The comparison of electrophysiologic findings of traumatic brachial plexopathies in a tertiary care center

Injury, Int. J. Care Injured 33 (2002) 591–595

The comparison of electrophysiologic findings of traumatic brachial plexopathies in a tertiary care center Tien-Yow Chuang a,∗ , Fang-Yao Chiu b , Yun-An Tsai a , Shu-Chiung Chiang c , Der-Jen Yen d , Henrich Cheng e a

d

Department of Physical Medicine and Rehabilitation, Taipei Veterans General Hospital, National Yang-Ming University, 201 Shih-Pai Road, Sec. 2, Peitou, Taipei 11217, Taiwan, ROC b Department of Orthopaedics, Taipei Veterans General Hospital, National Yang-Ming University, Taipei, Taiwan, ROC c Information Service Center, Taipei Veterans General Hospital, National Yang-Ming University, Taipei, Taiwan, ROC Department of Neurology, Neurological Institute, Taipei Veterans General Hospital, National Yang-Ming University, Taipei, Taiwan, ROC e Neural Regeneration Laboratory, Department of Neurosurgery, Neurological Institute, Taipei Veterans General Hospital, National Yang-Ming University, Taipei, Taiwan, ROC Accepted 5 March 2002

Abstract This study was undertaken to demonstrate the distribution of causative factors of brachial plexopathy (BP), to assess the association between the mechanism of injuries and the predominant level of the brachial plexus involved in the injuries, and to characterize the extent and degree of severity of injury in patients with BPI. It consisted of a cross-sectional, retrospective review of electrophysiological data of 5547 patients with 117 patients being identified as having BPI, of whom 86 patients were recruited into the study. The patients were divided into six subgroups according to the mechanism of the damage. The injury was subdivided according to the brachial plexus levels predominantly affected, and each component of the four major anatomical plexus levels—root, trunk, cord and nerve levels was analyzed. The affiliation between the type of injuries and the specified brachial plexus levels was calculated via a two-tailed Fisher’s exact test. These findings demonstrated that the type of brachial plexus injury (BPI) is significantly related to the brachial plexus level involved. The motorcycle and birth injury groups were affected at the trunk level, the fall group at the nerve level, the automobile group at the cord level, and the blunt injury group at the cord or nerve level. Moreover, the majority of patients in the motorcycle, fall, and pedestrian groups suffered from severe, incomplete lesions, while the neurophysiological results of the other groups varied. © 2002 Elsevier Science Ltd. All rights reserved.

1. Introduction Following a traumatic injury to the brachial plexus patients may suffer from differing degrees of tissue damage. Because of these variations, a great divergence in the appearance of the injury can be found in the same patient, ranging from neurapraxia with a conduction block to a complete anatomical interruption of continuity [9]. The intricacy of the plexus, anatomic variations, and possible combinations of motor and sensory deficits make it difficult to establish an accurate diagnosis, and to estimate the severity of the lesion [2]. Over the past decade, surgeons have endeavoured to advance surgical techniques on brachial plexus injury (BPI). All the efforts employed, however, have resulted in only limited general improvement [1,4,21,23,26,27,29]. Never∗ Corresponding author. Tel.: +886-2-28757296; fax: +886-2-28757359. E-mail address: [email protected] (T.-Y. Chuang).

theless, knowledge of the types of lesions, the severity and mechanism of injury, would have a major impact on preoperative decisions on surgical intervention and prognosis. Clinical assessment alone of the nature and territory of the lesion can be inaccurate. Therefore, electrodiagnostic measurements might help to define the lesion and provide a guide for intraoperative procedures [10,13,14,16,28]. In previous research, the authors have found that BPs in the different injuries (gunshot wounds versus motor vehicle accidents) is significantly associated with the level involved [5]. Since Taiwan has strict gun controls, the incidence of gunshot wounds is very low. In contrast, the incidence of motorcycle crashes is high. However, it appears that no data currently exists in Taiwan concerning the precise rate of occurrence, types, specified levels, severity, etc., of brachial plexus injuries. This study aims to increase knowledge of BPI by demonstrating the etiologic distribution of BPI, characterizing the extent and degree of severity, and by investigating the association between

0020-1383/02/$ – see front matter © 2002 Elsevier Science Ltd. All rights reserved. PII: S 0 0 2 0 - 1 3 8 3 ( 0 2 ) 0 0 0 9 4 - 3

592

T.-Y. Chuang et al. / Injury, Int. J. Care Injured 33 (2002) 591–595

mechanism of injuries and the predominant brachial plexus level involved.

