Click for next page ( 390


The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 389
Chapter IX RECOVERY FOLLOWING INJURY TO THE BRACHIAL PLEXUS Frank E. Nulsen and Harry W. Slade A. INTRODUCTION Because the resulting patterns of functional disturbance are so variable and complex, injuries to the brachial plexus seem most adequately approached via the case study and are given separate consideration here for that reason. Each brachial plexus case must be reviewed in detail if one is to draw any conclusion about the degree of motor or sensory recovery that has occurred, and whether this improvement may be attributed to therapy. In all, 117 cases were selected from the follow-up material as having sufficiently clear and consistent longitudinal data on motor and sensory status, and suffi- ciently detailed operative reports, to permit their inclusion for analyis. All cases were characterized by a sufficiently severe and lasting paralysis to result in the undertaking of surgical exploration, in the earliest at 6 weeks, and in 95 percent at 3 months or more, after injury. The indication for exploration, in retrospect, will be discussed but the point to be made here is that all cases in this study were obtained from surgical rosters and had therefore suffered major plexus injuries with sufficient, lasting dys- function to be subjected to surgery. In addition, an initial total paralysis of the arm was so universally reported, except in 3 stab wounds, that a classi- fication of cases according to nerve elements suffering major damage was achieved by arbitrarily considering an element uninvolved when it gave evidence of good motor and sensory function by 2 months after injury. Accordingly, the data on spontaneous recovery in this series can be expected to be less favorable than often reported, because all examples of very early recovery have been discarded from the analysis. In addition to these 3 stab wounds, trauma was by stretch in 5, and by high velocity missiles in the remaining 109 cases. In 10 of this last group a variable amount of initial nerve damage was compounded by subsequent pressure upon nerves by growing arterial aneurysms. Depending upon the location of the nerve lesions and the evidence for continuing dysfunction at 2 months, the patients were classified into the following groups: I. Damage confined to superior trunk or roots (C5, C6)—19 cases. II. Damage confined to middle or inferior trunk or roots (C7, C8* Tl)—3 cases. 389

OCR for page 389
III. Damage to all trunks or roots—11 cases. IV. Damage confined to lateral cord—8 cases. V. Damage confined to posterior cord—6 cases (3 cases IV plus V). VI. Damage confined to medial cord—15 cases (3 cases V plus VI). VII. Damage to all cords—49 cases. The disproportionately low number of cases in group II probably derives from the likelihood of fatality from arterial or pulmonary damage in associa- tion with such injuries. B. REGENERATION Group I Major involvement at the root or trunk level, which was essentially con- fined to the C5 and C6 roots, occurred in 19 cases. Only 5 of these patients showed a major permanent deficit in function secondary to loss of either shoulder abduction or elbow flexion. Suture of C5 or of upper trunk was responsible for restoration of shoulder abduction in 9 of 12 such cases. Suture of both C5 and C6 was associated with recovery of elbow flexion in 4 of 5 cases. Two failures resulted from grafts. Suture made a significant contribution in 10 of the 14 recoveries, while spontaneous regeneration appeared responsible for the other 4 good results. One might suppose that the C5, C6 root sensory loss would be disabling. Actually such loss was invariably incomplete early after injury and rapidly became inconsequential in hampering function of the hand, presumably as the result of overlapping sensory innervation. Group II In 3 cases damage was ultimately confined to the lower roots with good motor function above the elbow and none below. Sensation was preserved in at least the medial or working portion of the hand. Suture of lower roots in 2 cases and lysis in 1 did not result in the recovery of any finger move- ment, and the extremity remained virtually useless. Group III All roots showed prolonged evidence of dysfunction in 11 cases. Damage was by stretch in 5 instances. All of these were explored and resection of scar with suture of one or more nerves was done in 2. In no instance was there recovery below the shoulder level. The remaining 6 were injured by high velocity missiles. Among 5 who were subjected to suture of one or more elements, biceps and deltoid re- covery resulted with upper root sutures, but in every instance the absence of motor function below the elbow was by itself a total functional limitation. One satisfactory recovery occurred in a lysis case in which distal muscles began to function in the third month, and proximal muscles began 4 months after injury. 390

OCR for page 389
Group IV Eight cases had primarily lateral cord involvement. One showed late spontaneous recovery, beginning in biceps at 4 months, while of 7 sutures only 1, done with plasma glue, failed to give useful recovery in the biceps. This appraisal may be overly optimistic since strong elbow flexion can result purely from brachioradialis action and the degree of participation by the biceps is hard to define. At any rate, these sutures made some contribution to the biceps action which resulted in satisfactory function. Loss of the contribution of the lateral cord to median nerve function might be expected to result in some paralysis of median flexors and loss of median sensation. Actually such motor loss was incomplete—even initially each median flexor retained a useful degree of strength. Median sensory loss was also unlikely to be total and in only 2 instances was there a persistent deficit sufficient to cause inability to pick up small objects blindfolded. In pure lateral cord damage, therefore, the deficit is chiefly in biceps paralysis with a high expectation of recovery after suture when this does not occur spontaneously. In 3 cases where the posterior cord was also involved (but medial cord intact), good results were obtained from 2 lateral cord sutures, while the third recovered spontaneously. A real limitation in function remained in only 1 case on the basis of finger extensor paralysis. (See group V.) Group V Major damage limited to the posterior cord was seen in 6 instances, of which only 2 were subjected to suture. In both instances there was a failure to develop useful finger extension while wrist dorsiflexion, elbow extension, and deltoid function were good in 1 and absent in the other. A "pure" axillary lesion suture with plasma glue did not result in deltoid function. In the 3 cases subjected to lysis only, satisfactory recovery, including even the distal finger extensors, occurred in 2, while 1 failed to recover below the triceps level. In all 3 cases failing to develop useful extension of wrist and fingers, it should have been possible to utilize intact flexor muscles to provide this missing extensor function by tendon transfer. This was done in 1 case only and with a partially satisfactory result. Of the 3 cases with recovery from associated lateral cord involvement (discussed under group IV), 1 showed good spontaneous recovery while 2 were subjected to posterior cord suture. Failure in distal recovery only (finger extensors) was treated by flexor tendon transfers in 1 with an excellent result, while the second had the correctible permanent deficit of an inability to open his fingers. Three additional cases had posterior cord and medial cord damage. The latter by itself prevented the recovery of a useful hand so that it is of only academic interest that 2 posterior cord sutures resulted in good elbow and wrist extensor function. Tendon transfers were quite properly not under- taken. 391

