- Email: [email protected]
Somatic innervation of the human bulbocavernosus muscle
Clinical Neurophysiology 110 (1999) 412±418
Somatic innervation of the human bulbocavernosus muscle Claire C. Yang a,*, William E. Bradley a, b a
b
University of Washington, Department of Urology, Box 356510 Seattle, WA 98195 USA University of Washington, Department of Neurology, Box 356510 Seattle, WA 98195 USA Accepted 18 August 1998
Abstract Objective: To demonstrate the somatic re¯ex innervation of the bulbocavernosus muscle (BCM), the principal muscle for ejaculation. Methods: Genitourinary electrodiagnostic testing utilizing modi®cations of the standard bulbocavernosus re¯ex was performed in 13 healthy male volunteers ages 20±43. Results: Bulbocavernosus muscle contraction was elicited by stimulation of the dorsal nerve of the penis, from both the penile skin and from the anterior urethra, and following stimulation of the perineal nerve. Latencies were variable depending on the point of stimulation. Conclusions: All 3 afferent pathways synapse on pudendal motoneurons in the conus medullaris, and provide for peripheral re¯ex control of BCM contractions. Based on the latencies of the urethral evoked responses, urethral innervation differs from penile shaft innervation, each having a distinct population of the dorsal nerve of the penis (DNP) ®bers. The presence of an electrically-de®ned pathway from the anterior urethra to the BCM suggests that somatic afferents from the anterior urethra are involved with the ejaculatory re¯ex. These somatic re¯exes are components of normal ejaculatory function. The ®ndings contribute to understanding the neurophysiology of ejaculation, and may be applicable to the evaluation of ejaculatory disorders. q 1999 Elsevier Science Ireland Ltd. All rights reserved. Keywords: Bulbocavernosus muscle; Dorsal nerve of the penis; Perineal nerve; Bulbocavernosus re¯ex
1. Introduction The bulbocavernosus muscle (BCM) is a paired, striated muscle surrounding the bulbus urethra, arising from the median raphe on the inferior surface of the bulb of the corpus spongiosum, and inserting on the fascia between the corpus spongiosum and the corpora cavernosa. During the initial phase of ejaculation, known as emission, secretions from the periurethral glands, seminal vesicles and prostate, along with sperm from the vas, are deposited into the posterior urethra and drain into the anterior urethra. Clonic contractions of the bulbocavernosus and other perineal muscles expel semen from the urethra. The BCM is innervated by the perineal nerve, a division of the pudendal nerve (Fig. 1). This nerve containing afferent and efferent axons originates in a bundle coursing posterior to anterior towards the lateral aspect of the bulbocavernosus muscle, and sends branches across its surface. The perineal efferent innervation to the muscle is supplied from motoneurons in the pudendal nucleus located in the
* Corresponding author. Tel.: 1 1-206-543-3640; fax: 1 1-206-5433272.
sacral segments of the conus medullaris. In addition to the sensory axons of the perineal nerve, these motoneurons also receive synaptic input from two separate populations of axons, both from the dorsal nerve of the penis (DNP) (Yang and Bradley, 1998b). One set of axons is located on the dorsal mid-line of the penis, innervating the penile shaft and glans, and the other set of axons is comprised of the ventrolateral branches of the DNP innervating the anterior urethra. These pathways are arranged in re¯ex circuits controlling BCM contraction and ejaculation. This study was undertaken to de®ne the re¯ex innervation of the BCM by these axon populations.
2. Methods and materials Thirteen healthy, sexually potent males, ages 20±43 with no history or physical ®ndings of neurologic or urologic disease were enrolled in the study after approval from our hospital's human subjects committee and full informed consent was obtained. To facilitate the recording of penile nerve responses, a pharmacologic erection was induced by injecting the corpus cavernosum with PGE1 0.3±0.5 ml (Yang et al., 1997).
1388-2457/99/$ - see front matter q 1999 Elsevier Science Ireland Ltd. All rights reserved. PII: S13 88-2457(98)0002 5-X
CLINPH 98034
C.C. Yang, W.E. Bradley / Clinical Neurophysiology 110 (1999) 412±418
413
line at the base of the penis (Yang and Bradley, 1998b). The DNP branched exuberantly within the glans into discrete, smaller ®bers. Ventrolaterally radiating branches of the DNP coursed towards the anterior urethra. Based on these data, stimulation was delivered to the base of the penis, the glans, and the anterior urethra, as these sites represented different points of distribution of the branches of the DNP. 2.1. BCM contraction in response to DNP (dorsal branches) stimulation
Fig. 1. The bulbocavernosus muscle and its innervation by the perineal nerve. The muscle occurs bilaterally and attaches along the median raphe of the bulbus urethra. The nerve trunk originates dorsolateral to the muscle and branches across the ventral surface of the muscle.
Attempts at performing electrodiagnostic testing without a pharmacologic erection resulted in inconsistent results due to dif®culty in maintaining the catheter position within the urethra. After an adequate erection occurred, electrodiagnostic testing was performed with the subject in the supine position. Penile length was measured from the base of the penis to the corona. Previous human anatomic studies demonstrated the DNP to be a large, single trunk on either side of the dorsal mid-
Four disk electrodes were af®xed after skin preparation: two to the dorsal mid-line of the penis at the base, with the cathode 1 cm proximal to the anode; and two to the glans, along the dorsal mid-line with the cathode proximally. A 27 ga concentric needle electrode was placed into the bulbocavernosus muscle at the perineal-scrotal junction. BCM contractions were evoked by orthodromic stimulation of the DNP at the penile base and glans. A ground patch was applied to the inner thigh. Cutaneous stimulus parameters were set at intensity 20±25 mA, duration 0.1 ms, frequency 0.5 Hz, and bandpass ®lter at 5±1500 Hz. Glanular stimulation typically was 5±15 mA higher than at the penile base to achieve supramaximal stimulation. Twenty responses were recorded and the test repeated to ensure reproducibility. 2.2. BCM contraction in response to DNP (ventrolateral branches) stimulation The anatomic studies noted previously demonstrated DNP ®bers innervating the corpus spongiosum and urethra.
