluded, an infection of the contents may develop leading to the formation of a small abscess. As this abscess grows, it typically passes toward the anal verge where it may discharge either spontaneously or following surgical incision and drainage if the patient presents to hospital with an acute perianal fistula. In either situation, an abnormal fistulous communication may form in between 15% and 40% of cases between the anal canal (the internal opening) and the perineum (external opening). Alternatively, rather than being formed from occlusion or infection of glands, a fistula may form as a result of deep ulcerating diseases of the anorectum. The commonest cause in most Western Countries relates to Crohn disease, whereas in developing countries, such as in the Indian subcontinent, tuberculosis (TB) is much more common. Other less frequent causes include fistulation from adenocarcinoma or squamous cell carcinoma of the anorectum or less commonly as a complication of high-dose pelvic radiotherapy, whereas other infections are recognized to cause perianal fistulation including actinomycosis and human immunodeficiency virus (HIV) infection.
Classifying Fistula-In-Ano The most frequently used classification system used in surgical practice is the Parks Classification.1 This describes the anatomical location of the fistula. However, to understand the 313
D.J.M. Tolan
314 classification of fistulas using this and all other systems we must first evaluate the relevant anatomy of the perianal region and pelvic floor.
Anatomy for Fistula Classification The pelvic floor is a conical structure formed from the levator ani muscles (Fig. 1). As these pass medially and inferiorly from their attachment on the pelvic sidewall, they merge with the V-shaped or U-shaped puborectalis muscle sling. The puborectalis passes posterior to the anorectal junction and acts to pull the anorectal junction anteriorly toward the pubis to assist with continence. Inferiorly, the puborectalis blends with the continuous circular external anal sphincter forms, which passes inferiorly to the anal verge. Although anatomical studies have described 3 separate components of the external sphincter (deep, superficial, and subcutaneous), these are usually inseparable on imaging evaluation and act as a single functional unit. All of these pelvic floor muscles are striated, providing baseline tone in addition to voluntary squeeze or relaxation for the regulation of continence and defecation respectively. The internal sphincter lies medial to the external sphincter and is formed from smooth muscle, which is continuous with the muscularis propria of the rectum. This runs from the anorectal junction inferiorly to where it terminates as it interdigitates with the lowermost fibers of the external sphincter muscle. The relevance of the dentate line has already been described but this is not visible on radiological assessment. Instead, its position may be approximated to lie 2 cm superior to the anal verge, or alternatively caudal to the inferior border of the puborectalis muscle that is usually a position one-third to onehalf the distance of the anal canal from the anal verge. The anal
Figure 1 Normal anatomy of the pelvic floor and anal sphincters. The levator ani (lev ani) muscles blend with the puborectalis (pr), which in turn merges with the circular external sphincter (ext). The internal sphincter is continuous with the muscularis propria of the rectum, running the length of the anal canal. The intersphincteric space (white arrow) occupies the plane between the internal sphincter and the striated sphincter muscles. The dentate line, though not visible radiologically, lies just below puborectalis, approximately 2 cm from the anal verge (broken circular line).
glands arising at this location may penetrate into and beyond the internal anal sphincter muscle into a potential space called the intersphincteric plane. This is a critical potential space for perianal infection as such infections may track either inferiorly or superiorly along it. Alternatively, infection may pass from the intersphincteric plane through the external sphincter muscle into the fat of the ischioanal fossa, which lies below the levator ani muscles and lateral to the anal sphincters.
Parks Classification Sir Alan Parks from St Marks Hospital first produced a classification of perianal fistulas, based on the relationship of the fistula to the anal sphincter complex (Fig. 2).1 Superficial or submucous fistulas are rarely encountered in radiological practice but are the simplest to treat surgically. Intersphincteric fistulas arise in the anal canal before descending in the intersphincteric plane to reach the anal verge. A transsphincteric fistula may track for a variable distance in the intersphincteric plane before crossing the external sphincter muscle into the ischioanal fossa fat to reach the perineum or adjacent urogenital structures. Supralevator fistulas also originate in the intersphincteric plane but instead pass superiorly before they cross the levator ani muscles and descend through the ischioanal fossa to the perineum. The final group of fistulas are the extrasphincteric fistula; these usually arise from the rectum and entirely bypass the anal sphincter complex by crossing the levator ani muscle via the mesorectum before reaching the perineum through the ischioanal fossa.
