Arthroscopy: The Journal of Arthroscopic and Related Surgery 8(4):442-447 Published by Raven Press, Ltd. 0 1992Arthroscopy Association of North America
The Intraoperative Evaluation of the Neurosensory Function of the Anterior Cruciate Ligament in Humans Using Somatosensory Evoked Potentials Mark I. Pitman, M.D., Nahid Nainzadeh, M.D., David Menche, M.D., Richard Gasalberti, M.D., and Eun Kyoo Song, M.D.
Summary: Most of the investigation of the properties of the anterior cruciate ligament (ACL) has focused on its biomechanical functions. There has been increasing interest in the study of the possible neuroreceptor function of the ACL and its role in providing important proprioceptive feedback. Anatomic and histologic studies in humans have shown the presence of neuroreceptors within the anterior cruciate ligament. Indirect evidence exists that proprioception is diminished in the ACL-deficient knee. However, direct evidence in humans of the actual origin of the afferent impulses from within the ACL itself is lacking. Measurement of direct electrical afferent activity, occurring on stimulation of the ACL, should provide this evidence. Somatosensory evoked potentials (SEP) measure the electric potentials evoked in the cerebral cortex upon stimulation of a peripheral neuroreceptor. Carried by the posterior columns of the spinal cord, they reflect activity of the proprioceptive fibers. During arthroscopic procedures performed on nine patients, the normal ACL was stimulated by the use of electrodes applied to the femoral end, midsubstance, and tibia1 end, and cortical potentials thus evoked were recorded. In all cases, SEPs were recorded at the cerebral cortex upon stimulation of the ACL. The greatest potentials were reported upon stimulation of the midsubstance of the ligament. These findings provide direct evidence for, and strongly support the presence of, active proprioceptive receptors within the intact anterior cruciate ligament of the human knee. Key Words: Anterior cruciate ligamentNeuroreceptor function-Somatosensory evoked potentials-Proprioceptive feedback.
In the past,
of the many
of the anterior
cruciate
in the possible function of the anterior cruciate ligament as a neuroreceptor organ. There have been numerous anatomic studies attempting to determine the presence or absence of neuroreceptors in the ACL in humans. Kennedy et al. (1) noted the presence of free pain nerve endings and Golgi-like receptors in the synovial covering of the anterior cruciate ligament. Schultz et al. (2) found these receptors on the surface of the ligament beneath the synovial sheath, mostly at its femoral end. Schutte et al. (3) in a most detailed study observed the presence of an extensive network of sensory receptors in the ACL, mostly at the tibial end and least at the femoral end. These receptors con-
studies of the properties ligament (ACL), the vast
majority have focused on its biomechanical tion. There has, however, been increasing
funcinterest
From the Department of Sports Medicine (M.I.P., D.M., R.G., E.K.S.) and Department of Rehabilitation Medicine (N.N.), Hospital for Joint Diseases Orthopaedic Institute, New York, New York, U.S.A. Address correspondence and reprint requests to Mark I. Pitman, M.D., Department of Sports Medicine, Hospital for Joint Diseases Orthopaedic Institute, 301 East 17th Street, New York, New York IfKKl3, U.S.A. This article won the Richard O’Connor Award at the Annual Meeting of the Arthroscopy Association of North America in San Diego, April 1991, where it was presented.
442
EVALUATION
OF NEUROSENSORY
sisted of free nerve endings, and Ruffini and Pacinian corpuscles. The latter two respond to tension changes in the ligament and signify position, speed, and acceleration (4-6) of the joint and the limb. Many studies in animals have suggested that there is an electrical response to mechanical stimulation of the knee joint. It has been shown that raising intracapsular pressure or altering tension in joint ligaments and traction on the ligaments causes increased discharge of the medial articular nerve in cats. These sensory nerves respond to changes in position, movement, intracapsular pressure, and ligamentous tension (7-9). Schultz (2) suggested that a “physiological analysis of the stimuli that activate these receptors in human knee ligaments would contribute to an understanding of the dynamics of joint function.” In order to attempt to assess this function physiologically, we have used somatosensory evoked potentials. Somatosensory evoked potentials (SEPs) are defined as the cortical potentials evoked in response to mechanical or electrical stimulation of peripheral nerves of the lower or upper extremities (10). Mediated through large-diameter, fastconducting, myelinated (I-a) muscle afferent fibers and through group II cutaneous afferent fibers, through the posterior columns of the spinal cord to the cerebral cortex, SEPs monitor peripheral nerve and posterior column proprioception (11). SEP abnormalities are associated with disorders of joint positions, stereognosis, and vibration. The cortical evoked responses are complex waveforms representing the sensory impulse as it travels through the sensory pathways to the sensory cortex. The volley resulting from peripheral nerve stimulation has a variable degree of asynchrony reflecting a spectrum of peripheral conduction of the velocities, and this may be one of several factors accounting for the complexity of the SEP. Waves are labeled by polarity (P, positive; N, negative) and latency (expressed in ms). Latencies of interest for lower extremities are PI and Nl. The peripheral stimulus routinely used to elicit an SEP is usually electrical. An SEP may also be elicited through a variety of mechanical stimuli (12,13). SEPs are not associated with disorders of pain and temperature sensation. This finding has been confirmed in monkeys by surgically isolating or ablating the posterior columns and studying SEP changes (14). In humans with specific surgical lesions, SEPs correlate with cord lesions that impair position sense and awareness of passive joint mo-