Fig. 2. BCM EMG responses in a single subject following stimulation of the DNP at the (a) penile base and (b) glans penis. The characteristic response has an initial short duration component, followed by a broad depolarizing wave. Figures represent 20 averaged responses.
414
C.C. Yang, W.E. Bradley / Clinical Neurophysiology 110 (1999) 412±418
Table 1 BCM response latencies Response elicited from:
Mean latency (ms) (range)
No. of subjects responding
Penile base (cutaneous stimulation) Glans penis (cutaneous stimulation) Glanular urethra Penile urethra Bulbus urethra Perineal nerve
32.2 (27.9±38.4)
13
45.6 (35.6±65.4)
13
52.8 (37.7±81.4) 65.8 (53.1±81.7) 67.3 (57.8±81.5) 52.3 (30.7±67.4)
10 10 7 8
To stimulate these nerve endings, a 14 Fr urinary catheter with 4 ring electrodes mounted on the distal aspect was used (Bradley, 1972). The electrodes were separated from each other by 1 cm, with an electrode at the tip of the catheter. After the glans penis was prepped with povidone-iodine solution, the catheter tip was advanced into the bulbus urethra, using a small amount of water soluble lubricant to aid in insertion. Urethral stimulation of the DNP was performed through the distal-most electrodes on the catheter, with the cathode at the tip. Stimulation was applied to the bulbus urethra and subsequently at select positions within the urethra using the same stimulus parameters as for cutaneous stimulation, except for higher stimulus intensities (30±40 mA). BCM activity was recorded with the 27 ga bipolar electrode. At all stimulation points, 20 responses were recorded and repeated. 2.3. BCM contraction in response to perineal nerve stimulation To study the innervation of the BCM by the perineal
nerve, two monopolar needle electrodes were inserted: one as an anode into the contralateral BCM, and the second as a cathode 10 mm lateral to the anode, in an imaginary line connecting the electrodes to the ischial tuberosity. This alignment matched the course of the perineal nerve trunk. The thigh electrode was placed to ground. Stimulation was applied through the monopolar electrodes and recorded through the concentric electrode in the ipsilateral BCM. Stimulus parameters were the same as for cutaneous stimulation, except for stimulus intensities, which ranged from 3 to 10 mA. Stimulating and recording electrodes were then reversed to the opposite sides, so that the re¯ex pathway was demonstrated bilaterally.
3. Results 3.1. DNP (dorsal branches) to BCM pathway BCM EMG activity was recorded following stimulation of the penile base and glans in all of the subjects. In most subjects, the muscle was visibly contracting during the testing procedure. The BCM EMG recordings resembled the response ®rst described by Dick et al., with an initial early response, followed by a broad depolarizing wave (Dick et al., 1974) (Fig. 2). Responses between subjects were variable, however, and needle placement required close attention. With progressively higher stimulation rates, the contractile wave was attenuated, while the triphasic wave was retained. Mean onset latencies are recorded in Table 1. The shortest latency responses were obtained by stimulation of the main trunks of the DNP at the penile base, while those from the glans were of longer latency.
Fig. 3. BCM EMG responses following stimulation of the DNP from within the anterior urethra in a single subject. Point A 2 cm from urethral meatus (glanular urethra), point B 10 cm from urethral meatus (penile urethra), point C 14 cm from urethral meatus (bulbus urethra). Latencies of the responses are longer than those obtained with cutaneous stimulation. Each ®gure represents 20 averaged responses. Amplitude: A, C 20 mV/div., B 40 mV/div. (Figures redrawn from original data.)
C.C. Yang, W.E. Bradley / Clinical Neurophysiology 110 (1999) 412±418
Fig. 4. Left BCM response following stimulation of the right perineal nerve.
3.2. DNP (ventrolateral branches) to BCM pathway The BCM EMG responses varied depending on the point
415
of stimulation. Bulbus urethra responses were less consistently evoked compared to the responses evoked from the penile and glanular urethra, which occurred with more regularity (Fig. 3). Urethral responses, particularly from the bulb, had the characteristic two-phase appearance less uniformly than with cutaneous stimulation, with some EMG recordings having only a single monophasic negative wave, or generalized EMG spike activity. Latencies of the urethral responses were longer than the cutaneous responses, with a broad range for each subject. There was a decrease in the mean latencies along the length of the urethra, with the longest latencies measured from the bulbus urethra (Table 1). There was inter-subject variability in the amplitudes of all the responses, most likely a result of electrode placement within the muscle. In addition, there was some degree of needle movement within the muscle with repeated contractions, resulting in intra-subject variability as well. It was dif®cult to ®x needle position in the perineum during multiple examinations. For this reason, amplitude measurements were of secondary importance in this study. 3.3. Perineal nerve innervation of the BCM EMG responses were recorded in the BCM following
Fig. 5. Diagrammatic representation of the somatic re¯exes of the bulbocavernosus muscle. (A) DNP (dorsal branches) innervating the shaft and glans transiting centrally to synapse on motoneurons in the pudendal nucleus and exiting via the perineal nerve to terminate on the muscle ®bers of the BCM. (B) DNP (ventrolateral branches) innervating the anterior urethra, coursing centrally to synapse on motoneurons in the pudendal nucleus of the conus medullaris, and exiting through the perineal nerve to the BCM. (C) Afferent and efferent axons in the perineal nerve innervating the bulbocavernosus muscle. The intramedullary connections are most likely oligosynaptic, although the number of synapses is unknown.