Figure 2 Inter—intersphincteric fistula: confined to the intersphincteric plane. These may cross the lowermost fibers of the external sphincter (dotted arrow). Trans—transsphincteric fistula: may traverse the external sphincter high or low to pass some distance distally in the intersphincteric plane before crossing the external sphincter. Supra—supralevator fistula: sepsis tracks proximally in the intersphincteric plane before piercing the levator ani to descend through the ischioanal fossa. Extra—extrasphincteric fistula: pelvic pathology generates sepsis that tracks across the levator ani to the perineum.
MRI for perinatal fistula In addition to the main fistula, additional elements may extend from it that are commonly called extensions, secondary tracks, or branches. Although these extensions can be linear, they may also form a curved shape in either the intersphincteric plane, supralevator space, or within the ischioanal fossa fat, where they may be referred to as “horseshoe” extensions, which is a commonly used term in surgical practice.
St James’s University Classification This radiological classification system groups types of fistula based on outcomes from surgery and categorizes them from simple to progressively more complex using a combination of the fistula topography related to the sphincter complex and the presence or absence of extensions.2 Type 1 and type 2 refer to intersphincteric fistulas, the latter associated with additional secondary extensions or intersphincteric abscesses. Type 3 and type 4 are transsphincteric fistulas, the latter complicated by abscesses or secondary tracks in the ischioanal fossa. The final group type 5 includes supralevator and translevator fistulas (combining the suprasphincteric or extrasphincteric fistula classes from the Parks classification). As the class of fistula increases on this scale, the complexity of surgical intervention increases along as the risk of poorer outcome from surgery. Although this classification system has been adopted by some radiologists, and it has prognostic ability, most surgeons are unfamiliar with it which limits its applicability in day-today practice.
How Does MRI Make a Difference? MRI vs Other Modalities MRI offers superior contrast resolution to computed tomography in the assessment of perianal fistula, thereby limiting the role of computed tomography to the exclusion of pelvic organ pathology or abnormalities of the large or small bowel. Endoanal ultrasound is able to assess perianal fistula but because of the depth of penetration it is difficult to track the distribution of more complex features like deep extensions. However, it is better able to visualizing the internal opening of the fistula than MRI. Patients can find endoanal ultrasound assessments painful in the setting of active sepsis and in particular abscess formation. Contrast studies with fistulography no longer have a contributory role in the assessment of these patients.
Clinical Decision-Making With MRI MRI is better able to detect complex features than standard clinical evaluation.3 This added value has been shown to reduce the relapse rate in patients with recurrent perianal fistula after initial failed surgery by up to 75%.4 Surgical
315 guidelines have since been produced with recommendations for practice with the Association of Coloproctologists of Great Britain and Ireland advising MRI as the modality of choice for any patients with recurrent fistula or where clinical assessment indicates that this is a complex fistula at presentation.5 Guidance from the American Society of Colon and Rectal Surgeons does not give any specific recommendations on the value of imaging in the assessment of patients.6 MRI is also the method of choice for the assessment of patients with active perianal Crohn disease. The European Crohn’s and Colitis Organisation guidelines recommend MRI assessment at presentation in all patients, except for those with a simple fistula judged by a surgical expert, either with acute perianal sepsis or those presenting with a more indolent course to classify the fistula.7 The commonest cause for treatment failure is untreated or undetected fistula extensions and abscesses, and as MRI is the most accurate method of assessing perianal Crohn disease, surpassing examination under anesthesia, its role in initial assessment and in treatment planning is essential.8 MRI is then recommended in patients in remission at 12 months or in those patients with persistent disease despite surgery and medical therapy.9 In a follow-up of patients with Crohn disease on medical therapy, radiological healing lags behind the clinical response to treatment, and patients who are withdrawn from maintenance therapy can relapse in the same fistula distribution as seen on prior MRI scans where healing has been demonstrated in the past.10 A scoring system of signs of disease activity and assessment of severity has been developed by Van Assche et al,11 using a combination of an assessment of the extent of the fistula (the complexity of the fistula, the relationship to the sphincters, and whether there is supralevator extension) and the activity of the fistula (judged by the T2 signal intensity of the fistula, whether abscesses or cavities are present and the presence or absence of rectal wall involvement). Although these signs of disease activity are useful, a formal scoring of activity is not considered practical in routine clinical practice.8
Hardware Requirements for Imaging Endoanal coils are not required for assessment of perianal fistula. Multichannel phased array body coils obtain highquality imaging and allow a wider field of view to be obtained. They are also less uncomfortable for a patient than endoanal coils, who may have active perianal inflammation. Magnetic field strength is not an essential factor and 1.5-T MRI is more than adequate for diagnosis. In general, a slice thickness acquisition between 3-5 mm is required to allow detection of fistulas and extensions and the relationship of the fistula to the anal sphincter mechanism.