OCR for page 389
Group VI Lasting disturbance of medial cord function is of itself inconsistent with useful function of the hand and therefore of the extremity. A total ulnar nerve deficit is combined with loss of that component of median nerve supply necessary for intrinsic hand muscle function. The preservation of median sensation (through lateral cord) and of fair median forearm muscles avails little in the face of this total loss of intrinsic hand muscle action. Of 15 patients with damage confined to the medial cord, 7 had complete suture of this structure with no recovery of intrinsic muscle function in any instance. Despite varying degrees of action in proximal ulnar muscles and occasional recovery of ulnar sensation, functional ratings remained below 30 percent for this group. Four patients correctly subjected to lysis only (as proved by subsequent proximal motor recovery) again failed to develop useful intrinsic muscle action. More favorable results were seen, not as a result of treatment, but because only incomplete damage occurred to this element so vital for hand function. Damage was total for only the ulnar component of the medial cord in 3 patients in whom suture gave low-grade ulnar sensory return and fair proximal muscle function. In these high lesions with median sparing, clawing of the last 2 fingers was not likely to be pronounced and, in one instance, voluntary function returned to the abductor digiti quinti and first dorsal interosseous. A final case, subjected to lysis only, showed useful recovery with the assistance of a tendon transfer for opposition. In the face of good sensation in the median area or working part of the hand, and some action in the forearm flexors and intact extensors, some mechanical device should be provided for abducting fingers and opposing the thumb, functions absent with intrinsic muscle loss. Partial correction of this deficit to the point of useful function was shown to be possible in 4 instances (3 additional from group VII) where this motor loss was not total. Either the fingers functioned sufficiently well so that a concentrated effort could be made to achieve thumb opposition, utilizing an extensor tendon rather than a weakened flexor, or opposition was good and tendon work could be utilized to overcome clawing of fingers. Group VII Since medial cord damage by itself is, in general, an obstacle to the development of useful hand (and therefore arm) function, it can be an- ticipated that in this group with major damage to all 3 cords, suture will seldom be responsible for a practical degree of recovery. Such sutures of 1 or more cords were performed in 41 of 49 cases in this group and in only 6 of the sutured cases could function be rated above 30 percent (hand can be positioned as a weight and can hold some objects forced into it). The degree of recovery possible when all 3 cords are sutured is of some theoretical interest. In 1 case in which all 3 divided cords were sutured at 6 months after injury, final evaluation 5 years later assessed 392

OCR for page 389
deltoid at 50 percent, biceps at 30 percent, triceps at 25 percent, wrist extensors at 10 percent, and finger flexors at 10 percent. All of the muscles above the elbow would have functioned usefully (and strength could have been increased with training), had there been some mechanism for opening fingers and opposing thumb. However, with this major link in arm function missing, the recovery was of little practical advantage. The 6 good cases were all characterized by early recovery of some intrinsic hand muscle function as evidence of relatively intact medial cord function. In 1 case a suture of the lateral cord gave the usual good return in biceps with increasing strength in median flexors and improvement in median sensation (never totally lost), resulting in 80 percent overall function owing to coincident spontaneous recovery in medial and posterior cords. In 4 cases sutures of the posterior cord gave varying degrees of extensor function, always with triceps recovery (and deltoid, if affected). Two with no wrist or finger extension were subjected to flexor tendon transfers. In one, very good extension resulted with an overall functional rating of 80 percent, in the other, extension of fingers was weak because of incomplete recovery of the flexors utilized, but function was nevertheless considered 60 percent. The remaining 2 posterior cord sutures achieved fair wrist extension and overall function could probably have been raised from 40 percent to 70 percent or better had strong flexor muscles been utilized for extension. The final favorable result in this group of 6 had suture of both lateral and posterior cords with the usual optimal regenera- tion in each. Extension of wrist but not of fingers recovered and function was judged as 40 percent. Again the final assessment of flexor strength suggests finger extension could have been accomplished by tendon transfer. In the 8 cases in which operative treatment consisted in lysis only, 3 recovered useful arms spontaneously while 2 did so by virtue of tendon work. Both failed to recover finger extension and good thumb opposition, although intrinsic muscles were functioning and the fingers were not clawed. In each instance, flexors were sufficiently strong to permit establishment of finger extension and strengthening of opposition by tendon transfers with overall function of 70 or 80 percent. Three of the lysed cases failed to regain useful arms. In 1, a lesion in continuity of the medial cord was left alone and its failure to recover was sufficient to prevent useful function (suture would have accomplished no more). In 2, the only significant residual deficit was in extensor function of wrist and fingers, a situation which could presumably have been im- proved by flexor tendon transfers. The foregoing data are summarized in table 225 where results are classi- fied in terms of useful function of the arm and hand as a whole. The conclusion that only 28 of 89 sutures contributed to useful function dis- counts many "successful sutures" because the resulting motor recovery was negated in multiple nerve damage by a missing link necessary for use of the extremity. Further, the contribution of 8 of these sutures to useful function was made possible by the addition of a missing link through 403930—57 27 393

