Journal of Neurology, Neurosurgery, and Psychiatry, 1979, 42, 436-451 Nerve conduction, tactile sensibility, and the electromyogram after suture or compression of peripheral nerve: a longitudinal study in man F. BUCHTHAL AND V. KUHL From the Laboratory of Clinical Neurophysiology and the Research Laboratories 4112, Rigshospital, Copenhagen, Denmark SUMMARY In three patients sequential studies were performed of sensory and motor conduc- tion after complete section and suture of the median nerve at the wrist and in one patient after partial section of the nerve. The sensory potential evoked by stimuli to digits III and I and recorded proximal to the suture line at the wrist appeared after a delay of three to four months, corresponding to a growth rate of 1.5-2.0 mm per day. From early in the course of regeneration the sensory potential was dispersed in 40 components. In the adult patient the cumulative ampli- tude increased for two years slowly and thereafter at a two times faster rate. Amplitude and tactile sensibility were normal after 40 months, but the sensory potential was still five times more dispersed than normal. The overall increase in the amplitude of the sensory potentials in children aged 10 and 12 years was three times faster than in the adults. In the adults and in the children the maximum sensory conduction velocity was 10-25% of normal. It then increased at 3% per month during the first two years, and thereafter 10 times slower. Forty months after suture in the adults and 13-19 months after suture in the children the conduction velocity had reached 65-75% of normal. The pattern of discrete electrical activity during voluntary effort and the prolonged duration of motor unit potentials indicate persistent enlargement of the reinnervated motor units by peripheral sprouting. The sensory potential recovered five times faster after a compressive nerve lesion than after section and suture as seen in another patient with an affection of the ulnar nerve at the elbow. Normal tactile sensibility was attained 10 times faster than after section and suture. Maximum sensory and motor conduction velocity recovered within one year from 60-70% to 80-90% of normal. Regeneration of peripheral nerve after section occurs by the outgrowth of axons from the proximal stump (Cajal, 1928; Young, 1942; Guth, 1956; Schr6der, 1975). Animal studies indicate that in the early phase of regeneration the nerve fibres have small axons, thin myelin sheaths, a short internodal length, and a slow conduction velocity (Berry et al., 1944; Erlanger and Schoepfle, 1946; Sanders and Whitteridge, 1946; Cragg and Thomas, 1964; Jacobson and Guth, 1965; Gilliatt et al., 1975). The myelin sheath remains dispro- portionately thin and the short internodes persist even after the axons have attained near normal Address for reprint requests: Professor F. Buchthal, Rigshospital, Section 4112, Room 1220, Blegdamsvej 9, Copenhagen 2100, Denmark. Accepted 22 November 1978 diameters (Hiscoe, 1947; Vizoso and Young, 1948; Schr6der, 1972). In the cat the maximum con- duction velocity of regenerating fibres increases linearly with the external diameter of the largest myelinated fibres (Berry et al., 1944), as occurs during growth of the normal nerve (Hursh, 1939). The first study of conduction velocity in man during regeneration of motor nerve (Hodes et al., 1948) showed slowed maximum conduction velocity, prolonged motor latency, and an ampli- tude of the temporally dispersed muscle action potential which increased at first rapidly and thereafter more slowly, to at most 40-60% of normal 12-42 months after injury. Incomplete recovery of motor conduction was confirmed in later studies (Struppler and Huckauf, 1962; 436 Protected by copyright. on October 29, 2020 by guest. http://jnnp.bmj.com/ J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.42.5.436 on 1 May 1979. Downloaded from
Transcript
Journal ofNeurology, Neurosurgery, and Psychiatry, 1979, 42, 436-451
Nerve conduction, tactile sensibility, and theelectromyogram after suture or compression ofperipheral nerve: a longitudinal study in manF. BUCHTHAL AND V. KUHL
From the Laboratory of Clinical Neurophysiology and the Research Laboratories 4112,Rigshospital, Copenhagen, Denmark
SUMMARY In three patients sequential studies were performed of sensory and motor conduc-tion after complete section and suture of the median nerve at the wrist and in one patient afterpartial section of the nerve. The sensory potential evoked by stimuli to digits III and I andrecorded proximal to the suture line at the wrist appeared after a delay of three to four months,corresponding to a growth rate of 1.5-2.0 mm per day. From early in the course of regenerationthe sensory potential was dispersed in 40 components. In the adult patient the cumulative ampli-tude increased for two years slowly and thereafter at a two times faster rate. Amplitude andtactile sensibility were normal after 40 months, but the sensory potential was still five times moredispersed than normal. The overall increase in the amplitude of the sensory potentials in childrenaged 10 and 12 years was three times faster than in the adults. In the adults and in the childrenthe maximum sensory conduction velocity was 10-25% of normal. It then increased at 3% permonth during the first two years, and thereafter 10 times slower. Forty months after suture inthe adults and 13-19 months after suture in the children the conduction velocity had reached65-75% of normal. The pattern of discrete electrical activity during voluntary effort and theprolonged duration of motor unit potentials indicate persistent enlargement of the reinnervatedmotor units by peripheral sprouting. The sensory potential recovered five times faster after acompressive nerve lesion than after section and suture as seen in another patient with an affectionof the ulnar nerve at the elbow. Normal tactile sensibility was attained 10 times faster thanafter section and suture. Maximum sensory and motor conduction velocity recovered within oneyear from 60-70% to 80-90% of normal.
Regeneration of peripheral nerve after sectionoccurs by the outgrowth of axons from theproximal stump (Cajal, 1928; Young, 1942; Guth,1956; Schr6der, 1975). Animal studies indicatethat in the early phase of regeneration the nervefibres have small axons, thin myelin sheaths, ashort internodal length, and a slow conductionvelocity (Berry et al., 1944; Erlanger and Schoepfle,1946; Sanders and Whitteridge, 1946; Cragg andThomas, 1964; Jacobson and Guth, 1965; Gilliattet al., 1975). The myelin sheath remains dispro-portionately thin and the short internodes persisteven after the axons have attained near normalAddress for reprint requests: Professor F. Buchthal, Rigshospital,Section 4112, Room 1220, Blegdamsvej 9, Copenhagen 2100, Denmark.Accepted 22 November 1978
diameters (Hiscoe, 1947; Vizoso and Young, 1948;Schr6der, 1972). In the cat the maximum con-duction velocity of regenerating fibres increaseslinearly with the external diameter of the largestmyelinated fibres (Berry et al., 1944), as occursduring growth of the normal nerve (Hursh, 1939).The first study of conduction velocity in man
during regeneration of motor nerve (Hodes et al.,1948) showed slowed maximum conductionvelocity, prolonged motor latency, and an ampli-tude of the temporally dispersed muscle actionpotential which increased at first rapidly andthereafter more slowly, to at most 40-60% ofnormal 12-42 months after injury. Incompleterecovery of motor conduction was confirmed inlater studies (Struppler and Huckauf, 1962;
436
Protected by copyright.
