Median/ulnar premotor potential identification and localization

A small negative waveform is known to precede the median and ulnar compound muscle action potentials when recorded with surface or concentric needle electrodes. This investigation documents that there are two distinct waveforms preceding the median compound muscle action potential (CMAP) depending upon the type of recording electrodes used (concentric needle versus surface) and their respective locations. The negative waveform originally described with a concentric needle electrode positioned within the substance of the distal thenar eminence and having a restricted zone of detection is referred to as the intramuscular nerve action potential (INAP). This potential is shown to be distinct from the premotor potential (the small negative waveform preceding surface recorded ulnar and median CMAPs). Detection of the median and ulnar premotor potentials at multiple locations about the hand with the same respective onset/peak latencies and amplitudes substantiates that this potential is a far-field potential. The median and ulnar premotor potentials most likely originate from a dipolar moment imbalance generated by digital sensory nerve action potentials as they cross the first and fifth metacarpophalangeal junctions, respectively. Applying far-field principles permits the documentation of additional far-field potentials as they are generated at the second through fourth metacarpophalangeal junctions following median nerve stimulation. Also, because the premotor potential is a far-field potential, caution must be exercised with respect to its diagnostic utility as joint position and other unknown factors may affect amplitude and onset/peak latency. The INAP following median nerve excitation, however, is documented to be a near-field potential distinct from the premotor potential arising from the recurrent branch of the median nerve. Therefore, although the median intramuscular nerve action potential and premotor potentials both precede the compound muscle action potential, they are different potentials with unique generator sites. 0 1995 John Wiley & Sons, Inc.

Key words: premotor potential volume conduction far-field potential median nerve ulnar nerve MUSCLE & NERVE 18:518-525 1995 9

MEDIANAJLNAR PREMOTOR POTENTIAL IDENTIFICATION AND LOCALIZATION DANIEL DUMITRU, MD, and JOHN C. KING, MD

T h e median or ulnar nerve compound muscle action potential (CMAP) is usually preceded by a small (10-50 p,V) negative waveform originally referred to as the intramuscular nerve action potential (INAP) but now commonly known as the premotor potential (PMP).2,'7 Although the median

From the Department of Rehabilitation Medicine, The University of Texas Health Science Center, San Antonio, Texas. Address reprint requests to Daniel Dumitru, MD, Department of Rehabilitation Medicine, The University of Texas Health Science, Center at San Antonio, 7703 Floyd Curl Drive, San Antonio, TX 78284-7798. Accepted for publication November 16, 1994 CCC 0148-639)(/95/050518-08 0 1995 John Wiley & Sons, Inc

518

Premotor Potential

INAP was initially presumed to arise from the recurrent branch of the median nerve prior to entering the thenar muscle mass,2 investigations in persons with profound amyotrophic lateral sclerosis (ALS) and cervical preganglionic root avulsions demonstrated the presence of a PMP despite an absent CMAP.'p8,'5Also, direct activation of the sensory nerves in the first digit as recorded antidromically over the thenar eminence continued to generate a PMP.7 The palmar cutaneous branch of the median nerve has been suggested to be a source for the PhlP12316;however, this nerve's anatomy" and subsequent studies4 did not support this conclusion. Selective neural blockade of the palmar cutaneous a n d recurrent motor branches of the median nerve failed to abolish the

