- Email: [email protected]
Nonpenetrating subclavian artery trauma
Nonpenetrating subclavian artery trauma M. C. Costa, and J. V. Robbs,
Durban, Republic of So. Africa
Nonpenetrating injury to the subclavian vessels is uncommon. During a 6-year period we have treated 167 patients with injuries to the subclavian and superior mediasfinal arteries. Fifteen of these injuries (9%) occurred after blunt trauma. In 10 patients the proximal segment (first and second parts) of the artery was involved. N o patient had an isolated injury; the most frequent associated injuries were rib fractures (n = 11), with the first rib being involved in four of these. Total brachial plexus disruption was found in nine patients. All patients with distal artery involvement had a clavicular fracture. All had an absent radial pulse and eight had critical ischemia ofthe hand. Four patients were treated nonoperatively and the remainder were treated along standard lines. Brachial plexus reconstruction was not feasible in any patient. Within 2 weeks of operation, one patient died as a result of head injuries and one required amputation because of sepsis, During a 12-month period, five regained fifll function, one additional patient requested above-elbow amputation after 6 months, and seven had a flail anesthetic limb. Twelve of these patients were involved in automobile accidents, eight of whom were wearing lap~ shoulder harness seat belts with a loose-fitting shoulder strap component that created a characteristic abrasion pattern on the torso and chest. We conclude that the torsionshearing motion allowed by this situation contributed significantly to the pattern of injury and a plea is made for correctly fitted restraining devices. (J Vase SUgG 1988;8:71-5.)
Injury to the subclavian artery is uncommon in civilian and military surgical practicelS; of 7500 vascular injuries recorded in the Vietnam Vascular Registry less than 1% involved the subclavian artery. 1 Most of these injuries are due to penetrating trauma and outnumber blunt trauma to these vessels by a ratio of between 40 and 50 to 1.5 Blunt trauma may be associated with multiple injuries and presentation may be subtle. The consequences ofmisdiagnosis and late operative treatment of penetrating subclavian injuries have been well described, 24 with delayed hemorrhage, false aneurysm, and arteriovenous fistula being most frequently encountered. Operative management presents a diflicuk challenge and the reported mortality rate in various series ranges between 10% and 30%. 1"4Recent experience on our Vascular Service with several patients having blunt injury to these vessels has prompted the present review. PATIENTS A N D M E T H O D S Between 1981 and 1987, 167 patients have been treated on the Vascular Service for injuries to the subclavian and superior mediastinal arteries. O f these, 15 patients (9%) had injuries to the subclavian From the Department of Surgery, University of Natal Medical School. Reprint requests: ProfessorJ. V. Robbs, Dept. of Surgery, University of Natal Medical School, P.O. Box 17039, Congella, 4013, South Africa. 71
artery as a result of bltmt trauma. Details relating to the patients are summarized in Table I. Ten were male, and the average age of the group was 31.2 years (range 20 to 59 years). The injury w a s associated with motor vehicle accidents in 12 patients. Eight were either the driver or front seat passenger and were wearing lap-shoulder-harness seat belts. In every case the shoulder harness component of the belt had been loose with considerable play between the patient's shoulder and chest wall and the belt. Therefore at the time of impact the patient was,thrown against the belt, with considerable force being taken over the clavicle, shoulder, and anterior chest. In most cases this would apparently be followed by considerable lateral movement of the upper torso with the shoulder acting as a fixed point. This sequence resulted in a characteristic pattern of bruising and abrasions over the chest wall, neck, and shoulder. The other four cases comprised a pedestrian struck by a vehicle, one who fell from a moving truck, a young male thrown from a motorcycle, and one patient ejected from a vehicle after a collision. Three patients sustained more focal direct trauma to the supraclavicular area from a falling tree, assault with a club, and a heavy chain, respectively. The latter was an industrial accident in which a tow-chain parted while under tension. The second part of the artery was injured in seven
J oimTtad Oi"
72
VASCULAR SURGERY
Costa and Robb
Table I. Summary of 15 cases of nonpcnetrating trauma to the subclavian artery Patient
Age
Sex
1
24
M
2
43
3
Mechanism of injury
Artery
Associated injury
Presentation
Motor vehicle accident (seat belt)
(R) subclavian-second part
Total brachial plexus Multiple rib fractures Fracture humerus
Critical ischemia arm and hand Anesthetic flail limb
M
Motor vehicle accident (seat belt)
(R) subclavian-second part
Absent radial pulse Anesthetic flail limb
28
M
(R) subclavian-second part
4
30
F
