Results

The LSUHSC system for evaluating outcomes in plexus lesions was used (Box 1). This system

Box 1. System of grading by elements of brachial plexus

0: no muscle contraction

1 (poor): proximal muscles contract but not against gravity

2 (fair): proximal muscles contract against gravity, distal muscles do not contract; sensory grade, if applicable, was usually 2 or lower

3 (moderate): proximal muscles contract against gravity and some resistance; some distal muscles contract against gravity; sensory grade, if applicable, was usually 3

4 (good): all muscles contract against gravity and some resistance; sensory grade, if applicable, was 3 or 4

5 (excellent): all muscles contract against moderate resistance; sensory grade, if applicable, was 4 or better differs from the Medical Research Council (MRC) system in that LSUHSC grade 3 includes at least some muscle contraction against mild resistance and gravity, whereas a grade 2 is against gravity only, which would correspond to an MRC grade 3. Favorable outcomes were therefore determined in patients with elements recovering to an LSUHSC grade of 3 or better level not to an MRC grade 3.

Lacerations

Outcomes in which the mechanism was from presumed laceration by glass, knife, or other sharp object (sharp) versus propeller blades, chain saws, or auto metal (blunt) are seen in Table 2. Data combine unfavorable with favorable elements for repair by suture versus grafts, or neurolysis based on NAPs across contused but not lacerated elements. Thus, outcomes with lower plexus elements, such as C8, T1 spinal nerves and medial cord, were blended with those with more favorable elements for repair.

Table 2 shows reasonably favorable outcomes for suture: 81% of elements recovered to a grade 3 or better level if the laceration was sharp and the repair was done within 72 hours, and 70% of elements recovered to this level after delayed or secondary end-to-end suture. Delayed repair by grafts had grade 3 or better outcomes in 53% of elements. If sharp transection could be explored

Table 2

Surgical outcome in brachial plexus lacerationsa

Table 2

Surgical outcome in brachial plexus lacerationsa

Elements in continuity

Sharp transection

Blunt transection

Totals

Plexus cases

20

28

23

71

Plexus elements

57

83

61

201

Neurolysis (+NAPs)

24/26

0/0

0/0

24/26 (92%)

Primary suture

0/0

25/31

0/0

25/31 (81%)

Secondary suture

9/7

12/8

3/5

18/26 (70%)

Secondary graft

22/17

21/40

25/56

63/118 (53%)

Total elements

48/57

54/83

28/61

130/201 (65%)

Results are given as number of elements/total elements recovering to grade 3 or better (LSUHSC system). Primary = repair within 72 hours of injury; secondary = delayed repair, usually after several weeks. a N = 71.

Results are given as number of elements/total elements recovering to grade 3 or better (LSUHSC system). Primary = repair within 72 hours of injury; secondary = delayed repair, usually after several weeks. a N = 71.

acutely, end-to-end suture repair was usually possible and gave very good results. Blunt injuries had delayed repair so that the extent of trimming necessary to reach healthy tissue could be determined. Grafts therefore were usually necessary. Outcomes with blunt injuries were less favorable than with sharp injuries because these were more serious, with longitudinal injury more likely requiring grafts for repair rather than end-to-end suture. This series and the GSW series show that overall results are better with end-to-end suture than with graft repairs provided the circumstances are right so that this procedure can be performed without undue tension.

Despite a potentially transecting mechanism, some elements even with severe or complete loss in their distribution preoperatively were in continuity. They were managed, when possible, like other presumed plexus lesions in continuity by NAP recordings across the lesion; based on NAP findings, they were treated with either neurolysis or resection and repair. The latter was usually treated by grafts or less frequently by suture. If NAPs were recorded across the lesion, only a neurolysis was performed. The outcome in this subset of lesions in continuity associated with lacerating injury was favorable, with 47 of 58 elements recovering to grade 3 or better function.