FUNCTION
OF ACL BY SEP
tion. Ablation of the dorsal columns of the spinal cord abolishes SEPs. Thus, in laboratory animals and in humans, loss of dorsal column function is associated with abnormal SEPs. There has been no experimental or clinical evidence to suggest that SEPs travel in pathways other than the posterior column (15-17). SEPs, in addition to being used in the electrodiagnosis of neuromuscular disease (18,19), have been used in orthopedic surgery to monitor the status of nerve conduction during spinal cord surgery (20), hip arthroplasty (21), and shoulder arthroscopy (22). MATERIALS AND METHODS Nine patients with normal anterior cruciate ligaments were studied during arthroscopy of the knee for other conditions. All were anesthetized by use of nitrous oxide in oxygen, narcotics, and a muscle relaxant. Inhalation anesthetic agents, such as isofluorane , enfluorane , and halothane , were avoided because of their depressant effects upon cortical responses. No local anesthesia or epinephrine was used. The anterior cruciate ligament was electrically stimulated by a flush-tip monopolar electrode probe placed through the anteromedial portal. The probe is Teflon-insulated to its tip (Xomed-Treace, Jacksonville, FL, U.S.A.). The stimulus thus comes only from the extreme tip of the probe, is localized to the ACL, and does not spread through the synovial fluid or to other structures of the knee. The stimulus used was a square wave of 0.2 ms duration at an intensity of 8-10 mA and a rate of 2.9/s. The response was recorded over a bandwidth of 32-250 Hz for a duration of 200 ms. Two hundred epochs were averaged, and two responses were recorded for each trial and superimposed to ensure consistency. Cortical response was monitored by a platinum needle electrode inserted subcutaneously in the scalp at the CZ position of lo/20 (according to the international encephalography system) with FZ reference. The anterior cruciate ligament was stimulated at the tibia1 end, the midportion, and the femoral end (Fig. 1). In all nine cases, the posterior tibia1 nerve of the contralateral extremity was stimulated by a pair of disc electrodes-each 6 mm in diameter separated by a distance of 2 cm fitted in a plastic block. This pair of electrodes was placed behind the medial malleolus with the cathode proximal. A square curArthroscopy.
Vol. 8. No. 4, 1992
444
M. I. PITMAN
PI-528 Nl-598 P2-67 N2=79
1c
o-5
NZ
p
PI-
PI
568
rent of 18-20 mA of 0.2 ms in duration was delivered at a rate of 2.91s. Two hundred epochs were averaged twice. The two responses were superimposed to ensure consistency and also to confirm that the posterior tibia1 SEP was within normal limits; according to our laboratory’s normal subjects, the average posterior tibia1 SEP is 38 + 2 SD (range 3348 ms). The recording site was similar to that of the anterior cruciate ligament (Fig. 2). RESULTS In all nine patients, electrical stimulation of the anterior cruciate ligament produced measurable cerebral cortical evoked potentials. In all, the latency was shortest at the midportion of the anterior cruciate ligament. In seven it was longest at the femoral end and in two, at the tibia1 end (Tables 1 and 2). In all cases, the posterior tibia1 SEPs were within normal limits. DISCUSSION The current interest in the neuroreceptor function of the anterior cruciate ligament is not really new
~1-366 P2
NIPZNZ-
45.6 60 6 76 0
FIG. 2. Recording of the SEP of the contralateral posterior tibial nerve.
Arthroscopy,
Vol. 8, No. 4, 1992
P2
Nl’60.8 w-77
6
2 2
Nl=638 P2= 73 8 N2= 03 2
PI
ET AL.
FIG. 1. Recording of SEP of Case 1 upon stimulation of the tibial third of the ACL (A), the middle third of the ACL (B), and the femoral third of the ACL (C). The time-locked onset of the SEP is at Pl, and the latency is the time measured between the stimulus of the ACL and the cortical recording of the evoked potential. The amplitude is measured by the height of the spike between Pl and Nl.
but is actually a resurgence of interest in a function that was suggested many years ago. Ivar Palmar (23), in 1944, stated that “the function of certain ligaments is not only to restrict movement but to initiate muscular cooperation. . . . Inflow emanates from the ligament activating a muscular mechanism protecting the joint. . . . Ligaments are well supplied with nerve tissue. Actually, they constitute an exciting organ for the muscular defense reflexes of the joint. The tensing of a ligament activates a group of muscles through which it is connected functionally. It is through this mechanism, alone, that the bands are able to resist an often violent strain to which they are submitted through the course of a lifetime.” More recently, Kennedy et al. (1) also recognized this possibility and stated that the ligamentous structures “may also act to initiate a reflex which protects the joint by muscular splinting and in situations of abnormal stress.” It was their hypothesis that the ligaments provided a neurogenic contribution to knee stability and that ligament laxity caused a failure of mechanoreceptor feedback with a loss of reflex muscular splinting. Solomonow et al. (24) noted the presence of an apparent anterior cruciate ligament/hamstring reflex in cats. They mechanically applied tension to the ligament and observed electrical activity in the hamstring muscles of the animal. Pope et al. (25) have not been able to reproduce this effect. It is difficult to develop a model in the cat, and especially in the human, for accurate and significant mechanical stimulation. Mechanical loads are complex and may be affected by anesthesia or by the duration or type of load. Perhaps the lengthening of the
EVALUATION
OF NEUROSENSORY
ACL, as was done by Pope, and pulling it with a wire loop did not provide a correct and reproducible stimulus. There is as yet no model to determine which type of and how much mechanical stimulation is required to activate reproducibly the mechanoreceptors within the anterior cruciate ligament.