Motor vehicle accident (seat belt) Motor vehicle accident (seat belt)
Fracture first rib Multiple rib fractures Flail chest Total brachial plexus Fracture first rib
(L) subclavian-second part
Fracture clavicle Total brachial plexus Homer's syndrome
Absent radial pulse Anesthetic flail limb
5
59
F
(L) subclavian-second part
Multiple rib fractures
Absent radial pulse
6
23
M
Motor vehicle accident Passenger (seat belt) Motor vehicle accident (seat belt)
(R) subclavian-second part
Critical ischemia Anesthetic flail limb
7
30
M
(R) subclavian-first part
8
28
M
Motor vehicle accident (seat belt) Motor vehicle accident (seat belt)
Total brachial plexus Compound fracture humerus Fracture scapula Extensive soft tissue trauma Fracture first rib Multiple rib fractures
(L) subclavian-first part
Lung contusion Multiple rib fractures with flail segment
Absent radial pulse Ischemic hand
30
F
Motor vehicle accident (pedestrian)
(L) subclavian-second part
Critical ischemia arm and hand
10
5O
F
Motor vehicle accident (passenger)
(R) subclavian-third part
Dislocation sternoclavicular joint Lung contusion Diffuse brain injury Fracture clavicle Total brachial plexus
11
24
M
Motorcycle accident
(R) subclavian-first part
12
20
M
Fell off truck
(R) subclavian-third part
13
20
F
Hit by falling tree
(L) subclavian-third part
14
27
M
Assault with club
(L) subclavian-third part
15
29
M
Industrial-hit by chain
(L) subclavian-third part
Total brachial plexus Fracture humerus Multiple rib fractures Fracture clavicle Multiple rib fractures Total brachial plexus Head injury Fracture clavicle Multiple rib fractures Total brachial plexus Fracture ribs
Compound fracture clavicle Fracture first rib Multiple rib fractures Total brachial plexus
Absent radial pulse
Critical ischemia arm and hand
Critical ischemia arm and hand Pregangrenous fingers Flail anesthetic limb Absent radial pulse Flail anesthetic limb Absent radial pulse Flail anesthetic limb Absent radial pulse Flail anesthetic limb Critical ischemia arm and hand
Critical ischemia hand and arm Flail anesthetic limb
Volume 8 Number 1 luly 1988
Operative procedure
Nonpenetrating subclavian artery trauma
Pathologic features
Supraclavicular incision (L) carotidsubclavian bypass (vein) No operation
Intimal disruption and thrombosis Shredded brachial plexus
Sternotomy Interposition graft (Dacron) Supraclavicular approach - vessel ligated
Intimal tcar and thrombosis
Sternotomy Interposition graft (Dacron)
--
Total disruption artery Shredded brachial plexus trunks Intimal tear and thrombosis
Result 6 mo Pulses returned Anesthetic flail limb 3 mo Anesthetic flail limb Cansalgia 3 mo Pulse returned Normal function 6 mo Causalgia Above-elbow amputation 12 mo Pulse returnedfull function
Sternotomy Interposition graft (vein)
Total disruption artery Shredded brachial plexus trunks
1 wk Gross sepsis Above-elbow amputation
Sternotomy Interposition graft (Dacron) (L) Thoracotomy Balloon catheter Thrombectomy
Total disruption artery
3 mo Pulse returned Full fimction 6 wk No return of pulse--severe claudication of arm Carotid-subclavian bypass 6 mo Pulse returned-full fimction 2 wk Pulse restored Died--brain death 6 mo Pulse returned Flail limb
Intimal tear and thrombosis
Sternotomy Interposition graft (Dacron)
Total disruption artery
Supraclavicular approach Interposition graft
Total disruption artery Shredded brachial plexus
(vein) No operation
3 mo
Flail limb Causalgia 5 mo Flail limb
No operation
No operation
6 'mo
Flail limb Supraclavicular approach End-to-end anastomosis subclavian Sternotomy Interposition graft (vein)
Intimal tear and thrombosis
Total disruption artery Shredded brachial plexus
3 mo
Pulse returned Full function 3 mo
Pulse returned Flail limb
73
patients, the distal segment (third part) in five, and the proximal segment close to its origin (first part) in three patients. There was no predilection for any particular side, with eight involving the right. No patient had an isolated vascular injury. Multiple rib fractures were encountered in 10 patients and were the most common associated injuries, including the first rib in three; one additional patient had an isolated fracture of the first rib. Other bone and joint injuries included fractures of the clavicle (five patients), humerus (three), scapula (one), and facial bones (one). Dislocation of the sternoclavicular joint was encountered in one patient. Nine patients sustained total brachial plexus injuries. One of these (patient 4) had an obvious associated Homer's syndrome, suggesting avulsion of the nerves at root level. These patients had an anesthetic, totally paralyzed arm, four of the nine manifested critically ischemic changes. Four patients had a flail segment of the chest wall with associated lung contusion, whereas two patients had a significant head injury with an impaired level of consciousness. Clinically, a radial pulse was not palpable in all patients, with a discernible temperature difference compared with the opposite upper limb. Eight patients showed signs of critical ischemia with a cold, blue, mottled hand and forearm.
