- Email: [email protected]
Contribution of collateral sprouting to the sensory and sudomotor recovery in thehuman palm after peripheral nerve injury
British Journal of Plastic Surgery (1998), 51,436-443 9 1998 The British Association of Plastic Surgeons
BRITISH JOURNAL OF ~
PLASTIC SURGERY
Contribution of collateral sprouting to the sensory and sudomotor recovery in the human palm after peripheral nerve injury U. Ah~an*, Z. M. Arn6z*, E BajroviC and M. Janko*
* University Department of Plastic Surgery and Burns; ~University Institute of Clinical Neurophysiology, University Medical Centre; t Institute of Pathophysiology, Medical Faculty, Ljubljana, Slovenia SUMMARY. The contribution of collateral sprouting to the sensory and sudomotor recovery was studied in 52 patients aged 3-66 years (mean 35.5 years) from 2 to 9 years following nerve injury and repair. The study included three groups of patients: (1) patients with complete division of median and ulnar nerves (skin reinnervation exclusively due to axon regeneration), (2) patients with isolated division of ulnar or median nerve (skin reinnervation due to axon regeneration and possible collateral sprouting), and (3) patients in whom injured axons failed to regenerate (skin reinnervation exclusively due to collateral sprouting). The end stage of sensory and sudomotor recovery was studied by: clinical methods, sensory nerve action potential (SNAP) measurements, sympathetic skin response (SSR) and the ninhydrin test. We found that recovery of sensory and sudomotor function in groups 1 and 2 was similar. End-stage sudomotor and sensory recovery within the autonomous area of the nerve did not depend on possible collateral reinnervation. Collateral reinnervation from the uninjured nerve was limited to the border innervation area of the palm and ring finger. Adjacent uninjured nerve may contribute to sprouting of nociceptive axons providing a protective function.
Cutaneous sensory and sudomotor reinnervation following peripheral nerve injury is believed to be due to: regenerative growth of the axons from the proximal stump of injured nerve (axon regeneration), collateral sprouting of axons from the adjacent uninjured nerve and sprouting of axons from the adjacent injured nerve (expansive regenerative reinnervation). All three processes are well known from animal experiments. '-3 When injured axons fail to regenerate, collateral sprouting of adjacent uninjured axons is the only explanation for functional recovery. However, in the majority Of clinical cases the regeneration of axons from the proximal stump of the injured nerve is not prevented and functional recovery occurs mostly due to ingrowth of regenerating axons. 4 The recovery of cutaneous nerve function in the human palm after peripheral nerve injury is well documented, but the contribution of collateral sprouting of axons from the uninjured adjacent nerves to the recovery of sensory and sudomotor function in the partially denervated human palm has not been assessed. Collateral sprouting of uninjured axons in response to partial skin denervation has been studied extensively in animals. It has been shown that in neonatal animals all types of cutaneous axons have the ability to sprout and that different types of axons differ in their capacity to extend sprouts into adjacent denervated skin/,5 However, only high threshold nociceptive afferents (A~) 5'6 and sudomotor axons (C) 7 seem to have the capacity for collateral sprouting in adult animals. Despite the lack of histological proof, other evidence that collateral sprouting also occurs in humans is accumulating. Besides collateral sprouting of nociceptive and sudomotor axons, clinicals-'3 and
electrophysiologicaP* findings indicate that low threshold mechanoreceptors may also have the capacity for collateral reinnervation in humans. There is some evidence to suggest that collateral sprouting of sensory axons does occur in human palm t o o . 8,9A1A3 However, we are not aware of any report regarding possible collateral sprouting of sudomotor axons in that skin area. Devor and colleagues' reported that collateral sprouting of uninjured nerves plays an early role in reinnervation of rat foot skin after nerve injury. This leads to an initial reduction in anaesthetic area due to peripheral nerve injury and is followed by ingrowth of the injured axons. A similar pattern of recovery has been described after inferior alveolar nerve injury in man. '~ Because of intermingling of collateral sprouting of axons from the uninjured nerve and regeneration of axons from the injured one, we would expect a better functional recovery than in the case where both of the palmar nerves are injured and functional recovery occurs only due to ingrowth of regenerating axons. Some functional recovery might be expected even when injured axons failed to regenerate and collateral sprouting of adjacent uninjured nerve is the only possibility for functional recovery. The present study is based on three different groups of patients: (1) skin reinnervation exclusively due to axon regeneration (complete division of both median and ulnar nerves), (2) skin reinnervation due to axon regeneration and possible collateral sprouting (patients with isolated division of the ulnar or median nerve), and (3) skin reinnervation exclusively due to collateral sprouting (patients in whom injured axons failed to regenerate). 436
Contribution of collateral sprouting The study was undertaken in order to estimate the clinical importance of collateral sprouting in the h u m a n palm.
Material The study was approved by the National Committee for Medical Ethics and included three groups of patients.
Group 1. Patients with isolated sharp division and primary microsurgical repair of the ulnar (15 patients) or median nerve (20 patients), in whom collateral nerve regeneration from the uninjured side was possible (35 patients, 28 males and 7 females from 3 to 66 years - mean 29.8 years). Group 2. Injuries which precluded collateral sprouting: complete division and primary microsurgical repair of median and ulnar nerves (11 patients, 7 males and 4 females aged from 6 to 56 years - mean 28.2 years). Group 3. Six patients in whom injured axons failed to reinnervate the denervated area and collateral sprouting might be the only possibility for functional recovery. All six patients aged from 24 to 56 years (mean 42.7 years) visited our clinic 2-9 years after injury and treatment in other hospitals. In three patients we found neuromas at exploration. Another patient came to our clinic 9 years after injury and 4 years after a long non-functional neuroma was resected. In a 24-year-old patient with transection of the ulnar nerve in the axilla by knife, despite complete sensory loss in the ulnar nerve innervating area only skin was sutured. A 44-year-old patient with transection of the ulnar nerve in lower third of the forearm came after several neurolyses performed at the local hospital.
437 above the one producing m a x i m u m response. The skin temperature was kept above 31 ~ The latency and the amplitude of the response were measured.
Clinical testing of sensation. The patients were tested for the light touch, pain, static and dynamic two point discrimination in the zones of the palm and palmar surface of the finger using standard techniques (Fig. 1). For uniform assessment of clinical results we used modified Medical Research Council classification. 1~,16 We agree with most authors who believe that sensory return $3+ to $4 are needed ~5to have good to excellent clinical results. Evaluation of sudomotor nerve fibre regeneration Recording of SSRs. The SSRs were elicited and recorded using a standard technique.~7.~8 A two channel E M G machine Mystro, Medelec MS 20 was used. The stimulating electrode (13L 36, D A N T E C , Skovlunde, Denmark) was attached to the palmar surface of the wrist. An earthing electrode was applied to the opposite wrist. A single square electrical impulse of 80~150 V amplitude and 0.2 ms duration was delivered near the end of inspiration to evoke the SSR. Surface disc electrodes were attached at the centre of the distal phalanx of the little and index fingers (recording electrode) and dorsum of the same hand (reference electrode) bilaterally. In patients without regenerative growth of the
,,:.:.:,:..
....,,.,.,...,. ,,................
::::::::::::::::5:. ,....................
z,d e
Methods
Evaluation of sensory nerve fibre regeneration Recording of SNAPs. Antidromic sensory nerve action potentials (SNAPs) were recorded from the index and/or little finger after sutured ulnar and/or median nerve and from the corresponding digit of the uninjured hand. A pair of standard ring electrodes (Digital ring electrodes 16639, M E D E L E C , U K ) was placed around the proximal and distal interphalangeal joints. A stimulating electrode (Bipolar stimulating electrode 16893, M E D E L E C , U K ) was attached to the wrist 3 cm proximal to the distal crease of the wrist longitudinally along the course of the median or ulnar nerve. The ground electrode was attached to the same wrist close to the site of stimulation No. 1. SNAPs elicited by stimulation of the ulnar nerve at the wrist were recorded from the little finger, and SNAPs evoked by median nerve stimulation at the wrist were recorded on the index finger. Ten responses were averaged during each stimulation procedure. These recordings were made in a standard way on the E M G machine (Mystro, Medelec MS 20). The stimulus pulse duration was 0.1 ms and the stimulus frequency 0.5 Hz. The stimulation strength was slightly
Figure 1--Distribution of median and ulnar nerves, autonomous zones and points of the SSR and SNAP measurement. Note: Distribution of median and ulnar nerves of the palm and palmar surface of the fingers (line) divided into 7 zones (squares), areas of absolute sensory supply of the median and ulnar nerves autonomous zones (shadow). Position of surface disc electrode for the SSRs measurements (circles) and ring electrodes for the SNAPs recording (bold line).