on October 29, 2020 by guest.
http://jnnp.bmj.com
/J N
eurol Neurosurg P
sychiatry: first published as 10.1136/jnnp.42.5.436 on 1 May 1979. D
Ballantyne and Campbell, 1973) and agrees withexperimental findings (Berry et al., 1944; Erlangerand Schoepfle, 1946; Cragg and Thomas, 1964;Jacobson and Guth, 1965). Ballantyne and Camp-bell (1973) recorded sensory and motor actionpotentials in 20 patients each examined at adifferent time after suture of the nerve at thewrist. They used percutaneous recording, and asensory potential was not detected until 10months after suture. Similarly, Tallis et al (1978)did not find a sensory potential until 18 monthsafter a graft had been inserted. In both studiesthe maximum sensory and motor conductionvelocity recovered to 70-85% of normal; theamplitude of the sensory potential returned to15% of normal.Taking advantage of the increased resolution
obtained by recording from a near-nerve electrodeand by averaging (Buchthal and Rosenfalck, 1966,1971), sequential evaluations of motor and sensoryconduction velocities and of the amplitude ofevoked potentials from muscle and sensory nervefibres were performed in three patients duringrecovery after suture of the median nerve at thewrist, and in one patient with a partial transectionthat was not sutured. The electrophysiologicalfindings were related to the return of perceptionof light touch and pinprick. In addition, theelectrical activity during voluntary effort wasexamined at different times during reinnervationof the abductor pollicis brevis muscle. Finally, thetime course of the recovery after suture wascompared with that in a patient after a transientcompression of the ulnar nerve at the elbow.
Methods
The sensibility for light touch was examined bymeans of a soft brush, 0.75 mm in diameter. The
force which just bent the hairs was 0.3 g(0.8 g/mm radius in stiffness) near the normalthreshold (von Frey, 1896; Head, 1920; Buchthaland Rosenfalck, 1966). Perception of superficialpain was tested with a sharp pin.The force of the abductor pollicis brevis and the
abductor digiti minimi muscles was evaluatedaccording to a scale of 0 to 5 with subdivision ofgrade 4 (Medical Research Council, 1976).The method for evoking and recording sensory
potentials has already been described (Buchthaland Rosenfalck, 1966, 1971). The potentials wereevoked by stimuli to the distal and the proximalphalanx of digit III, and the distal phalanx ofdigits I and V. The stimulating current was0.2 ms in duration and 60 to 100 mA (surfaceelectrodes) or 20 to 30 mA (subcutaneouselectrodes).The recording electrode (bared tip 3 mm) was
placed near the median nerve proximal to thesuture line at the wrist. To record from the ulnarnerve, recording electrodes were placed near thenerve, 50 mm distal and 50 mm proximal to thesulcus nervi ulnaris at the elbow (Payan, 1969).The distances of conduction are given in Table 1.When the stimulus at the wrist failed to evokea muscle action potential the electrode was placednear the nerve by using the threshold of thepotential of mixed median nerve at the elbow as a
gauge.The sensory potentials were amplified (Andersen
and Buchthal, 1970), digitalised, and recorded viaa memory store on an ink-jet writer (Dahl andBuchthal, 1978). The criteria for identifying a
summated component of the potential were thatit remain similar in shape and increase with thenumber of responses roughly to the same extentas a calibration signal.The stimulating current was at least 10 times
Table 1 Distances of conduction (mm) to suture line and to recording electrode
Digit I to suture 106 - 102 102Digit I to recording electrode 128 110 142 135Distal digit III to suture 173 - 163 -
Distal digit III to recording electrode 195 172 183Proximal digit III to suture 144 - 123 136t$,Proximal digit III to recording electrode 166 145 143 169tM ABD§ to suture 58 - 55 55:M ABD§ to recording electrode 80 69 75 76Suture line to recording electrode 22 - 20 33$Digit V to distal to sulcus 344Digit V to proximal to sulcus 452Distal to proximal to sulcus 108
*Partial lesion, not sutured.tDigit II instead of digit III; to proximal suture line of graft.§Abductor pollicis brevis muscle.
437
Protected by copyright.
on October 29, 2020 by guest.
http://jnnp.bmj.com
/J N
eurol Neurosurg P
sychiatry: first published as 10.1136/jnnp.42.5.436 on 1 May 1979. D
the threshold of the current to evoke a muscleaction potential. The muscle action potentialswere recorded via a 70 ,xm non-insulated steelwire passed through the muscle at the level of theendplate zone or, when only few motor units were
active, through a coaxial needle electrode adjustedto give maximum response.The surface temperature of the extremity was
kept at 34-360C.To obtain a gross expression of the number of
sensory nerve fibres activated, the amplitudes ofthe 10 to 50 component potentials were addedand denoted "cumulated amplitude." In ad-dition to the maximum sensory conductionvelocity the velocities of the last component inthe main burst and of late solitary or late burstsof discharges were determined.The electromyogram was recorded via con-
centric electrodes (DISA 13L51) as described byDahl and Buchthal (1978).
Results
RECOVERY AFTER SUTURE OF THE MEDIAN NERVE ATTHE WRIST
Amplitude and maximum conduction velocity ofthe sensory potentialIn the adult (BM, Appendix, patient 1) a sensory
potential was recorded for the first time fourmonths and in the two children five and seven
months after suture (Table 2). The potentialevoked by stimuli to the distal phalanx of digitIII consisted of 30-40 components which added toabout 1.5 txV (Fig. 1). In the child (HJ, Appendix,patient 3) in whom a graft was inserted theamplitude was much smaller. The amplitude re-corded after partial section (JJ, Appendix, patient4) was similar to that after complete section andsuture when stimuli were applied to digit IIIwhereas stimuli to digit I evoked a potential ofnormal size. The cumulative amplitude evoked bystimuli to the proximal phalanx was twice thatevoked by stimuli to the distal phalanx. After
total section and suture the eight componentsevoked by stimulation of digit I summed to anamplitude which was only one-tenth of that evokedby stimuli to the proximal phalanx of digit IIIalthough the distance of conduction from digit Ito the wrist was shorter (Table 1).
