Evaluation of the anal sphincter by anal endosonography

Introduction

First described in 1989, anal endosonography (AES) was the first technique to visualize the anal sphincter complex with sufficient spatial resolution to resolve its individual muscular components. Despite the advent of magnetic resonance imaging (MRI), including endoanal receiver coils, AES generally remains the technique with the highest spatial resolution, and is also quick and easy to perform. The introduction of AES precipitated a significant reappraisal of the causes of anal incontinence (and its treatment), which had hitherto been thought to be mainly due to pelvic neuropathy. When incontinent patients were studied with AES, it rapidly became clear that unsuspected anal sphincter disruption was present in many cases. Patients with disrupted sphincters can be scheduled for surgical procedures that aim to restore integrity to the sphincter ring, whereas patients whose sphincters are intact, or whose muscles are thought to be of poor quality, can be directed toward conservative measures or alternative surgical approaches.

At present, AES is the pivotal examination in the clinical decision-making process for incontinent patients, and is usually performed together with anorectal physiologic testing. While AES is probably most useful to characterize obstetric sphincter injury, it has also facilitated characterization of other causes of anal incontinence. For example, with AES, the examiner can identify neurogenic incontinence by way of specific patterns of sphincter atrophy and can identify occult, iatrogenic sphincter damage following anal surgical procedures.

Equipment and examination technique

Although it is possible to perform AES using an echoendoscope, by far the best results are obtained using a dedicated anal ultrasound probe. The anus is a very superficial structure, and an echoendoscope is cumbersome and expensive when compared with a probe designed specifically for anal examination. AES first employed a 7.5-MHz transducer that had been designed initially for rectal cancer staging and prostatic imaging. The transducer was covered by a rubber balloon, then inserted through the anus into the rectum, the balloon inflated with degassed water, and the transducer rotated mechanically to produce 360-degree images of the rectal wall. Professor Clive Bartram of St. Mark’s Hospital, London, realized that by simply replacing the soft rubber balloon with a rigid plastic cone, the rotating transducer could be withdrawn into the anus safely. This maneuver was previously impossible because the balloon would be torn when compressed by the anus against the rotating metal transducer.

Modern probes utilize a 360-degree piezoelectric transducer element and are of higher frequency ( Fig. 18.1 ). Examples are the Hitachi EUP-R54AW-19/33 and the BK medical 20R3 transducer (3–20 MHz). If desired, three-dimensional images can be achieved by withdrawing the probe during image acquisition (Hitachi) or via automatic transducer movement along the probe Z -axis while the probe is held stationary within the anal canal (BK Medical).

The examination is simple, well tolerated by the patient, and very rapid when performed by an experienced operator. No special patient preparation is required. The patient is told that discomfort, if any, will be similar to having a small finger in the anus, and the procedure will likely be much less uncomfortable than digital anorectal examination by a doctor. To the patient, the probe is potentially quite frightening, so it is worth mentioning that only the distal few centimeters will enter the anus (as opposed to rectal endosonography, during which insertion is obviously deeper).

The author examines men in the left lateral patient position, but prefers to examine women prone. Placing women in the left lateral position can occasionally distort anterior perineal anatomy and can induce an asymmetrical image, which makes it difficult to distinguish perineal scarring from normal anatomic features. The probe tip is lubricated with ultrasound jelly to achieve acoustic coupling and is then covered with a condom, which is itself lubricated to facilitate coupling and insertion. The probe tip is then inserted into the anus, and image acquisition commenced. The aim is to insert the probe so that the transducer lies just into the distal rectum. The probe is then withdrawn gently and slowly to examine the anal sphincters.