438 injured nerve (axon regeneration) the disc electrodes were attached to the border innervation area on the palm (Fig. 1). Skin temperature in all patients was kept above 31~ The measurements were taken simultaneously on both hands. At each site four consecutive responses were recorded to ensure reproducibility. The peak-topeak amplitude and area (integrated surface below and above the baseline) of each response were measured. Left to right ratios of area of the responses were calculated and expressed as percentages.
Ninhydrin test. Spontaneous sweat secretion was observed and documented by the ninhydrin test. The patient placed his clean and dry palm and fingers on a sheet of white chromatographic paper and pressed gently. The paper was passed several times through a 1% ninhydrin solution in acetone. The sheet was then heat dried in an incubator at 110~ The results on the injured side were compared to those on the normal side. They were classified by the intensity of the print pattern: normal sweating (+ + +), diminished sweating (+ +) and deficient (+) or absent sweating (-). All patients were asked for history of post-traumatic cold intolerance. All studies were repeated on the corresponding nerves of the contralateral normal finger so that each parameter could be matched with a control value obtained from the same subject.
Results
Group 1 (patients where collateral nerve regeneration could be excluded) Five of 11 patients (in median nerve distribution) and 4 of 11 (in ulnar nerve distribution) had good results (sensory return of $3+ to $4 and diminished to normal sweating). SNAPs were present in 4 of 11 patients and the SSR area on the injured site reached 60% or more when compared to the uninjured site in 5 of 11 patients (Fig. 2).
Group 2 (patients with possible collateral nerve regeneration) Six of 15 patients with complete laceration of the ulnar nerve and 9 of 20 with complete laceration of the median nerve returned to $3+ or $4 and had diminished or normal sweating. SNAPs were present in 8 of 20 patients with median nerve injury and 6 of 15 patients with ulnar nerve injury. The SSR area on the injured site reached 60% or more when compared to the uninjured site in 10 of 20 patients with median nerve injury and 8 of 15 patients with ulnar nerve injury. A virtually normal pattern of sweating was observed in 6 of 15 ulnar nerve repairs and 10 of 20 median nerve repairs. Thirty of 35 patients could recognise different sensation on the ulnar or median half of the ring finger even when the final stage of nerve regeneration was reached. In a high percentage of patients post-traumatic cold intolerance was an important factor causing disability (Table 1).
British Journal of Plastic Surgery
Group 3 (patients in whom injured axons failed to regenerate) In all 6 patients only the ring finger and border innervation area of median or ulnar nerve territory on the palm (up to 15 mm) (Fig. 4), had good sensory regeneration ($3). In all remaining ulnar or median nerve distribution in the palm (especially in the autonomous zone) only deep pain sensibility (response to pin prick-S1) recovered. In the same area of the palm and fingers, the skin was dry, smooth and slippery. Spontaneous sweat secretion in those areas was not observed by the ninhydrin test. The lack of sweating corresponded well to the sensory deficit in all patients. No patients had recordable SNAPs or SSRs and all complained of post-traumatic cold intolerance (Figs 3 and 4, Table 1). Discussion
After peripheral nerve injury, the skin may be reinnervated by regenerating axons from injured nerve fibres and from adjacent uninjured nerves (collateral sprouting). Previous clinical studies of cutaneous reinnervation following nerve injury have mainly been performed to assess the regenerative capacity of injured cutaneous nerves. Our study was undertaken to estimate clinical importance of collateral sprouting of sensory and sudomotor axons for functional recovery in human palm in two situations: first, when injured nerve failed to regenerate and collateral sprouting of uninjured axons was the only possibility for functional recovery; and second, in cases where regeneration of injured nerve was not prevented and function recovered due to intermingling of both processes (regeneration of injured cutaneous nerves and collateral sprouting of uninjured nerves). We expected better sensory and sudomotor function recovery in patients with just one injured nerve (median or ulnar) since functional recovery of autonomous innervation territory was possible due to collateral sprouting and axon regeneration. However, in the present study, the results of sensory and sweating recovery in the autonomous area of the ulnar and median nerves were similar regardless of possible collateral sprouting. Furthermore, the borders of touch sensitive area in subjects where axon regeneration of injured ulnar or median nerve had failed, extended minimally over the boundaries of the normal innervation territories, determined clinically and electrophysiologically by the others, n,~3Therefore, we suggest that collateral reinnervation of low threshold receptors from the uninjured ulnar and median nerve does occur in the human palm, but is limited to the border innervation area of these two nerves. The lack of improvement of functional outcome in case of predicted collateral sprouting may be due either to a lack of collateral sprouting of low threshold sensory and sudomotor fibres, or withdrawal of collateral sprouts. Experimental studies of collateral reinnervation of the skin of mammals have yielded the rather surprising result that only high threshold pain afferents and sudomotor axons in the A5 and C fibre range, respectively, seem to have the capacity for collateral
Contribution of collateral sprouting
439
Sensory nerve action potential (SNAP) and sensation
1~---/' ~
~
s . AP r f o o
SNAP tkom the uninjured hand
f
1
...I"
SNAP from the injured hand
A
Sympathetic skin response (SSR)
SSR from the inJured hand SSR from the uninjured hand
SgR fiom the th/ured hand SSR from the uninjured hand
~
_ _
SSR fhom the m mredhand SSR from the uninjured hand
SSR from Ihe injured hand g g R from the uninjured hand
SSR from the injured hand SSR from th9 uninjured hand
SgR from the injured hand SSR from the uninjured hand
B Figure 2--(A, B) A 17-year-oldpatient with complete division of median and ulnar nerves by glass at mid third of the forearm. Five years after injury and primary epineural repair a complete sensory recovery($4) is noticed within ulnar innervation area and S3+ within median nerve area. The patient could recognise different sengation on the ulnar half of the ring finger. Sensory nerve action potentials were present. Normal pattern of sweating and normal SSRs show an excellent sudomotor recovery.
sprouting in adult animals? -7 Electrophysiological and histological studies of collateral sprouting in the skin of various mammals; cat, 19rabbit 2~and raC failed to demonstrate sprouting of low threshold mechanosensory cutaneous afferents (A[3) after partial denervation of skin in adult animals. In clinical studies we do not often encounter the favourable conditions present with laboratory animals. The traumatic wound is seldom as clean cut as the surgical one. Infection, delay of healing, extensive soft tissue damage, vascular injury, all promote scar formation. Nerve ends are often widely separated and occasionally require suture under greater than optimal
tension. The accidental wound may divide a nerve at the point where it divides into separate branches. 22 Furthermore, evaluation of axon regeneration in animal experimental studies mostly includes: nerve pinch test, skin pinch test, axon counting, immunohistochemical procedures and electron microscopy. These provide interesting morphological data, but poor information about functional results and abnormal sensory symptoms. In contrast to animal studies, there is some clinical evidence for collateral reinnervation of low threshold receptors occurring in the skin of the human palm ~s,H and other r e g i o n s J 2,23,24 Despite lack of histological
440
British J o u r n a l o f Plastic S u r g e r y
Table 1
Results after m e d i a n a n d / o r u l n a r n e r v e repair, a n d in p a t i e n t s w i t h n o a x o n r e g e n e r a t i o n
Injury
Sensation Recovery $3+ or $4 (in %) U M
Measurable SNAP (in %) U M
SSR Over 60% (in %) U M
Ninhydrin test ++ or +++ (in %) U M
CoM Intolerance (in %) U M
Division of M or U nerve Division of M and U nerve No axon regeneration
40
45
40
40
50
53
40
50
67
70
36
45
36
36
45
45
36
45
64
64
100
100
0
0
0
0
M = median, U = ulnar.