OCR for page 389
* 3 I ^ S § § § •a S i2 •2 a I fl. i 1 d 1 1 1 j 1 1 d 1 1 1" j 3 1 S .it 1 ill il^lllljllll tn i 2 ;§ •taa.v*o2o0'oo•»ooo':u. 9 £ 1 ill §1 gjll 111 g 111 § s ^ 3 ffi X 2; 13 H w H K H H « H h H K & c £ o g u S -O 2 CO CO -I v-* *.,..—* ^ -- ^ ^ £ C o § 2 ° £ •a-o TI c S 9 * * ' fl s .9 J3 J3 . . ^ .2 c- . . j) . . Z . -C s 1 liiiflillllllllil i illi!|ij!i|p!li 1 lii^llllllllllli^ I.ll -- * 2 •o g B en Sffi i •^g « 3 B 28 s 1 •-i +* * f •; •* $ csr- t-mn NO - ^ f h-a 9 05 L 11 & § 3fi | 1 " h 11 2 "S »^^ M2 ? £ i H o 3 •s a I 1 1 3- ^ hH 394

OCR for page 389
tendon transfers. Had such measures been more universally applied, only 8, rather than 75 of the 117 patients, need have had virtually useless extremities and only 24 of the 89 sutures would have been of no practical value. These possibilities for rehabilitation are discussed below. Table 226 has been prepared by combining the data on all 89 sutured cases, whenever an accurate follow-up observation for strength of movement or sensation was obtained, and such function was clearly the result of suture and not of regeneration through a neighboring element of the plexus. Multiple sutures in one patient were particularly useful in this^regard. In occasional single sutures, the recovery of certain muscles, not having dual innervation, was considered to have occurred by this avenue, when the timing of recovery was consistent with suture and inconsistent with spontaneous recovery. Data on median sensation are limited almost entirely to multiple lesions because of overlapping innervation with single nerve injury at any point above the formation of the median nerve. A greater median sensory deficit results from lateral cord than from medial cord division, but even in the former the loss was total in only two instances. Useful sensory recovery is considered present either when reduced pain and touch thresh- olds were described or when, for the median, the patient could pick up small objects blindfolded. Low-grade recovery indicates lesser but definite perception in the appropriate fingertips. Useful shoulder abduction implies the ability to abduct the arm 90°. When only the deltoid is in- volved, such shoulder abduction can occur without its recovery but the rating indicates function of deltoid at least sufficient to contribute to strong abduction. In the case of elbow flexion, the ability to raise the forearm against gravity plus resistance was considered useful. In root or upper trunk sutures all such flexion was the result of regeneration through suture. In lateral cord sutures considerable flexion could be provided by the unaffected brachioradialis when a "useful recovery" indicated the observa- tion of a strong contraction of the biceps as a major contribution to elbow flexion. Triceps and finger extensors offered fewer problems in isolated innervation—action against gravity plus resistance was considered useful. For the wrist and finger flexors the same criteria for useful function were used. Ulnar flexors could be considered the result of isolated medial cord regeneration through suture while median flexors, receiving contribu- tions from both lateral and medial cords, could only be included with dual lesions. The intrinsic hand muscles should behave as a group with medial cord damage and are separated only because of four examples in which only the median or only the ulnar portion of the medial cord was sutured. Useful median intrinsic function implies ability to pick up small objects between thumb and finger. Useful ulnar function indicates ability to abduct the fourth and fifth finger sufficiently to grasp objects. Study of table 226 suggests that suture of the upper trunk or its roots, of the lateral cord, and of the posterior cord, carries with it a reasonably high expectation of useful recovery in the appropriate muscles of the 395

OCR for page 389
shoulder and upper arm. The triceps does less well (46 percent) but even a low-grade recovery in this particular muscle can be of some benefit. It is worthy of note that useful recovery in wrist extensors and flexors does occur, but by no means presages extension of functional innervation to finger flexors and extensors—a very rare occurrence. With the limitation in distal recovery already so well defined, it is not unexpected that useful recovery in intrinsic hand muscles was never observed. Conceivably, low-grade recovery might have been found more often if electrical tests had been generally utilized to demonstrate minimal innervation. Table 226.—Motor and Sensory Recovery Attributable to Suture of a Component of the Brachial Plexus Recovery Specific motor or sensory involvement Total cases Useful Low- grade None Number Percent Median sensation 19 34 20 30 35 31 61 32 16 36 37 37 3 6 16 18 55 80 46 13 21 0 6 6 0 0 3 9 1 2 10 5 11 2 3 6 3 3 13 19 8 4 9 22 37 30 12 28 34 34 Ulnar sensation 11 24 16 4 13 0 1 2 0 0 Triceps Wrist extensors Wrist flexors Finger extensors Median finger flexors Ulnar finger flexors Median intrinsic hand muscles Ulnar intrinsic hand muscles The results of suture might have been better had it been accomplished earlier after injury. However, there is a sufficient number of relatively early sutures (38 by 4 months) to suggest that while the percentage of proximal muscles recovering and their actual strength might well be improved by more uniformly early suture, the limitations are absolute on the recovery of finger flexors and extensors and intrinsic hand muscles. The limitations apply even when regeneration occurs without suture. (See below.) The achievement of a reasonable degree of distal skin sensation through suture is worthy of note and illustrates the marked variability between limitations on sensory and motor recovery. The possibility of such re- covery through suture is of considerable importance in a pure median lesion but is of minor advantage in plexus lesions. Here, those lesions requiring suture to improve sensation include such a major permanent 396