As for all ultrasound examinations, interpretation is based on the image displayed in real time on the monitor screen (with the exception of three-dimensional acquisition, in which case the examination in its entirety can be replayed later). However, still 2D images are usually obtained for archival purposes, and the author finds it is convenient to obtain these at three levels: proximal, middle, and distal anal canal (see later). These three levels are imaged at both standard and higher magnifications, for a total of six images per patient. Additional images may be obtained to clarify any pathology. The probe is rotated so that anterior (i.e., the 12-o’clock position) is uppermost on the monitor, which mimics the generally accepted standard for axial medical imaging. The examination is normally very quick, perhaps only a minute or so for an experienced operator, and especially if the sphincters are normal.

Anal sphincter anatomy

Clearly, a sound understanding of basic anal anatomy is a prerequisite for accurate interpretation of endosonographic findings. There are two anal sphincters: the external anal sphincter (EAS) is composed of striated muscle, whereas the internal anal sphincter (IAS) is smooth muscle. These form two cylindrical layers, with the IAS innermost ( Fig. 18.2 ).

The EAS arises from the striated muscles of the pelvic floor and is composed of three successive cylindrical bundles (deep, superficial, and subcutaneous from cranial to caudal) that are pretty much impossible to separate in practice. The deep portion fuses with the puborectalis (or pubococcygeus) muscle, which itself merges with the levator plate of the pelvic floor. The EAS extends approximately 1 cm caudal to the IAS, to form the subcutaneous portion of the muscle. Anteriorly, the EAS is closely related to several surrounding structures, such as the superficial transverse muscle of the perineum and the perineal body. Posteriorly, it is continuous with the anococcygeal ligament, a structure that is often more prominent in men and should not be mistaken for a posterior sphincter defect. The craniocaudal (i.e., longitudinal) extent of the EAS is much shorter anteriorly in women than in men, and this feature should not be confused with a sphincter defect. The short anterior sphincter renders women generally much more susceptible to sphincter injury than men (childbirth notwithstanding!).

The IAS is the distal termination and condensation of the circular smooth muscle of the gut tube. It extends from the anorectal junction to approximately 1–1.5 cm below the dentate line (see Fig. 18.2 ). The longitudinal muscle of the gut tube also terminates in the anal canal, but is less apparent than the IAS. The anal longitudinal muscle interdigitates between the EAS and the IAS, and terminates in the subcutaneous EAS and subcutaneous anus. Its exact sphincteric action, if any, is much less clear than that of the EAS and IAS, and its main purpose is possibly to brace the anus and thus prevent its eversion during defecation. Lying between the EAS and the anal longitudinal muscle is a potential plane, the intersphincteric space, which often contains fat. The anal sphincter is surrounded by the ischioanal fossa (often referred to by surgeons as the ischiorectal fossa), which contains fat predominantly.

Directly anterior to the anal sphincter is the central perineal tendon or perineal body. In men, this lies posterior to the bulbospongiosus and corpus cavernosum and their related muscles, whereas in women, it lies within the anovaginal septum. Many structures insert fibers into the perineal body, such as the EAS, the deep and superficial transverse muscles of the perineum, the bulbocavernous muscle, and the puborectalis muscle. These structures should not be confused with sphincter defects. For example, normal variants of anal sphincter anatomy have been identified, such as differing relationships between the superficial transverse perineal muscle and the EAS.

The distal anal canal is lined with stratified squamous epithelium, richly supplied by sensory receptors. These receptors are most concentrated at the dentate line, which demarcates the junction with proximal columnar epithelium. The anal subepithelial tissues are relatively thick, and this lining and its underlining vascular spaces—the anal cushions—also play a role in maintaining continence.

Normal endosonographic findings

Because the sphincter muscles are cylindrical, a 360-degree field of view is optimal, and the axial plane is also the most relevant surgically when considering sphincter defects. As stated earlier, it is convenient to obtain baseline images at three levels at a minimum: the proximal, middle, and distal anal canal.