MUSCLE & NERVE

May 1995

PMP supporting the impression that the PMP originates preferentially from sensory fiber^.^ T h e PMP has recently been modeled to arise from a sensory nerve action potential (SNAP) preceding a CMAP derived from the superimposition of a first dorsal interosseous CMAP and radial nerve SNAP as recorded over the FDI? The PMP’s morphology appears quite similar to that of a SNAP. T h e above-noted study modeling the PMP as a SNAP preceding a CMAP suggests that it may indeed be a SNAP.3 This same investigation conjectured that the PMP may not be a near-field SNAP but a far-field or so-called junctional potential, however, offered no experimental data to support this supposition. The present investigation hypothesizes that the PMP is indeed a far-field or junctional potential as opposed to a near-field SNAP. Both near-field and far-field recording montages are utilized to document the PMP’s spatial distribution and localize the PMP’s origin for both the median and ulnar nerves. Also, generator sites for the INAP are investigated to determine any relationships between these similar potentials. MATERIALS AND METHODS Instrumentation. Investigations were performed on a Cadwell 5200A electrophysiologic instrument (Cadwell Laboratories, Inc., Kennewick, WA). Instrumentation parameters included amplifier sensitivities between 10 and 50 pV/cm, low and high filter settings of 10-10,000 Hz respectively, a supraniaximal stimulus with a pulse duration of 100 ps, and a sweep speed of 1 msicm. All recorded waveform amplitudes were measured from baseline-to-peak, onset latencies were measured to the waveform’s initial negative deflection, and peak latencies were designated by the waveform’s maximal negative peak. Sequential bipolar near-field SNAPS were recorded using the poles of a pediatric stimulator (Teca Corp., Pleasantville, NY) with an interelectrode separation of 1.25 cm. Because near-field and far-field recording montages are used in this investigation, the terms “active” and “reference” with respect to recording electrode locations become arbitrary. Therefore, the terms E-1 and E-2 replace the active and reference designations respectively. It is to be understood that a negative potential difference between the two recording electrodes produces an upward or negative waveform deflection.6

Eleven healthy adults (7 maled4 females) ranging in age from 31 to 40 years of age

Subjects.

Prernotor Potential

(mean 36.4 k 3.8 years) without symptoms or signs of neuromuscular disease provided informed consent to participate in this investigation. The right hand was investigated in all subjects. A surface temperature of 32-34°C on the volar and dorsal aspect of the examined hand was maintained throughout the study. Statistical analysis consisted of mean, standard deviation calculations, Pearson’s correlation coefficient, and two-tailed Student’s t-test. The median PMP was investigated by locating an E-1 electrode on the abductor pollicis brevis muscle’s motor point with E-2 positioned on the first digit’s dorsal surface immediately proximal to the thutnbnail. A median nerve stimulus was delivered 8 cm proximal to E-1 using a well-secured bar electrode (2.5 cm cathode/anode separation). Following the documentation of a median nerve thenar eminence PMP, E-1 was then located at various sites on the hand’s volar and dorsal aspect to document the PMP’s spatial distribution (Fig. 1A). A near-field SNAP recording montage using the pediatric stimulator as recording electrodes was then performed where this E- 1 electrode was serially positioned in 1 .O-cm increments along the anatomic course of the median nerve’s first digital branch (Fig. 1A). This recording montage provided a physiologic/anatomic representation of the first digit’s SNAP with respect to specific anatomic locations. T h e E-1 electrode (referenced to previously defined E-2 on first digit) was sequentially positioned along the first digit’s dorsal surface in 1.O-cm increments until the PMP’s amplitude no longer changed (Fig. 1B). Following documentation of the median PMP, the E-2 electrode was relocated to the hypothenar region in 5 individuals to verify the far-field nature of the PMP (Fig. 1B). Because far-field potentials can vary with anatomic alterations in the generator site, the median PMP was also qualitatively investigated in 5 individuals with the first digit 90”to the palm, adducted to the second metacarpal region, and passively maintained in maximal palmar adduction and flexion. In 5 individuals, an E-1 electrode was located on the hand’s dorsum over the second through fourth metacarpal bones midway between the wrist and metacarpophalangeal region, referenced to an E-2 electrode positioned immediately proximal to the respective digit’s proximal interphalangeal dorsal surface (Fig. 2A). This referential recording montage was utilized to detect any far-field potentials produced at the metacarpophalangeal joint

Median PMP.