MANAGEMENT Four patients were treated nonoperatively. All had total brachial plexus lesions, a viable limb, and severe associated chest wall injuries that required positive pressure ventilation. With the above four exceptions all patients in whom operative treatment was contemplated were subjected to arch angiograplay via the retrograde femoral route. Exploration of the neurovascular bundle was done in 11 patients. In six patients the proximal subclavian artery was approached via a limited median sternotomy and in the remainder with more distal injuries, a supraclavicular approach with division of the clavicle in mid-shaft: was used. These approaches have been described in detailY In one patient the proximal left subclavian artery was approached via a fourth intercostal space thoracotorny. In four patients the artery was found to be totally disrupted and in the remainder, intimal disruption with superimposed thrombosis was found. Of five patients with brachial plexus lesions subjected to exploration, four of whom had severe peripheral ischemia, all were found to have total disruption of the brachial plexus with shredding of the trunks in the supraclavicular fossa. In no patient was any attempt made at primary nerve repair and the nerve ends were marked with nonabsorbable suture material as best
74
Costaand Robb
possible to facilitate subsequent surgical exploration if required. To restore arterial continuity, seven patients required an interposition graft, with saphenous vein used in five and Dacron in two. On one occasion it was possible to achieve an end-to-end anastomosis without tension. In one patient (patient 4) in whom the brachial plexus was shredded, the vessel was ligated. Carotid-subclavian bypass on the left side was performed primarily in one patient (patient 1). One patient (patient 15) came to an outlying hospital after a motor vehicle accident, with severe chest injuries and an ischemic hand. Via a left thoracotomy a catheter thrombectomy of the left subclavian artery was attempted. The hand ischemia improved, although peripheral pulses were not restored. The patient was referred to the Vascular Service 6 weeks later with severe claudication of the hand and forearm. Angiography revealed total occlusion of the proximal left subclavian artery. Via a supraclavicular approach, carotid-subclavian bypass with reversed saphenous vein was performed; pulses were restored. All surviving patients were reassessed in the Vascuiar Clinic at monthly intervals for a period ranging between 3 and 12 months. All those with nerve injuries were referred to the peripheral nerve injury clinic, under the aegis of the Orthopaedic Department, at the time of discharge from hospital for longterm assessment and management. Details of this aspect of patient management are beyond the scope of this article. RESULTS One patient died (patient 9) as a result of severe diffuse brain injury. She was declared brain dead after 2 weeks and life support was discontinued. Two patients required amputation. One (patient 6) had severe soft tissue trauma and a total brachial plexus lesion. Gross infection mandated amputation within a week of operation at the subcapital level of the humerus despite a patent vascular repair. The other patient had severe cansalgia in an otherwise flail anesthetic limb and requested amputation 6 months after injury. This was also performed just below shoulder level. O f the remaining eight patients who had arterial reconstruction, all regained peripheral pulses; five have regained full function of the upper limb and three had associated brachial plexus lesions. The four patients who did not have initial exploratory surgery had viable limbs but with no neurologic function. To summarize the short-term outcome in the nine patients with brachial plexus disruption, two
Journal of VASCULAR SURGERY