Thus, over a distance of about 160 mm, a
sensory potential could be recorded four monthsafter suture. As estimated from extrapolation ofFig. 2, it might have been possible to distinguisha potential from noise as early as three monthsafter suture. After this time the cumulative ampli-tude in the adult increased by 0.3 ,uV per monthto the twenty-fifth month. Thereafter the increasewas at a steeper rate, by 0.7 ,uV per month, to thefortieth month, when the amplitude approachedthe range of normal. Forty months after suturethe potential still contained five times as manycomponents as with nerves from control subjects,the largest component being less than one-tenthof the amplitude in controls (Fig. 3).The rate of the increase in amplitude was three
times steeper in the two children (LS and HJ,Appendix, patients 2 and 3) than in the adult, thecumulative amplitude increasing by 1.0 ,uV permonth until one year after suture when it levelledoff and increased only slightly to the sixteenthmonth; at this time it was 40-60%, and the ampli-tude of the largest component 10% of that incontrol subjects (Figs. 4, 5). The sensory potentialwas markedly more dispersed than normal. At thethirteenth and nineteenth months the number ofcomponent potentials was twice that at five andseven months after suture.Four to seven months after suture the fastest
component of the sensory potential evoked bystimuli to the distal phalanx of digit III was
conducted at 10-18 m/s, 15-25% of the velocityin nerves from control subjects. The velocity fromthe proximal phalanx to the wrist and thevelocities in the two children were about 5 m/sfaster (Table 2). The slowest components wereconducted at 3-4 m/s. The maximum velocity
Table 2 Earliest sensory and motor responses recorded or evoked proximal to the suture at wrist
Sensory potentials Motor responses
Patient* Age Time after Conduction Cumulative Peak-to-peak Number of Maximum Conduction Peak-to-peak Latency to(yr) suture distance amplitude amplitude components conduction distance amplitude abd$
*See Appendix; tdistal phalanx of digit III to wrist; $abductor pollicis brevis muscle; §graft; Ildistal phalanx of digit II to wrist; **partiallesion, no suture; ¶concentric electrode.
438
Protected by copyright.
on October 29, 2020 by guest.
http://jnnp.bmj.com
/J N
eurol Neurosurg P
sychiatry: first published as 10.1136/jnnp.42.5.436 on 1 May 1979. D
Fig. 1 Sensory potentialsevoked by stimulation of theproximal phalanx of digit IIIand recorded from the mediannerve at wrist at different timesafter end-to-end suture of thenerve (mo=months, n denotesnumber of responses averaged).The horizontal bar at 17 momarks an artefact caused by amicroreflex (Buchthal et al.,1975). The twin bars at 22 and29 mo denote trains of latecomponents conducted at 2-3mls, probably generated byimmature myelinated fibres.Upper scales: latencies in ms.Lower scales: conductionvelocities in mls. Patient BM,20 years old (cf Fig. 2 for theplot of the cumulativeamplitude, Fig. 6 for the plot ofthe maximum conductionvelocity, and Table I for thedistances of conduction).
n375
750
1500
125512
1024
256
512
1024
2048
4096
256
1 024
2048
4 096
256
5 1 2
2048
4096
%30IS 10IS 6 5 4so030
E
439
'T
AA&Am,
Protected by copyright.
on October 29, 2020 by guest.
http://jnnp.bmj.com
/J N
eurol Neurosurg P
sychiatry: first published as 10.1136/jnnp.42.5.436 on 1 May 1979. D
Fig. 2 Cumulative amplitudeof the sensory potentials as afunction of time after end-to-endsuture of the median nerve atwrist (right, patient BM) andafter pressure on the ulnarnerve at elbow (left, patientEOH). Perception of tactilestimuli. A: the patient coulddistinguish light touch frompinprick and localise thestimulus correctly. B: the patientcould distinguish light touchfrom pinprick but was unableto localise the stimuluscorrectly. C: the patient couldperceive touch as anuncharacteristic stimulus, wasunable to distinguish touch frompinprick and to localise thestimulus correctly. The shadedbars denote the averagecumulative amplitude fromcontrol subjects matched for age(right) and from thecontralateral ulnar nerve (left).The conduction distances aregiven in Table 1. The curveswere drawn by eye.
from digit I was 6 m/s, 10% of that in controlsubjects though the distance over which regenera-tion had occurred was nearly 40 mm shorter thanto the proximal phalanx of digit III (Fig. 6).Unlike the recovery of the amplitude of thesensory potential, the rate of recovery of maximumconduction velocity was initially greater thanlater, both in the adults and in the children(Figs. 6, 4). In the adult the velocity increased by3% per month from four to 22 months aftersuture, and thereafter by 0.4% per month. At40 months the maximum velocity was 65-75% ofthat in control subjects. In the nerve repaired by agraft the velocity remained low from five to eightmonths after suture. At the last studies, 13 and 19months after suture when tactile sensibility was
normal, the maximum velocity had recovered to40 and 60% of normal (Fig. 4).
In the adult patient with partial division of thenerve the main component of the sensory
potential was initially conducted at 10 m/s (15%of normal), the velocity increasing by 10% per
month to 30 m/s (47% of normal) eight monthsafter the lesion, when tactile sensibility hadreturned to normal.
Tactile sensibilityFrom four months to 22 months after suture,when the cumulative amplitude had increased 3.5times, the adult patient (BM) could perceive a
tactile stimulus but could not distinguish touchfrom pinprick or localise the stimulus (Fig. 2).Twenty-nine months after suture, when the ampli-tude had begun to increase more steeply and hadreached 40-60% of normal, the patient coulddistinguish touch from pinprick but was stillunable to identify the site of the stimulus. Notuntil the cumulative amplitude was near normal,more than three years after suture, were tactilestimuli perceived and localised normally.The child (LS), in whom the nerve was sutured
end-to-end, could distinguish touch from pinprickbut could not localise a stimulus seven monthsafter suture-that is, 22 months earlier than theadult. The cumulative amplitude was at that timea quarter of that in the adult, when sensibilityhad recovered to the same extent. Recovery oftactile sensibility was delayed by two months inthe child in whom a graft was inserted. Eightmonths after suture he was still unable to dis-tinguish touch from pinprick and to localise the
40 r
5
440
Protected by copyright.
on October 29, 2020 by guest.
http://jnnp.bmj.com
/J N
eurol Neurosurg P
sychiatry: first published as 10.1136/jnnp.42.5.436 on 1 May 1979. D
0 10 20 30 40 50 60 70 msIII I I I II I I/ \20 15 10' 8 7 6 5 4 3 2.5 m/s50 30 9
Fig. 3 Sensory potential evoked by stimulation of the distal (top) and the proximal(bottom) phalanx of digit III, recorded at wrist 40 months after end-to-end suture ofthe median nerve at wrist (n denotes number of responses averaged). Upper scale:latencies in ms; lower scale: conduction velocities in mls. The conduction distances aregiven in Table 1. Patient BM, 20 years old.