MUSCLE & NERVE

May 1995

519

’.’+

FIGURE 1. (A) The right hand is displayed with multiple E-1 (E-1,-E-1 F) electrode locations (dashed circles represent dorsal recording sites) referenced to a common E-2 electrode (first digit). The median premotor potential remains constant irrespective of the E-1 recording site. Sequential 1-cm recording sites along the first digit reveal the SNAP at each location. The SNAP at the first metacarpophalangeal junction (5 cm) aligns with the premotor potential. (B) Sequential E-1, recording sites (each individual E-1, referreveal the premotor potential reaches its enced to E-2,) maximum amplitude at all locations distal to the first metacarpophalangeal joint. E-1 referenced to E-2, reveals the expected premotor potential. E-1 referenced to E-2,, however, demonstrates the premotor potential’s absence but continued CMAP presence.

region following median nerve stimulation. An E- 1 electrode located 2 cm proximal to the palmar third web space on the volar hand surface (E-2 positioned 1.25 cm distally) was used to further investigate the third digit’s possible far-field generator. This bipolar recording montage was formulated to correlate a near-field median SNAP with a possible far-field generator site. The E-1 electrode at the previously noted site was then referenced to an E-2 electrode placed on the distal aspect of the fifth digit’s dorsum creating a referential near-field montage (Fig. 2A).

520

Premotor Potential

FIGURE 2. (A) Locating a referential recording montage across each of the first four metacarpophalangeal joints reveals a far-field potential following median nerve stimulation. A near-field bipolar montage (dashed circles designated E-1 and E-2) located at the third metacarpophalangeal joint demonstrates an alignment between the far-field (3rd A) and near-field (3rd 6)potentials’ peaks. A referential recording (dashed circle E-1 referenced to E-2 on the fifth digit’s dorsum) demonstrates that the near-field potential’s positive deflection onset (3rd C) aligns with the far-field potential’s onset. (B) A concentric needle positioned in the abductor pollicis brevis’ motor point area fails to reproduce the premotor potential with a continued documentation of the median nerve’s CMAP. Relocating the concentric needle toward the base of the thenar eminence now detects the intramuscular nerve action potential.

In 10 subjects the median PMP was recorded with E-1 on the thenar muscles’ motor points and E-2 located distally on the first digits. A concentric needle electrode was inserted into the thenar muscle mass at the same location as the above-described E- 1 electrode (Fig. 2B). Additionally, the needle electrode was inserted 1-2 cm more distally toward the thenar eminence’s base with final needle placement determined by median nerve wrist excitation and the documentation of maximal INAP amplitude. The INAP and PMP onsedpeak latencies and amplitudes were compared. T w o digital ring electrodes were positioned on the first four digits with an interelectrode sepINAP vs. PYP.

MUSCLE & NERVE

May 1995

aration of 4 cm. T h e proximal ring electrode was located at 5 and 8 cm for the first and remaining three digits respectively. These two ring electrodes served as cathode and anode with a maximally tolerable digital excitation delivered while the concentric needle previously positioned to record the maximal INAP served as the recording electrode. In 10 subjects a monopolar needle (DMF 25, Teca Corp., Pleasantville, NY) was located at the same location as the concentric needle noted above. The most appropriate monopolar needle location was determined by maximizing the INAPs amplitude similar to the procedure used for the concentric needle location. This electrode then served as a cathode utilizing a 2-cm-diameter surface dispersive anode located on the hand’s dorsum and a stimulus pulse duration of 50 ps. A monopolar, as opposed to concentric, needle was used so that a safe amount of current could be delivered in proximity to neural tissue. A supramaximal surface recorded median CMAP was documented for both the surface (8 cm proximal) and needle cathode locations. Without moving the monopolar needle electrode, the above-described digital ring electrodes served as recording electrodes while a current capable of generating a supramaximal median CMAP was generated through the monopolar needle. The digital electrodes served to record any antidromic SNAPs generated by the monopolar cathode. Ulnar PMP. An E- I electrode was situated over the abductor digiti minimi muscle’s motor point with an E-2 electrode located on the fifth digit’s dorsal surface just proximal to the fingernail (Fig. 3A). Supramaximal ulnar nerve activation occurred 8 cm proximal to E-I. T h e E-1 electrode was then positioned at multiple sites on the volar and dorsal surface of the hand (Fig. 3A). A bipolar near-field SNAP montage was performed in 1.0-cm increments on the fifth digit and progressing into the palm along the lateral ulnar digital nerve branch similar to that performed for the median nerve. An E-1 electrode was then initially positioned adjacent to the fifth digit’s E-2 electrode and advanced proximally in 1-cm increments toward the wrist until the ulnar PMP’s amplitude maximized (Fig. 3B). RESULTS

