SACRAL MONITORING

Confusing anatomy is often encountered during operations on complex dysraphic lesions in the lumbosacral canal. It is common to see nerve roots embedded in lipoma or scar tissue, or they may not be easily distinguishable from fibrous adhesion bands. Sometimes nerve roots that are bundled tightly by abnormally thickened arachnoid can look like a thickened filum terminale. Also, the transition between a functional but structurally deformed conus and an intramedullary lipoma is not always visually apparent. Thus, some objective means to identify the sacral nerve roots and the conus is necessary to ensure preservation of these neuronal structures. In addition, in some cases of complex transitional lipomas, the tip of the conus is tautly suspended by low sacral roots that are short, stout, and fibrotic. An assessment of their functional integrity is useful for determining whether dividing them, in order to complete the untethering process, would lead to unacceptable loss of sphincter function.

The first sacral and lower lumbar roots are recognized readily by intraoperative nerve stimulation while palpating for contractions of the respective segmental muscle groups through the surgical drapes. Identification of the lower sacral roots and functional quantification of these roots and their corresponding medullary connections, however, require some objective assessment of perineal sensation and sphincter function.

The assessment of evoked responses generated by directly stimulating parts of the sex organs, urethra, and anal canal constitutes the mainstay of sensory monitoring of the lower sacral segments. Monitoring of such responses is most useful when the distal conus or dorsal nerve roots are being rather strenuously handled, as in certain difficult resections of large transitional lipomas or during removal of the median fibrous sleeve of a Type I split cord malformation. The latency and amplitudes of the waveforms are exquisitely sensitive to structural deformation and ischemic changes to the central sensory pathways. Sensory evoked response monitoring is less useful in the identification of sacral sensory roots, because the responses are generated by end organ stimulation. Cortical responses generated by direct dorsal root stimulation give much less predictable waveforms, which are not stable enough for foolproof identification purposes.

The peripheral nerves that supply the bladder, anal canal, and perineal skin, all potentially available for stimulation, are divided into three main groups.

1.The pudendal nerve is the primary somatic nerve to this region. The pudendal motor neurons innervating the external sphincter and pelvic floor originate from Onuf's nucleus in the anterior horn of the S2 to S4 cord segments. The sensory fibers come from the corresponding dorsal root ganglia. The mixed fibers course via the S2, S3, and S4 roots to exit the spinal canal through the sacral foramina (Figure-1). Somatosensory impulses travel in this nerve from receptors located in the skin of the genitalia and perineum, the pelvic floor, and bulbocavernosus muscles, as well as in the mucosa of the distal urethra and anus. Motor fibers in the pudendal nerve innervate the bulbocavernosus muscle, external urethral sphincter, external anal sphincter, and pelvic floor muscles. The pudendal nerve is the most easily accessible nerve for evoked response testing.

3.The pelvic splanchnic nerves supply the sacral parasympathetic innervation to the pelvic organs. The motor neurons in this nerve originate in the S2 to S4 cord segments, slightly more caudal than the pudendal motor neurons. The fibers are distributed to the pelvic organs via the S2 to S4 nerve roots and inferior epigastric plexus. The pelvic nerve carries sensory afferents from the proximal urethra, bladder wall, prostate, seminal vesicles, and rectum. Motor innervation is primarily to the detrusor muscles, the corpus cavernosus, the rectum, and probably the upper smooth-muscle portion of the external urethral sphincter. Evoked responses can be elicited on stimulation of the proximal urethra and bladder, presumably due to activation of the pelvic sensory fibers.

3.The hypogastric nerve plexuses carry autonomic (sympathetic) fibers from the intermediolateral cell column of the T11-L2 spinal cord segments. The preganglionic fibers course via the paravertebral sympathetic chain ganglia, inferior mesenteric plexus, superior hypogastric plexus, and finally the inferior hypogastric plexus. The postganglionic fibers are distributed to the smooth muscles of the bladder neck, the smooth-muscled internal urethral sphincter, the parasympathetic intramural ganglia of the detrusor muscles and probably the intrinsic portion of the external urethral sphincter. The postganglionic fibers also share connections with plexuses around the rectum and anal canal, seminal vesicles, ductus deferens, prostate, and corpus cavernosus in the male, and vagina in the female. It is uncertain how much the afferent component of the hypogastric nerves contributes to the evoked response in humans.