Table 1 Predominant levels involved in brachial plexus of six injured groups∗ Cases

2. Patients and method This study consisted of a retrospective review of 5547 patient charts. All patients were referred to the Department of Physical Medicine and Rehabilitation for electrodiagnostic tests from all over Taiwan from 1995 to 1999. During this time, 117 patients were diagnosed with BP. Any injury to the neck, shoulder, or chest comprised our inclusion criterion. The electrophysiologic findings were based on anatomic injuries to the brachial plexus which resulted in conduction block, demyelination, and/or axonal loss. Exclusion criteria were (1) peripheral nerve injuries caused by damage to arm or forearm, (2) data obtained from electrodiagnostic testing of BPI fewer than 2 weeks after an injury, (3) incomplete electromyography and nerve conduction studies (EMG/NCS), and (4) any other neuromuscular disorders in the affected limb. Eighty-six patients (74%) met the inclusion criteria and were admitted to the study. Sixty of the 86 patients (70%) were men. Age distribution was 1 month to 82 years, with an average age of 35.4 years ± 24.2 years. The time interval between injury and electrodiagnostic testing was 13.3 weeks ± 9.7 weeks. If the first EMG investigation did not show complete electrodiagnostic findings, repeat results from EMG examinations 3 months later were included in our database. The study protocol was approved by the Human Research Committee in this institution. Injury subtypes were also identified from reviewing the medical charts. These included 21 falls, 30 motorcycle crashes, eight automobile crashes, nine pedestrian hit by motor vehicles (MVA), 12 birth injuries and six blunt injuries. 2.1. Electrodiagnosis

Motorcycle Fall Automobile Pedestrian Birth Blunt

Supraclavicular lesions

Infraclavicular lesions

Root

Trunk

Cord

Nerve

4 0 1 3 4 0

17 7 1 1 8 1

4 2 3 2 0 2

5 11 2 2 0 2

“Cases” refers to the number of patients with such lesions at that level. Fisher’s exact test was used to investigate the association between the type of injuries and the predominant level involved. ∗ P < 0.05. The four double-level injury cases were omitted from statistical calculations to improve the accuracy of an approximate solution.

three portions (medial, lateral, posterior) were confined to the cord level; and five major nerves (axillary, musculocutaneous, median, ulnar, radial) from the brachial plexus. Each component of the four major levels of the plexus (i.e. roots, trunks, cords, nerves) was examined, including the extent and severity of the damage. Furthermore, the various degrees of impairment were graded using a modified version of Dumitru’s [11] and Wilbourn’s [30] scale as follows—mild: normal sensory nerve action potential (SNAP) amplitude and compound muscle action potential (CMAP) amplitudes of nerve conduction studies, with occasional fibrillation and normal motor unit recruitment on EMG; moderate: slight to profound decrease in SNAP amplitude and normal to slight decrease in CMAP amplitude, with moderate numbers of fibrillation in at least three muscle areas and normal to slight decrease in motor unit recruitment; severe: absent SNAP and profound decrease to absent CMAP amplitude, with marked fibrillation in all muscle regions and discrete-to-no motor unit recruited; and complete: absent SNAP and CMAP with no motor unit recruitment and profound fibrillation in all muscle regions.

The assistance of peripheral nerve sensory and motor studies is required to map out the components of BPs. For optimal assessment basic conduction techniques of the limb nerves as well as some proximal nerve root and plexus NCS may be executed [11]. After NCS, a needle electromyographic investigation is performed on the injured upper extremity using a disposable concentric needle. Muscles examined when BPI was suspected would be the trapezius, serratus anterior, rhomboid, infraspinatus, deltoid, biceps brachii, triceps, pectoralis major, pronator teres, extensor carpi radialis, flexor carpi radialis, abductor pollicis brevis, abductor digiti minimi, paraspinal C4–T1.