OCR for page 389
deficit in distal motor function, with so little possibility for substitutive tendon transfers, that such sensation contributes little to overall per- formance. Probability of spontaneous regeneration may be studied in table 227. Although all of these 117 cases were surgically explored, and at least 1 element sutured in 89, the assumption is made that lysis does not affect the already instituted regenerative process. Therefore, in both the lysis cases and the single suture cases with multiple nerve involvement, a large number of examples are available for an analysis of this spontaneous process. It has been arbitrarily assumed that a recovery manifest before 2 months speaks for a major degree of neurapraxia, or minor injury, and these frequent early recoveries are not included in the tabulation. When recovery is first seen more than 2 months after injury, a definite permanent residual deficit is the result, although the mechanism of such recovery as does occur may be unclear. Observations were not tabulated when the timing of recovery suggested a contribution by suture or when an absence of negative observations before 2 months failed to disprove the possibility of very early recovery. Table 227 shows, as might be anticipated, a higher recovery rate than obtains after suture. There remains the same serious limitation on distal motor recovery without a comparable infrequency of distal sensory return. This suggests that in partial lesions from missiles sensory axons fare better than motor, a discrepancy not observed in stretch injuries. While in most Table 227.—Motor and Sensory Recovery Occurring Spontaneously or After Lysis Following Brachial Plexus Injury * Total cases Recovery Specific motor or sensory involvement Useful Low- grade None Number Percent Median sensation 47 44 35 52 37 38 27 40 28 48 45 44 23 20 23 41 28 12 7 8 2 13 5 5 49 45 66 79 76 31 26 20 7 27 11 11 11 10 5 3 4 10 4 7 2 18 10 9 13 Ulnar sensation 14 7 8 5 16 16 25 24 17 30 30 Shoulder abductors . . Elbow flexors or biceps . * Triceps Wrist extensors Wrist flexors Finger extensors Median finger flexors . . Ulnar finger flexors 1 Functions recovering within 2 months of injury are not included in this analysis. 397

OCR for page 389
nerve injuries sensory recovery when clearly attributable to function of a given nerve suggests adequate motor recovery, in these high lesions it is frequently an accompaniment of recovery in proximal muscles only and does not preclude distal motor failure. As one proceeds distally, the major decrement in proportion of recoveries is seen to begin below the elbow but never results in the level of improb- ability defined for finger motions of any kind after suture. A marked, and doubtless artificial, difference between median and ulnar flexors arises from a difficulty in analysis. Since the median flexors receive innervation from both lateral and medial cords, observations about them enter this table only when they were totally involved by damage to both medial and lateral cords. Because one of these cords was so often sutured, such spontaneous recovery as was recorded for these muscles could derive from only one-half their potential nerve supply. A practical question concerns the timing of this spontaneous recovery— By what interval after injury does it occur if it is to occur at all? In the case of sensation, a large number of careful longitudinal observations is not available, nor is sensory recovery a primary concern in plexus injuries since it is never the limiting factor in function. It does appear that fingertip sensation can appear as late as a year after injury and still develop to a remarkable extent. In proximal muscles, down to and including the wrist flexors, recovery can begin as late as is consistent with regrowth of axons, e. g., up to 4 months for biceps and 7 months for wrist flexors, and still result in adequate function although an earlier trace of function carries a greater certainty of good strength. Much shorter time intervals obtain for the first evidences of recovery in the distal muscles, if such recovery is to progress to useful function. In none of these cases with useful recovery to finger flexors and extensors or to hand muscles was action in these muscles delayed beyond 3 months after injury. This constant behavior is useful in the early clarification of prognosis although the reasons for it are uncertain. Since one may assume that regeneration, which would require up to 18 months for hand muscles, cannot contribute significantly to the useful distal function that develops early, if ever, how is it possible for neurapraxia to persist so long? One explanation would be an "alienation" of voluntary function continuing long after the potential for neuromuscular function exists. With more frequently encountered distal nerve lesions, it is not unusual to see an absence of voluntary function in intrinsic muscles of hand and foot and in extensors of foot and toes for weeks after nerve stimulation results in muscle contraction. In the personal experience of one of us, 4 plexus injuries with hand muscle paralysis which ultimately showed adequate recovery in these muscles all showed good contractions of these muscles on median and ulnar nerve stimulation when first done from 3 weeks to 6 weeks after injury. Clinical recovery soon followed in this group, but could well have been delayed additional weeks had patients not seen this proof of function. A possibility is that the duration of a block in neuromuscular conduction 398