416
C.C. Yang, W.E. Bradley / Clinical Neurophysiology 110 (1999) 412±418
stimulation of the perineal nerve in 8 subjects (Fig. 4). The wave forms were similar to those obtained from the DNP evoked responses. Response latencies were measured at 30± 68 ms. This range of latencies was due to variability in stimulus locus and recording sites. 4. Discussion The bulbocavernosus muscle is a striated muscle located in the perineum, wrapped around the bulb of the urethra. Its function is to propel seminal ¯uid from the urethra during ejaculation. The original EMG studies of the BCM established this muscle as the one primarily responsible for ejaculation (Kollberg et al., 1962). Re¯ex BCM contraction can be elicited by stimulation of the dorsal nerve of the penis, either by squeezing the glans penis (Lapides and Bobbitt, 1956; Bors and Blinn, 1959), or through electrostimulation (Dick et al., 1974; Ertekin and Reel, 1976; Sarica and Karacan, 1987). In the clinical examination, re¯ex BCM contraction provides a general indication of anatomic sacral nerve pathway integrity. A more precise measurement of the bulbocavernosus re¯ex (BCR) was described by Dick et al., who measured the latency of the BCM contraction following a single stimulus to the DNP along the penile shaft (Dick et al., 1974). This technique proved to be more capable of identifying nerve pathology than the clinical examination, and has been utilized widely in the evaluation of neurologic sexual and bladder dysfunction (Ertekin and Reel, 1976; Sarica and Karacan, 1987; Porst et al., 1988). Elicitation of the re¯ex following urethral stimulation has not been elaborated upon since the initial studies (Dick et al., 1974; Yang and Bradley, 1998a). Based on these previous reports, this study was conducted to demonstrate through electrodiagnostic techniques, that re¯ex innervation of the BCM consists of 3 pathways providing for clonic contractions of the bulbocavernosus muscle during ejaculation (Fig. 5): (A) sensory axons of the dorsal division of the DNP ± conus medullaris ± perineal nerve ± BCM. This has been a commonly studied re¯ex arc in genitourinary electrodiagnostic studies. (B) Ventrolaterally radiating ®bers of the DNP ± conus medullaris ± perineal nerve ± BCM. This pathway has only been recently de®ned (Yang and Bradley, 1998a) and is further explored in this investigation. (C) Perineal nerve ± conus medullaris ± perineal nerve ± BCM. This pathway comprises the usual segmental innervation found in skeletal muscles, but had not been previously de®ned in the BCM using electrodiagnostic techniques. The most important re¯ex is that of the dorsal DNP ®bers innervating the shaft and glans penis to the perineal nerve innervating the BCM. The afferent impulses from these areas of the penis are critical for ejaculation (Hart, 1972; Larsson and SoÈdersten, 1973). However, the presence of an electrophysiologically-
demonstrated pathway from the urethra to the BCM suggests that sensory impulses from the urethra also have a role in ejaculatory function. Speci®cally, these afferent impulses may serve to facilitate the BCM contractile response. The stimulus of seminal ¯uid in the urethra, with afferent impulses mediated through the DNP, provides positive feedback to the pudendal nuclei in the sacral spinal cord during ejaculation. This sensory input acts to initiate and maintain BCM contractions until the urethra is emptied of semen. These contractions cannot be elicited by passage of urine through the urethra of an unaroused, ¯accid penis. Urine in the anterior urethra is not analogous to the passage of semen following sexual arousal and thus would not result in bulbocavernosus contraction. With sexual arousal, the presence of circulating neurohormonal agents modi®es synaptic transmission, resulting in coordinated clonic contractions of the bulbocavernosus muscles. The sensation of orgasm may be mediated through the same pudendal ®bers (Bergman et al., 1979). The BCR was evoked from all segments of the anterior urethra (bulbus, penile, glanular), as well as from the penile skin, and the responses varied depending on the point of stimulation. The response from the DNP at the penile base was rapid, consistent and had a characteristic morphology (Fig. 2). At the penile base, the DNP is a large, single trunk on either side of the mid-line with minimal branching, so stimulation applied at this point would be most likely to result in a clear response. The response at the glans was of longer latency, but still displayed the same wave pattern. The increase in latency can be attributed to the increase in nerve length from the penile base to the glans, and that there are smaller numbers of DNP ®bers being stimulated at the glans. In addition, the DNP within the glans is arborized into smaller branches, and may have a higher depolarization threshold. In all subjects, stimulation from within the urethra resulted in substantially longer latencies when compared with the cutaneous responses. There was a decrease in latencies from the bulbus to glanular urethra. If the DNP innervates the urethra from branches radiating ventrolaterally from the mid-line, then evoked responses from the distal urethra should have longer latencies than from the bulbus urethra. However, this is not apparent. We postulate that DNP axons innervating the bulbus urethra are small diameter axons with a high depolarization threshold and a resultant long latency response, and the axons innervating the distal urethra are larger diameter, faster conducting nerves. As previously noted, latency differences may be attributed to different numbers of depolarized ®bers at each stimulation point. Receptor characteristics may also be contributing to the variable responses from the urethra. These factors are currently under neuroanatomic investigation. Alternatively, the long latency responses to urethral stimulation may be due to excitation of a nerve other than the DNP, such as the hypogastric nerve. There is no
C.C. Yang, W.E. Bradley / Clinical Neurophysiology 110 (1999) 412±418