T h e multiple E-1 recording locations revealed PMP onsetipeak latencies as well as amplitudes that had no statistically significant differences (P > 0.3 and r > 0.9) with a power of

Median PMP.

Prernotor Potential

FIGURE 3. (A) Multiple E-1 (E-IA-E-l,) recording locations referenced to a common E-2 on the fifth digit demonstrate an ulnar prernotor potential with similar latencies and arnplitudes irrespective of recording location. A bipolar near-field recording montage along the fifth digit’s sensory branch reveals a number of ulnar SNAPs with the potential recorded at the fifth metacarpophalangeal joint aligning with the premotor potential. (6)Multiple E-1 electrodes each sequentially referenced to the E-2 electrode reveal the ulnar premotor potential to maximize distal to the fifth rnetacarpophalangeal joint.

greater than 30% for all waveform comparisons (Fig. 1A). The PMP, therefore, was isopotential at all recording locations. The mean PMP onset and peak latencies for the 11 subjects were 1.9 0.2 ms (range 1.4-2.5 ms) and 2.6 0.2 ms (range 2.2-2.9 ms), respectively, with a mean amplitude of 19.9 2 8.1 pV (range 10-34 kV) which were similar to those previously r e p ~ r t e d Recording .~ a SNAP in I-cm increments from the first digit’s tip to thenar base revealed a waveform that expectantly shortened in latency (Fig. 1A). I n all subjects, an antidromic SNAP with the same peak latency (P > 0.8, r > 0.8) to that of the PMP occurred when the E-1 recording electrode was positioned at the first metacarpophalangeal joint. Locating an E- 1 elec-

*

MUSCLE & NERVE

*

May 1995

521

trode in 1-cm increments on the first digit’s dorsum demonstrated a progressive PMP amplitude increase which no longer changed beyond 2 cm proximal to the first metacarpophalangeal joint for all subjects (Fig. 1B). Of note, in 2 persons a small negative potential appeared to be incorporated within the main PMP (Fig. 1A). The presence of this potential at all recording locations suggested that an additional far-field potential with a similar latency can occasionally be detected. The mechanism of production for this somewhat bifid response is unclear and may arise from an early branching of the first digital sensory nerve, differential volume conductor inhomogeneities, or some other volume conductor discontinuity and additional dipolar moment imbalance. It is also possible that the close approximation of the stimulus artifact and the PMP may result in some degree of interference distorting the PMP’s initiation. A complete absence of the median PMP resulted when the E-1 electrode was referenced to an E-2 electrode located on the hypothenar eminence for all participants (Fig. 1B). The disappearance of the median PMP confirms that both E-1 and E-2 recording electrodes are located in a region that is equipoten tial. Altering the first metacarpophalangeal joint angle produced a change in the PMP’s onsetipeak latencies and amplitude. For example, in 1 subject abducting the first digit 90” to the palm resulted in a PMP onsetipeak latency prolongation (2.2i2.8 ms) and amplitude reduction (23 kV) compared to the first digit adducted against the second metacarpal bone (onsedpeak latency: 2.112.6 ms; amplitude: 28 pV). Complete palmar adduction and flexion generated a median PMP with a shorter onset latency (1.9 ms) but same peak latency (2.6 ms) compared to the previous digit position; however, the amplitude markedly increased (40 kV). Similar but unpredictable changes occurred in all examined subjects. ‘I‘he first digit’s anatomic position, therefore, can influence the median PMP’s potential diagnostic parameters. Median nerve far-field potentials could be demonstrated to occur across each of the first four metacarpophalangeal joint regions for all 5 examined subjects (Fig. 2A). Additional investigations across the third metacarpophalangeal joint revealed a far-field potential with a peak (2.9 ms) that aligned with a near-field SNAP when E-1 was located 2 cm proximal to the third web space (Fig. 2A). This anatomic location corresponds to the metacarpophalangeal joint itself where the flat surface of the palm transitions to the cylindrical digit.