Urethral Evoked Response

Cortical evoked responses of very similar morphology and latencies can be obtained using stimulating electrodes embedded in a catheter inserted into the bladder. The catheter has a balloon at its tip, which can be pulled back snugly for anchorage. The location of the urethral electrodes can be kept reasonably constant to eliminate movement artifacts and interference.

Anal Evoked Response

Electrode-bearing catheters can also be inserted into the anal canal for measurement of anal evoked responses. The catheter is anchored by double balloons, the inner one within the anorectal junction and the outer one wedged at the anal verge. The cortical anal responses do not differ from the urethral responses or the pudendal dermatomal responses.

Spinal Evoked Response

Evoked responses can be recorded by electrodes placed on the skin over the spine in humans. They reflect the afferent volley traversing the dorsal columns. The responses progressively increase in latency at more rostral recording locations. Spinal evoked responses are relatively easy to obtain in children, but the amplitudes and waveform definition decrease with age, such that by mid-teenage years, these responses are more difficult to obtain, as in the case of adults. The response over the mid-tolower lumbar spine consists of an initially positive triphasic potential, representing the volley as it ascends the cauda equina. Over the caudal thoracic spine, the response consists of an initially positive, predominantly negative triphasic wave, the negative component of which has several peaks or inflections. The initial portion of this response arises in the intramedullary continuation of the dorsal root fibers, and the subsequent portion reflects synaptic activity concerned with local reflex mechanism rather than the propagation of the response to more rostral cord levels. From the mid-thoracic to the cervical levels, the response consists of small, triphasic potentials that are difficult to follow, presumably arising from multiple ascending pathways including the dorsal and dorsolateral columns.

The only consistent spinal pudendal response has been from stimulation of the dorsal nerve of the penis. The recording electrodes are usually fixed at the T12-L1 interspinous space. The response has a morphology comparable to the spinal response from the posterior tibial and peroneal nerves but with smaller amplitudes and a much shorter latency (Figure 3). The spinal pudendal evoked response is sometimes not measurable in overweight individuals, but its presence yields useful information concerning peripheral sensory conduction from the penis since it bypasses the central conduction pathway rostral to the thoracic levels. Because the cortical pudendal evoked response has similar latency with the cortical posterior tibial response, the central conduction time involved in the pudendal pathways must be considerably longer than that in the posterior tibial pathways.

Figure 3. Spinal pudendal evoked responses recorded over the T12-L1 interspinous space on stimulation of the dorsal nerve of the penis (upper), and spinal responses on stimulation of the posterior tibial nerve (lower). Note the much shorter latency of the pudendal response.

Modality for Motor Monitoring of the S2-4 Nerve Roots

Pudendal sensory evoked responses are useful in monitoring intraoperative injury to the conus and lower sacral sensory nerve roots, but they are neither qualitatively nor quantitatively suitable for the identification of the lower sacral roots (especially the motor roots) or conus from non-neural elements. Intraoperative identification requires some way of measuring the oneto-one stimulus-to-response coupling of end organ function when the nerve root in question is being stimulated. For the lower sacral roots, this means the assessment of sphincter function.

External Anal Sphincter Electromyography

External anal sphincter electromyography (EMG) has long been found useful as a qualitative tool for studying anorectal closure function and disorders. The EMG electrodes are either embedded in an anal plug or anal balloon, which is placed into the anal canal, or are in the form of needles inserted directly into the external anal sphincter transmucosally. The needle electrodes are more reliable because they are not subject to dislodgement or to having mechanical artifacts during contraction of the sphincter itself; however, accurate and secure placement of the needles requires some expertise and is initially best done by the neuro-urologist. The grounding plate is pasted on the patient's thigh, and EMG recordings are made using a standard bladder diagnostic unit. The sensitivity of the recording stylus is adjusted so that minimal deflection occurs at rest. With stimulation of the lower sacral motor roots, the stylus gives a discrete one-to-one spike-deflection, much different than the baseline.

External Anal Sphincter Pressure Monitor

The external anal sphincter EMG requires bulky and expensive equipment, as well as the availability of someone expert in the accurate placement of the needle electrodes. An alternate method of monitoring anal sphincter function is the direct measurement of the "squeeze pressure" induced by sacral root stimulation using a pressure-sensitive balloon inserted in the anal canal. This technique is simple and noninvasive, requires no special expertise, utilizes inexpensive, portable equipment, and produces easily interpretable pressure waves which are semi-quantitative and virtually unaffected by other electronic components in the operating room that are known to cause annoying baseline noise in an EMG recording.