441
Protected by copyright.
on October 29, 2020 by guest.
http://jnnp.bmj.com
/J N
eurol Neurosurg P
sychiatry: first published as 10.1136/jnnp.42.5.436 on 1 May 1979. D
Fig. 4 Recovery of theamplitude of the sensorypotential and of sensory andmotor conduction after sutureof the median nerve in a 12(left, patient LS) and a 10 (right,patient HJ) year old child.Below: cumulative amplitude ofthe sensory potentials at wrist;above: maximum sensoryconduction velocity andreciprocal motor latency as afunction of time after end-to-endsuture (left) and insertion of agraft (right). In control subjects
20 matched for age the amplitudeof the sensory potential at wristaveraged 16 ,uV, when thedistal phalanx, and 40 ,uV, whenthe proximal phalanx of digitsIII or II was stimulated. Forthe definition of A, B, and Csee legend to Fig. 2. Theconduction distances are givenin Table 1.
20
stimulus. At 10 and 12 months after suture tactilesensation had returned to normal in both children-that is, two years earlier than in the adult(Fig. 4).
Muscle action potential and motor conductionRecorded via a wire electrode, the actionpotential of the abductor pollicis brevis musclewas initially 1% of the average in control sub-jects; the latency was three to six times pro-longed (Table 2). The time course of recovery ofamplitude and of latency was similar to that ofthe sensory potential and of sensory conductionvelocity (Figs. 2, 4, 6). In the adult (BM) theamplitude reached the range of normal at 40months when the latency was still 50% longer thannormal and the force was nearly normal (gradedas 4+). When the abductor pollicis brevis musclehad reached normal force in the two children, theamplitude of the action potential was 25% ofnormal in the child whose nerve was sutured end-
to-end (HS), and 15% of normal in the child inwhom a graft was inserted (HJ). In both, thelatency was still twice as long as in control subjects(Fig. 4).
Electrical activity during voluntary effort of theabductor pollicis brevis muscleAt the first study after suture, when there was novisible contraction, the pattern of electrical activitywas discrete and of low amplitude in the adultand in the two children (Fig. 7). As the forceapproached normal, the pattern of activity re-mained discrete and its amplitude increased. Themean amplitude of randomly sampled motor unitpotentials increased by 60-300% and the meanduration by 40-125%, and remained so even afternormal force had been attained (Fig. 8). Theduration was prolonged whether or not polyphasicpotentials were excluded from the calculation.The incidence of polyphasic potentials was two tothree times the upper limit of normal early after
6U)-J
I 5Z -OzoJ LL
zowu 0 31003
D Z
Xw;LhJ
<LU
I
w0
J-
-j0DIw
442
_
Protected by copyright.
on October 29, 2020 by guest.
http://jnnp.bmj.com
/J N
eurol Neurosurg P
sychiatry: first published as 10.1136/jnnp.42.5.436 on 1 May 1979. D
Fig. 5 Sensory potential evoked bylOmo stimulation of digit I and recorded at
wrist, at different times after end-to-endsuture of the median nerve in a 12 yearold child. The arrows denote the positivepeak of the first component, identifiedwith certainty; the figures above the
13 mo traces denote the maximum conductionvelocity. The initial large deflection(latency about 2 ms) originated from theradial nerve. The number of responses (n)that were averaged is given to the lef t.Patient LS, cf Fig. 4, lef t. The conductiondistance is given in Table 1.
19mo
0 5 10 15ms
-o
3
0
0-0 DIGM DIST.-WRIST0-0 DIG.3 PROX.-WRIST
A-A Dl1G.I-WRIST
WRIST- M.ABD.POLL.BR.LATENCY'
Fig. 6 Maximum sensory conductionvelocity along the median nerve fromdigit III and digit I to wrist and reciprocalmotor latency from wrist to abductorpollicis brevis muscle as a function oftime after end-to-end suture. Patient BM,20 years old. The conduction distancesare given in Table 1. The curves were
drawn by eye.
20 30MONTHS AFTER SUTURE
80 r-J
70 _0z
o 60_wll
z 50_w
cr-w
400z0H- 300
0z820
- 10x
443
0 10 40
Protected by copyright.
on October 29, 2020 by guest.
http://jnnp.bmj.com
/J N
eurol Neurosurg P
sychiatry: first published as 10.1136/jnnp.42.5.436 on 1 May 1979. D
A_1]-bilALiim ,L.LAljjj]i5iTTr_ IV _f1pqpliT ff- 0.5 mV
0 0.5 1s+
B.,.
0 0.5 s17 r 1 r 1 1718 _ 2~~~~~~~~I 5mV
0 0.5 ls
] 3mV
FJ G
0 20 40ms 0 5 lOims
HJlm~~JV
I mI0 4 8 12 ms
I
J
AFTERSUTURE
ImV 4mo
17 mo
-_jlmV 22 mo
llmV 29 mo
o 0.5 i
0 0.5 is
K]3mV -J ]2mV 40 mo
0 4 8 12ms
Fig. 7 Recruitment pattern during full voluntary effort of the abductor pollicis brevis muscle (A-E) andaction potentials (F-K) evoked by maximal stimuli to the nerve at wrist at different times after end-to-endsuture of the median nerve. The first response (F), evoked four months after suture, was recorded with acoaxial needle electrode, placed at the site of maximum response; G was recorded with a faster sweepspeed; H-K were recorded with a wire electrode that traversed the cross-section of the muscle. Patient BM,20 years old; cf the plot of the reciprocal latencies in Fig. 3 and the conduction distances in Table 1.
suture and decreased with time after suture(Fig. 8). Fibrillation potentials were recorded infour to nine sites of the muscle at all studies up to40 months after suture.
RECOVERY AFTER COMPRESSION OF THE ULNARNERVE AT ELBOW (APPENDIX, PATIENT EOH)Amplitude of the sensory potential and sensoryconduction velocityThree weeks after the episode of compression thecumulated amplitude of the sensory potentialrecorded proximal to the sulcus was 20% and thatof the largest component 15% of the amplituderecorded from the contralateral nerve. The num-ber of components was as in the contralateralnerve. The maximum conduction velocity from
proximal to distal to the sulcus was 74% of thatin the contralateral nerve. Hereafter the amplitudeincreased (Fig. 2), and 10 months after theepisode the cumulated amplitude and theamplitude of the largest component were normal(Fig. 9). The number of components increased aswell, by 25% after three months, and was twicethat in the contralateral arm in the last study,20 months after the episode (Fig. 10). Within oneyear after compression the maximal conductionvelocity increased to 90% of normal.
Tactile sensibilityThree weeks after the episode of compression thepatient complained of paraesthesia; the perceptionof light touch and pinprick was diminished in the
444
Protected by copyright.
on October 29, 2020 by guest.
http://jnnp.bmj.com
/J N
eurol Neurosurg P
sychiatry: first published as 10.1136/jnnp.42.5.436 on 1 May 1979. D
duration and the amplitude of motor unitpotentials sampled in abductor pollicisbrevis muscle during voluntary effort atdifferent times after end-to-end suture ofthe median nerve at wrist. The stripedcolumns indicate the incidence of
29 mo polyphasic potentials. Below: 13 controlsubjects, matched for age. The dashedlines represent the upper 95% confidencelimits. Patient BM, 20 years old.