522

Premotor Potential

The SNAP’s onset, however, was noted to be 2.4 ms which did not correspond to the far-field potential’s onset of 2.1 ms. Referencing the E-l electrode located over the median sensory fibers in the palm to the fifth digit produced a triphasic SNAP with the initial positive deflection’s onset aligning with the far-field potential’s onset. The far-field potential’s and referentially recorded SNAP’s onsets should align since this is when the SNAP’s current first flows at the palmidigit transition. A small negative potential ( < 5 pV) can be observed to precede the primary potential of interest. The origin of this potential is unclear and was not investigated in this study. INAP vs. PMP. Locating a concentric needle within muscle tissue at the same surface E-1 location resulted in no detectable PMP (Fig. 2B). A negative potential preceding the CMAP could only be demonstrated when the concentric needle electrode was inserted 1-2 cm toward the thenar eminence’s base and had a very restricted recording zone.‘ The potential recorded with the needle electrode (INAP) had mean onsetipeak latencies of 1.5 2 0.2 ms and 2.0 0.2 ms, respectively, with an amplitude of 33.3 ? 8.4 kV. This potential had onset/ peak latencies and amplitude that were significantly different from the PMP ( P < 0.00 1; r < 0.1) but corresponded to values previously reported for the INAP.‘ Therefore, the INAP and PMP appear to be near-field and far-field potentials respectively; that is, dissimilar potentials with different generator sites. Occasionally, a small amplitude (10 k V or less) could be seen to precede the INAP. The origin of this potential remains unclear and was not investigated. Median nerve wrist stimulation resulted in a mean CMAY distal motor latency of 3.5 ? 0.4 ms (3.0-4.3 ms) and an amplitude of 7.4 2 2.1 mV (4.4-12.3 mV), while the palmar needle stimulus generated a motor latency of 1.3 ? 0.2 ms (1.0-1.7 mV) and an amplitude of 8.7 2 2.1 mV (5.5-12.8 mV). The IIL’APs recorded with the monopolar needle was similar to those obtained with the concentric needle (1.5 0.1; P > 0.8 and r > 0.8). An orthodromic SNAP (concentric needle recording) was not detected in any of the subjects. Similarly, an antidromic digital SNAP (monopolar needle stimulation) was not recorded from the first four digits in any subject despite the documentation of a supramaximal median CMAP.

*

*

Ulnar PMP. When the E-1 electrode was located over the hypothenar muscles’ motor point an ul-

MUSCLE & NERVE

May 1995

nar PMP could be recorded in all subjects. The ulnar PMP, unlike the median PMP, could not be recorded at every recording site about the hand in all subjects because of a superimposed ulnar CMAP prohibiting response observation. In only 4 of the 11 individuals could every location document an ulnar PMP (Fig. 3A).For those potentials present at more than a single location, there was noted to be no statistically significant difference for onsedpeak latencies or amplitudes ( P > 0.4and r > 0.9).The mean onsedpeak latency for the ulnar PMP detected over the hypothenar eminence were 2.0 ? 0.3ms (range 1.7-2.5 ms) and 2.6 f 0.2 ms (range 2.2-2.9 ms), respectively, with an amplitude of 14.6 2 9.0 pV (range 8-32.5 pV). Sequential bipolar E-1 electrodes positioned in 1-cm increments from the fifth digit’s tip into the palm revealed an ulnar SNAP that expectedly decreased in onsedpeak latency. At the fifth metacarpophalangeal region, a SNAP was produced with a peak latency corresponding to that of the ulnar PMP (Fig. 3A). Locating an E-1 electrode in 1-cm increments distal to an E-2on the fifth digit’s tip reveals a far-field potential that incrementally increased in amplitude reaching a maximum approximately 2 cm proximal to the fifth metacarpophalangeal joint (Fig. 3B). DISCUSSION