Gunshot wounds

An analysis of the plexus lesions injured by GSW was also interesting. Table 3 shows the large role played by graft repair. Again, less favorable elements for repair are grouped together with those more favorable for repair. End-to-end suture was superior in terms of restitution to grade 3 or better function in 19 of 27 elements (70%) versus grafts in 75 of 138 elements (54%). This finding showed that grafts were necessary in lengthier and thus more severe injuries, but again, if circumstances permitted, outcomes with end-to-end suture were superior to those with graft repair. NAP recording was extremely important in this group of patients because 46 of 202 (20%) elements, which were believed to have complete loss by clinical and EMG criteria preoperatively, transmitted operative NAPs across their lesion, indicating regeneration; therefore, these elements could receive neurolysis only with the expected good results. Of equal importance, nine elements believed to have incomplete injury did not transmit an NAP, and these had repair by either suture (six cases) or graft (three cases). Variations in anatomy found at the time of operation, particularly for some infraclavicular GSW injuries, accounted for most of these latter discrepancies.

The best results with grafts were for C5, C6, and upper trunk lesions, but also for C7 and middle trunk injuries. At an infraclavicular level,

Table 3

Surgical outcome in gunshot wound injuriesa

Table 3

Surgical outcome in gunshot wound injuriesa

Type of lesion

Neurolysis (+NAPs)

Suture

Graft

Elements with

42/46

14/21

73/135

complete loss

(91%)

(67%)

(54%)

(202 elements)

Elements with

78/82

5/6

2/3

incomplete loss

(95%)

(83%)

(67%)

(91 elements)

Total (293 elements)

120/128

19/27

75/138

(94%)

(70%)

(54%)

Results are given as number of elements/total elements recovering to grade 3 or better (see Box 1).

NAPs indicate increased conduction across a lesion

Results are given as number of elements/total elements recovering to grade 3 or better (see Box 1).

NAPs indicate increased conduction across a lesion in continuity. a N = 118.

repairs of the lateral cord and its outflows and the posterior cord to the axillary nerve did well. Repairs of the posterior cord to the radial nerve did less well, however. Furthermore, repairs of the medial cord to the median nerve also did well. Repairs to the C8 T1 to lower trunk and medial cord and medial cord to ulnar did poorly unless partially injured or shown to be regenerating by NAP recordings.

Stretch/contusion injuries

Earlier, before the authors began to add transfers to their armamentarium for repair of stretch injures, they would try for direct repair of as many elements as possible but would also use descending cervical plexus as a transfer but not accessory, medial pectoral branches or intercostals until later in the series. Figs. 1 through 3 depict the use of NAP recordings and some of the resultant direct repairs on C5, C6 (C7) cases.

Tables 4 through 6 show outcomes using grafts for direct plexus repair for various stretch/contusion patterns. Descending cervical plexus also had been used to add to the direct repairs in 13% of the C5 and C6 repairs and 27% of the C5, C6, and C7 repairs. Tables 4 and 5, which summarize outcomes for C5/C6 and C5, C6, and C7 stretches, show how the pattern of avulsions and whether they affected C5 or C6 or both changed outcomes. The additional value of transfer of accessory for shoulder and medial pectoral or other transfer for biceps in a smaller cohort of patients operated on between 1998 and 2000 can be seen by these tables. Nonetheless, it can also be appreciated that direct repairs by grafts had much to offer.

Since 2000, the authors have added transfers to most direct plexus repairs for stretch injuries. Accessory has been transferred to suprascapular nerve unless there is evidence of regeneration to it or C5 is usable as a lead-out for grafts. When this has been the case, most of the C5 grafts usually have been led to the posterior division of upper trunk. The authors favor, when possible, for C5, C6, and C7 stretch injuries, the use of medial pectoral branches to add to any direct repair they gain for elements leading to musculocutaneous

i=£> Neurolysis C5, C6, C7 to divisions/trunks

* 1/6th of patients with C5, C6 pattern of loss have a variable degree of injury to C7 to middle trunk (MT)

S = stimulus site

R = recording site for evoked Nerve Action Potentials (NAP's)

UT, MT, LT = upper, middle and lower trunks sen = suprascapular nerve men = musculocutaneous nerve

Antebr. cut. n. = antebrachial cutaneous nerve

Fig. 1. C5, C6 (C7) pattern of loss. One sixth of patients with C5, C6 pattern of loss have a variable degree of injury to C7 to middle trunk (MT). Antebr. cut. n., antebrachial cutaneous nerve; Lat cd, lateral cord; LT, lower trunk; Med cd, medial cord; mcn, musculocutaneous nerve; Post cd, posterior cord; R, recording site for evoked NAPs; S, stimulus site; scn, suprascapular nerve; UT, upper trunk.