had amputations and the remainder had a flail anesthetic upper limb 3 to 6 months later. In addition, two of these patients had intermittent episodes of unpleasant spontaneous phantom sensations (cansalgia). DISCUSSION The relative rarity of injury to the subclavian vessels is well documented in civilian and military surgical practice) s which probably reflects their anatomic inaccessibility. Nonpenetrating trauma resulting in damage to these vessels is even rarer. In the present series, most of the injuries resulted from motor vehicle accidents. In eight patients automobile passenger-restraining devices could be incriminated. In each case a shoulder-harness seat belt with a lap strap (three-point fixation) had been used. On close questioning of the individual patients, the diagonal component (shoulder strap) had been loosefitting, allowing considerable play between the chest wall and the belt. This had been a consequence of either faulty adjustment of the older (noninertia reel) belts, which require manual tightening, or the person had taken up the slack on the inertia reel by means of a peg or clip to relieve pressure across the shoulder. In two cases women were attempting to avoid creasing their dresses in this way. It is thought that at the moment of impact the forward acceleration of the torso, pivoting on the lap strap component, is suddenly stopped by the shoulder strap across the shoulder girdle, which being loose allows the torso to continue its movement anteriorly with the shoulder as a fixed point. The resultant distraction and shearing forces result in disruption of the neurovascular bundle and bony structures around the shoulder girdle. Similar stretchtraction-rotation stress mechanisms may be involved in injuries resulting from falls, such as from a moving vehicle. Previous reports have highlighted the skeletal complications of the use of automobile restraining devices6 and only isolated reports of arterial injuries consequent to their use have appeared. 7 Another possible mechanism of injury is the result of direct trauma with a fractured clavicle or first rib compressing the neurovascular bundle by a scissor mechanism. The degree of violence necessary to cause this type of injury is reflected in the fact that no patient had an isolated vascular injury. All patients had an absent radial pulse and more than half had an obviously ischernic limb, which reflects the degree of disruption of the collateral circulation. This is at variance with others who report the presence of pulses in most of their patients. ~'9All
Volume 8 Number 1 July 1988
patients with seat belt injuries had a characteristic pattern of abrasions over the shoulder region and across the torso. As previously described,2'3 we believe that all clinically stable patients in whom exploration of the superior mediastinal or shoulder girdle arteries is contemplated should have routine angiography to plan the operative approach. This assumes particular importance in the presence of severe blunt trauma to ensure that the aortic isthmus is intact; this approach is supported by others? ,9 There is little doubt that subclavian injuries resulting from penetrating trauma should be routinely explored and repaired in view of the complications that may follow the subsequent development of a false aneurysm, in particular brachial plexus compression injuries. 2,3,1°12 Similarly, in the situation in which there is critical ischemia of the extremity, coupled with possible neurologic injury, arterial repair should be performed. This would allow for a full assessment of the extent of the neurologic deficit subsequently. There is room for debate in the patient with a wellperfused flail limb and obvious total brachial plexus disruption. In all five of those with a total flail limb who had exploratory surgery, the brachial plexus was shredded beyond repair. In one of these patients clasSic Homer's syndrome was noted, which indicates avulsion of the nerve roots from the spinal cord. This sign is subtle and perhaps a more diligent assessment in these patients may reveal this as a frequent finding, which could be regarded as a contraindication to exploration. Detailed evaluation of the long-term neurologic status in these patients is beyond the scope of this article. However, brachial plexus repair at trunk level is generally unrewarding, although some success has been reported with penetrating injury? 3 It appears that the lesions found in this group of patients cannot be treated by reconstruction. Surgical approaches to the artery followed standard lines and have been fully described in earlier publications from this unit. 2,3The results for standard