OCR for page 389
is actually short in all cases—possibly less than 2 weeks—that neurapraxia is compounded by alienation of voluntary function, and that nerve stimula- tion would very early (in the first month) define those brachial plexus lesions which will develop finger movement. In other words, the evidence sug- gests that distal muscles can recover only in the presence of intact axons suffering from temporary dysfunction. C. DISCUSSION Intelligent management of the brachial plexus injury depends upon answers to these questions: 1. How much spontaneous recovery is probable in a given injury and by what time does the end result become clearly defined? 2. To what extent does surgery of the brachial plexus assist in defining prognosis? 3. Under what circumstances does suture of plexus elements contribute to function and when should it be undertaken? 4. When are tendon transfers of value and how early can the indications for them be defined? 5. What additional rehabilitative measures should be considered in a given case? 1. Spontaneous Recovery Most writers on this subject are in agreement that much of the initial total paralysis after brachial plexus injury can be a transient affair (neura- praxia) and that preservation of sensation is likely to be followed by early motor recovery (17). However, the time limits of neurapraxia are not stated beyond "up to a few weeks (70)." Nor has there been any clear definition of the time limit by which a further practical increment in func- tional recovery by any mechanism, including regeneration occurring spon- taneously or after suture, can no longer be expected. Brooks (10) has gone so far as to state that the end result of recovery can be clearly antici- pated by 9 months. On the other hand, Barnes (3) in a review of 63 cases, has made observations similar to ours; he concludes that, while proxi- mal motor recovery can begin as late as 6 months and progress to good strength, distal muscles not working by 2 months will never regain useful function. A review of the management of the present series shows that no time limitation for recovery was generally accepted. As a result, patients with useless hands were continued on elaborate programs of physiotherapy up to 2 or 3 years so as to have joints and muscles in good condition when recovery should finally occur. This impression of long intervals necessary for clarification of recoverability has led to the advice that all plexus injuries should be explored to secure an early definition of prognosis. With the general experience in more distal nerve lesions that paralysis of a muscle can exist for as long as 6 or even 9 months, and yet be replaced by a degree of action that contributes to function, it would seem logical to project such data to the brachial plexus injury and to assume that the same adequate regenerative process with longer distances to grow migh 399

OCR for page 389
require as much as 18 months to set up intrinsic hand muscle function— that paralysis here should not be written off as permanent until such elapsed time since injury or suture. Further basis for optimism was found in the concept that any function in a given area was likely to be followed by more function, regardless of time, so that even 5 years after injury electromyog- raphers who found motor units in a noncontracting muscle urged resump- tion of intensive physiotherapy. The data collected in this study seem sufficiently significant to disprove many of these viewpoints, although it must be recognized that they were not proposed on the basis of fact but reflect a sincere desire to postpone, beyond any possible time limit for recovery, the discontinuation of any therapy that might improve a disabling situation. It is apparent that there is an absolute limit to the distance below a lesion at which regeneration of axons can contribute a useful degree of motor recovery. Any finger move- ments developing to a useful extent, whether from forearm muscles or intrinsic hand muscles, have achieved an end point in recovery long before they could be affected by axon regrowth. From the data available, it is established that such distal muscle recovery is heralded by some motor function within 3 months of injury. It can be further proposed (but not proved) that more frequent early clinical observations among these cases might have shortened this time limit—that an even further shortening could be achieved by general use of nerve stimulation which has demonstrated (in a number of cases perhaps not significant statistically) the presence of distal muscle contractions by 1 month if useful function is to result. Cer- tainly the permanently useless hand, found as an end result in 65 of these cases, was defined at 3 months by the persistence of distal motor paralysis. Early clarification of the potentials for hand function should result in a realistic pessimism. It is precisely in this same area where nerve surgery will not alter the outcome. It is an area of indispensable function whose absence can largely negate the value of proximal motor function achieved spontaneously or by suture. The proximal muscles can, paradoxically, follow a slower schedule of recovery than distal muscles and progress to useful function because here the slow-going process of axonal regrowth can be effective. Examples of ultimately useful recovery begin as late as 4 months for deltoid, 5 months for biceps, 6 months for triceps, and 7 months for wrist extension. When the early definition of hand function has been favorable, these findings have an important bearing on management while the degree of proximal function has less significance in the face of a useless hand. 2. Surgery To Determine Prognosis Some writers (3, 17) maintain that even when surgery fails to improve nerve regeneration it has been useful in cases of severe paralysis as a means of defining prognosis for recovery so as to permit more intelligent manage- ment. However, Brooks (10) has demonstrated that, except in the case of gross nerve disruption, the surgeon's description of lesions in continuity 400

OCR for page 389
bears little relation to subsequent regeneration. Similar experience is described in the section of this monograph which deals with the neuroma in continuity. Furthermore, it has now become clear from our data that prognosis for hand function becomes established no later than 3 months by simple motor examination, which should ideally include electrical stimula- tion of the various peripheral nerves. The more slowly determined status of proximal muscles might well call for earlier definition by exploration, provided there will be a useful hand for these muscles to position. Since these are the same muscles whose function can be established by suture, it seems proper to dismiss this question by stating that the only necessary surgery is that which can improve function. Those elements which do not yield function through suture need not be inspected for their degree of continuity, even if this were always helpful, since their status is very clearly defined by distal physiologic events. A further means has been found for the demonstration of root avulsion in stretch injuries, namely, the performance of a cervical myelogram. When this demonstrates extravasation of the dye into extraspinal pockets, the impossibility for recovery is defined for certain roots. While this test is of negative value only, it permits definition of the hopeless prognosis at a very early date. 3. Surgery of the Plexus as a Therapeutic Measure. One question to be considered is whether neurolysis, or the release of external constricting scar, influences nerve function. When recovery follows such procedures, a cause and effect relationship is often presumed (24, 27, 76). Davis and Martin (17) have gone so far as to imply that lysis is important even with partial nerve function because "subsequent ex- tensive scar tissue formation tends to impair to various degrees many originally uninjured portions of the plexus and gives rise to disseminated and incomplete motor and sensory disturbances." Certainly this external scarring can be so dense as to make the surgeon believe he must be per- forming a useful function in liberating the encased nerve. However, the absence of one well-documented case where there has been a clear regression of function, without further external trauma or growth of an aneurysm, forces the conclusion that peripheral nerve function does not deteriorate from progressive scarring, logical as this idea might seem. The causal relationship of neuroloysis to subsequent recovery is more difficult to evaluate. However, whenever good longitudinal data on motor and sensory status are available, such increments in function have occurred at a time consistent with the usual schedule for regeneration obtaining with refer- ence to the date of injury rather than the date of surgery. One is therefore inclined to agree with the statement of Barnes (3): "It is important to appreciate it is intraneural damage and not extraneural scarring caused by associated soft tissue injury which is the barrier to recovery." The as- sumption is therefore made that, when surgery of the plexus is performed, only resection and suture alter the course of regeneration. 401