40mo.lilillo II.n/
CONTROL
500 1000 1500pV
distribution of the ulnar nerve. At three months,when the cumulated amplitude was one-third ofnormal, the only sensory impairment was a
diminished sensibility for pinprick confined to theulnar surface of digit V. At four months tactilesensibility had fully recovered. The amplitude wasat that time two-thirds of normal.
Muscle action potential and maximum motorconduction velocityAt three weeks, stimuli proximal to the sulcusevoked a potential in the abductor digiti minimiwhich was 26% of that in the normal nerve and40% of the amplitude evoked by stimuli distal tothe sulcus in the impaired nerve. The maximum
40-
30-
20-
1 0-
31SI21
31
21
3
21
2
ms
445
Protected by copyright.
on October 29, 2020 by guest.
http://jnnp.bmj.com
/J N
eurol Neurosurg P
sychiatry: first published as 10.1136/jnnp.42.5.436 on 1 May 1979. D
conduction velocity from proximal to distal to thesulcus was diminished to 60 or 70% of normal(42 m/s, m. abductor digiti minimi; 36 m/s, m.adductor pollicis). At three months the amplitudeof the muscle action potential was nearly normal,and at four months the amplitude evoked bystimuli proximal to the sulcus was the same as thatevoked by stimuli distal to the sulcus. At thistime the maximum conduction velocity had in-creased by 10 m/s to 80% of normal. The latencyfrom proximal to sulcus to the flexor carpiulnaris muscle was normal throughout the study(3 ms, distance of conduction 100 mm).
Electrical activity during voluntary effortAt three weeks the force of the abductor digitiminimi muscle was slightly diminished (graded as4+). The only electromyographic abnormality was
Fig. 9 Left: recovery of thesensory potential after pressureon the ulnar nerve in the sulcus.The potentials were evoked bymaximal stimuli to digit V andrecorded at the elbow 50 mmproximal to the sulcus of theulnar nerve. The figures to theright of the traces indicateweeks after the lesion. Thefigures above each trace givethe maximum conductionvelocity across the sulcus (mls).Right: sensory potentialsrecorded from the contralateralulnar nerve; the figure above thelower trace indicates themaximum conduction velocityacross the sulcus. Patient EOH,33 years old. The conductiondistances are given in Table 1.
a slight reduction in the pattern of recruitmentduring full effort; the amplitude was normal(4 mV). At four months the amplitude of therecruitment pattern had increased to 6 mV, themean duration of motor unit potentials was26% prolonged, and polyphasic potentials andfibrillation potentials were still absent.
Discussion
Regeneration of peripheral nerve is associated witha change in amplitude, conduction velocity, andshape of the evoked sensory and motor actionpotentials. The change in maximum conductionvelocity reflects properties of those fibres thatgrow fastest and that have the most rapid speedof conduction-that is, it gives information ononly a small proportion of the myelinated nerve
nNORMAL
446
Protected by copyright.
on October 29, 2020 by guest.
http://jnnp.bmj.com
/J N
eurol Neurosurg P
sychiatry: first published as 10.1136/jnnp.42.5.436 on 1 May 1979. D
fibres. The amplitude of the sensory potential, onthe other hand, in normal nerve and in nerves withpurely axonal neuropathy is a measure of thenumber of fibres of 7 ,tm or more in diameter thatcontribute to the main phase of the sensory poten-tial. Thus, the peak-to-peak amplitude of thesensory potential of the sural nerve was grosslyrelated to the number of large myelinated fibres(Buchthal and Behse, 1977; Behse and Buchthal,1978). In regenerating nerve the peak-to-peak
WEEKS amplitude of the sensory potential is a poor gaugeAFTER of the number of nerve fibres, since the sensoryONSET potential is split into many components of similar
3 amplitude for several years after suture of thenerve. This is illustrated by the finding in Ballan-tyne and Campbell's (1973) and in our patients,
5 that the peak-to-peak amplitude of the sensorypotential increased little with the passage of time.The cumulative amplitude obtained by adding thecomponent potentials seemed a better gauge of
7 the number of nerve fibres that are activated thanthe peak-to-peak amplitude. We have determinedthe amplitude of the spike components ratherthan the area under the spikes. The spike duration
9 varied less in regenerating than in normal nervesince there was less summation, and the amplitudewas easier to measure than the area. The cumu-
1 1 lative amplitude is, however, only a gross approxi-mation as a measure of the number of fibres.Relatively more fibres contribute to a givenamplitude of a slow component originating fromfibres of small diameter than to the amplitude of
16 a fast component originating from fibres of largediameter (Gasser and Erlanger, 1927). We did notcorrect for the relation of amplitude with thesquare of the diameter, since it is not knownwhether it is valid for incompletely myelinated
21 fibres during regeneration. In normal nerve theamplitude of the five or six slow components re-corded at the wrist is at most 2% of the ampli-tude of the main component. Hence, the averagecumulated amplitude is about the same as the
31 peak-to-peak amplitude.Neither the amplitude of the sensory potential
nor the maximum sensory conduction velocitywas related to the improvement in two-point dis-
4 8 crimination (Payan, 1969; Almquist and Eeg-Olofsson, 1970; Ballantyne and Campbell, 1973)which is considered to be an accurate test to
83 Fig. 10 Slow components of the sensory potentials.Above: recorded from the contralateral normal ulnarnerve proximal to the sulcus. Below: recorded atdifferent times after pressure at elbow (weeks, figuresto the right). Patient EOH, 33 years old.
447
Protected by copyright.
on October 29, 2020 by guest.
http://jnnp.bmj.com
/J N
eurol Neurosurg P
sychiatry: first published as 10.1136/jnnp.42.5.436 on 1 May 1979. D
quantitate sensory loss (Moberg, 1962, Onne,1962). The threshold to touch was related repro-ducibly to the amplitude and conduction velocityof the sensory potentials during recovery afterlocal anaesthesia (Buchthal and Rosenfalck, 1966).Therefore-in the present study-we used near-threshold stimuli for touch in addition to pinprickto assess tactile impairment.