Sensory nerve action potential (SNAP)
SNAP from the injured nerve
SNAP from the injured nerve
t
SNAP from the injured nerve
A
Sympathetic skin response (SSR) ....... .........~/'~
SSR from the injured nerve
SSR from the injured nerve
j
SSR from the uninjured nerve
SSR from the uninjured nerve SSR from the injured nerve
SSR from the uninjured nerve
SSR from the uninjured nerve SSR from the uninjured nerve
SSR from the injured nerwe
J~..~""~-'-~"
S S R f i ~ m the t m l n j u r e d nerve
ssR
B
fromth~~anj.~od,~e
.....
~.~,
_ / f_- - . . . ~ , ~ .
~
~
_......_.X~f/_~.._._~ . . . . . ---..__0
~ .
.
.
.
-
--
SSR from the uninjured nerve
~,...t,&~
SSR from the I,jurea , ~ , , e
~
SSR from the uninjured nerve
A, B, 1, 2, 3, 4, 5 = measurement points
Figure 3---(A, B) A 56-year-old patient with incomplete amputation at wrist level in 1988. Median nerve was repaired by primary suture and ulnar nerve was intact. Since the patient had complete sensory and sudomotor deficit within the median innervation area, a 6 cm long nonfunctional neuroma was resected and sural nerve grafts sutured in a year later. Six years after repair only sensation of deep pain recovered (S 1). The collateral sprouts reinnervated just the border innervation area o f the palm and ring finger. The sensory nerve action potentials on the first, second and third fingers were absent. The lack of sweating corresponded well to the sensory deficit. The SSR was absent on measurement points B, 1, 2 and 3, slightly diminished on measurement point 4 and normal on measurement points 5 and A. Upper traces are from injured hand and lower from the uninjured hand.
Contribution of collateral sprouting
441
iiiiiiii?
Figure 4~Area of collateral reinnervation in the palm and palmar surface of the ring finger. Note: In the group of patients in whom injured axons failed to reinnervate the denervated area and collateral sprouting might be the only possibility for functional recovery,the area of good sensation ($3) was limited to the border innervation area of the palm and ring finger (shadow). Reinnervation from the ulnar side - black arrows and from the median side - white arrows. proof, this collateral reinnervation is attributed to sprouting of adjacent uninjured nervesY '~-' On the other hand, synaptic unmasking and collateral sprouting in the spinal cord were suggested to explain such an effect. 25'26Central mechanisms cannot be definitely excluded? 7 When cutaneous nerve collateral sprouting was studied in adults in whom a forearm cutaneous nerve had been resected, the objective sensory test showed encroachment at margin by only 6 m m for light touch, 7 m m for sharp pain and 11.5mm for heat pain, with no significant change for warming or cooling. Collateral sprouting of sudomotor C fibres resulting in return of sweating was more extensive. 12 Transmedian cutaneous reinnervation in patients who had undergone trigeminal sensory root section was explained by collateral sprouting of sensory nerves. The mean extent of transmedian cutaneous innervation varied from about 2 m m in the middle of tongue to 9 m m on the lower lip. In all areas the extent was greater to pin prick than to light touch stimulus? 3 The extension of collateral sprouts was shown to be limited in these studies but the mechanism responsible for this limitation is not known. In some other adult clinical s t u d i e s ~2,23,z4 the collateral sprouting of low threshold sensory fibres is suggested to be more extensive, although the capacity of low-threshold afferents may differ with regard to anatomical location. It should be pointed out that in a traumatic wound or after nerve graft harvest, small branches or interconnections of neighbouring nerve are also injured, in addition to the principal nerve injury. In such circumstances, in which the nerves adjacent to the denervated area are themselves also injured, and during regenerative spread into the adjacent territory, the term 'expansive regenerative reinnervation' has
been proposed. 2,-'~ In this case, the reinnervated area was found to be larger and the term collateral regeneration is inappropriate. The ability of nerves to sprout and regenerate decreases with increasing age in experimental animals as well as in humans, where a significant decrease in the extent of transmedian innervation has also been observed with increasing age. 6'28 This could be due to deficient N G F production in the denervated skin, or its decreased utilisation by the uninjured nerves. Decreased density of nerve terminals seems to be due to decreased availability of trophic factors, such as NGF. 29 However, nerve fibres in old animals showed a decreased ability for collateral sprouting even into young transplanted target tissues, indicating a loss of neuronal plasticity with age. 3~ Remarkably complete return of skin sensation in children where digital nerves had not been repaired was attributed to the resumption by fibres of adjacent nerves in the intermediate zone and budding into the dermal plexus of the autonomous zone from these adjacent nerves, but the results of examination of patients before primary and secondary unilateral digital nerve repair show that significant overlap of adjacent digital nerves is c o m m o n and this has important implications for the interpretation of results after surgery. 3~ The distribution and extent of the skin area of sudomotor function in cases where regeneration of injured nerve was prevented, corresponded very well to the area of light touch sensitivity. In contrast to very scarce collateral reinnervation of sweat glands observed in our study, the collateral reinnervation of sweat glands reported by Healy and LeQuesnr 2 was more extensive. This could be due to different anatomical locations or two different types of sweat glands. Whereas tactile sensitivity and sudomotor function in the a u t o n o m o u s zone of injured nerves was limited to the border of the territories of the ulnar and median nerves, response to pin prick (S 1) was possible to elicit over the whole area of autonomous zone. This has also been observed by others. '1'13 Pollock and Davis 13attributed the gradual return of marginal sensibility they observed to a slow recovery of the threshold to stimulation in fibres already present in the zone of overlap. However, most observers now believe that normal overlap of high threshold afferents between neighbouring nerves in the human palm is much less extensive than Pollock and Davis maintained, and that the progressive shrinking of the sensory defect is largely due to an active ingrowth of sensory fibres from adjacent normally innervated skin areas. 9'~ On the other hand some regeneration may occur across a suture line regardless of neuroma formation and this could explain the return of deep pain sensibility (S1). The second reason for bad final functional outcome in case of predicted collateral sprouting could be due to withdrawal of collateral sprouts. I f the original transected nerve regenerates successfully, high-threshold nerve fibres of the original nerve have been shown to recapture the skin area collaterally innervated by sprouting axons of the adjacent nerves in rats, so that original peripheral nerve fields were restored, j~32 Regenerating low-threshold axons have also been
442 shown to recapture skin from sprouted nerves in y o u n g rats. 33 However, despite almost complete restoration o f the n o r m a l i n n e r v a t i o n territory o f the i n j u r e d nerve due to its successful regeneration, collateral sprouts in muscles 34,35a n d in t o o t h pulps 36 o f cats persisted if regeneration t o o k longer. Therefore, the time period d u r i n g which the collateral sprouts h a d been present before the regenerating fibres r e t u r n e d is a n i m p o r t a n t factor in deciding whether or n o t the collateral sprouts w o u l d be displaced. 34,35 I n h u m a n s , r e i n n e r v a t i o n of p a l m a r skin by axons from the m e d i a n a n d u l n a r nerves i n j u r e d proximal to the wrist, or in patients in w h o m injured axons failed to regenerate, the time taken for eventual collateral sprouts to m a t u r e w i t h o u t c o m p e t i t i o n with the regenerating axons is probably long enough. Hoffert a n d colleagues 24presented a case report, in which collateral sprouts in sural nerve skin d i s t r i b u t i o n persisted after a n ingrowth of regenerating nerves. Therefore, the possibility that collateral sprouts were replaced by regenerating nerves seems unlikely. I n conclusion we can say, that the end-stage sensory a n d sudomotor recovery within the a u t o n o m o u s area of median a n d u l n a r nerves do n o t depend on possible collateral sprouting of uninjured nerve. The most i m p o r t a n t factor in final functional outcome in that area is regeneration of injured a n d sutured nerve. We believe that in cases where skin is partially denervated and regeneration of injured nerve is n o t prevented, well-known factors affecting nerve recovery: (patient age, level of injury, specific nerve injured, mechanism o f injury a n d nerve defect, rehabilitation, repair techniques, associated injuries and delay to repair) 37a n d associated devascularising t r a u m a 38are far more i m p o r t a n t than intact neighbouring nerve as a possible source of collateral sprouting of axons from adjacent uninjured nerves. I n cases where regeneration of injured nerve is prevented, adjacent u n i n jured nerve may contribute to sprouting of nociceptive axons providing a protective function.