1. SSEP of the nine
TABLE
Latency ms Pl
Nl
P2
N2
Amplitude Pl Nl pV
38.6
45.6
60.0
76.8
1.0
tib mid fern
52.8 48.0 56.8
59.8 60.8 63.8
67.2 77.6 73.8
79.2 83.2
0.13 0.48 0.12
R
PT
44.8
58.0
76.0
85.6
0.5
L
ACL
tib mid fern
62.4 53.6 58.4
80.0 76.8 71.2
92.8 89.6 90.4
101.6 101.4 101.3
0.12 0.08 0.04
RPT L ACL tib mid fern
38.0
46.8
63.2
76.0
0.47
54.6 54.0 59.2
69.6 63.2 73.6
86.0 74.8 87.2
106.8 84.8 98.0
0.32 0.18 0.44
R
PT
38.0
45.2
53.6
62.6
0.19
L
ACL
tib mid fern
66.8 59.6 65.6
81.6 76.0 80.0
96.8 85.2 97.6
R
PT
39.6
50.4
60.0
79.2
0.68
L
ACL
tib mid fern
52.2 44.8 56.0
61.2 50.8 65.2
68.0 60.8 80.8
76.8 69.2 90.4
0.40 0.53 0.25
RPT L ACL tib mid fern
33.4
40.0
52.0
74.0
0.42
51.6 38.8 58.2
61.6 49.2 71.6
74.4 56.8 78.2
84.4 79.2 85.6
0.14 0.31 0.29
34.0
42.8
50.4
64.8
0.33
49.2 38.8 53.2
56.0 49.2 67.6
64.6 56.8 78.0
84.4 79.2 86.0
0.08 0.19 0.46
R PT L
ACL
RPT L ACL tib mid fern
0.43 0.23 0.24
RF-I L ACL tib mid fern
48.0
58.8
69.2
84.0
0.33
48.4 41.2 46.4
56.8 60.0 54.8
66.0 68.4 67.6
78.0 77.2 71.2
0.15 0.34 0.22
R
PT
43.6
61.6
75.2
95.6
0.27
L
ACL
46.9 35.6 48.8
54.4 44.8 62.8
64.0 58.0 76.0
89.2 66.0 92.0
0.27 0.31 0.05
tib mid fern
FUNCTION
OF ACL BY SEP
TABLE 2. The portion of Table I related to the PI
latency upon stimulation
of the nine subjects
Case
Tibia1
Midportion
Femoral
1 2 3 4 5 6 7 8 9
52.8 62.4 54.6 66.8 52.2 51.6 49.2 48.4 46.4
48.0 53.6 54.0 59.6 44.8 38.8 38.8 41.2 35.6
56.8 58.4 59.2 65.6 56.0 58.2 53.2 46.4 48.8
In addition, as Pope states, there may not be a direct ligament-muscular reflex. Indirectly, in humans, by the use of an isokinetic dynamometer, Solomonow et al. (24) also found that upon knee extension in ACL-deficient patients, there was a burst of hamstring EMG activity at about 45”, when subluxation usually occurs. Barrach et al. (26,27) showed that the threshold to the ability to determine passive knee motion and knee joint angulation was decreased in the ACL-deficient knee. Brand (28) and Wroble and Brand (29) also have questioned the purely mechanical view of ligament function. They speculated that although neurosensory deficits are difficult to diagnose and the mechanical effect has been of most concern to the clinician, neurosensory function may predominate during normal activity and there may be a proprioceptive defect that, if understood, could lead to improved therapy. There has recently been an argument for a sophisticated and more complex reflex effect on the 6 muscle spindle system that influences the primary muscle spindle afferents perhaps to regulate muscle stiffness around the knee joint and therefore, in a more complex manner, to regulate the stability of the knee (30). There is some question as to whether the stimulus we used stimulates the ACL alone and does not spread to other adjacent structures in the knee joint. This insulated probe was exposed only at its very tip, and the very small intensity of stimulus applied does not allow spread of the stimulus. No response to electrical stimulation could be obtained when the tip of the probe was placed either on articular cartilage or just in the saline itself without touching any intraarticular structures. It appears that the stimulus remains localized to the ACL. In our study, the latency from the anterior cruciArthroscopy,
Vol.
8, No. 4, 1992
M. I. PITMAN ET AL. ate ligament is longer than that from the posterior tibia1 nerve. One explanation for this difference is that fibers in the posterior tibia1 nerve contain very rapidly conducting muscle afferents, whereas the nerves supplied to the ACL consist of small fibers that conduct more slowly, both peripherally and centrally. Also, there is a much smaller amount of nerve fibers in the anterior cruciate ligament, and thus a smaller amount of fibers may be stimulated in the ACL than in the posterior tibial nerve. Our study provides direct evidence that stimulation of the ACL does produce reproducible cortical somatosensory evoked potentials. There are, of course, other structures within and about the knee, such as the synovium and the collateral ligaments, that also supply proprioceptive nerve endings and certainly provide sensory input. Although we have used electrical stimulus, the sensory stimulus in the active human knee is, of course, not electrical but mechanical. Further studies must be undertaken to attempt to define a reproducible technique to provide an accurate mechanical stimulus in the intact human ACL that will have an effect upon these sensory nerve endings. However, electrical stimulation is another tool and a further step in studying and determining the sensory function of the anterior cruciate ligament in humans.