Fig. 2. C5, C6 (C7) pattern of loss. Antebr. cut. n., antebrachial cutaneous nerve; Lat cd, lateral cord; LT, lower trunk; Med cd, medial cord; mcn, musculocutaneous nerve; MT, middle trunk; Post cd, posterior cord; R, recording site for evoked NAPs; S, stimulus site; scn, suprascapular nerve; UT, upper trunk.

Fig. 2. C5, C6 (C7) pattern of loss. Antebr. cut. n., antebrachial cutaneous nerve; Lat cd, lateral cord; LT, lower trunk; Med cd, medial cord; mcn, musculocutaneous nerve; MT, middle trunk; Post cd, posterior cord; R, recording site for evoked NAPs; S, stimulus site; scn, suprascapular nerve; UT, upper trunk.

Fig. 3. C5, C6 (C7) pattern of loss. Antebr. cut. n., antebrachial cutaneous nerve; Lat cd, lateral cord; LT, lower trunk; Med cd, medial cord; mcn, musculocutaneous nerve; MT, middle trunk; Post cd, posterior cord; R, recording site for evoked NAPs; S, stimulus site; scn, suprascapular nerve; UT, upper trunk.

Fig. 3. C5, C6 (C7) pattern of loss. Antebr. cut. n., antebrachial cutaneous nerve; Lat cd, lateral cord; LT, lower trunk; Med cd, medial cord; mcn, musculocutaneous nerve; MT, middle trunk; Post cd, posterior cord; R, recording site for evoked NAPs; S, stimulus site; scn, suprascapular nerve; UT, upper trunk.

Table 4

Operative results of C5, C6 stretch injuries with complete loss

Table 4

Operative results of C5, C6 stretch injuries with complete loss

No. patients Results

Operation

5 )

(averaged)

Grafts C5, C6

34

Grade 3 = 12;

grade 3-4 = 7;

grade 4 = 8

C5 grafts, C6 avulsed,

5

Cervical plexus

descendinga

used grade 2-3

C5 avulsed, C6 grafts,

2

Grade 3-4

descending cervical

plexus usedb

Neurolysis of C5 andC6 12

Grade 3-4

(positive NAPs)

C5 neurolysis, C6 grafts

2

Grade 3-4

a These cases might have done better had medial pectoral branch transfer to musculocutaneous nerve been added. Four cases done between 1998 and 2000 in this fashion had averaged outcomes of 4.

b These cases might have done better had accessory nerve transfer to suprascapular nerve been added. Two cases done between 1998 and 2000 in this fashion had averaged outcomes of 3.8.

nerve. Figs. 4 through 6 depict the use of nerve transfers to bolster potential outcomes from direct repairs in patients with C5, C6 (C7) patterns of loss.

Recently, one of the current authors (R.L.T.) has begun to use the Oberlin transfer of a fascicle of the ulnar nerve to the musculocutaneous nerve. This experience, although encouraging, is too nascent to report, but others have found this to be a favorable transfer. When there has been proximal direct repair with possible lead-in to musculocutaneous nerve, the latter is split in

Table 5

Operative results of C5, C6 and C7 stretch injuries with complete loss

Table 5

Operative results of C5, C6 and C7 stretch injuries with complete loss

No. patients

Results

Operation

7 )

(averaged)

Grafts of C5, C6 and C7

31

Grade >3-4

C5 grafts, C6 and C7

10

Grade 2-3

avulsed, descending

cervical plexus useda

C5 and C6 grafts,

10

Grade 3

C7 avulsed, descending

plexus used

C5 avulsed, C6 and

6

Grade 3-4

C7 grafts, descending

plexus usedb

Neurolysis C5, C6 and

18

Grade 4

C7 (positive NAPs)

a These cases might have done better had medial pectoral branch transfer to musculocutaneous nerve been added. Three cases done between 1998 and 2000 in this fashion had averaged outcome of 3.7.

b These cases might have done better had distal accessory nerve been transferred to suprascapular nerve as an additional step. Two cases done between 1998 and 2000 in this fashion had averaged outcome of 3.5.

a longitudinal direction. One distal half is then used for lead-in from medial pectoral branches or ulnar nerve, whereas the other half is left intact to receive proximal down-flow. If possible, either transfer is done without an intervening graft because the use of an intervening graft seems to decrease results. This type of "pants-over-vest" approach seems to make sense because it at least theoretically maximizes the number of axons brought to bear on a denervated structure.