Acknowledgements We are indebted to Mr Ciaran Healy MD, Consultant Plastic Surgeon at St Thomas' Hospital for useful discussion at the beginning of this work. We would like to thank Jo~e Trontelj, MD, PhD from University Institute of Clinical Neurophysiology for valuable suggestions during the electrophysiologic measurements. Many thanks to Mrs Jo~a Brili and Mrs Rupreht Ruth, Chem.Ing. for their help with ninhydrin tests, Mr Gregor Cerkvenik and Mr Rok Leskovec, Dipl. OEC. for their help with the computer work and Mrs Marija Bajc - "DOCA" for slides from our photo documentation. This work was supported by grants from PETROL Trgovina and Open Society Fund-Slovenia.
References 1. Devor M, Schonfeld D, Seltzer Z, Wall PD. Two modes of cutaneous reinnervation following peripheral nerve injury. J Comp Neurol 1979; 185: 211-20. 2. Wiesenfeld-HallinZ, Kinnman E, Aldskogius H. Expansion of innervation territory by afferents involved in plasma extravasation after nerve regeneration in adult and neonatal rats. Exp Brain Res 1989; 76: 88-96. 3. Weddell G, Guttmann L, Gutmann E. The local extension of nerve fibers into denervated areas of skin. J Neurol Psych 1941; 4: 206--25.
British Journal of Plastic Surgery 4. Mackinnon SE, Dellon AL. Anatomy and physiology of the peripheral nerve. In Mackinnon SE, Dellon AL, eds. Surgery of the peripheral nerve. New York: Thieme Medical Publishers, 1988. 5. Jackson PC, Diamond J. Temporal and spatial constraints of the collateral sprouting of low-threshold mechanosensory nerves in the skin of rats. J Comp Neurol 1984; 226: 33645. 6. Diamond J, Foerster A. Recovery of sensory function in skin deprived of its innervation by lesion of the peripheral nerve. Exp Neurol 1992; 115: 100-3. 7. Kennedy WR, Sakuta M. Collateral reinnervation of sweat glands. Ann Neurol 1984; 15: 73-8. 8. Leonard MH. Return of skin sensation in children without repair of nerves. Clin Orthop Rel Res 1973; 95: 273-7. 9. Livingstone WK. Evidence of active invasion of denervated areas by sensory fibres from neighbouring nerves in man. J Neurosurg 1947; 4: 140-5. 10. Robinson PR Observations on the recovery of sensation following inferior alveolar nerve injuries.Br J Oral Maxillofac Surg 1988; 26: 177-89. 11. Inbal R, Rousso M, Ashur H, Wall PD, Devor M. Collateral sprouting in skin and sensory recovery after nerve injury in man. Pain 1987; 28: 141-54. 12. Healy C, LeQuesne PM, Lynn B. Collateral sprouting of cutaneous nerves in man. Brain 1996; 119: 2063-72. 13. Polock LJ, Davies L. Pattern of sensory recovery of peripheral nerve injuries. In: Peripheral nerve injuries. New York: Hoeber, 1933. 14. Jfi~iskelfiinenSK, Petolta JK. Electrophysiologic evidence for extremely late sensory collateral reinnervation in humans. Neurology 1996;46: 1703-5. 15. Cooney WE Median nerve repairs: the results of treatment. In: Gelberman RH, ed. Operative nerve repair and reconstruction. Philadelphia: JB Lippincott, 1991. 16. Chassard M, Pham E, Comtet JJ. Two-pointdiscriminationtests versus functional sensory recovery in both median and ulnar nerve complete transections. J Hand Surg 1993; 18 B: 790-6. 17. Ah6an U, Arne~ ZM, Janko M, Dovgak D. Regeneration of sudomotor and sensory nerve fibres after digital replantation and microneurovasculartoe-to-hand transfer. Br J Plast Surg 1997; 50: 227-35. 18. Ah~an U, PlaningekF, Arne~ ZM. Measurement of sudomotor fibre regeneration by sympathetic skin response after complete division of peripheral nerves in children. Br J Plast Surg 1996; 49: 11-19. 19. Horch K. Absence of functional collateral sprouting of mechanoreceptor axons into denervated areas of mammalian skin. Exp Neurol 1981; 74: 313-7. 20. Jackson PC, Diamond J. Failure of intact cutaneous mechanosensory axons to sprout functional collaterals in skin of adult rabbits. Brain Res 1983; 273: 277-83. 21. Yasargil GM, Macintyre L, Doucette R, Visheau B, Holmes M, Diamond J. Axonal domains within shared touch domes in the rat: a comparison of their fate during conditions favoring collateral sprouting and following axonal regeneration. J Comp Neurol 1988; 270: 301-12. 22. McAllister RMR. Recovery of sensibility of the hand after nerve injuries. In: Boivie J, Hanson P, Lindblom VLS, eds. Touch, temperature and pain in health and disease: Mechanisms and assessments. Seattle: IASP Press, 1994. 23. Robinson PR Recessionof sensory loss from the midline following trigeminal sensory root section: collateral sprouting from the normal side? Brain Res 1983; 259: 177-80. 24. Hoffert MJ, Greenberg RE Wolskee PJ, et al. Abnormal and collateral innervations of sympatheticand peripheral sensory fields associatedwith a case of causalgia.Pain 1984;20:1 12. 25. Murray M, Goldberger ME. Restitution of function and collateral sprouting in the cat spinal cord: the partially hemisected animal. J Comp Neurol 1974; 158: 19-36. 26. Dostrovsky JO, Millar J, Wall PD. The immediate shift of afferent drive to dorsal column nucleus cells following deafferentation: a comparison of acute and chronic deafferentation in gracile nucleus and spinal cord. Exp Neurol 1976; 52: 480-95. 27. Rosen B, Lundborg G, Dahlin LB, Holmberg J, Karlson B. Nerve repair: correlation of restitution of functional sensibility with specific cognitive capacities. J Hand Surg 1994; 19B: 452-8.
C o n t r i b u t i o n of collateral sprouting 28. Kinnman E, Aldskogius H, Johansson O, Wiesenfeld-Hallin Z. Collateral reinnervation and expansive regenerative reinnervation by sensory axons into ~ denervated skin: an immunohistochemical study in the rat. Exp Brain Res 1992; 91: 61-72. 29. Gavazzi I, Andrews TJ, Thrasivoulou C, Cowen T. Influence of target tissues on their innervation in old age: a transplantation study. Neuro Report 1992; 3:717 20. 30. Gavazzi I. Collateral sprouting and responsiveness to nerve growth factor of ageing nearons. Neurosci Lett 1995; 189: 47 50. 31. Calder JS, McAllister RMR. Interpreting the results of unilateral digital nerve repair. J Hand Surg 1993; 18B: 797 9. 32. Beuerman RW, Rozsa AJ. Collateral sprouts are replaced by regenerating neurites in the wounded corneal epithelium. Neurosci Lett 1984; 44:99 104. 33. Jackson PC, Diamond I Regenerating axons reclaim sensory targets from collateral nerve sprouts. Science 1981; 214: 926 8. 34. Slack JR. Interaction between foreign and regenerating axons in axolotl muscle. Brain Res 1978; 146:172 6. 35. Thompson W. Reinnervation of partially denervated rat soleus muscle. Acta PhysioI Scand 1978; 103: 8I 91. 36. Robinson PR Regenerating nerve fibres do not displace the collateral reinnervation of cat teeth. Brain Res 1984; 310: 303-10. 37. Steinberg DR, Koman LA. Factors affecting the results of peripheral nerve repair. In: Gelberman RH, ed.
443 Operative nerve repair and reconstruction. Philadelphia: JB Lippincott, 1991. 38. Jupiter JB. Nerve injury associated with devascularizing trauma. In: Gelberman RH, ed. Operative nerve repair and reconstruction. Philadelphia: JB Lippincott, 1991.
The Authors Uro~ Ah~:anMD, Registrar,
Zoran M. Arne~ MD, PhD, Chief, University Department of Plastic Surgery and Burns, University Medical Centre, Ljubljana, Zalogka 7, 1000 Ljubljana, Slovenia.
Fajko Bajrovi~: MD, MSe Institute of Pathophysiology, Medical Faculty Ljubljana, Slovenia. Martin Janko MD, PhD University Institute of Clinical Neurophysiology, University Medical Centre, Ljubljana, Slovenia. Correspondence to Professor Zoran M. Arne~, University Department of Plastic Surgery and Burns, University Medical Centre, Ljubljana. Paper received 19 August 1997. Accepted 29 April 1998, after revision.