Extrapolated from the amplitude of the sen-sory potential four months after the lesion, ittook about 90 to 120 days before a sufficient num-ber of myelinated fibres had regenerated to give aresponse that could be distinguished from noise(>0.02 ,uV). This extrapolation seems justifiedbecause the same patient (BM) had been suturedpreviously, and four months after the first suturea sensory potential was absent. In nerves withnear-normal conduction velocity an amplitude ofthe sensory potential of 0.02 ,uV corresponds tothe activation of about 10 fibres of more than7 um in diameter. Assuming that the maximumconduction velocity of regenerating nerve de-creases linearly with the diameter of the largestfibres, as shown in the cat by Berry et al. (1944),and, assuming that the amplitude of the actionpotential in regenerated fibres is related to thecross-sectional area of the nerve fibres as in normalnerve (Gasser and Erlanger, 1927), it would takeabout 40 fibres to produce a response of 0.02 ,uVthat is conducted at 10-20 m/s. A delay of 90 to120 days for the first response to be measuredcorresponds to a rate of recovery of 2.0 to 1.5 mmper day assuming that the distance of regenerationextends 20 mm into the proximal stump (170+20 mm). This is in agreement with previous esti-mates (the advance of Tinel's sign and of return-ing sensibility, Seddon et al., 1943; Bowden andSholl, 1954). With percutaneous recording it took2.5-4 times longer before the first sensory potentialcould be distinguished from noise (Ballantyne andCampbell, 1973; Tallis et al., 1978).At the time when a sensory potential was just
recorded some degree of tactile perception hadbeen re-established. When light touch or pinprickwas applied to digit III the patient perceived astimulus but was unable to identify its type or siteof application. That tactile stimuli were perceivedat all, although the number of excitable largemyelinated fibres was presumably about 1% ofnormal, is consistent with the perception of astimulus applied to a single cutaneous afferentnerve fibre in man (Hensel and Boman, 1960) andwith the return of the perception of touch duringrecovery from local anaesthesia when the ampli-tude of the sensory potential is still markedlydiminished (Buchthal and Rosenfalck, 1966).
When a sensory potential could be recorded, themotor latency from just proximal to the site ofsuture to the abductor pollicis brevis muscle wasprolonged more than three times and the ampli-tude of the muscle action potential was less than1% of the average in control subjects. Electro-myography during voluntary effort showed evi-dence of regeneration as well, although clinicaltesting did not reveal activity in the muscle. Themarked increase in mean duration of motor unitpotentials and in the incidence of polyphasicpotentials reflects reinnervation.
Until two years after suture the cumulatedamplitude of the sensory potential increased slowlyby 2% per month reflecting the appearance ofadditional fibres and the increase in diameter ofregenerating fibres (Young, 1942). During thistime the perception of tactile stimuli remainedmarkedly impaired. At 28 months after suture thecumulated amplitude had recovered to half thatof normal nerve and tactile perception improvedsuch that touch could be distinguished from pin-prick but the site of the stimulus could still not berecognised. From two to three years after suturethe cumulated amplitude increased more rapidly(3% per month) and approached normal. Thepeak-to-peak amplitude was, however, only 6-7%of normal, consistent with the persisting markedtemporal dispersion of the sensory potential. Therewere 30-60 components compared to 5-7 withnormal nerve. The patient was now able to localisealso the site of the stimulus.The slow increase in the amplitude of the
evoked muscle action potential with time aftersuture reflects an incomplete synchronisation ofthe component potentials as well; at 40 months theamplitude was only half of that in control subjects,although the force was nearly normal.The maximum rate of sensory and motor con-
duction increased at first rapidly, by 3% permonth, and thereafter more slowly, by 0.3% permonth, similar to findings in cat (Berry et al.,1944) and rabbit nerve (Sanders and Whitteridge,1946). As the velocity of the fastest fibres in-creased, the slower components also moved tofaster velocities, and new slow components ap-peared that were absent in previous recordings.Even 40 months after suture recovery of the con-duction velocity of the fastest fibres was incom-plete, to 80% of normal, in agreement with mostprevious findings in animals (Berry et al., 1944;Cragg and Thomas, 1964) and in man (Ballantyneand Campbell, 1973). The distal motor latencywas at that time 1.5 times normal. The improve-ment in tactile perception occurred while therewas little change in maximum sensory conduction
448
Protected by copyright.
on October 29, 2020 by guest.
http://jnnp.bmj.com
/J N
eurol Neurosurg P
sychiatry: first published as 10.1136/jnnp.42.5.436 on 1 May 1979. D
velocity and confirms findings in patients investi-gated five years or longer after the suture (Alm-quist and Eeg-Olofsson, 1970).
Denervated muscle fibres are reinnervated eitherby regeneration or by sprouting of intramuscularnerve branches-that is, by the incorporation ofdenervated muscle fibres into motor units that al-ready have become reinnervated. Which of thetwo processes prevails depends on the speed atwhich regeneration proceeds. In rat soleus muscle,if regeneration occurs early during the sprouting,the synapses established by the sprouts are elimin-ated within a few weeks and the enlarged motorunits again become smaller. When regeneration isdelayed the motor units innervated by sprouts re-tain their increased size (Thompson and Jansen,1977; Thompson, 1978). The fact that full vol-untary effort in the reinnervated muscles of thepatients was associated with a pattern of discretedischarges, even after the force of the muscle hadbecome nearly normal, indicates that the nervefibres, regenerated earliest, had sprouted andthat the enlargement of the motor units persists.Because of the larger distance of regeneration,reinnervation is delayed in man and peripheralsprouting is of greater importance than in the rat.The fivefold increase in amplitude of the dis-charges during the first year of follow-up and theprolonged mean duration of motor unit potentialsare consistent with the assumption that the in-crease in force is due to an enlargement of theinitially reinnervated motor units by peripheralsprouting. The high incidence of polyphasic poten-tials which decreased with the passage of timeis presumably the result of slow conduction alongimmature sprouts, the temporal dispersion decreas-ing as the sprouts grow in diameter and becomefully myelinated. Possibly the transient lesser in-crease in the duration of motor unit potentials ob-served 10-17 months after suture reflects a periodof hyperinnervation (Guth, 1962).Hodes et al. (1948) observed that partially tran-
sected nerve, which did not require suture, oftenrecovered faster than completely transected andsutured nerve. In our patient with partial transec-tion of the median nerve the peak-to-peak ampli-tude increased three times faster with time afterthe lesion than did the amplitude of the sensorypotential during regeneration after complete sec-tion and suture. The increase in cumulative ampli-tude was, however, similar after partial and aftercomplete transection because after partial tran-section the temporal dispersion decreased whileit remained constant or increased after completetransection.
449
In the child in whom end-to-end suture was per-formed the delay until the first sensory and muscleaction potential appeared was similar to that in theadults (growth rate 2 mm per day); in the childin whom a nerve transplant was inserted and theregenerating fibres had to traverse the barrier of asecond suture, the delay was twice as long. Afterthe initial delay, the cumulative amplitude and themaximum sensory conduction velocity increasedthree times faster than in the adult nerve. Tactilefunction was normalised much earlier than in theadult. This confirms the findings of Onne (1962)and of Almquist and Eeg-Olofsson (1970). Thedistal motor latency was initially increased six toseven times; it recovered at the same fast rate asthe maximum sensory conduction velocity. Normalmuscle power was attained twice as fast as inadults.The five times faster rate of recovery in ampli-
tude of the sensory potential after transient localcompression than after transection and suture re-flects the faster repair of demyelination than offibre loss by regeneration. Nevertheless, it tookabout one year after compression until the cumu-lative amplitude reached normal and recovery wasstill proceeding, since there were twice as manyslow components in the sensory potential of thepreviously compressed nerve than of the normalnerve. Whether some of the slow componentsoriginate from regeneration of fibres that weredestroyed by the compression cannot be decided.They were, however, conducted at a faster velocitythan the slowest components, one to three yearsafter transection and suture. The maximum sen-sory conduction velocity across the region of com-pression was initially diminished by 25%; itremained at this level for eight months and seemeda poor indicator of recovery. Tactile perceptionhad fully recovered four months after the com-pression when the cumulative and the peak-to-peak amplitude of the sensory potential hadreached two-thirds of normal. The much earlierrecovery of tactile perception after compressionthan after transection is probably due to the ab-sence of neuronal branching and of misalignmentof regenerating fibres.