The proximal anal canal is primarily identified by the puborectalis and transverse perineal muscles ( Fig. 18.3 A). The puborectalis slings around the anorectal junction posteriorly and can be distinguished from the EAS because its anterior ends splay outward as they travel toward their fusion with the pubic arch (see Fig. 18.3 A). The IAS is visible as a continuous ring and is generally the easiest individual structure to differentiate from other sphincter components because it is normally very hyporeflective (i.e., “dark”). The subepithelial tissues, EAS, and longitudinal muscle all normally display varying degrees of hyperreflectivity, and their margins can often be difficult to define precisely, although direct comparisons with endoanal MRI have helped tremendously. Increases in transducer frequency that improve spatial resolution have also helped to clarify sonographic anatomy, as has three-dimensional imaging.

In their seminal early studies, Sultan et al. carefully imaged cadaveric specimens following sequential histologic dissection of anal layers to validate sonographic appearances. These investigators found that the echogenicity of normal muscular components changed as their orientation was altered with respect to the transducer. Thus normal variant striated muscle slips may appear hypoechoic, depending on their orientation to the transducer, and should not be confused with sphincter tears or scars.

If the probe is withdrawn just a centimeter or so from the proximal anal canal position, the anterior ends of the normal puborectalis muscle will converge anteriorly as they segue imperceptibly into the adjacent deep EAS. The midanal canal is thus defined where the EAS forms a complete ring anteriorly, at superficial EAS level (see Fig. 18.3 B). The IAS is also normally thickest and best demonstrated at this level. At this level, the intersphincteric plane and longitudinal muscle may also be resolved as two distinct layers, with the longitudinal muscle forming distinct bundles of hyporeflective smooth muscle fibers.

Withdrawing the probe slightly more will move the field of view into the subcutaneous EAS (see Fig. 18.3 C). This lies below the termination of the IAS, which is either not visualized here or only partially visualized if its termination is irregular (a common normal variant). It is usually impossible to visualize the longitudinal muscle reliably at this level because it has thinned out as it interdigitates into the EAS, and it is predominantly composed of fibroelastic tissue rather than the smooth muscle found more proximally.

Correct interpretation of AES demands the operator has a firm grasp of normal sonographic anatomy. Pathology is defined by either muscular discontinuity (i.e., from sphincter tears/lacerations, secondary to a variety of causes) or abnormal muscular quality (which is usually caused by neuromuscular atrophy or degeneration). To appreciate muscular quality correctly, it is important to realize that normal sonographic appearances are contingent on both age and sex. Frudinger et al. examined 150 nulliparous women with AES to define normal age-related differences in sphincter morphology and found a highly significant positive correlation between IAS thickness and increasing age. In contrast, EAS thickness showed a highly significant negative correlation with increasing age. Some evidence also suggested that IAS reflectivity increased with age. No significant correlation was noted between age and thickness of subepithelial tissues, the longitudinal muscle, or the puborectalis muscle.

On average, the IAS measures 2–3 mm thick in normal adults (measured at either the 3-o’clock or 9-o’clock position at midanal canal level). A thin IAS has more significance in an older person with symptoms (see later sections). In addition, although the IAS can be measured easily because it contrasts with adjacent structures, other muscles are frequently more difficult to identify precisely and measurements suffer from greater interobserver variation. Gold et al. measured anal canal structures in 51 consecutive referrals: intraobserver agreement was superior to interobserver agreement and the 95% limits of agreement for EAS measurements spanned 5 mm versus 1.5 mm for IAS measurements. Interobserver agreement for diagnosis of sphincter disruption and IAS echogenicity was very good, suggesting that AES has generalizable diagnostic utility ( κ = 0.80 and 0.74, respectively).

Clear sonographic differences exist between men and women regarding the dimensions of anal canal structures and their sonographic appearances. Most importantly, the anterior complete ring of the EAS is shorter in women in the craniocaudal (longitudinal) direction. This difference has been widely appreciated for some time: Williams et al. used three-dimensional AES to show that craniocaudal EAS length was approximately 17 mm in women versus 30 mm in men. A short anterior EAS in a woman should not be misinterpreted as a sphincter defect. In addition, in men, the various muscular components generally show a more striated appearance ( Fig. 18.4 ).