Physiology

The relationship between EMG and contractile strength in a longitudinal muscle was first defined by Lippold, who found a linear relationship between the integrated action potentials on the EMG and the tension generated by voluntary isometric contractions of the human gastrocnemius. This linearity was explained by the fact that an increase in contractile strength of a muscle is brought about either by a spatially random increase in the number of contracting motor units or by random increments of discharge frequencies of the active units; in both situations, the integrated electrical output of the muscle would increase proportionately. The same linear relationship was also demonstrated in the external anal sphincter by Schweiger, who made simultaneous recordings of sphincter EMG and anal canal pressures with an anal balloon. These data support the validity of using squeeze pressure, instead of sphincter EMG, to monitor the functional status of the lower sacral motor neurons.

In order for the anal pressure monitor to be operational, some sphincter function must be present. Theoretically, a severely damaged motor nerve with only enough viable axons to generate a barely visible EMG would produce no measurable squeeze pressure; in such a situation, the EMG might be more sensitive. However, such a nerve would not provide useful sphincter function for the patient, and its preservation is of doubtful value. In the author's experience, any external anal sphincter that could generate enough voluntary or reflex (as in the bulbocavernosus or anal wink reflex in the infant) contractions to be appreciable by preoperative digital examination should produce recognizable pressure spikes on the anal pressure monitor. This anal balloon monitor is therefore sufficiently sensitive for the practical purpose of sacral root and conus identification.

Anatomy

The external anal sphincter consists of a bulky deep part, a fusiform superficial part, and a subcutaneous part decussating behind and in front of the anus. It encloses the lower part of the levator ani, the anorectal junction, and the anal canal in the shape of a funnel. The internal anal sphincter arises from the muscular coats of the rectum and insinuates itself between the rectal mucosa and the upper portion of the funnel.

The external anal sphincter is innervated by the pudendal nerve. This arises from the anterior division of S2 and S3 and both divisions of S4, enters the pudendal (Alcock's) canal through the lesser sciatic foramen, and divides into two main branches just proximal to the urogenital diaphragm. The proximal branch, the inferior hemorrhoidal nerve, supplies the striated muscles of the external anal sphincter; the distal branch, the perineal nerve, supplies the external urethral sphincter. The internal anal sphincter, composed of smooth muscles, is innervated by the hypogastric nerve, derived from the intermediolateral (sympathetic) columns of L1 and L2. Stimulation of the S2, S3, and S4 roots, therefore, activates only the external and not the internal anal sphincter. Furthermore, unless there is localized disease or trauma to the pudendal branches at the urogenital diaphragm, activity of the external anal sphincter reflects function of the external urethral sphincter.

The anal pressure balloon described here is an elongated ellipsoid selected specifically to pick up activities from all three parts of the external sphincter funnel. Its elongated span also minimizes the possibility of accidental dislodgement by contractions of the pelvic musculature induced intraoperatively. Although the elongated balloon will also pick up contractions of the internal anal sphincter, the latter is never activated by the nerve stimulator or by manipulation of the lower sacral spinal cord or nerve roots because its nerve supply is from L1 and L2. However, being made up of smooth muscles, the internal sphincter does have spontaneous rhythmic contractions that will be registered by the balloon, and these must be distinguished from the stimuli-generated pressure spikes from the external anal sphincter.

Equipment

The pressure sensor is a double-lumen balloon catheter ordinarily used for intraluminal angioplasty (Figure 4). The ellipsoidal balloon is made of treated polyethylene, which does not stretch or deform at high inflation pressures, so that a high degree of sensitivity to circumferential squeezing can be maintained. The central infusion catheter concentric with the balloon is not actually being used in the pressure measurement but functions effectively as a stent for easy balloon insertion. The balloon comes in different sizes, but the 3 x 0.8-cm (inflated diameter) balloon should fit almost any patient, from infants to large adults.

Figure 4. Double-lumen polyethylene balloon catheter. The infusion lumen is not involved in the pressure measurement but merely serves as a stent.