A small, negative, easily observed waveform preceding the median and ulnar nerves’ CMAP can typically be detected when recorded at relatively high amplifier sensitivities. T h e origin of this potential is the source of considerable discussion in the literature extending over 30 years. ‘The most recent area of contention concerns whether the PMP is a near-field (SNAP) or far-field (junctional) p ~ t e n t i a l .If~ the PMP is a near-field SNAP, it should be localized to a restricted anatomic location, change latency as the E-l electrode is moved along the course of the presumed nerve, and dramatically decline in amplitude when moved off the nerve. A far-field potential, however, should have a widespread spatial distribution with little change in amplitude and latency despite widely disparate E-1 recording locations. T h e differentiation between a near-field and far-field potential is accomplished by employing a combination of bipolar and referential recording montages.6 These same recording montages can also be used to identify a far-field potential’s generator site. Multiple referential recording montages with E-1 located at various positions on the volar and

Prernotor Potential

dorsal hand surface clearly establishes the median PMP as a far-field and not near-field potential. This conclusion is validated by not only the PMP presence, but constancy of latency and amplitude irrespective of E-1’sposition about a large area of the hand or digits (Fig. 1A). When the E-1 approaches the E-2 electrode from the hand’s dorsum, the median PMP does not change in amplitude until the E-1 electrode approaches and then crosses the first metacarpophalangeal joint after which the PMP amplitude precipitously declines (Fig. 1B). This suggests that all regions proximal to the first metacarpophalangeal region are equipotential. This metacarpophalangeal joint transition implies there is an action potential dipolar moment imbalance generated across this anatomic junction. In other words, the PMP is a junctional potential created by a volume conductor discontinuity at the first metacarpophalangeal joint. Confirmation of this supposition is provided by coincidence of the near-field SNAP at the first metacarpophalangeal joint having the same peak latency as the PMP (Fig. 1A). Locating an E-1 (thenar eminence) and E-2 (hypothenar eminence) electrode on the same side of the dipolar moment imbalance completely eliminates the median PMP supporting the contention that this aspect ofthe hand is isopotential (Fig. 1B). The median PMP, therefore, is a far-field potential generated by the first digit’s sensory action potentials traversing the first metacarpophalangeal joint region. This finding has been previously documented for the radial nerve which appeared to generate a far-field potential at the wrist and first metacarpophalangeal joint region. I s The above description of far-field potential generation can be understood with the leadingtrailing dipole model in which an action potential is considered to be a quadrupole consisting of a leading and trailing dipole (Fig. 4).5,6As the action potential crosses the first metacarpophalangeal junction, the leading dipole travels from a large flat volume (palm of hand) into a small cylindrical volume (first phalanx). Additionally, there is a change in action potential propagation direction. The trailing dipole located in the palm now has a different dipole moment than the leading dipole thereby generating a far-field potential. It is likely that both the size transition and directional change combine to create the relatively large amplitude far-field potential known as the PMP. Also, the small cylindrical digit creating an increased electric field density per unit volume compared to the palm combined with E-1located on the thenar em-

MUSCLE & NERVE

May 1995

523

Differential Amplification { LD ( 101 - (10) I

5) 1 -5 - (5)= -10

+ { TD (51 - (

A

FIGURE 4. An action potential depicted as a quadrupole (leadinghrailing dipole) reaches its maximum dipolar moment imbalance when the leading dipole is contained in the cylindrical digit while the trailing dipole remains in the flat hand volume. The smaller cylindrical digit increases the leading dipole’s dipolar moment compared to that of the trailing dipole. The above recording montage results in a monophasic negative potential. The arbitrary numbers included depict the manner in which differential amplification ({E-l}-{E-2}) processes the leading dipole (LD)and trailing dipole’s (TD) respective voltages as recorded at the electrodes.