OCR for page 389
Before any discussion of what can be accomplished by suture, it should be stressed that this procedure should never be elected at surgical exploration in the absence of gross evidence for nerve discontinuity, unless sufficient time has elapsed to establish that the lesion is inconsistent with recovery. In early resections of lesions in continuity only histologic study will deter- mine whether this was done in the face of a high degree of axonal continuity. Resection has clearly been unwisely undertaken when it destroys partial distal function which will never return (4 cases in this series). Except for those who believe nerve function is favorably influenced by lysis, it is generally agreed that plexus surgery is of no benefit in stretch injuries (10, 27, 76). Neither has suture of disrupted nerve ends nor re- section and suture of the segment most damaged ever been reported to yield function. Such was the case in two stretch injuries in this series in which suture was undertaken. Opinions concerning the efficacy of nerve suture in the plexus after open wounding show considerable variance. Most enthusiastic reports deal with isolated cases subjected to early operation and suture of one ele- ment only, where subsequent improvement was not clearly attributable to the suture. Wider experience has led to conclusions such as that of Brooks (10) that "exploration of open wounds of the brachial plexus is rarely profitable or justifiable." These conclusions from the British experience with World War II injuries actually are based on only 11 sutured cases of which only 1 derived useful recovery consisting in shoulder abduc- tion and elbow flexion after suture of the upper trunk. In this series the possible contribution of suture to function is defined by noting those muscles capable of recovery (table 226). Suture of the following elements carries a high expectation of useful innervation: upper trunk or either of its roots for supraspinatus, deltoid, and biceps; lateral cord for biceps and a partial contribution to median forearm flexors; posterior cord for deltoid, triceps, and, with luck, wrist extensors. Suture of both lateral and posterior cords has resulted in a useful extremity when followed by tendon transfers for extension. While recovery of sensation to fingertips can occur through suture, the distal motor loss with those plexus injuries causing total sensory loss is so disabling as to negate the value of sensation gained by suture. Suture of the isolated median nerve injury below the plexus for its sensory contribution is another matter. Suture of the lower roots or of the medial cord is not capable of con- tributing to useful function except when the damage is partial. If either the ulnar half or the median half of the cord is functioning, suture of the damaged element has been of value. In considering the individual patient or his extremity as a whole, the value of suture, undertaken to achieve proximal function, is questionable in the face of permanent distal loss. If one reviews the 89 cases subjected to suture, designating those who achieved a functioning arm rather than just some new muscles after surgery, the number is discouragingly small— 28. At the same time, one has the feeling that rehabilitation possibilities 402

OCR for page 389
were missed, measures whose application would have turned more of these proximal recoveries to practical use in the overall function. Despite the certain limitations on recovery from suture of the inferior portions of the plexus, the situations where surgery can contribute are clearly defined, and in these same situations the contribution is probable rather than possible. An early time for plexus surgery is urged by some writers (17, 27) on the basis that: (a) earlier suture is accomplished when badly damaged nerves are discovered (with better regeneration as a result); and (b) exploration is technically easier and less likely to result in damage to normal nerves before dense scarring has occurred. At the other pole are those (76) who feel that suture has so little to offer, in comparison with spontaneous recovery, that no part of the plexus should be even handled until its po- tential for recovery has been clarified by 6 to 9 months of observation. The material under consideration presents one urgent argument for early exploration of injuries at the cord (rather than the root) level, namely, the rather frequent finding of a traumatic aneurysm of the axillary artery in proximity to nerve elements and causing an increasing involvement of axons with the passage of time. Ten such aneurysms were found in this series. The nerves showing dysfunction were doubtless affected by the initial wound but were directly impinged upon and often thinned out by the aneurysm. In 3 of these cases a partial nerve deficit progressed with the passage of time but in the remaining 7 the paralysis was present from the time of wounding. However, there is still evidence in the similar operative findings for the two groups that the growing aneurysm caused structural damage to nerves, quite probably superimposing axonal dis- ruption upon an initial neurapraxia. In the absence of progressing paralysis, because of masking neurapraxia, these aneurysms were still recognized by increasing axillary mass or bruit in 5 cases. However, in the remaining 2, the aneurysm was an unanticipated finding when surgery was undertaken solely for purposes of plexus exploration. The medial cord was involved in 8 cases and served by itself to prevent functional recovery, whether sutured or simply lysed. The lateral head of the median was involved in 5 of these cases. Surprisingly, in 5 instances the posterior cord was affected. In 2, where the medial cord was intact, posterior cord suture permitted useful recovery with the assistance of tendon transfers. In summary, 8 of the 10 aneurysms ended up with useless arms, while 2 had serious residual deficits. In the 2 cases with poor end results who developed paralysis late (at 10 days and 17 days) earlier intervention would clearly have been "arm saving." It is certainly possible that the other 8, none of whom suffered actual nerve division from the original wounding, would have fared far better with early attention to the aneurysm, which was delayed from 1 to 7 months. Since aneurysm was found unexpectedly in 2 instances (both at 5 months) there may be an argument for exploring all cord lesions early on the possibility of finding this progressively destruc- tive lesion. However, the unexpected aneurysm remains a rarity and is 403