BRITISH JOURNAL OF ~
PLASTIC SURGERY
Contribution of collateral sprouting to the sensory and sudomotor recovery in the human palm after peripheral nerve injury U. Ah~an*, Z. M. Arn6z*, E BajroviC and M. Janko*
* University Department of Plastic Surgery and Burns; ~University Institute of Clinical Neurophysiology, University Medical Centre; t Institute of Pathophysiology, Medical Faculty, Ljubljana, Slovenia SUMMARY. The contribution of collateral sprouting to the sensory and sudomotor recovery was studied in 52 patients aged 3-66 years (mean 35.5 years) from 2 to 9 years following nerve injury and repair. The study included three groups of patients: (1) patients with complete division of median and ulnar nerves (skin reinnervation exclusively due to axon regeneration), (2) patients with isolated division of ulnar or median nerve (skin reinnervation due to axon regeneration and possible collateral sprouting), and (3) patients in whom injured axons failed to regenerate (skin reinnervation exclusively due to collateral sprouting). The end stage of sensory and sudomotor recovery was studied by: clinical methods, sensory nerve action potential (SNAP) measurements, sympathetic skin response (SSR) and the ninhydrin test. We found that recovery of sensory and sudomotor function in groups 1 and 2 was similar. End-stage sudomotor and sensory recovery within the autonomous area of the nerve did not depend on possible collateral reinnervation. Collateral reinnervation from the uninjured nerve was limited to the border innervation area of the palm and ring finger. Adjacent uninjured nerve may contribute to sprouting of nociceptive axons providing a protective function.
Cutaneous sensory and sudomotor reinnervation following peripheral nerve injury is believed to be due to: regenerative growth of the axons from the proximal stump of injured nerve (axon regeneration), collateral sprouting of axons from the adjacent uninjured nerve and sprouting of axons from the adjacent injured nerve (expansive regenerative reinnervation). All three processes are well known from animal experiments. '-3 When injured axons fail to regenerate, collateral sprouting of adjacent uninjured axons is the only explanation for functional recovery. However, in the majority Of clinical cases the regeneration of axons from the proximal stump of the injured nerve is not prevented and functional recovery occurs mostly due to ingrowth of regenerating axons. 4 The recovery of cutaneous nerve function in the human palm after peripheral nerve injury is well documented, but the contribution of collateral sprouting of axons from the uninjured adjacent nerves to the recovery of sensory and sudomotor function in the partially denervated human palm has not been assessed. Collateral sprouting of uninjured axons in response to partial skin denervation has been studied extensively in animals. It has been shown that in neonatal animals all types of cutaneous axons have the ability to sprout and that different types of axons differ in their capacity to extend sprouts into adjacent denervated skin/,5 However, only high threshold nociceptive afferents (A~) 5'6 and sudomotor axons (C) 7 seem to have the capacity for collateral sprouting in adult animals. Despite the lack of histological proof, other evidence that collateral sprouting also occurs in humans is accumulating. Besides collateral sprouting of nociceptive and sudomotor axons, clinicals-'3 and
electrophysiologicaP* findings indicate that low threshold mechanoreceptors may also have the capacity for collateral reinnervation in humans. There is some evidence to suggest that collateral sprouting of sensory axons does occur in human palm t o o . 8,9A1A3 However, we are not aware of any report regarding possible collateral sprouting of sudomotor axons in that skin area. Devor and colleagues' reported that collateral sprouting of uninjured nerves plays an early role in reinnervation of rat foot skin after nerve injury. This leads to an initial reduction in anaesthetic area due to peripheral nerve injury and is followed by ingrowth of the injured axons. A similar pattern of recovery has been described after inferior alveolar nerve injury in man. '~ Because of intermingling of collateral sprouting of axons from the uninjured nerve and regeneration of axons from the injured one, we would expect a better functional recovery than in the case where both of the palmar nerves are injured and functional recovery occurs only due to ingrowth of regenerating axons. Some functional recovery might be expected even when injured axons failed to regenerate and collateral sprouting of adjacent uninjured nerve is the only possibility for functional recovery. The present study is based on three different groups of patients: (1) skin reinnervation exclusively due to axon regeneration (complete division of both median and ulnar nerves), (2) skin reinnervation due to axon regeneration and possible collateral sprouting (patients with isolated division of the ulnar or median nerve), and (3) skin reinnervation exclusively due to collateral sprouting (patients in whom injured axons failed to regenerate). 436
Contribution of collateral sprouting The study was undertaken in order to estimate the clinical importance of collateral sprouting in the h u m a n palm.
Material The study was approved by the National Committee for Medical Ethics and included three groups of patients.
Group 1. Patients with isolated sharp division and primary microsurgical repair of the ulnar (15 patients) or median nerve (20 patients), in whom collateral nerve regeneration from the uninjured side was possible (35 patients, 28 males and 7 females from 3 to 66 years - mean 29.8 years). Group 2. Injuries which precluded collateral sprouting: complete division and primary microsurgical repair of median and ulnar nerves (11 patients, 7 males and 4 females aged from 6 to 56 years - mean 28.2 years). Group 3. Six patients in whom injured axons failed to reinnervate the denervated area and collateral sprouting might be the only possibility for functional recovery. All six patients aged from 24 to 56 years (mean 42.7 years) visited our clinic 2-9 years after injury and treatment in other hospitals. In three patients we found neuromas at exploration. Another patient came to our clinic 9 years after injury and 4 years after a long non-functional neuroma was resected. In a 24-year-old patient with transection of the ulnar nerve in the axilla by knife, despite complete sensory loss in the ulnar nerve innervating area only skin was sutured. A 44-year-old patient with transection of the ulnar nerve in lower third of the forearm came after several neurolyses performed at the local hospital.
437 above the one producing m a x i m u m response. The skin temperature was kept above 31 ~ The latency and the amplitude of the response were measured.
Clinical testing of sensation. The patients were tested for the light touch, pain, static and dynamic two point discrimination in the zones of the palm and palmar surface of the finger using standard techniques (Fig. 1). For uniform assessment of clinical results we used modified Medical Research Council classification. 1~,16 We agree with most authors who believe that sensory return $3+ to $4 are needed ~5to have good to excellent clinical results. Evaluation of sudomotor nerve fibre regeneration Recording of SSRs. The SSRs were elicited and recorded using a standard technique.~7.~8 A two channel E M G machine Mystro, Medelec MS 20 was used. The stimulating electrode (13L 36, D A N T E C , Skovlunde, Denmark) was attached to the palmar surface of the wrist. An earthing electrode was applied to the opposite wrist. A single square electrical impulse of 80~150 V amplitude and 0.2 ms duration was delivered near the end of inspiration to evoke the SSR. Surface disc electrodes were attached at the centre of the distal phalanx of the little and index fingers (recording electrode) and dorsum of the same hand (reference electrode) bilaterally. In patients without regenerative growth of the
,,:.:.:,:..
....,,.,.,...,. ,,................
::::::::::::::::5:. ,....................
z,d e
Methods
Evaluation of sensory nerve fibre regeneration Recording of SNAPs. Antidromic sensory nerve action potentials (SNAPs) were recorded from the index and/or little finger after sutured ulnar and/or median nerve and from the corresponding digit of the uninjured hand. A pair of standard ring electrodes (Digital ring electrodes 16639, M E D E L E C , U K ) was placed around the proximal and distal interphalangeal joints. A stimulating electrode (Bipolar stimulating electrode 16893, M E D E L E C , U K ) was attached to the wrist 3 cm proximal to the distal crease of the wrist longitudinally along the course of the median or ulnar nerve. The ground electrode was attached to the same wrist close to the site of stimulation No. 1. SNAPs elicited by stimulation of the ulnar nerve at the wrist were recorded from the little finger, and SNAPs evoked by median nerve stimulation at the wrist were recorded on the index finger. Ten responses were averaged during each stimulation procedure. These recordings were made in a standard way on the E M G machine (Mystro, Medelec MS 20). The stimulus pulse duration was 0.1 ms and the stimulus frequency 0.5 Hz. The stimulation strength was slightly
Figure 1--Distribution of median and ulnar nerves, autonomous zones and points of the SSR and SNAP measurement. Note: Distribution of median and ulnar nerves of the palm and palmar surface of the fingers (line) divided into 7 zones (squares), areas of absolute sensory supply of the median and ulnar nerves autonomous zones (shadow). Position of surface disc electrode for the SSRs measurements (circles) and ring electrodes for the SNAPs recording (bold line).