inence accounts for the negative polarity of the PMP (Fig. 4). T h e ulnar PMP is generated by a similar mechanism to that described for the median PMP. A mobile E-1 referenced to a stationary E-2 on the tip of the fifth digit demonstrates that the ulnar PMP has a widespread distribution about the hand substantiating it as a far-field potential (Fig. 3A). The near-field sequential recording of the ulnar nerve’s SNAP to the fifth digit localizes the volume conductor transition at the fifth metacarpophalangeal joint as the likely site for the ulnar PMP. Specifically, as the leading dipole enters the fifth digit, its dipole moment is increased compared to the trailing dipole moment which is still contained in the larger volume of the palm similar to the situation described for the first digit’s median sensory fibers. These suppositions are supported by finding no further change in the generated far-field after the E-1 electrode is beyond the fifth metacarpophalangeal joint (Fig. 3B). Near-field SNAP recordings localize the generation site for the median and ulnar PMPs to the first and fifth metacarpophalangeal regions re-

524

Premotor Potential

spectively. T h e far-field potential hypothesis is consistent with all of the previously reported clinical findings. Specifically, the median and ulnar PMP are generated by digital sensory nerves crossing a volume conductor boundary and explains why the PMP remains in persons with lesions affecting the motor but not sensory fibers. Also, the median PMP arises from the digital sensory fibers destined to innervate the first digit as suggested by different neural techniques. This explains why persons with carpal tunnel syndrome or other neuropathies have a delayed or absent PMP. l 1 Any pathology affecting the digital sensory nerves to the first digit should result in an alteration of the PMP. Caution must be exercised, however, whenever far-field potentials are used for diagnostic purposes as they are subject to any alteration in volume conductor characteristics and directional changes.14 This is illustrated qualitatively by recording the median PMP from the thenar eminence with different first digit anatomic positions. Because a far-field potential is not solely an intrinsic property of the nerve but the result of an interaction between the neural impulse and its enveloping volume conductor, anatomic factors may adversely affect the far-field potential yielding uncertain diagnostic consequences. Thus, anatomic position must be standardized when recording far-field positions. Applying the above far-field principles suggests that the median sensory nerves crossing the metacarpophalangeal joint regions to the second through fourth digits should also generate farfield potentials. This supposition is substantiated by qualitatively demonstrating a far-field potential for each of the four metacarpophalangeal joints following median nerve activation in 5 subjects (Fig. 2A). A more detailed investigation in a single individual was performed for the third metacarpophalangeal joint. The far-field potential at this region is found to align with a volarly recorded median near-field SNAP as the nerve traverses the hand/digit transition. The far-field potential’s onset is found to align with the action potential’s initial positive source current. This investigation demonstrates a significant difference between the PMP and INAP suggesting these potentials have different origins. As originally described, the INAP has a very restricted recording zone suggesting it is a near-field potential.* Locating a concentric needle electrode in the same location as a surface electrode fails to document a PMP which is anticipated for a far-field