anatomic evidence, however, that hypogastric innervation extends into the anterior urethra. In the cat, the anterior urethra is innervated by the pudendal nerve and the posterior urethra by the hypogastric and pelvic nerves (Bradley et al., 1973); this is presumed to be true for humans. Another possibility is that the perineal nerve, a branch of the pudendal nerve shown to innervate the ventral aspect of the penile shaft (Kaneko and Bradley, 1987), may be the source of the evoked responses. Human urethral innervation has not been fully de®ned by anatomic and electrodiagnostic methods. The BCM EMG recordings resembled the response described by Dick et al. (1974), with an initial traveling wave indicating the arrival of an axonal response in the nerve innervating the BCM, followed by a broad depolarizing wave, most likely signifying motion from nearby muscle contraction (Fig. 2). This response was consistent with what would be anticipated if the recording needle was in close proximity to a branch of the perineal nerve. Another explanation may be that the initial portion of the response represents the activity of motor units in proximity to the concentric needle. With progressively higher stimulation rates, the contractile wave was attenuated while the early component was retained, demonstrating the minimal habituation of this portion of the bulbocavernosus re¯ex (Vodusek and Janko, 1990). Further wave form analysis will require neuroanatomic dissections to de®ne the perineal nerve innervation of the BCM, particularly to identify the variability of muscle innervation. As a result of the relative proximity of the populations of axons being stimulated, every attempt was made in this study to optimize nerve selectivity. This was done in the following ways: (1) as previously noted, neuroanatomic dissections demonstrated DNP ®bers radiating around the penile shaft to innervate the corpus spongiosum and urethra, which are distinct from DNP ®bers innervating the glans (Yang and Bradley, 1998b). This ®nding established the presence of two separate populations of axons in the DNP. (2) Low amplitude, antidromic stimulation of the DNP at the penile base resulted in standing potentials along the length of the anterior urethra (Yang and Bradley, 1998a), indicating that selective recording and thus, stimulation, of DNP ®bers to the urethra was possible. (3) A pharmacologic erection was induced to enlarge the penis and thereby stretch the dorsal nerve of the penis (DNP), thereby maximizing the separation between the axons of the glans and the axons of the urethra. (4) Stimulating electrodes were placed in close proximity to each other to con®ne the stimulus to as small an area as possible. This was particularly pertinent in the perineum. Electrical ®elds created by cutaneous surface electrodes may not selectively stimulate the perineal nerve, since the ventrolateral branches of the DNP are in very close proximity; therefore, needle electrodes were used in this study. Needle placement was based on the knowledge of the neuroanatomy of the perineal region, gained from the author's experience in perineal operations. With this infor-
417
mation, it was reasonable to consider that the placement of stimulating electrodes as described in this study was adequate to selectively stimulate different populations of axons, even at the higher stimulus intensities. However, the anatomic course and variability of the perineal nerve, and its innervation of the BCM, have not been well-documented by dissection in the human. The nerve trunk runs deep to the skin, and the needles were better able to access this nerve. Further clari®cation of the anatomy of this re¯ex pathway will require direct visualization of the nerve and muscle, such as in an operating room setting. Electrodiagnostic measurement of the afferent and efferent innervation of the bulbocavernosus muscle via the perineal nerve, the only nerve to directly innervate the muscle, has not been previously described. Demonstration of this pathway completes the somatic re¯exes known to result in BCM contraction. How these 3 re¯exes interact in human ejaculatory function remains to be investigated; they are relevant to the evaluation of ejaculatory disorders, such as premature ejaculation, retrograde ejaculation and anejaculation. Additionally, de®ning 3 distinct peripheral re¯exes controlling ejaculation suggests the possibility of different central nervous system pathways, with conus medullaris control of ejaculation more complex than originally believed. Acknowledgements The authors gratefully acknowledge Alisa Malloch, RN for her assistance in this study. References Bergman B, Nilsson S, Petersen I. The effect on erection and orgasm of cystectomy, prostatectomy and vesiculectomy for cancer of the bladder: a clinical and electromyographic study. Br J Urol 1979;51:114±120. Bors E, Blinn KA. Bulbocavernosus re¯ex. J Urol 1959;82:128±130. Bradley W, Grif®n D, Teague C, Timm G. Sensory innervation of the mammalian urethra. Invest Urol 1973;10:287±289. Bradley WE. Urethral electromyelography. J Urol 1972;108:563±564. Dick HC, Bradley WE, Scott FB, Timm GW. Pudendal sexual re¯exes. Urology 1974;3:376±379. Ertekin C, Reel F. Bulbocavernosus re¯ex in normal men and in patients with neurogenic bladder and/or impotence. J Neuro Sci, 1976;28:1±15. Hart BL. Sexual re¯exes in the male rat after anesthetization of the glans penis. Behav Biol 1972;7:127±130. Kaneko S, Bradley WE. Penile electrodiagnosis: penile peripheral innervation. Urology 1987;30:210±212. Kollberg S, Petersen I, Stener I. Preliminary results of an electromyographic study of ejaculation. Acta Chir Scand 1962;123:478±482. Lapides J, Bobbitt JM. Diagnostic value of bulbocavernous re¯ex. J Am Med Assoc 1956;162:971±972. Larsson K, SoÈdersten P. Mating in male rats after section of the dorsal penile nerve. Physiol Behav 1973;10:567±571. Porst H, Tackmann W, van Ahlen H. Neurophysiological investigations in potent and impotent men. Br J Urol 1988;61:445±450. Sarica Y, Karacan I. Bulbocavernosus re¯ex to somatic and visceral nerve stimulation in normal subjects and in diabetics with erectile impotence. J Urol 1987;138:55±58.
418
C.C. Yang, W.E. Bradley / Clinical Neurophysiology 110 (1999) 412±418
Vodusek DB, Janko M. The bulbocavernosus re¯ex. A single motor neuron study. Brain 1990;113:813±820. Yang CC, Bradley WE, Berger RE. The effect of pharmacologic erection on the dorsal nerve of the penis. Muscle Nerve 1997;20:1439±1444.
Yang CC, Bradley WE. Innervation of the human anterior urethra by the dorsal nerve of the penis. Muscle Nerve 1998;21:514±518. Yang CC, Bradley WE. Peripheral distribution of the human dorsal nerve of the penis. J Urol 1998;159:1912±1917.