Nonpenetratmg subclavian artery trauma 75
arterial repair techniques are rewarding as borne out in the present series, although the prospects are bleak for restoration of nel-ce function at the brachial plexus. The merits and demerits of automobile restraining devices have been debated at great l e n g t h / A t present they are a legal requirement in the Republic of South Africa and our general impression is that in terms of survival and patterns of injury in automobile accidents, the advantages outweigh the hazards. However, it must be emphasized that these seat belts should be correctly and snugly applied to avoid the catastrophic injuries described herein. REFERENCES 1. Rich NM, Hobson RW, Jarsffer BS, et al. Subclavian artery trauma. J Trauma 1973;13:485-96. 2. Robbs IV~ Baker LW, Human RR, et al. Cervico-mediastinal arterial injuries: a surgical challenge. Arch Surg 1981; 116:663-8. 3. Robbs JV, Baker LW. Cardiovascular trauma. Curr Probl Surg, Chicago: Year Book Press, April 1984. 4. Graham JM, Feliciano DV, Mattox KL, et al. Management of subclavian vascular injuries. I Trauma 1980;20:537-44. 5. DeBakey ME, Simeone FA. Battle injuries of the arteries in World War H. An analysis of 2471 cases. Ann Surg 1946; 123:534-79. 6. Williams JS, Kirkpatrick JR. The nature of seat belt injuries. J Trauma 1971;11:207-18. 7. Woelfel GF, Moore EE, GoghiU TH, et al. Severe thoracic and abdominal injuries associated with lap-harness seatbeks. J Trauma 1984;24:166-7. 8. Zelenock GB~ Kazmers A, Graham LM, et al. Nonpenetrating subclavian artery injuries. Arch Surg 1985;120:685-91. 9. Lira LT, Saletra ~D, Flanigan DP, et al. Subclavian and innominate artery trauma. Surgery 1979;86:890-6. 10. Andersen SK, Lorentzen JE, Rohr N. Arterial injuries of the upper extremities. Acta Chir Scand 1983;149:473-7. 11. Tomaszek DE. Combined subdavian artery and brachial plexus injuries from blunt upper extremity trauma. J Trauma 1984;24:161-3. 12. Robbs JV, Naidoo KS. Nerve compression injuries due to traumatic false aneurysm. Ann Surg 1984;200:80-2. 13. Sarkin TL. The diagnosis and management ofbrachial plexus lesions. S Afr I Surg t975;13:107-11.
Durban, Republic of So. Africa
Nonpenetrating injury to the subclavian vessels is uncommon. During a 6-year period we have treated 167 patients with injuries to the subclavian and superior mediasfinal arteries. Fifteen of these injuries (9%) occurred after blunt trauma. In 10 patients the proximal segment (first and second parts) of the artery was involved. N o patient had an isolated injury; the most frequent associated injuries were rib fractures (n = 11), with the first rib being involved in four of these. Total brachial plexus disruption was found in nine patients. All patients with distal artery involvement had a clavicular fracture. All had an absent radial pulse and eight had critical ischemia ofthe hand. Four patients were treated nonoperatively and the remainder were treated along standard lines. Brachial plexus reconstruction was not feasible in any patient. Within 2 weeks of operation, one patient died as a result of head injuries and one required amputation because of sepsis, During a 12-month period, five regained fifll function, one additional patient requested above-elbow amputation after 6 months, and seven had a flail anesthetic limb. Twelve of these patients were involved in automobile accidents, eight of whom were wearing lap~ shoulder harness seat belts with a loose-fitting shoulder strap component that created a characteristic abrasion pattern on the torso and chest. We conclude that the torsionshearing motion allowed by this situation contributed significantly to the pattern of injury and a plea is made for correctly fitted restraining devices. (J Vase SUgG 1988;8:71-5.)