MUSCLE & NERVE

May 1995

potential when both recording electrodes (needle core and cannula) are positioned in an isopotential zone (Fig. 2B). However, relocating the concentric needle 1-2 cm toward the base of the thenar eminence results in the appearance of a potential with parameters similar to that described for the INAP and a restricted recording location. The INAP and PMP, therefore, are not the same potential as is presently assumed. The INAP’s origin may be the first digit’s sensory fibers or the mixed nerve innervating the thenar muscles. Documentation of an easily obtainable median CMAP from a needle cathode located at the same site as INAP observation confirms the earlier impression that at least some of the fibers comprising the INAP are motor in origin. Failure to produce a median antidromic digital SNAP despite the detection of a median CMAP suggests that there are few if any sensory fibers in the vicinity of the INAP generator. This is supported by failure to detect an orthodromic SNAP at the INAP location following digital median nerve excitation. The combination of a median CMAP but no SNAP with stimulation at the INAP recording location and absent INAP with digital stimulation suggests that the INAP arises from median nerve fibers destined to innervate the muscles of the thenar eminence with little if any contribution from digital sensory fibers; that is, the INAP likely arises from the recurrent branch of the median nerve as originally assumed. Of course, there may be some “sensory” fibers contained in the INAP arising from those afferent fibers innervating the muscle spindles. Therefore, despite the qualitative resemblance between the INAP and PMP, they are distinct waveforms. Specifically, the INAP has shorter mean onsedpeak latencies (1.5/2.0 ms versus 1.912.6 ms) with a larger amplitude (33.5 versus 19.9 pV). Also, the INAP has a very restricted zone of detection with a concentric needle electrode, whereas the PMP completely disappears when re-

Premotor Potential

corded with the concentric needle electrode. The INAP, therefore, is a near-field potential, while the PMP is a sensory junctional far-field potential.

REFERENCES 1 . Baba M, Narita S, Matsunaga M, et al.: Intramuscular

nerve action potential derived from sensory fibers. Electromyogr Clin Neurophysiol 1979;19:46747 1. 2. Buchthal F, Rosenfalck A: T h e intramuscular nerve action potential. Brain. Res 1966;3:72-73. 3 . Del Toro DR, Park T A , Wertsch JJ: Development of a model of the premotor potential. Arch Phys Med Rehabil 1994;75:493-497. 4. Dumitru D, Walsh NE, Ramaniurthy S: The premotor potential. Arch Phys Med Rehabil 1989;70:537-540. 5. Dumitru D, King JC: Far-field potential production by quadrupole generators in cylindrical volume conductors. Electroencephalogr Clin Neurophysiol 1993;88:421 4 3 1. 6. Dumitru D, Jewett DL: Far-field potentials. Muscle Nerue 1993;16:237-254. 7. Fiaschi A: Observations of the sensory nature of the intramuscular nerve action potential. J Neurol Neuro.surg Psychiatry 1973;36:509-513. 8 . Gutmann L: T h e intramuscular nerve action potential. J Neurol Meurosurg Psychiatq 1969;32: 193-196. 9. Gutmann L, Hopf HC, Roeder K: Origin of intramuscular nerve action potential. J Neurol Neurosurg Psychiatv 1987; 50: 1669-1670. 10. Hobbs RA, Magnussen PA, Tonkin MA: Palmar cutaneous branch of the median nerve.] Hand Surg 1990;15A:38-43. 1. Johnsen JA, MacLean IC: T h e intrapalmar nerve action potential: its value in diagnosing polyneuropathy. Muscle Nerve 1990;13:867. 2. Kanakamedala RV, Hong CZ: Electrophysiologic studies of the median nerve and its palmar cutaneous branches after nerve grafting. Arch I’hys Med Rehabil 1988;69:344-347. 3. Kimura J , Mitsudome A, Yamada T, Dickins QS: Stationary peaks from a moving source in far-field recording. Electroencephalogr Clin Neurophysiol 1984;58:35 1-36 1. 4. KimuraJ, Kimura A, Ishida T, et al.: What determines the latency and amplitude of stationary peaks in far-field recordings? A n n Neurol 1986;19:479486. 5. Lambert EH: Electromyogrdphy in amyotrophic lateral sclerosis, in Norris FH, Kurland LT (eds): Motor Neuron Diseases. New York, Grune & Stratton, 1969. 6. Lum PB, Kanakamedala RV: Conduction of the palmar cutaneous branch of the median nerve. Arch Phys Med Rehabil 1986;67:805-806. 7. Simpson JA: Fact and fallacy in measurement of conduction velocity in motor nerves. J Neurol Neurosurg Psychiatry 1964;27:381-385.

MUSCLE &

NERVE

May 1995

525