Somatic innervation of the human bulbocavernosus muscle Claire C. Yang a,*, William E. Bradley a, b a
b
University of Washington, Department of Urology, Box 356510 Seattle, WA 98195 USA University of Washington, Department of Neurology, Box 356510 Seattle, WA 98195 USA Accepted 18 August 1998
Abstract Objective: To demonstrate the somatic re¯ex innervation of the bulbocavernosus muscle (BCM), the principal muscle for ejaculation. Methods: Genitourinary electrodiagnostic testing utilizing modi®cations of the standard bulbocavernosus re¯ex was performed in 13 healthy male volunteers ages 20±43. Results: Bulbocavernosus muscle contraction was elicited by stimulation of the dorsal nerve of the penis, from both the penile skin and from the anterior urethra, and following stimulation of the perineal nerve. Latencies were variable depending on the point of stimulation. Conclusions: All 3 afferent pathways synapse on pudendal motoneurons in the conus medullaris, and provide for peripheral re¯ex control of BCM contractions. Based on the latencies of the urethral evoked responses, urethral innervation differs from penile shaft innervation, each having a distinct population of the dorsal nerve of the penis (DNP) ®bers. The presence of an electrically-de®ned pathway from the anterior urethra to the BCM suggests that somatic afferents from the anterior urethra are involved with the ejaculatory re¯ex. These somatic re¯exes are components of normal ejaculatory function. The ®ndings contribute to understanding the neurophysiology of ejaculation, and may be applicable to the evaluation of ejaculatory disorders. q 1999 Elsevier Science Ireland Ltd. All rights reserved. Keywords: Bulbocavernosus muscle; Dorsal nerve of the penis; Perineal nerve; Bulbocavernosus re¯ex
1. Introduction The bulbocavernosus muscle (BCM) is a paired, striated muscle surrounding the bulbus urethra, arising from the median raphe on the inferior surface of the bulb of the corpus spongiosum, and inserting on the fascia between the corpus spongiosum and the corpora cavernosa. During the initial phase of ejaculation, known as emission, secretions from the periurethral glands, seminal vesicles and prostate, along with sperm from the vas, are deposited into the posterior urethra and drain into the anterior urethra. Clonic contractions of the bulbocavernosus and other perineal muscles expel semen from the urethra. The BCM is innervated by the perineal nerve, a division of the pudendal nerve (Fig. 1). This nerve containing afferent and efferent axons originates in a bundle coursing posterior to anterior towards the lateral aspect of the bulbocavernosus muscle, and sends branches across its surface. The perineal efferent innervation to the muscle is supplied from motoneurons in the pudendal nucleus located in the
* Corresponding author. Tel.: 1 1-206-543-3640; fax: 1 1-206-5433272.
sacral segments of the conus medullaris. In addition to the sensory axons of the perineal nerve, these motoneurons also receive synaptic input from two separate populations of axons, both from the dorsal nerve of the penis (DNP) (Yang and Bradley, 1998b). One set of axons is located on the dorsal mid-line of the penis, innervating the penile shaft and glans, and the other set of axons is comprised of the ventrolateral branches of the DNP innervating the anterior urethra. These pathways are arranged in re¯ex circuits controlling BCM contraction and ejaculation. This study was undertaken to de®ne the re¯ex innervation of the BCM by these axon populations.
2. Methods and materials Thirteen healthy, sexually potent males, ages 20±43 with no history or physical ®ndings of neurologic or urologic disease were enrolled in the study after approval from our hospital's human subjects committee and full informed consent was obtained. To facilitate the recording of penile nerve responses, a pharmacologic erection was induced by injecting the corpus cavernosum with PGE1 0.3±0.5 ml (Yang et al., 1997).
1388-2457/99/$ - see front matter q 1999 Elsevier Science Ireland Ltd. All rights reserved. PII: S13 88-2457(98)0002 5-X
CLINPH 98034
C.C. Yang, W.E. Bradley / Clinical Neurophysiology 110 (1999) 412±418
413
line at the base of the penis (Yang and Bradley, 1998b). The DNP branched exuberantly within the glans into discrete, smaller ®bers. Ventrolaterally radiating branches of the DNP coursed towards the anterior urethra. Based on these data, stimulation was delivered to the base of the penis, the glans, and the anterior urethra, as these sites represented different points of distribution of the branches of the DNP. 2.1. BCM contraction in response to DNP (dorsal branches) stimulation
Fig. 1. The bulbocavernosus muscle and its innervation by the perineal nerve. The muscle occurs bilaterally and attaches along the median raphe of the bulbus urethra. The nerve trunk originates dorsolateral to the muscle and branches across the ventral surface of the muscle.
Attempts at performing electrodiagnostic testing without a pharmacologic erection resulted in inconsistent results due to dif®culty in maintaining the catheter position within the urethra. After an adequate erection occurred, electrodiagnostic testing was performed with the subject in the supine position. Penile length was measured from the base of the penis to the corona. Previous human anatomic studies demonstrated the DNP to be a large, single trunk on either side of the dorsal mid-
Four disk electrodes were af®xed after skin preparation: two to the dorsal mid-line of the penis at the base, with the cathode 1 cm proximal to the anode; and two to the glans, along the dorsal mid-line with the cathode proximally. A 27 ga concentric needle electrode was placed into the bulbocavernosus muscle at the perineal-scrotal junction. BCM contractions were evoked by orthodromic stimulation of the DNP at the penile base and glans. A ground patch was applied to the inner thigh. Cutaneous stimulus parameters were set at intensity 20±25 mA, duration 0.1 ms, frequency 0.5 Hz, and bandpass ®lter at 5±1500 Hz. Glanular stimulation typically was 5±15 mA higher than at the penile base to achieve supramaximal stimulation. Twenty responses were recorded and the test repeated to ensure reproducibility. 2.2. BCM contraction in response to DNP (ventrolateral branches) stimulation The anatomic studies noted previously demonstrated DNP ®bers innervating the corpus spongiosum and urethra.