We are indebted to the late Mrs Ulla Scheel and toMr K. Dahl for their able help during the study,and to Dr S. Horowitz, Professor A. Rosenfalck,and Dr W. Trojaborg for helpful criticism of themanuscript. The work was supported by grantsfrom the Muscular Dystrophy Associations ofAmerica, New York, and the Michaelsen Founda-tion, Copenhagen.
Protected by copyright.
on October 29, 2020 by guest.
http://jnnp.bmj.com
/J N
eurol Neurosurg P
sychiatry: first published as 10.1136/jnnp.42.5.436 on 1 May 1979. D
PATIENTS WITH INJURY OF THE MEDIAN NERVE AT THEWRISTCase I A 20 year old female accountant (BM) in-curred, in a traffic accident, a fracture of the radius,the ulna, and the triquetrum at the wrist with luxationof the joint. End-to-end sutures of the ulnar andmedian nerves were performed a few hours after theaccident. Two weeks later a neuroma was removedfrom the ulnar nerve. Inspection of the median nervedid not show a neuroma. At four months neithertouch nor pinprick was perceived in the area of themedian nerve, the force of the abductor pollicis brevismuscle was graded as 0, motor unit potentials wereabsent during voluntary effort, and there were fibril-lation potentials and positive sharp waves in five sites.At five months the median nerve was re-explored andfound to be buried in fibrous tissue. A neuroma wasremoved and a new perineurial end-to-end suture wasperformed. Three months later the pain had subsided.The patient noticed when the hand and the proximalphalanx of the median-innervated digits were touchedbut could describe neither the site nor the type of thestimulus. The force in the abductor pollicis brevismuscle was graded as 0. The time course of recoverywas followed from the time of the second suture until40 months later.Case 2 A 12 year old girl (LS) fell over a tent poleand cut her wrist. Two days later the wrist was ex-plored, the median nerve was found to be severed,and a perineurial end-to-end suture was performed.Her recovery was followed over 19 months. At sixweeks she could perceive a stimulus in the palm of thehand, but not on the digits innervated by the mediannerve. There were paraesthesias but no pain. The forceof the abductor pollicis brevis muscle was graded as 0.At six months she still complained of paraesthesiain digits II and III. She perceived a stimulus on theproximal phalanx of digits II and III but was unableto distinguish touch and pinprick and to localise thestimulus. The force of the abductor pollicis brevismuscle was graded as 0 to 1. Ten and 13 months aftersuture, touch of the digits elicited pricking paraes-thesia lasting for several seconds; there was no pain.The force of the abductor pollicis brevis muscle wasgraded as 1. At 19 months pinprick elicited prickingand tingling, less than one second in duration. Theforce of the abductor pollicis brevis muscle was nearlynormal (graded as 4+).Case 3 A 10 year old boy (HJ) fell on a glass doorand injured his right wrist. Tactile stimuli were notperceived in the distribution of the median nervealthough the nerve showed continuity one day afterthe injury. Seven months after the injury tactile sen-sibility was absent, and touch of the scar elicited painirradiating to the digits innervated by the mediannerve. A neuroma was removed and a 40 mm longgraft from the sural nerve was inserted by epineurialsutures. The proximal suture line was 150 mm fromthe tip of digit II. The pain subsided. His recovery wasfollowed for 13 months after the suture.
Case 4 A 19 year old assistant nurse (JJ) incurreda 10 mm deep incision at the wrist when she fellagainst a window. The tendon of the palmaris longusmuscle was sutured. The median nerve was not exposedsince tactile sensibility of the hand and movement ofthe thumb seemed unimpaired. One week after thelesion the sensibility for touch was diminished in thedistribution of the median nerve; pinprick was nor-mally perceived. There was neither pain nor paraes-thesia. Five months after the lesion the sensibility fortouch and pinprick was diminished in the distributionof the median nerve, and tactile stimuli elicited prick-ing paraesthesia in digits I to IV but no pain. Theforce and the electromyogram of the abductor pollicisbrevis muscle were normal. Eight months after thelesion the paraesthesia had subsided. The perceptionof touch and pinprick was diminished in the distalphalanges of digits II, III, and IV (radial side). Electro-physiological studies were obtained five and eightmonths after the injury.
PATIENT WITH COMPRESSION OF THE ULNAR NERVE ATTHE ELBOWA 33 year old carpenter (EOH) was seen three weeksafter he had "slept" on his right elbow in a drunkenstupor. He was not an habituated alcoholic. He com-plained of paraesthesia in digits IV and V. The per-ception of touch and pinprick was diminished in digitV and on the ulnar side of digit IV. The force of theabductor digiti minimi muscle slightly reduced (gradedas 4+). Five weeks after the episode of compression,the paraesthesia in digits IV and V and the diminutionin sensibility of touch and pinprick were less marked.The force of the abductor digiti minimi muscle was stillslightly reduced. At seven weeks the paraesthesia wasconfined to digit V and the decrease in sensibility fortouch and pinprick to the ulnar surface of digit V.At nine and 11 weeks there was slight paraesthesia indigit V, and the diminution in tactile sensibility on theulnar surface of digit V was less than at seven weeks.The force of the abductor digiti minimi muscle wasnormal. At four months the paraesthesia had subsided,and the only abnormality was a slight diminution inthe perception of pinprick on the ulnar surface of thedistal phalanx of digit V. At five, seven, 12, and 21months the tactile sensibility and the force of theulnar-innervated muscles were normal.
References
Almquist, E., and Eeg-Olofsson, 0. (1970). Sensorynerve conduction velocity and two-point discrimina-tion in sutured nerves. Journal of Bone and JointSurgery, 52-A, 791-796.
Andersen, V. O., and Buchthal, F. (1970). Low noisealternating current amplifier and compensator to re-duce stimulus artefact. Medical and BiologicalEngineering, 8, 501-508.
Ballantyne, J. P., and Campbell, M. J. (1973). Electro-physiological study after surgical repair of sectionedhuman peripheral nerves. Journal of Neurology,Neurosurgery, and Psychiatry, 36, 797-805.