Choosing the Right Plane for Imaging Scanning with a “true” anatomical plane produces images acquired obliquely through the sphincter complex, which can
D.J.M. Tolan
316
A
B
C
D
Figure 3 High transsphincteric fistula (white arrow). (A) oblique coronal, (B) axial FS T2, (C) oblique coronal, and (D) axial postcontrast FS T1. Oblique coronal image shows a long track anterior to the external sphincter. Oblique axial image more clearly shows the fistula lying anterior to the external sphincter muscle. Note the dark dot in the center of the fistula denoting a seton suture.
render accurate assessment of these features impossible, and this should be avoided. Instead, by following the same conventions as used for imaging the rectum and gynecological structures for assessing malignancy, it is important to perform imaging in planes that are orthogonal to the anal canal itself. As the anal canal passes craniocaudally, it moves posteriorly at an angle of approximately 451. It is important to scan using this plane to allow evaluation of the height of the internal opening relative to the anal canal and also the relationship of the fistula to the sphincter muscles. It is a matter of personal preference which scan planes are acquired. Although the axial and coronal planes are essential, it is the author’s personal experience that the sagittal plane is less useful, except for patients where there is clinical suspicion of fistula-to-urogenital structures or involvement of the natal cleft. This can usually be deduced before the scan at the time the examination is protocolled. It is also important to evaluate the whole of the perineum and to include an assessment of the lower pelvis or mesorectum as these are potential sites of fistula extension. This is particularly important in patients with complex fistulas, for example, from Crohn disease.
Sequence Selection for Interpretation Different authors have used a range of sequences for imaging evaluation of perianal fistula. These may be broadly divided into the following categories:
“Standard” Pelvic Imaging Sequences (T1 and T2 Spin Echo) Some authors advocate standard T1 or T2 (or a combination of both) for assessment of fistula.12,13 These provide good demonstration of sphincter anatomy. Fistula tracks appear dark on T1-weighted sequences (similar to adjacent muscle) and so this reminds of the dark fistula being visible on the bright background of fat in the soft tissues. A healed fibrotic fistula track would look indistinguishable from an active fistula on this sequence alone. Similarly, a highsignal fistula on T2-weighted sequences may have similar signal to adjacent intersphincteric or ischioanal fossa fat that may limit conspicuity.
MRI for perinatal fistula
Sequences Assessing Edema (STIR or T2 With Fat Saturation T2) An advantage of these sequences is an increase in conspicuity of inflammatory changes related to an active fistula on the scan. The fistula may have a dark wall, whereas the fistula contents would appear bright from fluid or inflammatory granulation tissue (Fig. 3A and B).13-15 Inflammatory changes may extend beyond the fistula itself, particularly in the context of a contained abscess. Typically, the anatomy of the anal canal is clearly visible enough to allow identification of the internal and external sphincter muscles. The internal sphincter smooth muscle is intermediate-to-high signal compared with the lower-signal striated external sphincter muscle and puborectalis in a normal subject. Fibrosis appears dark, as sometimes seen at the site of a healed fistula.13
Sequences Assessing Contrast Enhancement (T1 With or Without Fat Saturation Post Gadolinium) Post–gadolinium T1-weighted MRI assess abnormal enhancement related to inflammation (Fig. 3 C and D). This is made more conspicuous where fat saturation is applied and can be derived from 2-dimensional or 3-dimensional sequences. Although contrast injection adds cost and time, it allows differentiation of abscess or fluid within a fistula track from granulation tissue, as an abscess demonstrates rim enhancement with central nonenhancement (from the fluid within it) (Fig. 4).16 This can be particularly important in surgical planning for complex cases or where patients are being treated with powerful immunosuppressants, such as anti–TNF agents, where it may assist in deciding to pause treatment and offer antibiotic therapy or surgery. In the author’s and others’ experience, this can also make small fistulas more conspicuous and increase diagnostic confidence.13 Healed fibrotic fistula tracks conversely either demonstrate no enhancement or only low-level enhancement similar to adjacent muscle.