438 injured nerve (axon regeneration) the disc electrodes were attached to the border innervation area on the palm (Fig. 1). Skin temperature in all patients was kept above 31~ The measurements were taken simultaneously on both hands. At each site four consecutive responses were recorded to ensure reproducibility. The peak-topeak amplitude and area (integrated surface below and above the baseline) of each response were measured. Left to right ratios of area of the responses were calculated and expressed as percentages.
Ninhydrin test. Spontaneous sweat secretion was observed and documented by the ninhydrin test. The patient placed his clean and dry palm and fingers on a sheet of white chromatographic paper and pressed gently. The paper was passed several times through a 1% ninhydrin solution in acetone. The sheet was then heat dried in an incubator at 110~ The results on the injured side were compared to those on the normal side. They were classified by the intensity of the print pattern: normal sweating (+ + +), diminished sweating (+ +) and deficient (+) or absent sweating (-). All patients were asked for history of post-traumatic cold intolerance. All studies were repeated on the corresponding nerves of the contralateral normal finger so that each parameter could be matched with a control value obtained from the same subject.
Results
Group 1 (patients where collateral nerve regeneration could be excluded) Five of 11 patients (in median nerve distribution) and 4 of 11 (in ulnar nerve distribution) had good results (sensory return of $3+ to $4 and diminished to normal sweating). SNAPs were present in 4 of 11 patients and the SSR area on the injured site reached 60% or more when compared to the uninjured site in 5 of 11 patients (Fig. 2).
Group 2 (patients with possible collateral nerve regeneration) Six of 15 patients with complete laceration of the ulnar nerve and 9 of 20 with complete laceration of the median nerve returned to $3+ or $4 and had diminished or normal sweating. SNAPs were present in 8 of 20 patients with median nerve injury and 6 of 15 patients with ulnar nerve injury. The SSR area on the injured site reached 60% or more when compared to the uninjured site in 10 of 20 patients with median nerve injury and 8 of 15 patients with ulnar nerve injury. A virtually normal pattern of sweating was observed in 6 of 15 ulnar nerve repairs and 10 of 20 median nerve repairs. Thirty of 35 patients could recognise different sensation on the ulnar or median half of the ring finger even when the final stage of nerve regeneration was reached. In a high percentage of patients post-traumatic cold intolerance was an important factor causing disability (Table 1).
British Journal of Plastic Surgery
Group 3 (patients in whom injured axons failed to regenerate) In all 6 patients only the ring finger and border innervation area of median or ulnar nerve territory on the palm (up to 15 mm) (Fig. 4), had good sensory regeneration ($3). In all remaining ulnar or median nerve distribution in the palm (especially in the autonomous zone) only deep pain sensibility (response to pin prick-S1) recovered. In the same area of the palm and fingers, the skin was dry, smooth and slippery. Spontaneous sweat secretion in those areas was not observed by the ninhydrin test. The lack of sweating corresponded well to the sensory deficit in all patients. No patients had recordable SNAPs or SSRs and all complained of post-traumatic cold intolerance (Figs 3 and 4, Table 1). Discussion
After peripheral nerve injury, the skin may be reinnervated by regenerating axons from injured nerve fibres and from adjacent uninjured nerves (collateral sprouting). Previous clinical studies of cutaneous reinnervation following nerve injury have mainly been performed to assess the regenerative capacity of injured cutaneous nerves. Our study was undertaken to estimate clinical importance of collateral sprouting of sensory and sudomotor axons for functional recovery in human palm in two situations: first, when injured nerve failed to regenerate and collateral sprouting of uninjured axons was the only possibility for functional recovery; and second, in cases where regeneration of injured nerve was not prevented and function recovered due to intermingling of both processes (regeneration of injured cutaneous nerves and collateral sprouting of uninjured nerves). We expected better sensory and sudomotor function recovery in patients with just one injured nerve (median or ulnar) since functional recovery of autonomous innervation territory was possible due to collateral sprouting and axon regeneration. However, in the present study, the results of sensory and sweating recovery in the autonomous area of the ulnar and median nerves were similar regardless of possible collateral sprouting. Furthermore, the borders of touch sensitive area in subjects where axon regeneration of injured ulnar or median nerve had failed, extended minimally over the boundaries of the normal innervation territories, determined clinically and electrophysiologically by the others, n,~3Therefore, we suggest that collateral reinnervation of low threshold receptors from the uninjured ulnar and median nerve does occur in the human palm, but is limited to the border innervation area of these two nerves. The lack of improvement of functional outcome in case of predicted collateral sprouting may be due either to a lack of collateral sprouting of low threshold sensory and sudomotor fibres, or withdrawal of collateral sprouts. Experimental studies of collateral reinnervation of the skin of mammals have yielded the rather surprising result that only high threshold pain afferents and sudomotor axons in the A5 and C fibre range, respectively, seem to have the capacity for collateral
Contribution of collateral sprouting
439
Sensory nerve action potential (SNAP) and sensation
1~---/' ~
~
s . AP r f o o
SNAP tkom the uninjured hand
f
1
...I"
SNAP from the injured hand
A
Sympathetic skin response (SSR)
SSR from the inJured hand SSR from the uninjured hand
SgR fiom the th/ured hand SSR from the uninjured hand
~
_ _
SSR fhom the m mredhand SSR from the uninjured hand
SSR from Ihe injured hand g g R from the uninjured hand
SSR from the injured hand SSR from th9 uninjured hand
SgR from the injured hand SSR from the uninjured hand
B Figure 2--(A, B) A 17-year-oldpatient with complete division of median and ulnar nerves by glass at mid third of the forearm. Five years after injury and primary epineural repair a complete sensory recovery($4) is noticed within ulnar innervation area and S3+ within median nerve area. The patient could recognise different sengation on the ulnar half of the ring finger. Sensory nerve action potentials were present. Normal pattern of sweating and normal SSRs show an excellent sudomotor recovery.
sprouting in adult animals? -7 Electrophysiological and histological studies of collateral sprouting in the skin of various mammals; cat, 19rabbit 2~and raC failed to demonstrate sprouting of low threshold mechanosensory cutaneous afferents (A[3) after partial denervation of skin in adult animals. In clinical studies we do not often encounter the favourable conditions present with laboratory animals. The traumatic wound is seldom as clean cut as the surgical one. Infection, delay of healing, extensive soft tissue damage, vascular injury, all promote scar formation. Nerve ends are often widely separated and occasionally require suture under greater than optimal
tension. The accidental wound may divide a nerve at the point where it divides into separate branches. 22 Furthermore, evaluation of axon regeneration in animal experimental studies mostly includes: nerve pinch test, skin pinch test, axon counting, immunohistochemical procedures and electron microscopy. These provide interesting morphological data, but poor information about functional results and abnormal sensory symptoms. In contrast to animal studies, there is some clinical evidence for collateral reinnervation of low threshold receptors occurring in the skin of the human palm ~s,H and other r e g i o n s J 2,23,24 Despite lack of histological
440
British J o u r n a l o f Plastic S u r g e r y
Table 1
Results after m e d i a n a n d / o r u l n a r n e r v e repair, a n d in p a t i e n t s w i t h n o a x o n r e g e n e r a t i o n
Injury
Sensation Recovery $3+ or $4 (in %) U M
Measurable SNAP (in %) U M
SSR Over 60% (in %) U M
Ninhydrin test ++ or +++ (in %) U M
CoM Intolerance (in %) U M
Division of M or U nerve Division of M and U nerve No axon regeneration
40
45
40
40
50
53
40
50
67
70
36
45
36
36
45
45
36
45
64
64
100
100
0
0
0
0
M = median, U = ulnar.
Sensory nerve action potential (SNAP)
SNAP from the injured nerve
SNAP from the injured nerve
t
SNAP from the injured nerve
A
Sympathetic skin response (SSR) ....... .........~/'~
SSR from the injured nerve
SSR from the injured nerve
j
SSR from the uninjured nerve
SSR from the uninjured nerve SSR from the injured nerve
SSR from the uninjured nerve
SSR from the uninjured nerve SSR from the uninjured nerve
SSR from the injured nerwe
J~..~""~-'-~"
S S R f i ~ m the t m l n j u r e d nerve
ssR
B
fromth~~anj.~od,~e
.....