CONCLUSION This study has proposed a method to study physiologically the sensory function of the anterior cruciate ligament. We have obtained reproducible SEPs upon electrical stimulation of the intact anterior cruciate ligament. This method provides direct evidence of the presence of a proprioceptive function of the anterior cruciate ligament and represents a means to study this function physiologically in humans. Determination of the significance of the neurosensory function of the anterior cruciate ligament may have important consequences on the prevention, the surgery, and especially the rehabilitation of injuries to this ligament.
REFERENCES 1. Kennedy JC, Alexander IJ, Hayes KC. Nerve supply of the human knee and its functional importance. Am J Sports Med 1982;10:329-35. 2. Schultz RA, Miller DC, Kerr CS, Micheli L. Mechanorecep-
Arthroscopy, Vol. 8, NO. 4, 1992
tom in human cruciate ligaments:
a histological
study. J
Bone Joint Surg 1984;66-A:1072-76.
3. Schutte MJ, Dabezies EJ, Zimny ML, Happel LT. Neural anatomy of the human anterior cruciate ligament. J Bone Joint Surg 1987;69-A:243-7. 4. Halata Z, Haus J. The ultrastructure sensory nerve endings in human anterior cruciate ligament. Anat Embryo1 1989; 179:415-21. 5. Grigg T, Hoffman AH. Ruflini mechanoreceptors in isolated joint capsule response: correlated with strain energy density. Somatosens Res 1984;2: 149-62. 6. Grigg P, Hoffman AH, Fogerty KE. Properties of GolgiMazzoni: afferents in cat knee joint capsule as revealed by mechanical studies of isolated joint capsule. J Neurophysio/ 1982;47:31-40. I. Andrew BL, Dodt E. The deployment of sensory nerve endings at the knee joint of the cat. Acta Physioi Stand 1953; 281287-96. 8. Grigg P. Mechanical factor influencing response of joint afferent neurons from cat knee. J Neurophysiol1975;38-147384. 9. Grigg P, Greenspan B. The response of primate joint afferent neurons to mechanical stimulation of knee joint. J Neurophysiot 1977;40:1-8. 10. Dawson GP. The cerebral response to electrical stimulation of the peripheral nerve in man. J Neurol Neurosurg Psychiatry 1947;10:134-40. 11. Nainzadeh N. Somatosensory evoked potentials: assessment and management in the lumbar spine surgery. In: Camins M, O’Leary P, eds. The lumbar spine. New York: Raven Press, 1987:275. 12. Pratt H, Starr A, Amrie RN, Politoshe D. Mechanically and electrically evoked somatosensory potentials in normal humans. Neurology 1979;29: 1236. 13. Cohen L, Starr A. Effects of vibration and muscular contraction and several types of somatosensory evoked potentials. Neurology 1985;35:691. 14. Cusick JF, Myklebust J, Larson SJ, Sances A. Spinal evoked potentials in the primate: neuralsubstrate. J Neurosurg 1978;49:551-7. 15. Chiappa KH. Evokedpotentials in clinical medicine. 2nd ed. New York: Raven Press, 1990:371-92. 16. Cracco R, Bodis-Wollner I. Evoked potentials. New York: Alan R. Liss, 1986:379-85. 17. Aminoff MJ. Efectrodiagnosis of clinical neuroiogy. 2nd ed. New York: Churchill Livingston, 1986:535-8. 18. Goodgold J, Eberstein A. Electrodiagnosis of neuromuscular disease. 3rd ed., Baltimore: Williams and Wilkins, 1983:282-305. 19. Kimura J. Electrodiagnosis in diseases of nerve and muscle. Philadelphia: F.A. Davis, 1983:399. 20. Bunch WH, Scharff TB, Trimbie J. Current concepts review: spinal cord monitoring. J Bone Joint Surg 1983;65: 707-10.
21. Stowe RG, Weeks LE, Hajdu M, Stinchtield FE. Evaluation of sciatic nerve compromise during total hip arthroplasty. Clin Orthop Rel Res 198.5;201:26. 22. Pitman MI, Nainzadeh N, Ergas E, Springer S. The use of somatosensory evoked potentials for detection of neuropraxia during shoulder arthroscopy. Arthroscopy 1988;4(4): 250-5.
23. Palmar Ivar. Plastic surgery of the ligaments of the knee. Acta Chir Stand
24 Solomonow
1944;91:37-48.
M, Baratta
R, Zhou H, Bose W, Beck C,
EVALUATION
OF NEUROSENSORY
D’Ambrosia R. The synergistic action of the anterior cruciate ligament and thigh muscles in maintaining joint stability. Am J Sports Med 1987;15:207-13. 25. Pope PE, Kelly JC, Brand RA. Physiologic loading of the anterior cruciate ligament does not activate quadriceps or hamstrings in the anesthetized cat. Am J Sports Med 1990; 18:595-9. 26. Barrach RL, Skinner HB, Cook SD. Proprioception of the knee joint: paradoxical effects of training. Am J Sports Med 1984;63: 175-80.
FUNCTION
OF ACL BY SEP
447
21. Barrach RL, Skinner HB, Buckley SL. Proprioception in the anterior cruciate deficient knee. Am J Sports Med 1989;17: l-6. 28. Brand RA. Knee ligaments: a new view. Truns ASME 1986; 108: 10610. 29. Wroble RR, Brand RA. Function of knee ligaments: an historical review of two perspectives. Iowa Orthop J 1988;8: 62-l.