450
Protected by copyright.
on October 29, 2020 by guest.
http://jnnp.bmj.com
/J N
eurol Neurosurg P
sychiatry: first published as 10.1136/jnnp.42.5.436 on 1 May 1979. D
Behse, F., and Buchthal, F. (1978). Sensory actionpotentials and biopsy of the sural nerve in neuro-pathy. Brain, 101, 473-493.
Berry, C. M., Grunfest, H., and Hinsey, J. C. (1944).The electrical activity of regenerating nerves in thecat. Journal of Neurophysiology, 7, 103-115.
Bowden, R. E. M., and Sholl, D. A. (1954). InPeripheral Nerve Injuries, Medical ResearchCouncil, Special Report Series, no. 282, pp. 16-24.Edited by Her Majesty's Stationery Office: London.
Buchthal, F., and Rosenfalck, A. (1966). Evokedaction potentials and conduction velocity in humansensory nerves. Brain Research, 3, 1-122.
Buchthal, F., and Rosenfalck, A. (1971). Sensorypotentials in polyneuropathy. Brain, 94, 241-262.
Buchthal, F., and Behse, F. (1977). Peroneal muscularatrophy (PMA) and related disorders. I. Clinicalmanifestations as related to biopsy findings, nerve,conduction and electromyography. Brain, 100,41-66.
Buchthal, F., Rosenfalck, A., and Behse, F. (1975).Sensory potentials of normal and diseased nerves.In Peripheral Neuropathy, chapter 21, pp. 442-464.Edited by P. 0. Dyck, P. K. Thomas and E. H.Lambert. W. B. Saunders: Philadelphia, London,Toronto.
Cajal, Ramon y, S. (1928). Degeneration and Regen-eration in the Nervous System, pp. 1-769.Humphrey Milford: London.
Cragg, B. G., and Thomas, P. K. (1964). The conduc-tion velocity of regenerated peripheral nerve fibres.Journal of Physiology (London), 171, 164-175.
Dahl, K., and Buchthal, F. (1978). Digital memory re-corder in electromyography and nerve conductionstudies. Electroencephalography and Clinical Neuro-physiology, 45, 538-544.
Erlanger, J., and Schoepfle, G. M. (1946). A study ofnerve degeneration and regeneration. AmericanJournal of Physiology, 147, 550-581.
Frey, M. von (1896). Untersuchungen uber die Sin-nesfunktionen der menschlichen Haut. Abhandlun-gen der Koniglich Sachsischen Gesellschaft derWissenschaften, Mathematisch-Physikalische Klasse,3, 218-238.
Gasser, H. S., and Erlanger, J. (1927). The roleplayed by the constituent fibres of a nerve trunkin determining the form of its action potentialwave. A merican Journal of Physiology, 80, 522-547.
Gilliatt, R. W., Fowler, T. J., and Rudge, P. (1975).Peripheral neuropathy in baboons. In Advances inNeurology, volume 10, pp. 253-272. Edited by B. S.Meldrum and C. D. Marsden. Raven Press: NewYork.
Guth, L. (1956). Regeneration in the peripheral ner-vous system. Physiological Reviews, 36, 441-478.
Head, H. (1920). Studies in Neurology, pp. 1-262.Oxford Medical Publications: London.
Hensel, H., and Boman, K. K. A. (1960). Afferentimpulses in cutaneous sensory nerves in humansubjects. Journal of Neurophysiology, 23, 564-578.
Hiscoe, H. B. (1947). Distribution of nodes and in-cisures in normal and regenerated nerve fibres.Anatomical Record, 99, 447-475.
Hodes, R., Larrabee, M. C., and German, W. (1948).The human electromyogram in response to nervestimulation and the conduction velocity of motoraxons. Archives of Neurology and Psychiatry(Chicago), 60, 340-365.
Hursh, J. B. (1939). Conduction velocity and diameterof nerve fibres. A merican Journal of Physiology,127, 131-139.
Jacobson, S., and Guth, L. (1965). An electrophysio-logical study of the early stages of peripheral nerveregeneration. Experimental Neurology, 11, 48-60.
Medical Research Council (1976). Aids to the Examin-ation of the Peripheral Nervous System, pp. 1-62.Her Majesty's Stationery Office: London.
Moberg, E. (1962). Criticism and studv of methodsfor examining sensibility in the hand. Neurology(Minneapolis), 12, 8-19.
Onne, L. (1962). Recovery of sensibility and sudomotoractivity in the hand after suture. Acta ChirurgicaScandinavica, Supplement 300, 1-69.
Payan, J. (1969). Electrophysiological localization ofulnar nerve lesions. Journal of Neurology, Neuro-surgery, and Psychiatry, 32, 208-220.
Sanders, F. K., and Whitteridge, D. (1946). Con-duction velocity and myelin thickness in regenerat-ing nerve fibres. Journal of Physiology (London),105, 152-174.
Schroder, J. M. (1972). Altered ratio between axondiameter and myelin sheath thickness in regeneratednerve fibres. Brain Research, 45, 49-65.
Schroder, J. M. (1975). Degeneration and regenerationof myelinated nerve fibres in experimental neur-opathies. In Peripheral Neuropathy, chapter 16,pp. 337-362. Edited by P. J. Dyck, P. K. Thomas,and E. H. Lambert. W. B. Saunders: Philadelphia,London, Toronto.
Seddon, H. J., Medawar, P. B., and Smith, H. (1943).Rate of regeneration of peripheral nerves in man.Journal of Physiology (London), 102, 191-215.
Struppler, A., and Huckauf, H. (1962). Propagationvelocity in regenerated motor nerve fibres. Electro-encephalography and Clinical Neurophysiology, Sup-plement, 22, 58-60.
Tallis, R., Staniforth, P., and Fisher, T. R. (1978).Neurophysiological studies of autogenous suralnerve grafts. Journal of Neurology, Neurosurgery,and Psychiatry, 41, 677-683.
Thompson, W. (1978). Reinnervation of partially de-nervated rat soleus muscle. A cta PhysiologicaScandinavica, 103, 81-91.
Thompson, W., and Jansen, J. K. S. (1977). Theextent of sprouting of remaining motor units inpartly denervated immature and mature rat soleusmuscle. Neuroscience, 2, 523-535.
Vizoso, A. D., and Young, J. Z. (1948). Internodallength and fibre diameter in developing and regen-erating nerves. Journal of Anatomy, 82, 110-134.
Young, J. Z. (1942). The functional repair of nervoustissue. Physiological Reviews, 22, 318-374.
451
Protected by copyright.
on October 29, 2020 by guest.
http://jnnp.bmj.com
/J N
eurol Neurosurg P
sychiatry: first published as 10.1136/jnnp.42.5.436 on 1 May 1979. D