Table 6

Overall postoperative grades on patients with supraclavicular plexus stretch injuries patientsa

Postoperative grade

Table 6

Overall postoperative grades on patients with supraclavicular plexus stretch injuries patientsa

Postoperative grade

Initial loss

0

1

1-2

2

2-3

3

3-4

4

4-5

Totals

C5/C6 (C)

1

0

0

0

6

14

18

9

7

55

C5/C6/C7 (C)

1

0

0

2

14

23

20

9

6

75

C5 to T1 (C)

21

8

15

29

62

40

20

13

0

208

C5 to C8 (C)

0

0

0

0

0

1

1

0

0

2

C6/C7/C8/T1 (C)

0

0

0

1

0

2

1

0

0

4

C7 to T1 (C)

0

0

0

0

1

1

0

0

0

2

C8 to T1 (C)

2

3

2

2

0

2

0

0

0

11

C8 to T1 (I)

0

0

0

0

0

0

2

2

3

7

C7/C8/T1 (I)

0

0

0

0

0

1

1

0

0

2

Total

25

11

17

34

83

84

63

33

16

366

C = complete or nearly complete loss; I = incomplete loss. a N = 366.

C = complete or nearly complete loss; I = incomplete loss. a N = 366.

Fig. 4. C5, C6 (C7) pattern ofloss. Antebr. cut. n., antebrachial cutaneous nerve; Lat cd, lateral cord; LT, lower trunk; Med cd, medial cord; mcn, musculocutaneous nerve; MT, middle trunk; Post cd, posterior cord; R, recording site for evoked NAPs; S, stimulus site; scn/SCN, suprascapular nerve; UT, upper trunk.

Fig. 4. C5, C6 (C7) pattern ofloss. Antebr. cut. n., antebrachial cutaneous nerve; Lat cd, lateral cord; LT, lower trunk; Med cd, medial cord; mcn, musculocutaneous nerve; MT, middle trunk; Post cd, posterior cord; R, recording site for evoked NAPs; S, stimulus site; scn/SCN, suprascapular nerve; UT, upper trunk.

Fig. 5. C5, C6 (C7) pattern ofloss. Antebr. cut. n., antebrachial cutaneous nerve; Lat cd, lateral cord; LT, lower trunk; Med cd, medial cord; mcn, musculocutaneous nerve; MT, middle trunk; Post cd, posterior cord; R, recording site for evoked NAPs; S, stimulus site; scn/SCN, suprascapular nerve; UT, upper trunk.

Fig. 5. C5, C6 (C7) pattern ofloss. Antebr. cut. n., antebrachial cutaneous nerve; Lat cd, lateral cord; LT, lower trunk; Med cd, medial cord; mcn, musculocutaneous nerve; MT, middle trunk; Post cd, posterior cord; R, recording site for evoked NAPs; S, stimulus site; scn/SCN, suprascapular nerve; UT, upper trunk.

Fig. 6. C5, C6 (C7) pattern ofloss. Antebr. cut. n., antebrachial cutaneous nerve; Desc Cx pl, descending cervical plexus; Lat cd, lateral cord; LT, lower trunk; Med cd, medial cord; mcn, musculocutaneous nerve; MT, middle trunk; Post cd, posterior cord; R, recording site for evoked NAPs; S, stimulus site; scn/SCN, suprascapular nerve; UT, upper trunk.

Fig. 6. C5, C6 (C7) pattern ofloss. Antebr. cut. n., antebrachial cutaneous nerve; Desc Cx pl, descending cervical plexus; Lat cd, lateral cord; LT, lower trunk; Med cd, medial cord; mcn, musculocutaneous nerve; MT, middle trunk; Post cd, posterior cord; R, recording site for evoked NAPs; S, stimulus site; scn/SCN, suprascapular nerve; UT, upper trunk.