Diffusion-Weighted Imaging Although some authors propose a potential routine role for diffusion-weighted imaging,17 and in particular for the diagnosis of abscess complicating fistula-in-ano that might obviate the requirement for contrast administration, there are downsides to diffusion-weighted imaging assessment of the perianal region. In particular, there are difficulties in interpretation resulting from artifacts close to air–soft-tissue interfaces, which can degrade image quality.18
Sequence Combinations for MRI In arriving at a standard technique for fistula assessment, a combination of different imaging sequences is typically preferred. Although a combination of T2 and postcontrast 3-dimensional gradient-echo T1-weighted sequences with fat saturation has been proposed as optimal for the accurate assessment of fistula,13 the author’s own experience prefers a combination of a fat-saturated T2-weighted sequence, to look
317 for edema, in combination with postcontrast fat-saturated T1-weighted sequences to look for abnormal enhancement.
When Is a Fistula Not a Fistula? The Differential Diagnosis Although fistula-in-ano is a clinical diagnosis, other perianal conditions can mimic the symptoms and imaging appearances. Pilonidal sinus results from infection of ingrowing hairs, most commonly in the subcutaneous tissues of the natal cleft. These can produce a variety of appearances ranging from focal inflammation in the subcutaneous fat to complex branching sinuses. Sinuses may extend superiorly to the posterior aspect of the sacrum and inferiorly they may reach the outer aspect of the external anal sphincter. However, the sphincter mechanism itself is only rarely affected.19 Sparing of the sphincter complex is typical and allows differentiation from true fistula-in-ano (Fig. 5). Hidradenitis suppurativa is an apocrine gland disorder producing axillary and genitofemoral abscesses. Abscesses, sinuses, and fistulas in chronic cases may extend posteriorly to the perirectal area.20,21 Although it is less common, isolated anorectal involvement is recognised, with perianal fistula and abscess formation.21 In such cases, Crohn disease is a recognized association and these 2 conditions may coexist.20 Atypical infection can produce perianal fistula and is recognized in HIV, TB, and actinomycosis.22-24 The diagnosis is usually indicated either by a failure of surgical therapy, prior known infection, or a high prevalence in the local population (in the case of TB). Patients with HIV can develop perianal symptoms unrelated to fistula-in-ano as 2 or more of fissures, tumors, condylomas, and abscesses may be present in up to 66% of cases.22 Immunocompromised patients, for example, neutropenic patients having chemotherapy for solid-organ malignancy or patients with hematological malignancy, are a final group for consideration. These are a difficult group of patients to assess clinically and surgeons are understandably reluctant to perform examination under anesthesia without an intention to intervene to drain an abscess. MRI typically demonstrates generalized and quite-extensive perianal edema, but without abscess formation. Only a minority have an abscess requiring surgical intervention.25
Recommendations for Structured Reporting The purpose of the MRI is to determine confirmation of the presence of abnormality and to provide a surgical road map to guide treatment. In particular, the report needs to accurately map the distribution of the fistula and the location of any complex features, as these can be difficult to detect at examination under anesthesia and particularly in more complex cases.26 To achieve this, the following points should be included in reports to make them clinically useful.
D.J.M. Tolan
318
A
B
Figure 4 Recurrent intersphincteric fistula due to clinically occult left supralevator extension (arrow), a strength of MRI assessment. (A) Note the uniform high signal on FS T2 and (B) high-signal rim with central nonenhancement from pus on postcontrast FS T1.
on the scan, such as rectal inflammation may suggest underlying Crohn disease as a cause for fistula formation.