The balloon is held vertically with the tip down, and is maximally deflated and inflated several times with water to expel all air bubbles. It is then connected to a Bentley Model D240 pressure transducer, which displays the pressure tracing on a two-channel Datascope Model 870 monitor (Figure 5). Although the baseline pressure of the balloon, which can be adjusted by varying the amount of water used, does not affect the actual pressure measurement, it should be kept within a range that allows the monitor to give good-sized pressure waves in the usual sensitivity setting. The optimal condition is when the balloon is rendered just turgid (with 0.8 ml water for the 3 x 0.8-cm balloon) and when the sensitivity on the Datascope monitor is set at 25 (1 cm on the screen is calibrated to 25 torr). The balloon is inserted into the anal canal until its posterior end is just visible at the mucocutaneous junction and then taped securely to the gluteal skin.

One cutaneous electrocardiography (ECG) electrode is pasted over each iliac crest and a third on the right upper thigh. The ECG tracing is displayed continuously on the second channel of the Datascope screen (Figure 6).

Figure-5. Simple assembly, consisting of the balloon catheter, the Bentley pressure transducer, the injecting syringe and stop-cocks and the 2-channel Datascope monitor.

Figure 6. ECG electrodes are placed over the iliac crests and the thigh to register the stimulus artifact.

Technical Points

Intraoperative nerve stimulation is done with a disposable monopolar nerve locator-stimulator using 3 V and three variable current intensities: 0.5, 1, and 2 mA. The monopolar stimulator is chosen over the bipolar stimulator because only the former will generate sufficient volume-conducted current to produce an obvious stimulus artifact on the ECG when any tissue is touched by the monopolar electrode. When a lower sacral root is stimulated, the combined ECG stimulus artifact and the pressure spike from the external anal sphincter form an easily recognizable electromechanical couple on the monitor (Figure 7).

There are two advantages in having this electromechanical couple. 1) The stimulus artifact eliminates the possibility of a faulty stimulator or faulty stimulation technique because even at 0.5-mA current, the monopolar electrode always produces a prominent spike on the ECG tracing. If an artifact is present without a corresponding pressure spike at high current intensity, the tissue stimulated does not innervate the external anal sphincter; if neither stimulus artifact nor pressure wave is obtainable, then the nerve stimulator is faulty (Figure 8).

Figure 7. Monopolar stimulations of the S3 root in a 5-year-old child. Note the prominent stimulus artifact on the ECG tracing (arrows) coupled with large pressure spike waves measured at 50-torr peak values. Scale in torr.

Figure 8. Use of the EGG. A) Stimulation of a nonfunctional element (fibrous adhesion band) showing only EGG stimulus artifacts (arrows) but no pressure response. Scale in torr. B) Recognition of a faulty nerve stimulator when neither an EGG stimulus artifact nor a pressure response is detectable. Scale in torr.

2) Involuntary, rhythmic activity of the internal anal sphincter has been noted as spontaneous 5-to10-torr waves with a frequency of 10 to 30 per minute, which may be confused with external anal sphincter activities except for the fact that these spontaneous waves are completely out of phase with the ECG stimulus artifacts (Figure 9).

During nerve stimulation, the cerebrospinal fluid must be continuously suctioned away from the stimulation site to prevent current dispersion. With this precaution, supramaximal stimulation of the small sacral roots of infants and young children can usually be accomplished with 0.5 mA. The larger roots of adults sometimes require higher amperage, as does direct conus stimulation. Unilateral S2, S3, or S4 stimulation consistently generates a peak pressure of 40 to 70 torr, in line with recordings reported by Lane of 60 to 125 torr pressures with voluntary (bilateral) contraction in normal adults. Even in young infants, peak pressure responses are generally above 40 torr. S3 stimulation produces the strongest and most consistent response in the external anal sphincter. Direct stimulation of the conus also results in waves of comparable peak values but usually with a wider base, probably because of multilevel and bilateral recruitment of anterior horn cell units (Figure 10).

Figure 9. Use of the EGG. Spontaneous lower pressure waves «10 torr) from the internal anal sphincter are completely out of phase with the EGG stimulus artifacts (arrows). Scale in torr.

Figure 10. Direct stimulation of the conus in a 3-year-old boy, generating pressure waves with a wide base and irregular blunted peaks. Stimulus artifacts are indicated by the arrows. Scale in torr.

Occasionally, a small pressure wave of less than 7 torr follows S1 stimulation (which does not innervate the sphincters) because of compression on the protruding proximal portion of the balloon by the medial inferior fibers of the gluteus maximus. Although such "ripple waves" are easily differentiated from the tall spike waves of healthy lower sacral roots, they could be mistakenly construed as the subdued responses seen with partially damaged S2, S3, and S4 roots. This confusion is eliminated if care is taken to secure the posterior end of the balloon just above the mucocutaneous junction.