Table 6 gives the postoperative grades for stretch injuries with various preoperative distributions. Almost all of the data in this table on flail arms come from direct graft repair. The difficulties in obtaining satisfactory results in the flail arm (C5 to T1 loss) where loss is complete are evident. It is also evident that outcomes with direct repair of many stretch injuries are less than optimal. This situation has prompted the current authors to add nerve transfers to direct repair whenever possible in recent years.

In patients with C5-6 stretch injuries, medial pectoral nerve transfer added to musculocutaneous nerve improved grades to an average of 4. Accessory to suprascapular transfers improved the grade in that subset of patients to 3.8 (see Table 4).

In patients with C5, C6, and C7 stretch injuries, medial pectoral transfer to musculocutaneous nerve helped achieve an average grade of 3.7, whereas XI (necessary nerve) to suprascapular nerve improved average grades to 3.5 (see Table 5).

Since 1998, another hundred or so patients have had nerve transfers added to whatever direct repair the authors could obtain. Detailed outcomes of these patients will be the subject of another article, but experience to date with such an approach has been encouraging.

In the C5, C6 or C5, C6 and C7 cases, accessory has been placed to suprascapular and usually medial pectoral branches to a split half of the musculocutaneous nerve. Direct graft repairs, when possible because of sufficient lead-out, usually have been placed to the divisions of upper and, if possible, middle trunks.

In the flail arm cases in which medial pectoral or other nerves are not available for transfer, intercostal nerves have been transferred end-to-end to musculocutaneous nerve, which has been sectioned proximally and then curved down distally to meet the intercostals at the level of the axilla. The accessory nerve has been mobilized so that it can be sutured directly to the suprascapular nerve, which, whenever possible, is sectioned well prox-imally. Any available direct outflow from C5, and less frequently C6 and rarely C7, has been led to the posterior division of the upper trunk, posterior cord, or the portion of lateral cord going to the median nerve. When direct repair of C6 outflow is judged to be feasible, the musculocuta-neous nerve is split longitudinally so that inter-costals are sewn into one half of the nerve. Figs. 7 through 10 depict some of the repairs used in the patients with flail arms.

Discussion

Open injuries

The direct repair of plexus elements injured by lacerations and GSWs is not only possible but

All 5 spinal nerves with either pre and post-gangliotic injury or pre-gangliotic injury i-

XI to Suprascapular nerve

Intercostal nerves(ICN) to Musculocutaneous nerve Descending Cervical Plexus to posterior division or uper trunk

Fig. 7. C5 through T1 loss (flail arm). Antebr. cut. n., antebrachial cutaneous nerve; Desc Cx pl, descending cervical plexus; Icn, intercostal nerve; Lat cd, lateral cord; Med cd, medial cord; mcn, musculocutaneous nerve; Post cd, posterior cord; scn, suprascapular nerve.

gives acceptable results. End-to-end repair is usually, although not always, possible for most transections caused by laceration or GSWs. Where either mechanism gives lesions in continuity, results with direct repair, usually by grafts, is also acceptable. These results are possible if intraoperative electrophysiologic studies (usually NAP recordings) are used to differentiate those lesions not needing repair from those that do. It is possible that the addition of nerve transfers under certain circumstances might add to these outcomes, but the authors have had limited experience with such repairs in these injury categories.

Lacerations

The advantages of relatively acute repair for sharply transected elements are evident in the data

Fig. 8. Another pattern of C5 through T1 loss (flail arm). Antebr. cut. n., antebrachial cutaneous nerve; Desc Cx pl, descending cervical plexus; Icn, intercostal nerve; Lat cd, lateral cord; Med cd, medial cord; mcn, musculocutaneous nerve; Post cd, posterior cord; R, recording site for evoked NAPs; S, stimulus site; scn, suprascapular nerve; UT, upper trunk.

Fig. 8. Another pattern of C5 through T1 loss (flail arm). Antebr. cut. n., antebrachial cutaneous nerve; Desc Cx pl, descending cervical plexus; Icn, intercostal nerve; Lat cd, lateral cord; Med cd, medial cord; mcn, musculocutaneous nerve; Post cd, posterior cord; R, recording site for evoked NAPs; S, stimulus site; scn, suprascapular nerve; UT, upper trunk.