Classification of the Abnormality Fistula-In-Ano or Alternative Diagnosis? Sometimes the radiologist will be the first to suggest an alternative diagnosis such as pilonidal sinus or hidradenitis, for example, if there are inflammatory changes in the groins, or only if they are localized to the natal cleft. Additional features
A
Single Fistula or Multiple Fistulas? If a single fistula is present, then the structured report is usually straightforward. If multiple fistulas are present, then these
B
Figure 5 Extensive pilonidal sinus (black arrow). FS T2 sagittal (left) and axial (right). Note the typical sparing of the external sphincter (thick arrow) to differentiate from fistula-in-ano.
MRI for perinatal fistula
319
should be described in sequence and in a logical order (for example, clockwise around the circumference of the anal canal according to the internal opening).
What Fistula Type? The Parks classification of the fistula should be provided depending on the relationship of the fistula track to the sphincter complex, as described earlier.1 In some cases, a suture called a “Seton” may have been already placed within the fistula following examination under anesthesia. This would appear as a continuous low-signal loop within the fistula track (Fig. 6). This is sometimes best appreciated on coronal sequences. Figure 7 Axial FS T2 midcanal. The internal opening of a fistula at 11o’clock seen as an oval high-signal area crossing the internal sphincter (black arrow head). The internal opening is not often visible in this way. Clock face position common surgical terminology and should be used routinely for reporting.
Describing the Path of the Fistula Internal Opening The internal opening position should be given both in the height within the anal canal and the “clock position,” which is the standard surgical terminology, using the anal clock face with 12-o’clock anteriorly, 3-o’clock to the left, 6-o’clock posteriorly, and 9-o’clock to the right (Fig. 7). The internal opening is not always visible as a clear defect. Most commonly, its position is adjacent to the epicenter of the inflammation in the intersphincteric plane and so the position can be “predicted” from this. It is important to provide this prediction, as the surgeons will rely on this to guide their initial exploration.
sphincter,as this may alter the therapeutic options for the surgeon. In particular, the surgeon may choose to avoid a sphincterotomy (cut the sphincter) if the fistula is deemed to be high. External Opening The external opening should be provided with an estimated clock position and distance from the anal verge. This is particularly important where more than 1 fistula is present.
Path of the Fistula The course of the fistula should be described from the internal opening. In particular, details need to be provided of the clock position that the fistula passes inferiorly either within the intersphincteric plane (for an intersphincteric fistula) or where it crosses the external sphincter (in a transsphincteric fistula). For a transsphincteric fistula, it is important to provide an estimate of the height that the fistula traverses the external
A
Complex Features Position of extensions. The position of extensions from the main fistula should now be provided if these are present. The location should be clearly stated, for example, whether it is within the intersphincteric plane or within the supralevator compartment or ischioanal fossa and the height relative to the
B
Figure 6 (A) Oblique coronal and (B) axial FS T2. Chronic high transsphincteric fistula containing a seton (thin white arrow). The fistula is surrounded by low-signal fibrosis (thick white arrow). Ongoing high-signal inflammation is present in surrounding ischioanal fat (thin black arrow).
D.J.M. Tolan
320
A
B
Figure 8 Perianal pain in Crohn disease. (A) FS T2 and (B) postcontrast FS T1-weighted image shows a 2.5-cm intersphincteric abscess with avid rim enhancement (thin white arrow), The external sphincter (thick white arrow) is thinned out between 7 and 10-o’clock position. Note the eccentric displacement of the anal canal (asterisk).
anal canal and the clock position that the abnormality is located. Where a horseshoe extension is present, the clock positions of the arc of the extension should be provided (for example, “between 2-7–o’clock in the intersphincteric plane in the mid canal”).
T2-weighted); postcontrast FS T1-weighted sequences allow more easy differentiation, as fluid will not enhance within the fistula. Abscesses will often also have a surrounding “halo” of edema in the surrounding tissues as the inflammation spreads beyond the wall of the abscess itself.