~.~,
_ / f_- - . . . ~ , ~ .
~
~
_......_.X~f/_~.._._~ . . . . . ---..__0
~ .
.
.
.
-
--
SSR from the uninjured nerve
~,...t,&~
SSR from the I,jurea , ~ , , e
~
SSR from the uninjured nerve
A, B, 1, 2, 3, 4, 5 = measurement points
Figure 3---(A, B) A 56-year-old patient with incomplete amputation at wrist level in 1988. Median nerve was repaired by primary suture and ulnar nerve was intact. Since the patient had complete sensory and sudomotor deficit within the median innervation area, a 6 cm long nonfunctional neuroma was resected and sural nerve grafts sutured in a year later. Six years after repair only sensation of deep pain recovered (S 1). The collateral sprouts reinnervated just the border innervation area o f the palm and ring finger. The sensory nerve action potentials on the first, second and third fingers were absent. The lack of sweating corresponded well to the sensory deficit. The SSR was absent on measurement points B, 1, 2 and 3, slightly diminished on measurement point 4 and normal on measurement points 5 and A. Upper traces are from injured hand and lower from the uninjured hand.
Contribution of collateral sprouting
441
iiiiiiii?
Figure 4~Area of collateral reinnervation in the palm and palmar surface of the ring finger. Note: In the group of patients in whom injured axons failed to reinnervate the denervated area and collateral sprouting might be the only possibility for functional recovery,the area of good sensation ($3) was limited to the border innervation area of the palm and ring finger (shadow). Reinnervation from the ulnar side - black arrows and from the median side - white arrows. proof, this collateral reinnervation is attributed to sprouting of adjacent uninjured nervesY '~-' On the other hand, synaptic unmasking and collateral sprouting in the spinal cord were suggested to explain such an effect. 25'26Central mechanisms cannot be definitely excluded? 7 When cutaneous nerve collateral sprouting was studied in adults in whom a forearm cutaneous nerve had been resected, the objective sensory test showed encroachment at margin by only 6 m m for light touch, 7 m m for sharp pain and 11.5mm for heat pain, with no significant change for warming or cooling. Collateral sprouting of sudomotor C fibres resulting in return of sweating was more extensive. 12 Transmedian cutaneous reinnervation in patients who had undergone trigeminal sensory root section was explained by collateral sprouting of sensory nerves. The mean extent of transmedian cutaneous innervation varied from about 2 m m in the middle of tongue to 9 m m on the lower lip. In all areas the extent was greater to pin prick than to light touch stimulus? 3 The extension of collateral sprouts was shown to be limited in these studies but the mechanism responsible for this limitation is not known. In some other adult clinical s t u d i e s ~2,23,z4 the collateral sprouting of low threshold sensory fibres is suggested to be more extensive, although the capacity of low-threshold afferents may differ with regard to anatomical location. It should be pointed out that in a traumatic wound or after nerve graft harvest, small branches or interconnections of neighbouring nerve are also injured, in addition to the principal nerve injury. In such circumstances, in which the nerves adjacent to the denervated area are themselves also injured, and during regenerative spread into the adjacent territory, the term 'expansive regenerative reinnervation' has
been proposed. 2,-'~ In this case, the reinnervated area was found to be larger and the term collateral regeneration is inappropriate. The ability of nerves to sprout and regenerate decreases with increasing age in experimental animals as well as in humans, where a significant decrease in the extent of transmedian innervation has also been observed with increasing age. 6'28 This could be due to deficient N G F production in the denervated skin, or its decreased utilisation by the uninjured nerves. Decreased density of nerve terminals seems to be due to decreased availability of trophic factors, such as NGF. 29 However, nerve fibres in old animals showed a decreased ability for collateral sprouting even into young transplanted target tissues, indicating a loss of neuronal plasticity with age. 3~ Remarkably complete return of skin sensation in children where digital nerves had not been repaired was attributed to the resumption by fibres of adjacent nerves in the intermediate zone and budding into the dermal plexus of the autonomous zone from these adjacent nerves, but the results of examination of patients before primary and secondary unilateral digital nerve repair show that significant overlap of adjacent digital nerves is c o m m o n and this has important implications for the interpretation of results after surgery. 3~ The distribution and extent of the skin area of sudomotor function in cases where regeneration of injured nerve was prevented, corresponded very well to the area of light touch sensitivity. In contrast to very scarce collateral reinnervation of sweat glands observed in our study, the collateral reinnervation of sweat glands reported by Healy and LeQuesnr 2 was more extensive. This could be due to different anatomical locations or two different types of sweat glands. Whereas tactile sensitivity and sudomotor function in the a u t o n o m o u s zone of injured nerves was limited to the border of the territories of the ulnar and median nerves, response to pin prick (S 1) was possible to elicit over the whole area of autonomous zone. This has also been observed by others. '1'13 Pollock and Davis 13attributed the gradual return of marginal sensibility they observed to a slow recovery of the threshold to stimulation in fibres already present in the zone of overlap. However, most observers now believe that normal overlap of high threshold afferents between neighbouring nerves in the human palm is much less extensive than Pollock and Davis maintained, and that the progressive shrinking of the sensory defect is largely due to an active ingrowth of sensory fibres from adjacent normally innervated skin areas. 9'~ On the other hand some regeneration may occur across a suture line regardless of neuroma formation and this could explain the return of deep pain sensibility (S1). The second reason for bad final functional outcome in case of predicted collateral sprouting could be due to withdrawal of collateral sprouts. I f the original transected nerve regenerates successfully, high-threshold nerve fibres of the original nerve have been shown to recapture the skin area collaterally innervated by sprouting axons of the adjacent nerves in rats, so that original peripheral nerve fields were restored, j~32 Regenerating low-threshold axons have also been
442 shown to recapture skin from sprouted nerves in y o u n g rats. 33 However, despite almost complete restoration o f the n o r m a l i n n e r v a t i o n territory o f the i n j u r e d nerve due to its successful regeneration, collateral sprouts in muscles 34,35a n d in t o o t h pulps 36 o f cats persisted if regeneration t o o k longer. Therefore, the time period d u r i n g which the collateral sprouts h a d been present before the regenerating fibres r e t u r n e d is a n i m p o r t a n t factor in deciding whether or n o t the collateral sprouts w o u l d be displaced. 34,35 I n h u m a n s , r e i n n e r v a t i o n of p a l m a r skin by axons from the m e d i a n a n d u l n a r nerves i n j u r e d proximal to the wrist, or in patients in w h o m injured axons failed to regenerate, the time taken for eventual collateral sprouts to m a t u r e w i t h o u t c o m p e t i t i o n with the regenerating axons is probably long enough. Hoffert a n d colleagues 24presented a case report, in which collateral sprouts in sural nerve skin d i s t r i b u t i o n persisted after a n ingrowth of regenerating nerves. Therefore, the possibility that collateral sprouts were replaced by regenerating nerves seems unlikely. I n conclusion we can say, that the end-stage sensory a n d sudomotor recovery within the a u t o n o m o u s area of median a n d u l n a r nerves do n o t depend on possible collateral sprouting of uninjured nerve. The most i m p o r t a n t factor in final functional outcome in that area is regeneration of injured a n d sutured nerve. We believe that in cases where skin is partially denervated and regeneration of injured nerve is n o t prevented, well-known factors affecting nerve recovery: (patient age, level of injury, specific nerve injured, mechanism o f injury a n d nerve defect, rehabilitation, repair techniques, associated injuries and delay to repair) 37a n d associated devascularising t r a u m a 38are far more i m p o r t a n t than intact neighbouring nerve as a possible source of collateral sprouting of axons from adjacent uninjured nerves. I n cases where regeneration of injured nerve is prevented, adjacent u n i n jured nerve may contribute to sprouting of nociceptive axons providing a protective function.