Fig. 9. C5 through T1 loss (flail arm). Antebr. cut. n., antebrachial cutaneous nerve; Desc Cx pl, descending cervical plexus; Icn, intercostal nerve; Lat cd, lateral cord; Med cd, medial cord; mcn, musculocutaneous nerve; MT, middle trunk; Post cd, posterior cord; R, recording site for evoked NAPs; S, stimulus site; scn, suprascapular nerve; UT, upper trunk.

Fig. 9. C5 through T1 loss (flail arm). Antebr. cut. n., antebrachial cutaneous nerve; Desc Cx pl, descending cervical plexus; Icn, intercostal nerve; Lat cd, lateral cord; Med cd, medial cord; mcn, musculocutaneous nerve; MT, middle trunk; Post cd, posterior cord; R, recording site for evoked NAPs; S, stimulus site; scn, suprascapular nerve; UT, upper trunk.

I-Neurolysis or

(-) i-[> grafts to UT divisions plus XI to Suprascapular nerve

Fig. 10. C5 through T1 loss (flail arm). Antebr. cut. n., antebrachial cutaneous nerve; Desc Cx pl, descending cervical plexus; Lat cd, lateral cord; Med cd, medial cord; mcn, musculocutaneous nerve; MT, middle trunk; Post cd, posterior cord; R, recording site for evoked NAPs; S, stimulus site; scn, suprascapular nerve; UT, upper trunk.

i [> Descending Cervical Plexus to MT divisions

Fig. 10. C5 through T1 loss (flail arm). Antebr. cut. n., antebrachial cutaneous nerve; Desc Cx pl, descending cervical plexus; Lat cd, lateral cord; Med cd, medial cord; mcn, musculocutaneous nerve; MT, middle trunk; Post cd, posterior cord; R, recording site for evoked NAPs; S, stimulus site; scn, suprascapular nerve; UT, upper trunk.

presented, because end-to-end repair is more likely under these circumstances than had repair been delayed. Transection by blunt mechanisms was managed by delayed repair so that resection to relatively healthy tissue could be done. Despite laceration as a mechanism of injury, some elements with loss in their distribution had lesions in continuity. NAP recordings were used on these elements to determine the need for resection and then end-to-end suture or a graft was done for those elements without NAPs across their lesions.

Gunshot wounds

Some gunshot injuries involving the plexus improve enough in the early months so that surgery on the plexus is not needed [12,13]. Only 15% of the serious injuries transect one or more plexus elements, so lesions in continuity are the most common pathology. NAP recordings then play an important role in determining the need for resection and direct repair. Thus, 46 of 202 elements (23%) with complete loss clinically by EMG at 3 or more months post injury transmitted an NAP and were regenerating. One hundred fifty-six other injured elements did not transmit an NAP and had direct repair either by end-to-end suture or by grafts. Most of the 91 elements believed to have incomplete loss preoperatively transmitted NAPs and had neurolysis, but importantly, nine elements did not and they required repair.

Stretch/contusion injuries, including avulsions

Operative inspection of the injured portion of the supraclavicular plexus is also important in patients with flail arm (C5-T1). A group of 208 patients with flail arm underwent preoperative clinical observation and myelographic examination and EMG workup of each root, matched with results of operative inspection of spinal nerves and plexus elements [14]. Of 1040 root levels studied, 470 had irreparable proximal spinal nerve or nerve root lesions. Approximately 35% of the avulsions involved C7, C8, and T1. Another 35% involved C7 and C8, whereas 20% had other combinations, which included C6. Only 4% of patients had avulsion of all five levels, including C5, and only 10% of patients had avulsion of C5 as part of the pattern of loss. Thus, repairs, usually by grafts, were possible in about half of the levels inspected. Grafts were often long, so good or better results were only achieved in 35%. Nonetheless, it was shown that direct repair was feasible in some patients and gave good results, especially if C5

and C6 were usable for outflow. In the patients with flail arm, C5 can frequently and C6 can sometimes be a source of fascicular lead-out for direct repair by grafts and should be inspected and tested in most cases.