Data was processed with SAS software modules for descriptive statistics and contingency tables. The two-tailed Fisher’s exact test was used to detect the association between categorical variables and the predominant plexus level involved (Table 1). The four double-level injury cases were omitted from Table 1 for statistical calculations to improve the accuracy of an approximate solution. The criterion for significance was a P value <0.05.

2.2. Analysis of plexus lesion

3. Results

The following makes up the component parts—five portions (C5–T1) were confined to the root level; three portions (upper, middle, lower) were confined to the trunk level;

Table 1 lists the predominant plexus level involved among six different categories. There were significant differences across six subtypes on each anatomical level (Fisher Exact,

2.3. Statistical analysis

T.-Y. Chuang et al. / Injury, Int. J. Care Injured 33 (2002) 591–595

593

Table 2 Electrodiagnostic results of all groups by components Location of injury

Root

UT

MT

LT

LC

MC

PC

MCN

Med

Uln

Rad

Axil

Group A

Com (n) Mild (%) Mod (%) Sev (%) Cpl (%)

20 0 0 25 75

17 6 6 41 47

11 0 0 45 55

6 0 0 50 50

4 0 0 75 25

3 0 0 100 0

4 0 0 75 25

2 50 0 50 0

0

0

0

5 100 0 0 0

Group B

Com (n) Mild (%) Mod (%) Sev (%) Cpl (%)

0

7 0 0 71 29

3 0 0 67 33

2 0 0 50 50

2 0 0 100 0

2 0 0 100 0

3 33 0 67 0

3 0 0 100 0

1 0 0 100 0

2 0 50 50 0

3 0 33 67 0

11 19 9 36 36

Group C

Com (n) Mild (%) Mod (%) Sev (%) Cpl (%)

5 0 0 100 0

1 0 0 0 100

1 0 0 0 100

1 0 0 0 100

1 0 0 100 0

2 0 50 50 0

1 100 0 0 0

1 0 0 0 100

2 0 0 0 100

3 33 0 0 67

4 0 0 50 50

Group D

Com (n) Mild (%) Mod (%) Sev (%) Cpl (%)

13 0 0 100 0

1 0 0 100 0

0

0

2 0 0 100 0

1 0 0 100 0

2 0 50 50 0

0

0

0

0

4 0 0 50 50

Group E

Com (n) Mild (%) Mod (%) Sev (%) Cpl (%)

14 36 50 14 0

8 25 13 37 25

2 50 50 0 0

1 0 100 0 0

0

0

0

0

0

0

0

0

Group F

Com (n) Mild (%) Mod (%) Sev (%) Cpl (%)

0

1 0 100 0 0

0

0

0

2 0 50 50 0

1 0 100 0 0

1 0 0 0 100

0

0

0

3 0 33 67 0

3 0 33.3 33.3 33.3

Group: A motorcycle; B fall; C automobile crashes; D pedestrian; E birth injuries; F blunt injuries. Abbreviations: Com components; Mod moderate; Sev severe; Cpl complete; UT upper trunk; MT middle trunk; LT lower trunk; LC lateral cord; MC medial cord; PC posterior cord; MCN musculocutaneous; Med median; Uln ulnar; Rad radial; Axil axillary.

P < 0.05). Of the fall group, 55% (11 of 20) sustained lesions at the nerve level. For the 30 patients with motorcycle crashes, 56% (17 of 30) had damage at the trunk level, whereas 43% (three of seven) of automobile crash victims had lesions at the cord level, and 67% (8 of 12) of birth injuries had trunk damage, and 80% (four of five) of blunt injuries patients suffered from cord or nerve lesions. There were four cases presenting multilevel plexopathies: a two cords and one nerve injury in the fall group; a one cord and three nerves injury in the automobile crash group; a two cords and one nerve injury in the pedestrian group; and a one cord and one nerve injury in the blunt injury group. The electrodiagnostic results of six categories are shown in Table 2. Most patients in the fall, motorcycle, and pedestrian groups sustained severe, incomplete lesions. The majority of the injuries caused by automobile crashes were severe, with equal distribution among patients in complete/incomplete interruption of axonal continuity. However, the neurophysiological results revealed varying degrees of severity secondary to birth and blunt injuries.