Fig. 2. BCM EMG responses in a single subject following stimulation of the DNP at the (a) penile base and (b) glans penis. The characteristic response has an initial short duration component, followed by a broad depolarizing wave. Figures represent 20 averaged responses.
414
C.C. Yang, W.E. Bradley / Clinical Neurophysiology 110 (1999) 412±418
Table 1 BCM response latencies Response elicited from:
Mean latency (ms) (range)
No. of subjects responding
Penile base (cutaneous stimulation) Glans penis (cutaneous stimulation) Glanular urethra Penile urethra Bulbus urethra Perineal nerve
32.2 (27.9±38.4)
13
45.6 (35.6±65.4)
13
52.8 (37.7±81.4) 65.8 (53.1±81.7) 67.3 (57.8±81.5) 52.3 (30.7±67.4)
10 10 7 8
To stimulate these nerve endings, a 14 Fr urinary catheter with 4 ring electrodes mounted on the distal aspect was used (Bradley, 1972). The electrodes were separated from each other by 1 cm, with an electrode at the tip of the catheter. After the glans penis was prepped with povidone-iodine solution, the catheter tip was advanced into the bulbus urethra, using a small amount of water soluble lubricant to aid in insertion. Urethral stimulation of the DNP was performed through the distal-most electrodes on the catheter, with the cathode at the tip. Stimulation was applied to the bulbus urethra and subsequently at select positions within the urethra using the same stimulus parameters as for cutaneous stimulation, except for higher stimulus intensities (30±40 mA). BCM activity was recorded with the 27 ga bipolar electrode. At all stimulation points, 20 responses were recorded and repeated. 2.3. BCM contraction in response to perineal nerve stimulation To study the innervation of the BCM by the perineal
nerve, two monopolar needle electrodes were inserted: one as an anode into the contralateral BCM, and the second as a cathode 10 mm lateral to the anode, in an imaginary line connecting the electrodes to the ischial tuberosity. This alignment matched the course of the perineal nerve trunk. The thigh electrode was placed to ground. Stimulation was applied through the monopolar electrodes and recorded through the concentric electrode in the ipsilateral BCM. Stimulus parameters were the same as for cutaneous stimulation, except for stimulus intensities, which ranged from 3 to 10 mA. Stimulating and recording electrodes were then reversed to the opposite sides, so that the re¯ex pathway was demonstrated bilaterally.
3. Results 3.1. DNP (dorsal branches) to BCM pathway BCM EMG activity was recorded following stimulation of the penile base and glans in all of the subjects. In most subjects, the muscle was visibly contracting during the testing procedure. The BCM EMG recordings resembled the response ®rst described by Dick et al., with an initial early response, followed by a broad depolarizing wave (Dick et al., 1974) (Fig. 2). Responses between subjects were variable, however, and needle placement required close attention. With progressively higher stimulation rates, the contractile wave was attenuated, while the triphasic wave was retained. Mean onset latencies are recorded in Table 1. The shortest latency responses were obtained by stimulation of the main trunks of the DNP at the penile base, while those from the glans were of longer latency.
Fig. 3. BCM EMG responses following stimulation of the DNP from within the anterior urethra in a single subject. Point A 2 cm from urethral meatus (glanular urethra), point B 10 cm from urethral meatus (penile urethra), point C 14 cm from urethral meatus (bulbus urethra). Latencies of the responses are longer than those obtained with cutaneous stimulation. Each ®gure represents 20 averaged responses. Amplitude: A, C 20 mV/div., B 40 mV/div. (Figures redrawn from original data.)
C.C. Yang, W.E. Bradley / Clinical Neurophysiology 110 (1999) 412±418
Fig. 4. Left BCM response following stimulation of the right perineal nerve.
3.2. DNP (ventrolateral branches) to BCM pathway The BCM EMG responses varied depending on the point
415
of stimulation. Bulbus urethra responses were less consistently evoked compared to the responses evoked from the penile and glanular urethra, which occurred with more regularity (Fig. 3). Urethral responses, particularly from the bulb, had the characteristic two-phase appearance less uniformly than with cutaneous stimulation, with some EMG recordings having only a single monophasic negative wave, or generalized EMG spike activity. Latencies of the urethral responses were longer than the cutaneous responses, with a broad range for each subject. There was a decrease in the mean latencies along the length of the urethra, with the longest latencies measured from the bulbus urethra (Table 1). There was inter-subject variability in the amplitudes of all the responses, most likely a result of electrode placement within the muscle. In addition, there was some degree of needle movement within the muscle with repeated contractions, resulting in intra-subject variability as well. It was dif®cult to ®x needle position in the perineum during multiple examinations. For this reason, amplitude measurements were of secondary importance in this study. 3.3. Perineal nerve innervation of the BCM EMG responses were recorded in the BCM following
Fig. 5. Diagrammatic representation of the somatic re¯exes of the bulbocavernosus muscle. (A) DNP (dorsal branches) innervating the shaft and glans transiting centrally to synapse on motoneurons in the pudendal nucleus and exiting via the perineal nerve to terminate on the muscle ®bers of the BCM. (B) DNP (ventrolateral branches) innervating the anterior urethra, coursing centrally to synapse on motoneurons in the pudendal nucleus of the conus medullaris, and exiting through the perineal nerve to the BCM. (C) Afferent and efferent axons in the perineal nerve innervating the bulbocavernosus muscle. The intramedullary connections are most likely oligosynaptic, although the number of synapses is unknown.