Outcomes in C5, C6, and C7 lesions, although acceptable before 1998, have improved since then by the addition of nerve transfers. This situation also is true for patients with flail arm (C5 to T1). Recovery of proximal shoulder and upper arm function was acceptable but forearm recovery was poor or nonexistent, so the overall grade for recovery in the C5, C6, and C7 distribution in the flail arm limbs with direct repairs was only grade 3 in 35% of patients. When transfers were added to direct repair of one or more spinal nerve levels, 45% of patients achieved an LSUHSC grade 3 or better level. Furthermore, where only transfers without direct repair were done, usually for C5 through T1 avulsions, grade 3 or better levels were only achieved in 30% of patients.

Shoulder abduction

Shoulder abduction is one of the two most necessary proximal functions for a useful arm. Initiation of abduction is most important and that is provided by the supraspinatus for the first 30° to 40°. External rotation of the shoulder is also important and that is provided by infraspinatus. Therefore, neurotization of the suprascapular nerve is a high priority with stretch/contusive plexus injuries. The use of the accessory nerve to do this was devised some years ago and has proved effective, especially if transferred end to end rather than with the interposition of grafts [15-18].

Elbow flexion

Of equal importance to usable function is the restoration of elbow flexion. This result can be achieved in some patients with supraclavicular stretch injuries by direct repair using grafts, even though these grafts are several inches long. Neu-rotization by nerve transfer more frequently produces usable elbow flexion [19,20], especially if C8-T1 and its outflows are functional, because then either the medial pectoral branches or fasicles of the ulnar nerve can be transferred to the musculocutaneous nerve [21]. Both methods can be effective, although proponents of the Oberlin technique prefer it because placement of an ulnar fasicle into the motor portion of the musculocuta-neous nerve is more selective [22]. If C8-T1 and its outflows are damaged, as they are in patients with flail arm, then either intercostal transfers or the use of contralateral C7 has been recommended. The original intercostal transfers were done with the help of interposed grafts [23-26]. Experience has shown that intercostals placed directly to muscu-locutaneous nerve provide more frequent and better results than those with interposed grafts [27-29].

Other transfers

The current authors have not routinely used phrenic nerve as a source of axons for transfer, although others have had good results [30]. In the authors' experience, the loss of phrenic function during supraclavicular brachial plexus dissections is not without secondary complications, such as atelectasis and pneumonia and occasional complaints from patients about shortness of breath. The authors therefore have not purposely sectioned this nerve for transfer nor have they taken the chance, admittedly small, of endangering function of the contralateral limb by using contralateral C7 as a source of axons. Unless the operation is performed at the level of the spinal canal, the use of C7 for outflow to neurotize contralateral plexus nerves also requires relatively lengthy grafts [31]. Although they have fewer motor fibers than accessory or contralateral C7, the C3, C4 spinal nerves or their outflows are also a nonplexus source of axons [32-34]. Thus, the descending cervical plexus can provide some fibers to either the upper or middle trunk posterior division to the posterior cord. If some direct lead-out

Box 2. Repair preference for C5/C6

stretch or C5/C6/C7 stretch injuries

1. Neurolysis if NAPs indicate regeneration

2. Direct repair if NAPs are negative (flat) and proximal fascicular structure is found on sectioning spinal nerve plus addition of medial pectoral to split musculocutaneous nerve (for C6 avulsion) or accessory to suprascapular nerve (for C5 avulsion)

3. Accessory and medial pectoral transfers and, less frequently, descending cervical plexus transfers if NAP studies indicate preganglionic lesion or sectioning indicates pre- and postganglionic lesions of C5 and C6.

to the posterior divisions by direct repair is possible, then the descending plexus can be placed to the anterior division of the middle trunk.