4. Discussion Studies on the consequences of BPIs are particularly relevant for young adults because a lifelong disability is often the consequence of these injuries. Disruption in these formative years can result in an inability to take advantage of educational and occupational opportunities to minimize damage [3]. Unfortunately, BPI was still identified in 117 of 5547 (2.1%) patients referred for electrophysiological testing as a result of various upper extremity problems in our laboratory. Of the given causative factors, motorcycle crashes were the most frequent cause overall (36%). In the past few years, the number of patients with BPI in Taiwan has increased, in parallel with the number of motorcycle accidents. Falls were the second most common cause (24%) of BPI in this group of patients. It was not surprising that most of the older-aged (52.8 years ± 23.2 years), male subjects (characteristics of a veterans’ hospital) were recruited into this subtype. Falls carrying high risks for proximal humeral fractures and/or shoulder dislocations affecting the shoulder girdle region were particularly common [15,24]. Indeed, 70% of the fall

594

T.-Y. Chuang et al. / Injury, Int. J. Care Injured 33 (2002) 591–595

patients in our study were suffering from humeral fracture or shoulder dislocation. The potential causes of BPI are diverse, as exemplified in the study. The differences in frequency between the present study and many of the previous reports probably reflect the populations studied. In Midha’s study [24], motorcycle and snowmobile crashes were the most frequent causes overall; furthermore, perioperative damage to the plexus was also reported by Wilbourn [31]. Based on the location of the injury, we referred to the brachial plexus lesion as either being supraclavicular or infraclavicular [20]. Two components, root and trunk, are contained in the supraclavicular plexus. The infraclavicular portion involves cords and terminal nerves. This classification is of paramount importance in both surgical repair and outcome predictions [22]. However, this classification by level indicates the center of the lesion, but not necessarily the extent of it [19]. As seen in Table 1, nine motorcycle cases involved division to cord and terminal nerve levels. Thus, for the most part, lesions were infraclavicular. Twenty-one cases involved root and trunk levels. The largest category occurred at the trunk level, thus making the lesion supraclavicular. By this categorization, we can find that birth and blunt injuries had lesions confined to supraclavicular and infraclavicular levels, respectively. It is interesting to note that most automobile crash victims had lesions confined to the infraclavicular level. This result was quite different from the authors’ previous research carried out in the US, which demonstrated that MVA patients sustained damage at supraclavicular level [5]. It is possible that our sample in this subgroup was so small that the skewing of several cases may have altered the results. The other possibility for the difference is a methodological defect. In a previous study, we put drivers, passengers, and pedestrians all together in one group under the term of MVA [5]. In this study, however, the MVA victims were divided into automobile crashes and pedestrian versus motor vehicle accident subgroups. Thus the characteristics of these subtypes may be described specifically. Two issues warrant discussion in relation to the electrodiagnostic measurements we used for the patients with BPI. Firstly, prior studies have documented the intimate relation between the type and anatomical region of injuries in gunshot wounds and motor vehicle crashes. Given this, perhaps more attention should be paid to the possible damaged brachial plexus levels with respect to their specific types. As revealed in our patient groups, the type and level were strongly related except in the victims of the pedestrian group, who were equally affected in the supraclavicular and infraclavicular regions. The possible theory for how this occurs is that the distribution of injury is mainly determined by the position of the arm in relation to the body at the time of impact [22]. Therefore, various types of collisions and protective reflexes from the victims may contribute to such outcome variability. The second important issue is the numbers of components damaged after BPI. Multiple element involvement clearly re-