OCR for page 389
probably diagnosable if looked for repeatedly in every plexus cord lesion by combining careful palpation and auscultation of the wounded area with the neurological evaluation in all patients routed to the "nerve specialists." If suture will make a significant contribution to function, its early per- formance will both shorten the period of dysfunction and improve the ultimate result, although any proximal muscles capable of useful recovery after suture do not appear to show a critical reduction in strength up to 6 months of delay in suture. Good proximal recovery was seen after 8 months delay but the proportion of recoveries decreased at 6 months. In addition, the technical difficulty of exploration increases with the pas- sage of time. The following views as to timing of brachial plexus explora- tion seem tenable* (a) If paralysis is limited to proximal muscles, explore the appropriate roots or cords early because the finding of any gross discontinuity and its correction by suture will give the best possible result. However, any lesion in gross continuity must be left alone with the neces- sity for a second operation to resect and suture if recovery does not occur in 3 to 6 months, depending on the distance for regrowth to muscle* (b) If paralysis involves those distal muscles which will be uninfluenced by surgery, and proximal muscles whose isolated recovery will contribute little, sur- gery can be deferred for 3 months to determine whether distal recovery, even with tendon transfers, is consistent with hand function. In other words, the question must be settled whether a gain in proximal motor function will be useful to the patient. 4. The Use of Tendon Transfers The literature on tendon transfers for the improvement of motor deficits is extensive and is particularly well reviewed by Bunnell (12), Littler (43), and Brooks (9). Nevertheless, this often all-important link in therapy is seldom stressed in writings about the overall management of nerve injuries. Failure to institute these important rehabilitation measures in this series stemmed from such factors as the extensiveness of motor deficit, false opti- mism concerning the possibility of late improvement in function from nerve regrowth until elapsed time forced a halt to all therapy by closing of hos- pitals or discouragement of patients, plus lack of realization of the neuro- surgeon concerning feasible alternatives to nerve function, coupled with unavailability of surgeons skilled in these procedures in many centers where nerve injuries were treated. Examples of the situations where tendon transfers are of benefit have been given and are summarized in table 225. The clearest indication is the isolated failure of wrist and finger extensor recovery. This indication existed in 3 cases of primary posterior cord damage and in 12 more cases of multiple nerve involvement where reasonable recovery occurred in the other ele- ments. Actually, only 6 of these 15 cases were given the benefit of this procedure with good results in 5. The remaining cases might have been given useful hands with the occasional requirement that the procedure be 404

OCR for page 389
accompanied with a wrist fusion, where weakened flexor tendons were inadequate to extend both wrist and fingers. The other opportunities for striking improvement of hand function have occurred with partial loss of intrinsic hand function in partial medial cord lesions. Three cases with such incomplete loss and inadequate thumb opposition were given this function by tendon transfer while an apparent indication was neglected in 2 others. In 2 cases with good forearm flexors and sensation, both extensor and opposition transfers were done with achievement of both satisfactory finger extension and thumb opposition. Two additional men with partial medial cord lesions, with clawing of fourth and fifth fingers, were enabled to open these fingers by tendon transfers to stabilize the metacarpophalangeal joints. So much for the clear indications for tendon transfer. In 10 pure medial cord lesions and in 5 multiple lesions, the residual deficit was total loss of intrinsic hand muscle function in the face of good sensation and working forearm muscles. However, too few of these forearm muscles were con- sidered sufficiently strong to provide both opposition of thumb and stabiliza- tion of curled fingers to oppose the thumb. In these situations it would appear that appropriate joint fusions combined with tendon transfers would have resulted in at least a thumb-to-finger pincers action for handling objects, as a marked improvement over the useless clawed hand which could be employed only as a paper weight. In considering the possibilities for tendon substitutions, sources for failure should be stressed. Experience with procedures proven feasible, and with technical details of such surgery, is essential to success, as well as the recogni- tion that minimal residual weakness in muscles serving as motors causes a major reduction in their effectiveness. In three failures in this series, either another strong muscle should have been chosen or, failing such choice, the weakened muscle might have contributed to function with its load reduced by joint fusions. As to timing, where such a good guess can be made as to permanent deficits in hand function at 3 months, and partial function reaches a limit in im- provement soon thereafter, there seems little reason why these procedures cannot be recognized as essential to function and undertaken by 6 months. 5. Other Rehabilitative Measures Maintenance of unrestricted joint movement by physiotherapy and avoid- ance of overstretching of inert muscles by splinting are necessary considera- tions in management during a temporary period of motor paralysis. The great difficulty in many plexus injuries has been the institution of these measures without any plan by doctor or patient as to their duration. On the one hand is seen the frequent patient whose major energies over a 2-year period or more were devoted to the nursing of what was to be always a useless extremity. In such individuals the attention of well-wishing ther- apists served only to increase frustration and pain complaints and to block the direction of energy to a realistic adjustment and compensation for a 405