Direct repair supplemented by transfers

When some direct repair is possible, the current authors have tried to preserve routes for down-flow to more distal plexus elements, and especially the musculocutaneous nerve, by splitting them longitudinally and placing the transferred nerve into half of the recipient nerve rather than into the whole element or nerve. This approach has been done most often after direct repair of one or more proximal elements leading to the musculocutaneous nerve. The musculocuta-neous nerve is then split longitudinally and one half is anastomosed to the medial pectoral branches and the other half is left alone for receipt of any down-flow from the proximal repair. When the thoracodorsal nerve can be transferred into the axillary nerve to provide some recovery of the deltoid muscle, it has been done to half of the recipient nerve [35]. This approach works best if

Box 3. Repair preferences for C5

through T1 stretch injuries or flail arm

1. Neurolysis where NAPs indicate regeneration. (This finding is infrequent but does occur.)

2. Direct repair if NAPs are negative and yet, on sectioning, fascicular structure is found proximally. Grafts are from proximal spinal nerves to divisions or cords. Added in are the following: (1) accessory to suprascapular nerve, (2) descending cervical plexus to posterior division of upper trunk or middle trunk and its divisions, (3) intercostals (three or four) to a longitudinally split portion of musculocutaneous nerve.

3. If all five plexus roots are avulsed, the following repairs are preferred: (1) XI (necessary nerve) to suprascapular nerve, (2) intercostal nerves to musculocutaneous nerve, and (3) either descending cervical plexus or XI input to sternocleidomastoid muscle placed to posterior division of upper trunk.

Table 7

Surgical outcome in infraclavicular plexus stretch injuries3

No. elements (average grade of recovery achieved) Plexus elements Neurolysisb Suture Grafts Split repair

Cords

Cords

Table 7

Surgical outcome in infraclavicular plexus stretch injuries3

No. elements (average grade of recovery achieved) Plexus elements Neurolysisb Suture Grafts Split repair

Lateral

12

(4.5)

3 (4.3)

6 (3.8)

3 (4)

Medial

16

(3.9)

2 (2.2)

7 (1.2)

4 (3.6)

Posterior

14

(4.1)

2 (3.6)

6 (3.0)

3 (3.5)

Cords to nerves

Lateral to

20

(4.4)

0 (0)

35 (3.8)

0 (0)

musculocutaneous

Lateral to median

29

(4.1)

1 (4)

19(3)

0 (0)

Medial to median

24

(4.3)

1 (4)

17(3)

0 (0)

Medial to ulnar

33

(3.6)

1 (0)

13 (1.4)

1 (2.3)

Posterior to radial

29

(4.1)

1 (0)

32 (2.7)

3 (3.5)

Posterior to axillary

28

(4.7)

1 (3)

48 (3.5)

1 (4)

Totals

Elements/results

205

(4.2)

12 (2.3)

183 (2.8)

b Neurolysis based on a positive NAP.

b Neurolysis based on a positive NAP.

some down-flow to the axillary nerve can be gained by more direct repair proximally, such as C5 by means of grafts to the posterior division of the upper trunk. This approach has not been used with the suprascapular nerve, which is either repaired directly by grafts from C5 and C6 or by neurotization through transfer of accessory to the nerve without other input.

To date, this "pants-over-vest" approach has paid dividends, although outcomes still leave much to be desired in this difficult category of injury, especially for the patients with flail arm for whom eventual usefulness of the limb may be quite limited [36,37]. The authors' current preferences for repair of C5-6, C5, C6, C7, and C5 through T1 (flail arm) stretch injuries are outlined in Boxes 2 and 3.

Infraclavicular stretch injuries

Although some infraclavicular stretch injuries improve spontaneously, many do not and require surgery for correction [38].

Most of the infraclavicular stretch injuries managed before and since 1998 in this series had direct repair rather than nerve transfers. Outcomes were best for lateral and posterior cord lesions and their outflows. Medial cord to median repair also gave good results (Table 7).

Other techniques

The current authors applaud the concepts but do not at this time perform other more recently described reconstructive techniques, such as neu-rotized and vascularized free-muscle transfers for plexus stretch/contusion/avulsive lesions [39]. They also have not undertaken spinal cord implantation for avulsed roots and prefer to await more definitive data about outcomes and possible spinal cord complications [40]. Certainly, various repair techniques will continue to evolve, especially for the supraclavicular stretch palsies needing reconstruction [41-43].

Biofeedback Mastery

Biofeedback Mastery

Have you ever wondered what Biofeedback is all about? Uncover these unique information on Biofeedback! Are you in constant pain? Do you wish you could ever just find some relief? If so, you are not alone. Relieving chronic pain can be difficult and frustrating.

Get My Free Ebook


Post a comment