mains a challenging problem facing surgeons repairing BPI. For example, a motorcycle injury is a high-energy trauma; therefore, it is not surprising that these victims sustain multiple cervical root avulsions [18]. In our study, however, there were significantly higher numbers of components injured at the trunk level (average = 1.13 components, median = 1). These findings suggest that the speed of the motorcycle could be a cause of the difference in the affected levels [8]. The majority of motorcycles in Taiwan are the small-sized scooters (90–120 cc), and traffic congestion in both the countryside and downtown would not enable riders to increase their speed. A decrease in speed and the duration of impact might reduce the magnitude of the forces exerted on the body; moreover, the victims would have more time to perform immediate critical protection from the injury. Infants with obstetrical BPI would mostly manifest Erb’s or Klumpke’s palsy [7,12]. There is not much difference here between the present study and previous reports. Although the appearance of fibrillation potentials in needle EMG performs an essential role in identifying the levels of BPI, our experience and previous research have revealed that denervation does not always appear in young infants, despite later illustration of BPI [7,25]. Five major terminal nerve branches occur in the brachial plexus. All these nerves are in close proximity in the axilla. As a result, all elements are subject to being assaulted simultaneously by damage near the shoulder joints. Nevertheless, multiple nerve implications associated with humeral fractures and tearing of the rotator cuff were frequent in the fall group. With regard to the terminal nerves, isolated axillary nerve lesions may commonly occur or combine with musculocutaneous and/or radial nerve injury in this group. With respect to the blunt group, all subjects, with the exception of one, sustained interpersonal violence on the chests and shoulders. There is no doubt that 80% of them would have infraclavicular plexopathies. These findings all demonstrate that establishing the type, severity, and sites of BPs is challenging. Because incomplete lesions are not uncommon, it is mandatory to investigate each part of the plexus. For diffuse plexopathies, the same principles are used to examine the whole plexus, but the evaluation is often lengthy and sometimes puzzling [14]. In view of the complete EMG findings, the differences between the present study and many of the previous reports probably reflect the differing characteristics of the people studied. Since Veterans General Hospital is a large tertiary care facility in Taiwan, only those who were severely injured are referred to this center. This may partly explain the presentation to the EMG of patients with severe, incomplete lesions, in the majority of motorcycle, fall, automobile, and pedestrian injuries. Midha, in his valuable research, has mentioned that when suprascapular plexopathies have been associated with shoulder disorders, the injuries have always been severe, but the infraclavicular lesions have often been minor [24]. However, both the supraclavicular and infraclavicular regions were severely damaged in the fall group of our study. More devastation and less regeneration might

T.-Y. Chuang et al. / Injury, Int. J. Care Injured 33 (2002) 591–595

be one of the factors in this old-aged group in comparison to Midha’s young-aged population. The prevalence of plexopathies in birth injury was unequally distributed. Four subjects incurred 14 components of root injuries with the most common lesions at C5 and C6; five had isolated upper trunk, two had combined upper and middle trunk, and one had three trunk lesions. The severity of the injuries varied, with most of the axonal continuity preserved. In addition, the EMG presentations varied in blunt injuries: one subject hit by a crane digger suffered from a complete musculocutaneous nerve lesion associated with clavicular fracture. The various classifications of BPI in previous articles are confusing [14]. The etiologic categorization used in this study attempts to clarify the relationship between mechanisms of injuries, damaged numbers of components, and anatomical levels. It is valuable to recognize the significant association between causative factors and anatomical levels. Therefore, this information will impact significantly on surgical planning and patients’ intraoperative positioning. When combined, clinical research and basic science studies [6,17,31], we are sure to steadily increase our understanding of the problems related to BPI patients, and to better determine what future services can be provided. References [1] Ben-David B, Stahl S. Prognosis of intraoperative brachial plexus injury: a review of 22 cases. Br J Anaesth 1997;79:440–5. [2] Brooks DM. Open wounds of the brachial plexus. J Bone Joint Surg Br 1949;31:17–33. [3] Choi PD, Novak CB, Mackinnon SE, Kline DG. Quality of life and functional outcome following brachial plexus injury. J Hand Surg 1997;22A:605–12. [4] Chuang DCC, Lee GW, Hashem F, Wei FC. Restoration of shoulder abduction by nerve transfer in avulsed brachial plexus injury: evaluation of 99 patients with various nerve transfers. Plast Reconstr Surg 1995;96:122–8. [5] Chuang TY, Chiou-Tan FY, Vennix MJ. Brachial plexopathy in gunshot wounds and motor vehicle accidents: comparison of electrophysiologic findings. Arch Phys Med Rehabil 1998;79:201–4. [6] Chuang TY, Huang MC, Chang WC, Lee LS, Chang AC, Hoffer B, et al. Electrophysiological changes following spinal cord implantation of avulsed spinal nerve root in rat. In: Proceedings of the 29th Annual Meeting. Soc Neurosci Abstr 1999;25:1513. [7] Clarke HM, Curtis CG. An approach to obstetrical brachial plexus injuries. Hand Clin 1995;11(4):563–81. [8] Coert JH, Dellon AL. Peripheral nerve entrapment caused by motor vehicle crashes. J Trauma 1994;37(2):191–4. [9] Davis L, Martin J, Perret G. The treatment of injuries of the brachial plexus. Ann Surg 1947;125:647–57.