30. Johannson H, Sjolander P, Sojka P. A sensory role for the cruciate ligaments. Chin Urthop Re/ Res 1991;268: 161-78.
Arthroscopy,
Vol. 8, No. 4, I992
The Intraoperative Evaluation of the Neurosensory Function of the Anterior Cruciate Ligament in Humans Using Somatosensory Evoked Potentials Mark I. Pitman, M.D., Nahid Nainzadeh, M.D., David Menche, M.D., Richard Gasalberti, M.D., and Eun Kyoo Song, M.D.
Summary: Most of the investigation of the properties of the anterior cruciate ligament (ACL) has focused on its biomechanical functions. There has been increasing interest in the study of the possible neuroreceptor function of the ACL and its role in providing important proprioceptive feedback. Anatomic and histologic studies in humans have shown the presence of neuroreceptors within the anterior cruciate ligament. Indirect evidence exists that proprioception is diminished in the ACL-deficient knee. However, direct evidence in humans of the actual origin of the afferent impulses from within the ACL itself is lacking. Measurement of direct electrical afferent activity, occurring on stimulation of the ACL, should provide this evidence. Somatosensory evoked potentials (SEP) measure the electric potentials evoked in the cerebral cortex upon stimulation of a peripheral neuroreceptor. Carried by the posterior columns of the spinal cord, they reflect activity of the proprioceptive fibers. During arthroscopic procedures performed on nine patients, the normal ACL was stimulated by the use of electrodes applied to the femoral end, midsubstance, and tibia1 end, and cortical potentials thus evoked were recorded. In all cases, SEPs were recorded at the cerebral cortex upon stimulation of the ACL. The greatest potentials were reported upon stimulation of the midsubstance of the ligament. These findings provide direct evidence for, and strongly support the presence of, active proprioceptive receptors within the intact anterior cruciate ligament of the human knee. Key Words: Anterior cruciate ligamentNeuroreceptor function-Somatosensory evoked potentials-Proprioceptive feedback.
In the past,
of the many
of the anterior
cruciate
in the possible function of the anterior cruciate ligament as a neuroreceptor organ. There have been numerous anatomic studies attempting to determine the presence or absence of neuroreceptors in the ACL in humans. Kennedy et al. (1) noted the presence of free pain nerve endings and Golgi-like receptors in the synovial covering of the anterior cruciate ligament. Schultz et al. (2) found these receptors on the surface of the ligament beneath the synovial sheath, mostly at its femoral end. Schutte et al. (3) in a most detailed study observed the presence of an extensive network of sensory receptors in the ACL, mostly at the tibial end and least at the femoral end. These receptors con-
studies of the properties ligament (ACL), the vast
majority have focused on its biomechanical tion. There has, however, been increasing
funcinterest
From the Department of Sports Medicine (M.I.P., D.M., R.G., E.K.S.) and Department of Rehabilitation Medicine (N.N.), Hospital for Joint Diseases Orthopaedic Institute, New York, New York, U.S.A. Address correspondence and reprint requests to Mark I. Pitman, M.D., Department of Sports Medicine, Hospital for Joint Diseases Orthopaedic Institute, 301 East 17th Street, New York, New York IfKKl3, U.S.A. This article won the Richard O’Connor Award at the Annual Meeting of the Arthroscopy Association of North America in San Diego, April 1991, where it was presented.
442
EVALUATION
OF NEUROSENSORY
sisted of free nerve endings, and Ruffini and Pacinian corpuscles. The latter two respond to tension changes in the ligament and signify position, speed, and acceleration (4-6) of the joint and the limb. Many studies in animals have suggested that there is an electrical response to mechanical stimulation of the knee joint. It has been shown that raising intracapsular pressure or altering tension in joint ligaments and traction on the ligaments causes increased discharge of the medial articular nerve in cats. These sensory nerves respond to changes in position, movement, intracapsular pressure, and ligamentous tension (7-9). Schultz (2) suggested that a “physiological analysis of the stimuli that activate these receptors in human knee ligaments would contribute to an understanding of the dynamics of joint function.” In order to attempt to assess this function physiologically, we have used somatosensory evoked potentials. Somatosensory evoked potentials (SEPs) are defined as the cortical potentials evoked in response to mechanical or electrical stimulation of peripheral nerves of the lower or upper extremities (10). Mediated through large-diameter, fastconducting, myelinated (I-a) muscle afferent fibers and through group II cutaneous afferent fibers, through the posterior columns of the spinal cord to the cerebral cortex, SEPs monitor peripheral nerve and posterior column proprioception (11). SEP abnormalities are associated with disorders of joint positions, stereognosis, and vibration. The cortical evoked responses are complex waveforms representing the sensory impulse as it travels through the sensory pathways to the sensory cortex. The volley resulting from peripheral nerve stimulation has a variable degree of asynchrony reflecting a spectrum of peripheral conduction of the velocities, and this may be one of several factors accounting for the complexity of the SEP. Waves are labeled by polarity (P, positive; N, negative) and latency (expressed in ms). Latencies of interest for lower extremities are PI and Nl. The peripheral stimulus routinely used to elicit an SEP is usually electrical. An SEP may also be elicited through a variety of mechanical stimuli (12,13). SEPs are not associated with disorders of pain and temperature sensation. This finding has been confirmed in monkeys by surgically isolating or ablating the posterior columns and studying SEP changes (14). In humans with specific surgical lesions, SEPs correlate with cord lesions that impair position sense and awareness of passive joint mo-