416
C.C. Yang, W.E. Bradley / Clinical Neurophysiology 110 (1999) 412±418
stimulation of the perineal nerve in 8 subjects (Fig. 4). The wave forms were similar to those obtained from the DNP evoked responses. Response latencies were measured at 30± 68 ms. This range of latencies was due to variability in stimulus locus and recording sites. 4. Discussion The bulbocavernosus muscle is a striated muscle located in the perineum, wrapped around the bulb of the urethra. Its function is to propel seminal ¯uid from the urethra during ejaculation. The original EMG studies of the BCM established this muscle as the one primarily responsible for ejaculation (Kollberg et al., 1962). Re¯ex BCM contraction can be elicited by stimulation of the dorsal nerve of the penis, either by squeezing the glans penis (Lapides and Bobbitt, 1956; Bors and Blinn, 1959), or through electrostimulation (Dick et al., 1974; Ertekin and Reel, 1976; Sarica and Karacan, 1987). In the clinical examination, re¯ex BCM contraction provides a general indication of anatomic sacral nerve pathway integrity. A more precise measurement of the bulbocavernosus re¯ex (BCR) was described by Dick et al., who measured the latency of the BCM contraction following a single stimulus to the DNP along the penile shaft (Dick et al., 1974). This technique proved to be more capable of identifying nerve pathology than the clinical examination, and has been utilized widely in the evaluation of neurologic sexual and bladder dysfunction (Ertekin and Reel, 1976; Sarica and Karacan, 1987; Porst et al., 1988). Elicitation of the re¯ex following urethral stimulation has not been elaborated upon since the initial studies (Dick et al., 1974; Yang and Bradley, 1998a). Based on these previous reports, this study was conducted to demonstrate through electrodiagnostic techniques, that re¯ex innervation of the BCM consists of 3 pathways providing for clonic contractions of the bulbocavernosus muscle during ejaculation (Fig. 5): (A) sensory axons of the dorsal division of the DNP ± conus medullaris ± perineal nerve ± BCM. This has been a commonly studied re¯ex arc in genitourinary electrodiagnostic studies. (B) Ventrolaterally radiating ®bers of the DNP ± conus medullaris ± perineal nerve ± BCM. This pathway has only been recently de®ned (Yang and Bradley, 1998a) and is further explored in this investigation. (C) Perineal nerve ± conus medullaris ± perineal nerve ± BCM. This pathway comprises the usual segmental innervation found in skeletal muscles, but had not been previously de®ned in the BCM using electrodiagnostic techniques. The most important re¯ex is that of the dorsal DNP ®bers innervating the shaft and glans penis to the perineal nerve innervating the BCM. The afferent impulses from these areas of the penis are critical for ejaculation (Hart, 1972; Larsson and SoÈdersten, 1973). However, the presence of an electrophysiologically-
demonstrated pathway from the urethra to the BCM suggests that sensory impulses from the urethra also have a role in ejaculatory function. Speci®cally, these afferent impulses may serve to facilitate the BCM contractile response. The stimulus of seminal ¯uid in the urethra, with afferent impulses mediated through the DNP, provides positive feedback to the pudendal nuclei in the sacral spinal cord during ejaculation. This sensory input acts to initiate and maintain BCM contractions until the urethra is emptied of semen. These contractions cannot be elicited by passage of urine through the urethra of an unaroused, ¯accid penis. Urine in the anterior urethra is not analogous to the passage of semen following sexual arousal and thus would not result in bulbocavernosus contraction. With sexual arousal, the presence of circulating neurohormonal agents modi®es synaptic transmission, resulting in coordinated clonic contractions of the bulbocavernosus muscles. The sensation of orgasm may be mediated through the same pudendal ®bers (Bergman et al., 1979). The BCR was evoked from all segments of the anterior urethra (bulbus, penile, glanular), as well as from the penile skin, and the responses varied depending on the point of stimulation. The response from the DNP at the penile base was rapid, consistent and had a characteristic morphology (Fig. 2). At the penile base, the DNP is a large, single trunk on either side of the mid-line with minimal branching, so stimulation applied at this point would be most likely to result in a clear response. The response at the glans was of longer latency, but still displayed the same wave pattern. The increase in latency can be attributed to the increase in nerve length from the penile base to the glans, and that there are smaller numbers of DNP ®bers being stimulated at the glans. In addition, the DNP within the glans is arborized into smaller branches, and may have a higher depolarization threshold. In all subjects, stimulation from within the urethra resulted in substantially longer latencies when compared with the cutaneous responses. There was a decrease in latencies from the bulbus to glanular urethra. If the DNP innervates the urethra from branches radiating ventrolaterally from the mid-line, then evoked responses from the distal urethra should have longer latencies than from the bulbus urethra. However, this is not apparent. We postulate that DNP axons innervating the bulbus urethra are small diameter axons with a high depolarization threshold and a resultant long latency response, and the axons innervating the distal urethra are larger diameter, faster conducting nerves. As previously noted, latency differences may be attributed to different numbers of depolarized ®bers at each stimulation point. Receptor characteristics may also be contributing to the variable responses from the urethra. These factors are currently under neuroanatomic investigation. Alternatively, the long latency responses to urethral stimulation may be due to excitation of a nerve other than the DNP, such as the hypogastric nerve. There is no
C.C. Yang, W.E. Bradley / Clinical Neurophysiology 110 (1999) 412±418