Stimulation of the filum terminale and nonneural tissues always produces a stimulus artifact but not a pressure wave. Thus, the S2, S3, and S4 roots and the conus can be distinguished from the S1 and lumbar roots, the filum, lipoma, fibrous adhesions, and other nonfunctional fibroneural bands, such as an occult myelomeningocele. The ECG artifact and pressure wave relationships are summarized in Table 1.

Table 1. Interpretation of Stimulus (ECG) Artifact and Pressure Response Relationship
Stimulus (ECG) Artifact	Pressure Response	Interpretation
-	-	Faulty stimulator
+	-; spike waves 40-75 torr	S2,S3,S4, conus medullaris
+	- or ripple waves (<7 torr) (and plantar flexion)	S1
+	-	Lumbar roots, filum, non-neural tissues
No stimulation	Rhythmic waves 10-30/min < 10 torr	Internal anal sphincter spontaneous activity

Clinical Use of the Anal Sphincter Function Monitors

The anal sphincter function monitors (EMG or pressure balloon) have been found useful in the following circumstances.

1. Functional sacral nerve roots embedded in large lipomas may be detected and traced through a sometimes aberrant course to their exit foramina. This is particularly useful in transitional lipomas that involve the dorsal as well as the ventral portions of the conus

2. Sacral nerve roots can be distinguished from fibrous adhesion bands.

3. Atrophic, fibrous distal nerve roots in long-standing myelodysplastic cases can be holding the conus tautly against the dura and can thus prevent complete release of the tethering. If these roots can be shown by the monitors to have no contribution to sphincteric functions, they should be cut.

4. Occasionally, the junction between functional conus and fat is not well demarcated in cases of large transitional lipomas or the type of terminal lipoma not having an intervening filum terminale. Direct stimulation proceeding from the obviously normal portion of the conus in a caudal direction will identify the lowest extent of pudendal motor neurons, beyond which sphincter contractions can no longer be elicited by the nerve stimulator (Figure 11).

Figure 11. Progressively caudal stimulation of the extremely stretched-out conus of a 48-year-old patient with adult tethered cord syndrome. Stimulation on the obviously normal portion of the conus generated tall spike waves (A), stimulation at the junctional zone between the conus and the filum produced smaller waves with a wide base (B), and stimulation just beyond the caudal extent of the conus elicited a minimal pressure response (C). Arrows indicate stimulus artifacts. Scale in torr.

Pelvic Floor EMG

Needle recording electrodes can be percutaneously inserted into the "extrinsic" portion of the external urethral sphincter to monitor activity of this sphincter. This technique is routinely used by neurourologists to correlate simultaneous measurements of bladder pressure, urethral pressure, and external urethral sphincter activities. Pelvic floor EMG can thus be used for intraoperative sacral root identification in the same manner as external anal sphincter EMG.

Modality for Sacral Reflex Monitoring

Two reflexes with centers in the sacral cord can be utilized to assess the integrity of both the sensory and motor roots as well as their interconnecting intramedullary components.

Bulbocavernosus Reflex

The reflex response of the bulbocavernosus muscle to stimulation of penile nerves can be studied using square wave electrical stimuli applied through ring electrodes on the penis, and recorded either by needle electrodes in the muscle or by surface electrodes fixed to the midline of the perineum, between the base of the penis and the anus. The averaged response from 50 to 100 stimuli is usually biphasic with an initial negative peak. The latency for most healthy adults is 24 to 42 msec but varies with age and maturation in young children. The waveform is also distorted significantly in most cases of myelodysplasia and tends to become "unstable" with very minor manipulations of the conus. The use of this monitoring modality is therefore limited and is feasible only in patients with virtually normal sphincter function preoperatively.

Urethral to Anal Sphincter Reflex Response

The urethral to anal sphincter reflex can be measured using stimulating electrodes similar to those used in eliciting urethral cortical evoked responses and recording electrodes used in recording external anal sphincter EMG. The latency is considerably longer (50 to 70 msec) than the bulbocavernosus reflex, although their morphologies are similar. The long latency in the urethral-anal sphincter reflex is due partly to the slower conducting velocity of autonomic afferent fibers and partly to a more complex central polysynaptic reflex organization.