Undrained fluid and abscesses. Where a blind-ending extension contains fluid that has not been drained, the estimated diameter of this collection should be provided, in addition to the position, as this would allow the clinician to determine whether it requires specific intervention (Fig. 4). This can be particularly important in the setting of Crohn disease where patients are on biological therapy, where patients are potentially at risk of sepsis (Fig. 8), or in patients with complex fistula-in-ano where the presence of an undrained extension might increase the risk of failed surgery. Differentiating undrained fluid from granulation tissue can be very difficult without contrast agent administration, as both show high signal on edema sequences (STIR or FS
Bone edema and osteomyelitis. In patients with established complex fistulation, particularly with Crohn disease, fistulas may extend to involve the bony pelvis and particularly the coccyx and lower sacrum. In such cases, the addition of a sagittal plane is very useful to cover the whole of the sacrum to determine the extent of the abnormality. This would appear as localized high-signal bone edema on either STIR or FS T2-weighted sequences and may be associated with disruption of the cortex when osteomyelitis develops. The supervising clinician will need urgent notification of this finding, as it is usually clinically occult and may require cessation of biological therapy and long-term antibiotics in addition to surgical exploration to drain the fistula in question.
A
B
Figure 9 Recurrent urinary tract infection with history of radiotherapy for prostate cancer. (A) Sagittal FS T2 and (B) postcontrast FS T1 acquisition. Foley catheter (thin arrow) and erectile device (asterisk). Small fistula with defect visible only on the post–contrast-enhanced image (thick white arrow).
MRI for perinatal fistula
A
321
B
Figure 10 Axial T2. History of fistula-in-ano. Pain and discharge. (A) Dark fibrosis related to chronic intersphincteric fistulas (thin arrow). (B) Mucin filling a fistula track (thick arrow) and displacing the vagina (V) anteriorly from mucinous fistula adenocarcinoma.
Imaging follow-up is usually required to see a reduction in edema in response to treatment. Anogenital fistulation. Depending on the length, distribution, and severity of fistulas, the vagina and base of penis can be involved by fistulation. This is not always very symptomatic or clinically suspected but obviously important for surgical management. In some cases, symptomatic patients may have defunctioning of the rectum, particularly where surgical intervention has failed. Fistulas may also be detected between the rectum and the bladder or vagina. These are often caused by other diseases, such as diverticular disease, Crohn disease, and as a delayed complication from radiotherapy (Fig. 9). Malignancy in chronic fistulas. Few patients will develop malignancy as a consequence of chronic fistula inflammation.27 This is usually where a fistula has been present for 10 or more years. Both squamous cell carcinoma and adenocarcinoma tumor types are recognized. Identification of these tumors can be very challenging. Signs include development of a new soft-tissue signal component when compared with previous assessments or alternatively, where mucinous adenocarcinoma develops, a lobulated pattern with a more complex internal architecture to the fistula than a usual more uniform fluid or granulation tissue– related content (Fig. 10). In such cases, biopsy is required to confirm the diagnosis and it would be directed by the imaging appearances.
Conclusion MRI is an essential tool for the assessment of patients with fistula-in-ano, and particularly in more complex disease. Radiologists must be aware of the anatomy of the anal canal,
the classification of fistulas and able to produce a succinct report detailing the course of the fistula to allow planning for surgical intervention. A thorough understanding of the implications of more complex features of fistulas is required, in particular for the management of patients with perianal fistulas from Crohn disease on biological therapy, as well as an ability to differentiate conditions mimicking fistula-in-ano, such as hidradenitis and pilonidal sinus.