anatomic evidence, however, that hypogastric innervation extends into the anterior urethra. In the cat, the anterior urethra is innervated by the pudendal nerve and the posterior urethra by the hypogastric and pelvic nerves (Bradley et al., 1973); this is presumed to be true for humans. Another possibility is that the perineal nerve, a branch of the pudendal nerve shown to innervate the ventral aspect of the penile shaft (Kaneko and Bradley, 1987), may be the source of the evoked responses. Human urethral innervation has not been fully de®ned by anatomic and electrodiagnostic methods. The BCM EMG recordings resembled the response described by Dick et al. (1974), with an initial traveling wave indicating the arrival of an axonal response in the nerve innervating the BCM, followed by a broad depolarizing wave, most likely signifying motion from nearby muscle contraction (Fig. 2). This response was consistent with what would be anticipated if the recording needle was in close proximity to a branch of the perineal nerve. Another explanation may be that the initial portion of the response represents the activity of motor units in proximity to the concentric needle. With progressively higher stimulation rates, the contractile wave was attenuated while the early component was retained, demonstrating the minimal habituation of this portion of the bulbocavernosus re¯ex (Vodusek and Janko, 1990). Further wave form analysis will require neuroanatomic dissections to de®ne the perineal nerve innervation of the BCM, particularly to identify the variability of muscle innervation. As a result of the relative proximity of the populations of axons being stimulated, every attempt was made in this study to optimize nerve selectivity. This was done in the following ways: (1) as previously noted, neuroanatomic dissections demonstrated DNP ®bers radiating around the penile shaft to innervate the corpus spongiosum and urethra, which are distinct from DNP ®bers innervating the glans (Yang and Bradley, 1998b). This ®nding established the presence of two separate populations of axons in the DNP. (2) Low amplitude, antidromic stimulation of the DNP at the penile base resulted in standing potentials along the length of the anterior urethra (Yang and Bradley, 1998a), indicating that selective recording and thus, stimulation, of DNP ®bers to the urethra was possible. (3) A pharmacologic erection was induced to enlarge the penis and thereby stretch the dorsal nerve of the penis (DNP), thereby maximizing the separation between the axons of the glans and the axons of the urethra. (4) Stimulating electrodes were placed in close proximity to each other to con®ne the stimulus to as small an area as possible. This was particularly pertinent in the perineum. Electrical ®elds created by cutaneous surface electrodes may not selectively stimulate the perineal nerve, since the ventrolateral branches of the DNP are in very close proximity; therefore, needle electrodes were used in this study. Needle placement was based on the knowledge of the neuroanatomy of the perineal region, gained from the author's experience in perineal operations. With this infor-
417
mation, it was reasonable to consider that the placement of stimulating electrodes as described in this study was adequate to selectively stimulate different populations of axons, even at the higher stimulus intensities. However, the anatomic course and variability of the perineal nerve, and its innervation of the BCM, have not been well-documented by dissection in the human. The nerve trunk runs deep to the skin, and the needles were better able to access this nerve. Further clari®cation of the anatomy of this re¯ex pathway will require direct visualization of the nerve and muscle, such as in an operating room setting. Electrodiagnostic measurement of the afferent and efferent innervation of the bulbocavernosus muscle via the perineal nerve, the only nerve to directly innervate the muscle, has not been previously described. Demonstration of this pathway completes the somatic re¯exes known to result in BCM contraction. How these 3 re¯exes interact in human ejaculatory function remains to be investigated; they are relevant to the evaluation of ejaculatory disorders, such as premature ejaculation, retrograde ejaculation and anejaculation. Additionally, de®ning 3 distinct peripheral re¯exes controlling ejaculation suggests the possibility of different central nervous system pathways, with conus medullaris control of ejaculation more complex than originally believed. Acknowledgements The authors gratefully acknowledge Alisa Malloch, RN for her assistance in this study. References Bergman B, Nilsson S, Petersen I. The effect on erection and orgasm of cystectomy, prostatectomy and vesiculectomy for cancer of the bladder: a clinical and electromyographic study. Br J Urol 1979;51:114±120. Bors E, Blinn KA. Bulbocavernosus re¯ex. J Urol 1959;82:128±130. Bradley W, Grif®n D, Teague C, Timm G. Sensory innervation of the mammalian urethra. Invest Urol 1973;10:287±289. Bradley WE. Urethral electromyelography. J Urol 1972;108:563±564. Dick HC, Bradley WE, Scott FB, Timm GW. Pudendal sexual re¯exes. Urology 1974;3:376±379. Ertekin C, Reel F. Bulbocavernosus re¯ex in normal men and in patients with neurogenic bladder and/or impotence. J Neuro Sci, 1976;28:1±15. Hart BL. Sexual re¯exes in the male rat after anesthetization of the glans penis. Behav Biol 1972;7:127±130. Kaneko S, Bradley WE. Penile electrodiagnosis: penile peripheral innervation. Urology 1987;30:210±212. Kollberg S, Petersen I, Stener I. Preliminary results of an electromyographic study of ejaculation. Acta Chir Scand 1962;123:478±482. Lapides J, Bobbitt JM. Diagnostic value of bulbocavernous re¯ex. J Am Med Assoc 1956;162:971±972. Larsson K, SoÈdersten P. Mating in male rats after section of the dorsal penile nerve. Physiol Behav 1973;10:567±571. Porst H, Tackmann W, van Ahlen H. Neurophysiological investigations in potent and impotent men. Br J Urol 1988;61:445±450. Sarica Y, Karacan I. Bulbocavernosus re¯ex to somatic and visceral nerve stimulation in normal subjects and in diabetics with erectile impotence. J Urol 1987;138:55±58.
418
C.C. Yang, W.E. Bradley / Clinical Neurophysiology 110 (1999) 412±418
Vodusek DB, Janko M. The bulbocavernosus re¯ex. A single motor neuron study. Brain 1990;113:813±820. Yang CC, Bradley WE, Berger RE. The effect of pharmacologic erection on the dorsal nerve of the penis. Muscle Nerve 1997;20:1439±1444.
Yang CC, Bradley WE. Innervation of the human anterior urethra by the dorsal nerve of the penis. Muscle Nerve 1998;21:514±518. Yang CC, Bradley WE. Peripheral distribution of the human dorsal nerve of the penis. J Urol 1998;159:1912±1917.