FUNCTION
OF ACL BY SEP
tion. Ablation of the dorsal columns of the spinal cord abolishes SEPs. Thus, in laboratory animals and in humans, loss of dorsal column function is associated with abnormal SEPs. There has been no experimental or clinical evidence to suggest that SEPs travel in pathways other than the posterior column (15-17). SEPs, in addition to being used in the electrodiagnosis of neuromuscular disease (18,19), have been used in orthopedic surgery to monitor the status of nerve conduction during spinal cord surgery (20), hip arthroplasty (21), and shoulder arthroscopy (22). MATERIALS AND METHODS Nine patients with normal anterior cruciate ligaments were studied during arthroscopy of the knee for other conditions. All were anesthetized by use of nitrous oxide in oxygen, narcotics, and a muscle relaxant. Inhalation anesthetic agents, such as isofluorane , enfluorane , and halothane , were avoided because of their depressant effects upon cortical responses. No local anesthesia or epinephrine was used. The anterior cruciate ligament was electrically stimulated by a flush-tip monopolar electrode probe placed through the anteromedial portal. The probe is Teflon-insulated to its tip (Xomed-Treace, Jacksonville, FL, U.S.A.). The stimulus thus comes only from the extreme tip of the probe, is localized to the ACL, and does not spread through the synovial fluid or to other structures of the knee. The stimulus used was a square wave of 0.2 ms duration at an intensity of 8-10 mA and a rate of 2.9/s. The response was recorded over a bandwidth of 32-250 Hz for a duration of 200 ms. Two hundred epochs were averaged, and two responses were recorded for each trial and superimposed to ensure consistency. Cortical response was monitored by a platinum needle electrode inserted subcutaneously in the scalp at the CZ position of lo/20 (according to the international encephalography system) with FZ reference. The anterior cruciate ligament was stimulated at the tibia1 end, the midportion, and the femoral end (Fig. 1). In all nine cases, the posterior tibia1 nerve of the contralateral extremity was stimulated by a pair of disc electrodes-each 6 mm in diameter separated by a distance of 2 cm fitted in a plastic block. This pair of electrodes was placed behind the medial malleolus with the cathode proximal. A square curArthroscopy.
Vol. 8. No. 4, 1992
444
M. I. PITMAN
PI-528 Nl-598 P2-67 N2=79
1c
o-5
NZ
p
PI-
PI
568
rent of 18-20 mA of 0.2 ms in duration was delivered at a rate of 2.91s. Two hundred epochs were averaged twice. The two responses were superimposed to ensure consistency and also to confirm that the posterior tibia1 SEP was within normal limits; according to our laboratory’s normal subjects, the average posterior tibia1 SEP is 38 + 2 SD (range 3348 ms). The recording site was similar to that of the anterior cruciate ligament (Fig. 2). RESULTS In all nine patients, electrical stimulation of the anterior cruciate ligament produced measurable cerebral cortical evoked potentials. In all, the latency was shortest at the midportion of the anterior cruciate ligament. In seven it was longest at the femoral end and in two, at the tibia1 end (Tables 1 and 2). In all cases, the posterior tibia1 SEPs were within normal limits. DISCUSSION The current interest in the neuroreceptor function of the anterior cruciate ligament is not really new
~1-366 P2
NIPZNZ-
45.6 60 6 76 0
FIG. 2. Recording of the SEP of the contralateral posterior tibial nerve.
Arthroscopy,
Vol. 8, No. 4, 1992
P2
Nl’60.8 w-77
6
2 2
Nl=638 P2= 73 8 N2= 03 2
PI
ET AL.
FIG. 1. Recording of SEP of Case 1 upon stimulation of the tibial third of the ACL (A), the middle third of the ACL (B), and the femoral third of the ACL (C). The time-locked onset of the SEP is at Pl, and the latency is the time measured between the stimulus of the ACL and the cortical recording of the evoked potential. The amplitude is measured by the height of the spike between Pl and Nl.
but is actually a resurgence of interest in a function that was suggested many years ago. Ivar Palmar (23), in 1944, stated that “the function of certain ligaments is not only to restrict movement but to initiate muscular cooperation. . . . Inflow emanates from the ligament activating a muscular mechanism protecting the joint. . . . Ligaments are well supplied with nerve tissue. Actually, they constitute an exciting organ for the muscular defense reflexes of the joint. The tensing of a ligament activates a group of muscles through which it is connected functionally. It is through this mechanism, alone, that the bands are able to resist an often violent strain to which they are submitted through the course of a lifetime.” More recently, Kennedy et al. (1) also recognized this possibility and stated that the ligamentous structures “may also act to initiate a reflex which protects the joint by muscular splinting and in situations of abnormal stress.” It was their hypothesis that the ligaments provided a neurogenic contribution to knee stability and that ligament laxity caused a failure of mechanoreceptor feedback with a loss of reflex muscular splinting. Solomonow et al. (24) noted the presence of an apparent anterior cruciate ligament/hamstring reflex in cats. They mechanically applied tension to the ligament and observed electrical activity in the hamstring muscles of the animal. Pope et al. (25) have not been able to reproduce this effect. It is difficult to develop a model in the cat, and especially in the human, for accurate and significant mechanical stimulation. Mechanical loads are complex and may be affected by anesthesia or by the duration or type of load. Perhaps the lengthening of the
EVALUATION
OF NEUROSENSORY
ACL, as was done by Pope, and pulling it with a wire loop did not provide a correct and reproducible stimulus. There is as yet no model to determine which type of and how much mechanical stimulation is required to activate reproducibly the mechanoreceptors within the anterior cruciate ligament.