References 1. Parks AG, Gordon PH, Hardcastle JD: A classification of fistula-in-ano. Br J Surg 63:1-12, 1976 2. Spencer JA, Chapple K, Wilson D, et al: Outcome after surgery for perianal fistula: Predictive value of MR imaging. Am J Roentgenol 171:403-406, 1998 3. Buchanan GN, Halligan S, Bartram CI, et al: Clinical examination, endosonography, and MR imaging in preoperative assessment of fistula in ano: Comparison with outcome-based reference standard. Radiology 233:674-681, 2004 4. Buchanan GN, Halligan S, Williams AB, et al: Magnetic resonance imaging for primary fistula in ano. Br J Surg 90:877-881, 2003 5. Williams JG, Farrands PA, Williams AB, et al: The treatment of anal fistula: ACPGBI position statement. Colorectal Dis 9:18-50, 2007 6. Steele SR, Kumar R, Feingold DL, et al: Practice parameters for the management of perianal abscess and fistula-in-ano. Dis Colon Rectum 54:1465-1474, 2011 7. Van Assche G, Dignass A, Reinisch W, et al: The second European evidence-based Consensus on the diagnosis and management of Crohn’s disease: Special situations. J Crohns Colitis 4:63-101, 2010 8. Panes J, Bouhnik Y, Reinisch W, et al: Imaging techniques for assessment of inflammatory bowel disease: Joint ECCO and ESGAR evidence-based consensus guidelines. J Crohns Colitis 7:556-585, 2013 9. Avaliable at: http://www.e-guide.ecco-ibd.eu/algorithm/treatment-fistula ting-perianal-disease. Accessed March 30, 2016. 10. Tozer P, Ng SC, Siddiqui MR, et al: Long-term MRI-guided combined anti-TNF-α and thiopurine therapy for Crohn’s perianal fistulas. Inflamm Bowel Dis 18:1825-1834, 2012 11. Van Assche G, Vanbeckevoort D, Bielen D, et al: Magnetic resonance imaging of the effects of infliximab on perianal fistulizing Crohn’s disease. Am J Gastroenterol 98:332-339, 2003
322 12. Koelbel G, Schmeidl U, Majer MC, et al: Diagnosis of fistulae and sinus tracts in patients with Crohn’s disease: Value of MR imaging. Am J Roentgenol 152:999-1003, 1989 13. Torkzad MR, Ahlström H, Karlbom U: Comparison of different magnetic resonance imaging sequences for assessment of fistula-in-ano. World J Radiol 6:203-209, 2014 14. Barker PG, Lunniss PJ, Armstrong P, et al: Magnetic resonance imaging of fistula-in-ano: Technique, interpretation, and accuracy. Clin Radiol 49: 7-13, 1994 15. Haggett PJ, Moore NM, Shearman JD, et al: Pelvic and perianal complications of Crohn’s disease: Assessment using magnetic resonance imaging. Gut 36:407-410, 1995 16. Spencer JA, Ward J, Beckingham IJ, et al: Dynamic contrast-enhanced MR imaging of perianal fistulas. Am J Roentgenol 167:735-741, 1995 17. Hori M, Oto A, Orrin S, et al: Diffusion-weighted MRI: A new tool for the diagnosis of fistula in ano. J Magn Reson Imaging 30:1021-1026, 2009 18. Dohan A, Eveno C, Oprea R, et al: Diffusion-weighted MR imaging for the diagnosis of abscess complicating fistula-in-ano: Preliminary experience. Eur Radiol 24:2906-2915, 2014 19. Taylor SA, Halligan S, Bartram CI: Pilonidal sinus disease: MR imaging distinction from fistula in ano. Radiology 226:662-667, 2003
D.J.M. Tolan 20. Kelly AM, Cronin P: MRI features of hidradenitis suppurativa and review of the literature. Am J Roentgenol 185:1201-1204, 2005 21. Griffin N, Williams AB, Anderson S, et al: Hidradenitis suppurativa: MRI features in anogenital disease. Dis Colon Rectum 57:762-771, 2014 22. Barrett WL, Callahan TD, Orkin BA: Perianal manifestations of human immunodeficiency virus infection: Experience with 260 patients. Dis Colon Rectum 41:606-611, 1998 23. Sultan S, Azria F, Bauer P, et al: Anoperineal tuberculosis: Diagnostic and management considerations in seven cases. Dis Colon Rectum 45: 407-410, 2002 24. Coremans G, Margaritis V, Van Poppel HP, et al: Actinomycosis, a rare and unsuspected cause of anal fistulous abscess: Report of three cases and review of the literature. Dis Colon Rectum 48:575-581, 2005 25. Plumb AA, Halligan S, Bhatnagar G, et al: Perianal sepsis in hematologic malignancy: MR imaging appearances and distinction from cryptoglandular infection in immunocompetent patients. Radiology 276:147-155, 2015 26. Buchanan G, Halligan S, Williams A, et al: Effect of MRI on clinical outcome of recurrent fistula-in-ano. Lancet 360:1661-1662, 2002 27. Egan L, D’Inca R, Jess T, et al: Non-colorectal intestinal tract carcinomas in inflammatory bowel disease: Results of the 3rd ECCO Pathogenesis Scientific Workshop (II). J Crohns Colitis 8:19-30, 2014