Acknowledgements We are indebted to Mr Ciaran Healy MD, Consultant Plastic Surgeon at St Thomas' Hospital for useful discussion at the beginning of this work. We would like to thank Jo~e Trontelj, MD, PhD from University Institute of Clinical Neurophysiology for valuable suggestions during the electrophysiologic measurements. Many thanks to Mrs Jo~a Brili and Mrs Rupreht Ruth, Chem.Ing. for their help with ninhydrin tests, Mr Gregor Cerkvenik and Mr Rok Leskovec, Dipl. OEC. for their help with the computer work and Mrs Marija Bajc - "DOCA" for slides from our photo documentation. This work was supported by grants from PETROL Trgovina and Open Society Fund-Slovenia.
References 1. Devor M, Schonfeld D, Seltzer Z, Wall PD. Two modes of cutaneous reinnervation following peripheral nerve injury. J Comp Neurol 1979; 185: 211-20. 2. Wiesenfeld-HallinZ, Kinnman E, Aldskogius H. Expansion of innervation territory by afferents involved in plasma extravasation after nerve regeneration in adult and neonatal rats. Exp Brain Res 1989; 76: 88-96. 3. Weddell G, Guttmann L, Gutmann E. The local extension of nerve fibers into denervated areas of skin. J Neurol Psych 1941; 4: 206--25.
British Journal of Plastic Surgery 4. Mackinnon SE, Dellon AL. Anatomy and physiology of the peripheral nerve. In Mackinnon SE, Dellon AL, eds. Surgery of the peripheral nerve. New York: Thieme Medical Publishers, 1988. 5. Jackson PC, Diamond J. Temporal and spatial constraints of the collateral sprouting of low-threshold mechanosensory nerves in the skin of rats. J Comp Neurol 1984; 226: 33645. 6. Diamond J, Foerster A. Recovery of sensory function in skin deprived of its innervation by lesion of the peripheral nerve. Exp Neurol 1992; 115: 100-3. 7. Kennedy WR, Sakuta M. Collateral reinnervation of sweat glands. Ann Neurol 1984; 15: 73-8. 8. Leonard MH. Return of skin sensation in children without repair of nerves. Clin Orthop Rel Res 1973; 95: 273-7. 9. Livingstone WK. Evidence of active invasion of denervated areas by sensory fibres from neighbouring nerves in man. J Neurosurg 1947; 4: 140-5. 10. Robinson PR Observations on the recovery of sensation following inferior alveolar nerve injuries.Br J Oral Maxillofac Surg 1988; 26: 177-89. 11. Inbal R, Rousso M, Ashur H, Wall PD, Devor M. Collateral sprouting in skin and sensory recovery after nerve injury in man. Pain 1987; 28: 141-54. 12. Healy C, LeQuesne PM, Lynn B. Collateral sprouting of cutaneous nerves in man. Brain 1996; 119: 2063-72. 13. Polock LJ, Davies L. Pattern of sensory recovery of peripheral nerve injuries. In: Peripheral nerve injuries. New York: Hoeber, 1933. 14. Jfi~iskelfiinenSK, Petolta JK. Electrophysiologic evidence for extremely late sensory collateral reinnervation in humans. Neurology 1996;46: 1703-5. 15. Cooney WE Median nerve repairs: the results of treatment. In: Gelberman RH, ed. Operative nerve repair and reconstruction. Philadelphia: JB Lippincott, 1991. 16. Chassard M, Pham E, Comtet JJ. Two-pointdiscriminationtests versus functional sensory recovery in both median and ulnar nerve complete transections. J Hand Surg 1993; 18 B: 790-6. 17. Ah6an U, Arne~ ZM, Janko M, Dovgak D. Regeneration of sudomotor and sensory nerve fibres after digital replantation and microneurovasculartoe-to-hand transfer. Br J Plast Surg 1997; 50: 227-35. 18. Ah~an U, PlaningekF, Arne~ ZM. Measurement of sudomotor fibre regeneration by sympathetic skin response after complete division of peripheral nerves in children. Br J Plast Surg 1996; 49: 11-19. 19. Horch K. Absence of functional collateral sprouting of mechanoreceptor axons into denervated areas of mammalian skin. Exp Neurol 1981; 74: 313-7. 20. Jackson PC, Diamond J. Failure of intact cutaneous mechanosensory axons to sprout functional collaterals in skin of adult rabbits. Brain Res 1983; 273: 277-83. 21. Yasargil GM, Macintyre L, Doucette R, Visheau B, Holmes M, Diamond J. Axonal domains within shared touch domes in the rat: a comparison of their fate during conditions favoring collateral sprouting and following axonal regeneration. J Comp Neurol 1988; 270: 301-12. 22. McAllister RMR. Recovery of sensibility of the hand after nerve injuries. In: Boivie J, Hanson P, Lindblom VLS, eds. Touch, temperature and pain in health and disease: Mechanisms and assessments. Seattle: IASP Press, 1994. 23. Robinson PR Recessionof sensory loss from the midline following trigeminal sensory root section: collateral sprouting from the normal side? Brain Res 1983; 259: 177-80. 24. Hoffert MJ, Greenberg RE Wolskee PJ, et al. Abnormal and collateral innervations of sympatheticand peripheral sensory fields associatedwith a case of causalgia.Pain 1984;20:1 12. 25. Murray M, Goldberger ME. Restitution of function and collateral sprouting in the cat spinal cord: the partially hemisected animal. J Comp Neurol 1974; 158: 19-36. 26. Dostrovsky JO, Millar J, Wall PD. The immediate shift of afferent drive to dorsal column nucleus cells following deafferentation: a comparison of acute and chronic deafferentation in gracile nucleus and spinal cord. Exp Neurol 1976; 52: 480-95. 27. Rosen B, Lundborg G, Dahlin LB, Holmberg J, Karlson B. Nerve repair: correlation of restitution of functional sensibility with specific cognitive capacities. J Hand Surg 1994; 19B: 452-8.
C o n t r i b u t i o n of collateral sprouting 28. Kinnman E, Aldskogius H, Johansson O, Wiesenfeld-Hallin Z. Collateral reinnervation and expansive regenerative reinnervation by sensory axons into ~ denervated skin: an immunohistochemical study in the rat. Exp Brain Res 1992; 91: 61-72. 29. Gavazzi I, Andrews TJ, Thrasivoulou C, Cowen T. Influence of target tissues on their innervation in old age: a transplantation study. Neuro Report 1992; 3:717 20. 30. Gavazzi I. Collateral sprouting and responsiveness to nerve growth factor of ageing nearons. Neurosci Lett 1995; 189: 47 50. 31. Calder JS, McAllister RMR. Interpreting the results of unilateral digital nerve repair. J Hand Surg 1993; 18B: 797 9. 32. Beuerman RW, Rozsa AJ. Collateral sprouts are replaced by regenerating neurites in the wounded corneal epithelium. Neurosci Lett 1984; 44:99 104. 33. Jackson PC, Diamond I Regenerating axons reclaim sensory targets from collateral nerve sprouts. Science 1981; 214: 926 8. 34. Slack JR. Interaction between foreign and regenerating axons in axolotl muscle. Brain Res 1978; 146:172 6. 35. Thompson W. Reinnervation of partially denervated rat soleus muscle. Acta PhysioI Scand 1978; 103: 8I 91. 36. Robinson PR Regenerating nerve fibres do not displace the collateral reinnervation of cat teeth. Brain Res 1984; 310: 303-10. 37. Steinberg DR, Koman LA. Factors affecting the results of peripheral nerve repair. In: Gelberman RH, ed.
443 Operative nerve repair and reconstruction. Philadelphia: JB Lippincott, 1991. 38. Jupiter JB. Nerve injury associated with devascularizing trauma. In: Gelberman RH, ed. Operative nerve repair and reconstruction. Philadelphia: JB Lippincott, 1991.
The Authors Uro~ Ah~:anMD, Registrar,
Zoran M. Arne~ MD, PhD, Chief, University Department of Plastic Surgery and Burns, University Medical Centre, Ljubljana, Zalogka 7, 1000 Ljubljana, Slovenia.
Fajko Bajrovi~: MD, MSe Institute of Pathophysiology, Medical Faculty Ljubljana, Slovenia. Martin Janko MD, PhD University Institute of Clinical Neurophysiology, University Medical Centre, Ljubljana, Slovenia. Correspondence to Professor Zoran M. Arne~, University Department of Plastic Surgery and Burns, University Medical Centre, Ljubljana. Paper received 19 August 1997. Accepted 29 April 1998, after revision.