OCR for page 389
permanent handicap. If we can now accept the evidence that the prognosis for function, at least for the all-important distal muscles, becomes defined by 3 months, there should be little reason for inability to define what will be useful therapy at a relatively early period. On the other hand, one finds patients (at least eight in this study) in whom concentration on such rehabilitation measures might have allowed real use of an arm, either by themselves or with the assistance of tendon transfers. In particular, the less frequent recovery of shoulder abduction in com- parison with elbow flexion appears not to depend upon any less frequent motor recovery. Rather, restricted shoulder joint movement coupled with overstretching of paralyzed muscles appears to account for this difference in results, presumably remediable by greater attention to shoulder mobility and splinting in abduction. Pain rarely constituted a barrier to rehabilitation in these patients with major nerve damage. True causalgia, relieved by sympathectomy, was found in only two cases with minimal involvement of medial or lateral cord. Prolonged and distressing pain, often requiring narcotics and not relieved by sympathetic blocking, was an occasional but by no means constant manifestation in total nerve involvement. Such discomfort frequently had the characteristics of phantom limb pain, with referral to totally anesthetic areas of the extremity. Late follow-up indicated that this pain had gradually subsided over a period of months without relation to any specific therapy except in two instances where amputation was followed by marked pain relief. Loss of shoulder abduction was an isolated handicap in 4 cases. (See table 225.) Considerable improvement in the ability to position the arm has been obtained by both shoulder fusions and tendon procedures (2, 16, 48, 58), none of which was utilized in these cases. Loss of elbow flexion as observed in four cases might have been corrected by a number of procedures. British authors (2) indicate that the Steindler operation has not worked favorably. Seddon (69) has had dramatic success in securing flexion by utilizing the lower portion of the pectoralis major as a motor for the biceps. Fortunately, this is usually spared in superior plexus injuries involving biceps. Hendry (34) suggests another alternative which he claims is superior to elbow fusion; the performance of a posterior bone block at the elbow which checks extension at a given limit but still allows a range of further passive flexion for use in placing the hand on a table while sitting, etc. Again such measures did not receive serious consideration in these patients. There were some 32 patients in this group who had sufficient proximal function to provide good movement at shoulder and elbow but such total or near total paralysis of forearm and hand that no combination of tendon transfers and fusions could have given even a useful pincers. Accordingly, the arm was of little use despite this excellent proximal action. One possibility for achieving function of the arm would have been to amputate the hand with substitution of a prosthesis having a pincers action controlled 406

OCR for page 389
by the opposite shoulder. Certainly patients with forearm amputation at initial injury ended up with a far more useful arm simply because this amputation, which physicians disliked suggesting and patients were loath to accept, had occurred fortuitously. As a result, prompt training in the use of a prosthesis was immediately undertaken. Another alternative for the flail hand, provided there are mobile joints and enough proximal forearm innervation to allow at least weak pronation and supination, is suggested by Hendry (34). Flexor and extensor tendons are fused to bone above the wrist in such a way that when supination causes the hand to dorsiflex (by gravity) the fingers close, while turning the arm into pronation drops the wrist with opening of the fingers. It is claimed that patients can attain fair facility in handling light objects and will prefer such a mechanism to a hand prosthesis. Finally, there were 8 patients (5 were stretch injuries) with such total paralysis of entire arm and shoulder that amputation below the shoulder joint might actually have resulted in better rehabilitation. Even if one neglects the long period during which active treatment to regain arm func- tion was fruitlessly pursued and considers the end result, he is still impressed by the amount of energy which the patient must devote to nursing this dead weight, which not only has no function but interferes with many activities by being in the way. The 4 patients who accepted such ampu- tation expressed no regrets, but rather satisfaction at being freed of this encumbrance and, in 2 instances, attributed to the procedure a marked relief in pain. Such then are some of the rehabilitation measures that might have been more widely applied had there been a clearer realization of the early ability to predict the limitations in nerve regeneration in each case and the avail- able measures for at least partially substituting for nerve function. Refer- ence to table 225 indicates that, of the 75 poor results, 20 could have achieved relatively useful arms through tendon transfers and joint fusions, 15 might have had at least a pincers action for grasping objects instead of clawed hand, and 32 more could have put their wounded extremity to some use by substituting a prosthesis for an inert forearm. This leaves only 8 un- salvageable situations, 5 of them stretch injuries where high amputation might be indicated, and was actually of considerable benefit in 4. It must be remembered that how much a patient will use his hand can be vastly affected by providing only one missing link in its function. D. CONCLUSIONS 1. Suture of upper plexus elements carries a high probability of re- storing useful function to muscles of shoulder and upper arm, except in stretch injuries. 2. The major disability in brachial plexus lesions occurs when its lower roots or medial cord elements are involved with loss of hand function. 3. Such deficits are not improved by surgery of the plexus. Recovery of hand function must depend upon spontaneous recovery which becomes 407

OCR for page 389
defined by 3 months, since it cannot be assisted by axonal regeneration. 4. When such deficits are incomplete, yet disabling, tendon transfers, sometimes in combination with joint fusions, may permit useful function. These procedures, when indicated, can be correctly planned and under- taken as early as 3 months and never later than 6 months after injury. 5. An understanding of the time limitations upon useful recovery should lead to earlier utilization of all possible accessory rehabilitative measures with improvement in results and in 75 percent curtailment of time necessary for active treatment. 403