1. SSEP of the nine
TABLE
Latency ms Pl
Nl
P2
N2
Amplitude Pl Nl pV
38.6
45.6
60.0
76.8
1.0
tib mid fern
52.8 48.0 56.8
59.8 60.8 63.8
67.2 77.6 73.8
79.2 83.2
0.13 0.48 0.12
R
PT
44.8
58.0
76.0
85.6
0.5
L
ACL
tib mid fern
62.4 53.6 58.4
80.0 76.8 71.2
92.8 89.6 90.4
101.6 101.4 101.3
0.12 0.08 0.04
RPT L ACL tib mid fern
38.0
46.8
63.2
76.0
0.47
54.6 54.0 59.2
69.6 63.2 73.6
86.0 74.8 87.2
106.8 84.8 98.0
0.32 0.18 0.44
R
PT
38.0
45.2
53.6
62.6
0.19
L
ACL
tib mid fern
66.8 59.6 65.6
81.6 76.0 80.0
96.8 85.2 97.6
R
PT
39.6
50.4
60.0
79.2
0.68
L
ACL
tib mid fern
52.2 44.8 56.0
61.2 50.8 65.2
68.0 60.8 80.8
76.8 69.2 90.4
0.40 0.53 0.25
RPT L ACL tib mid fern
33.4
40.0
52.0
74.0
0.42
51.6 38.8 58.2
61.6 49.2 71.6
74.4 56.8 78.2
84.4 79.2 85.6
0.14 0.31 0.29
34.0
42.8
50.4
64.8
0.33
49.2 38.8 53.2
56.0 49.2 67.6
64.6 56.8 78.0
84.4 79.2 86.0
0.08 0.19 0.46
R PT L
ACL
RPT L ACL tib mid fern
0.43 0.23 0.24
RF-I L ACL tib mid fern
48.0
58.8
69.2
84.0
0.33
48.4 41.2 46.4
56.8 60.0 54.8
66.0 68.4 67.6
78.0 77.2 71.2
0.15 0.34 0.22
R
PT
43.6
61.6
75.2
95.6
0.27
L
ACL
46.9 35.6 48.8
54.4 44.8 62.8
64.0 58.0 76.0
89.2 66.0 92.0
0.27 0.31 0.05
tib mid fern
FUNCTION
OF ACL BY SEP
TABLE 2. The portion of Table I related to the PI
latency upon stimulation
of the nine subjects
Case
Tibia1
Midportion
Femoral
1 2 3 4 5 6 7 8 9
52.8 62.4 54.6 66.8 52.2 51.6 49.2 48.4 46.4
48.0 53.6 54.0 59.6 44.8 38.8 38.8 41.2 35.6
56.8 58.4 59.2 65.6 56.0 58.2 53.2 46.4 48.8
In addition, as Pope states, there may not be a direct ligament-muscular reflex. Indirectly, in humans, by the use of an isokinetic dynamometer, Solomonow et al. (24) also found that upon knee extension in ACL-deficient patients, there was a burst of hamstring EMG activity at about 45”, when subluxation usually occurs. Barrach et al. (26,27) showed that the threshold to the ability to determine passive knee motion and knee joint angulation was decreased in the ACL-deficient knee. Brand (28) and Wroble and Brand (29) also have questioned the purely mechanical view of ligament function. They speculated that although neurosensory deficits are difficult to diagnose and the mechanical effect has been of most concern to the clinician, neurosensory function may predominate during normal activity and there may be a proprioceptive defect that, if understood, could lead to improved therapy. There has recently been an argument for a sophisticated and more complex reflex effect on the 6 muscle spindle system that influences the primary muscle spindle afferents perhaps to regulate muscle stiffness around the knee joint and therefore, in a more complex manner, to regulate the stability of the knee (30). There is some question as to whether the stimulus we used stimulates the ACL alone and does not spread to other adjacent structures in the knee joint. This insulated probe was exposed only at its very tip, and the very small intensity of stimulus applied does not allow spread of the stimulus. No response to electrical stimulation could be obtained when the tip of the probe was placed either on articular cartilage or just in the saline itself without touching any intraarticular structures. It appears that the stimulus remains localized to the ACL. In our study, the latency from the anterior cruciArthroscopy,
Vol.
8, No. 4, 1992
M. I. PITMAN ET AL. ate ligament is longer than that from the posterior tibia1 nerve. One explanation for this difference is that fibers in the posterior tibia1 nerve contain very rapidly conducting muscle afferents, whereas the nerves supplied to the ACL consist of small fibers that conduct more slowly, both peripherally and centrally. Also, there is a much smaller amount of nerve fibers in the anterior cruciate ligament, and thus a smaller amount of fibers may be stimulated in the ACL than in the posterior tibial nerve. Our study provides direct evidence that stimulation of the ACL does produce reproducible cortical somatosensory evoked potentials. There are, of course, other structures within and about the knee, such as the synovium and the collateral ligaments, that also supply proprioceptive nerve endings and certainly provide sensory input. Although we have used electrical stimulus, the sensory stimulus in the active human knee is, of course, not electrical but mechanical. Further studies must be undertaken to attempt to define a reproducible technique to provide an accurate mechanical stimulus in the intact human ACL that will have an effect upon these sensory nerve endings. However, electrical stimulation is another tool and a further step in studying and determining the sensory function of the anterior cruciate ligament in humans.