595

[10] Deletis V, Morota N, Abbott IR. Electrodiagnosis in the management of brachial plexus surgery. Hand Clin 1995;11(4):555–61. [11] Dumitru D. Brachial plexopathies and proximal mononeuropathies. In: Dumitru D, editor. Electrodiagnostic Medicine. Philadelphia: Hanley & Belfus, 1995. p. 585–642. [12] Eng GD, Binder H, Getson P, O’Donnell R. Obstetrical brachial plexus palsy (OBPP) outcome with conservative management. Muscle Nerve 1996;19:884–91. [13] Ferrante MA, Wilbourn AJ. The utility of various sensory nerve conduction responses in assessing brachial plexopathies. Muscle Nerve 1995;18:879–89. [14] Goldstein B. Applied anatomy and electrodiagnosis of brachial plexopathies. Phys Med Rehabil Clin North Am 1994;5:477–93. [15] Guven O, Akbar Z, Yalcin S, Gundes H. Concomitant rotator cuff tear and brachial plexus injury in association with anterior shoulder dislocation: unhappy triad of the shoulder. J Orthop Trauma 1994;8(5):429–30. [16] Hallikainen H, Partanen J, Mervaala E. The importance of neurophysiological evaluation of plexus brachialis injury caused by open heart surgery. Electromyogr Clin Neurophysiol 1993;33:67–71. [17] Huang MC, Chen KC, Chuang TY, Chang WC, Lee YS, Lee LS, et al. Functional recovery after repair of avulsed cervical root. In: Proceedings of the 29th Annual Meeting. Soc Neurosci Abstr 1999;25:494. [18] Ihara K, Doi K, Sakai K, Kuwata N, Kawai S. Restoration of sensibility in the hand after complete brachial plexus injury. J Hand Surg 1996;21A:381–6. [19] Kline DG. Civilian gunshot wounds to the brachial plexus. J Neurosurg 1989;70(2):166–74. [20] Kline DG, Hackett ER, Happel LH. Surgery for lesions of the brachial plexus. Arch Neurol 1986;43:170–81. [21] Kline DG, Judice DJ. Operative management of selected brachial plexus lesions. J Neurosurg 1983;58:631–49. [22] Leffert RD. Brachial plexus injuries. New Engl J Med 1974;14:1059– 66. [23] Liu J, Pho RWH, Kour AK, Zhang AH, Ong BKC. Neurologic deficit and recovery in the donor limb following cross-C7 transfer in brachial plexus injury. J Reconstr Microsurg 1997;13(4):237–43. [24] Midha R. Epidemiology of brachial plexus injuries in a multitrauma population. Neurosurgery 1997;40(6):1182–9. [25] Narakas AO. Brachial plexus surgery. Orthop Clin North Am 1981;12:303. [26] Ochiai N, Nagano A, Sugioka H, Hara T. Nerve grafting in brachial plexus injuries: results of free grafts in 90 patients. J Bone Joint Surg Br 1996;78(B):754–8. [27] Richardson PM. Recovery of biceps function after delayed repair for brachial plexus injury. J Trauma 1997;42(5):791–2. [28] Smith SJM. The role of neurophysiological investigation in traumatic brachial plexus lesions in adults and children. J Hand Surg 1996;21B(2):145–7. [29] Songcharoen P. Brachial plexus injury in Thailand: a report of 520 cases. Microsurgery 1995;16:35–9. [30] Wilbourn AJ. Electrodiagnosis of plexopathies. Neurol Clin 1985;3: 511–29. [31] Wilbourn AJ. Thoracic outlet syndrome surgery causing severe brachial plexopathy. Muscle Nerve 1988;11:66–74.