CONCLUSION This study has proposed a method to study physiologically the sensory function of the anterior cruciate ligament. We have obtained reproducible SEPs upon electrical stimulation of the intact anterior cruciate ligament. This method provides direct evidence of the presence of a proprioceptive function of the anterior cruciate ligament and represents a means to study this function physiologically in humans. Determination of the significance of the neurosensory function of the anterior cruciate ligament may have important consequences on the prevention, the surgery, and especially the rehabilitation of injuries to this ligament.
REFERENCES 1. Kennedy JC, Alexander IJ, Hayes KC. Nerve supply of the human knee and its functional importance. Am J Sports Med 1982;10:329-35. 2. Schultz RA, Miller DC, Kerr CS, Micheli L. Mechanorecep-
Arthroscopy, Vol. 8, NO. 4, 1992
tom in human cruciate ligaments:
a histological
study. J
Bone Joint Surg 1984;66-A:1072-76.
3. Schutte MJ, Dabezies EJ, Zimny ML, Happel LT. Neural anatomy of the human anterior cruciate ligament. J Bone Joint Surg 1987;69-A:243-7. 4. Halata Z, Haus J. The ultrastructure sensory nerve endings in human anterior cruciate ligament. Anat Embryo1 1989; 179:415-21. 5. Grigg T, Hoffman AH. Ruflini mechanoreceptors in isolated joint capsule response: correlated with strain energy density. Somatosens Res 1984;2: 149-62. 6. Grigg P, Hoffman AH, Fogerty KE. Properties of GolgiMazzoni: afferents in cat knee joint capsule as revealed by mechanical studies of isolated joint capsule. J Neurophysio/ 1982;47:31-40. I. Andrew BL, Dodt E. The deployment of sensory nerve endings at the knee joint of the cat. Acta Physioi Stand 1953; 281287-96. 8. Grigg P. Mechanical factor influencing response of joint afferent neurons from cat knee. J Neurophysiol1975;38-147384. 9. Grigg P, Greenspan B. The response of primate joint afferent neurons to mechanical stimulation of knee joint. J Neurophysiot 1977;40:1-8. 10. Dawson GP. The cerebral response to electrical stimulation of the peripheral nerve in man. J Neurol Neurosurg Psychiatry 1947;10:134-40. 11. Nainzadeh N. Somatosensory evoked potentials: assessment and management in the lumbar spine surgery. In: Camins M, O’Leary P, eds. The lumbar spine. New York: Raven Press, 1987:275. 12. Pratt H, Starr A, Amrie RN, Politoshe D. Mechanically and electrically evoked somatosensory potentials in normal humans. Neurology 1979;29: 1236. 13. Cohen L, Starr A. Effects of vibration and muscular contraction and several types of somatosensory evoked potentials. Neurology 1985;35:691. 14. Cusick JF, Myklebust J, Larson SJ, Sances A. Spinal evoked potentials in the primate: neuralsubstrate. J Neurosurg 1978;49:551-7. 15. Chiappa KH. Evokedpotentials in clinical medicine. 2nd ed. New York: Raven Press, 1990:371-92. 16. Cracco R, Bodis-Wollner I. Evoked potentials. New York: Alan R. Liss, 1986:379-85. 17. Aminoff MJ. Efectrodiagnosis of clinical neuroiogy. 2nd ed. New York: Churchill Livingston, 1986:535-8. 18. Goodgold J, Eberstein A. Electrodiagnosis of neuromuscular disease. 3rd ed., Baltimore: Williams and Wilkins, 1983:282-305. 19. Kimura J. Electrodiagnosis in diseases of nerve and muscle. Philadelphia: F.A. Davis, 1983:399. 20. Bunch WH, Scharff TB, Trimbie J. Current concepts review: spinal cord monitoring. J Bone Joint Surg 1983;65: 707-10.
21. Stowe RG, Weeks LE, Hajdu M, Stinchtield FE. Evaluation of sciatic nerve compromise during total hip arthroplasty. Clin Orthop Rel Res 198.5;201:26. 22. Pitman MI, Nainzadeh N, Ergas E, Springer S. The use of somatosensory evoked potentials for detection of neuropraxia during shoulder arthroscopy. Arthroscopy 1988;4(4): 250-5.
23. Palmar Ivar. Plastic surgery of the ligaments of the knee. Acta Chir Stand
24 Solomonow
1944;91:37-48.
M, Baratta
R, Zhou H, Bose W, Beck C,
EVALUATION
OF NEUROSENSORY
D’Ambrosia R. The synergistic action of the anterior cruciate ligament and thigh muscles in maintaining joint stability. Am J Sports Med 1987;15:207-13. 25. Pope PE, Kelly JC, Brand RA. Physiologic loading of the anterior cruciate ligament does not activate quadriceps or hamstrings in the anesthetized cat. Am J Sports Med 1990; 18:595-9. 26. Barrach RL, Skinner HB, Cook SD. Proprioception of the knee joint: paradoxical effects of training. Am J Sports Med 1984;63: 175-80.
FUNCTION
OF ACL BY SEP
447
21. Barrach RL, Skinner HB, Buckley SL. Proprioception in the anterior cruciate deficient knee. Am J Sports Med 1989;17: l-6. 28. Brand RA. Knee ligaments: a new view. Truns ASME 1986; 108: 10610. 29. Wroble RR, Brand RA. Function of knee ligaments: an historical review of two perspectives. Iowa Orthop J 1988;8: 62-l.
30. Johannson H, Sjolander P, Sojka P. A sensory role for the cruciate ligaments. Chin Urthop Re/ Res 1991;268: 161-78.
Arthroscopy,
Vol. 8, No. 4, I992