Injury to the subclavian artery is uncommon in civilian and military surgical practicelS; of 7500 vascular injuries recorded in the Vietnam Vascular Registry less than 1% involved the subclavian artery. 1 Most of these injuries are due to penetrating trauma and outnumber blunt trauma to these vessels by a ratio of between 40 and 50 to 1.5 Blunt trauma may be associated with multiple injuries and presentation may be subtle. The consequences ofmisdiagnosis and late operative treatment of penetrating subclavian injuries have been well described, 24 with delayed hemorrhage, false aneurysm, and arteriovenous fistula being most frequently encountered. Operative management presents a diflicuk challenge and the reported mortality rate in various series ranges between 10% and 30%. 1"4Recent experience on our Vascular Service with several patients having blunt injury to these vessels has prompted the present review. PATIENTS A N D M E T H O D S Between 1981 and 1987, 167 patients have been treated on the Vascular Service for injuries to the subclavian and superior mediastinal arteries. O f these, 15 patients (9%) had injuries to the subclavian From the Department of Surgery, University of Natal Medical School. Reprint requests: ProfessorJ. V. Robbs, Dept. of Surgery, University of Natal Medical School, P.O. Box 17039, Congella, 4013, South Africa. 71
artery as a result of bltmt trauma. Details relating to the patients are summarized in Table I. Ten were male, and the average age of the group was 31.2 years (range 20 to 59 years). The injury w a s associated with motor vehicle accidents in 12 patients. Eight were either the driver or front seat passenger and were wearing lap-shoulder-harness seat belts. In every case the shoulder harness component of the belt had been loose with considerable play between the patient's shoulder and chest wall and the belt. Therefore at the time of impact the patient was,thrown against the belt, with considerable force being taken over the clavicle, shoulder, and anterior chest. In most cases this would apparently be followed by considerable lateral movement of the upper torso with the shoulder acting as a fixed point. This sequence resulted in a characteristic pattern of bruising and abrasions over the chest wall, neck, and shoulder. The other four cases comprised a pedestrian struck by a vehicle, one who fell from a moving truck, a young male thrown from a motorcycle, and one patient ejected from a vehicle after a collision. Three patients sustained more focal direct trauma to the supraclavicular area from a falling tree, assault with a club, and a heavy chain, respectively. The latter was an industrial accident in which a tow-chain parted while under tension. The second part of the artery was injured in seven
J oimTtad Oi"
72
VASCULAR SURGERY
Costa and Robb
Table I. Summary of 15 cases of nonpcnetrating trauma to the subclavian artery Patient
Age
Sex
1
24
M
2
43
3
Mechanism of injury
Artery
Associated injury
Presentation
Motor vehicle accident (seat belt)
(R) subclavian-second part
Total brachial plexus Multiple rib fractures Fracture humerus
Critical ischemia arm and hand Anesthetic flail limb
M
Motor vehicle accident (seat belt)
(R) subclavian-second part
Absent radial pulse Anesthetic flail limb
28
M
(R) subclavian-second part
4
30
F
Motor vehicle accident (seat belt) Motor vehicle accident (seat belt)
Fracture first rib Multiple rib fractures Flail chest Total brachial plexus Fracture first rib
(L) subclavian-second part
Fracture clavicle Total brachial plexus Homer's syndrome
Absent radial pulse Anesthetic flail limb
5
59
F
(L) subclavian-second part
Multiple rib fractures
Absent radial pulse
6
23
M
Motor vehicle accident Passenger (seat belt) Motor vehicle accident (seat belt)
(R) subclavian-second part
Critical ischemia Anesthetic flail limb
7
30
M
(R) subclavian-first part
8
28
M
Motor vehicle accident (seat belt) Motor vehicle accident (seat belt)
Total brachial plexus Compound fracture humerus Fracture scapula Extensive soft tissue trauma Fracture first rib Multiple rib fractures
(L) subclavian-first part
Lung contusion Multiple rib fractures with flail segment
Absent radial pulse Ischemic hand
30
F
Motor vehicle accident (pedestrian)
(L) subclavian-second part
Critical ischemia arm and hand
10
5O
F
Motor vehicle accident (passenger)
(R) subclavian-third part
Dislocation sternoclavicular joint Lung contusion Diffuse brain injury Fracture clavicle Total brachial plexus
11
24
M
Motorcycle accident
(R) subclavian-first part
12
20
M
Fell off truck
(R) subclavian-third part
13
20
F
Hit by falling tree
(L) subclavian-third part
14
27
M
Assault with club
(L) subclavian-third part
15
29
M
Industrial-hit by chain
(L) subclavian-third part
Total brachial plexus Fracture humerus Multiple rib fractures Fracture clavicle Multiple rib fractures Total brachial plexus Head injury Fracture clavicle Multiple rib fractures Total brachial plexus Fracture ribs
Compound fracture clavicle Fracture first rib Multiple rib fractures Total brachial plexus
Absent radial pulse
Critical ischemia arm and hand
Critical ischemia arm and hand Pregangrenous fingers Flail anesthetic limb Absent radial pulse Flail anesthetic limb Absent radial pulse Flail anesthetic limb Absent radial pulse Flail anesthetic limb Critical ischemia arm and hand
Critical ischemia hand and arm Flail anesthetic limb
Volume 8 Number 1 luly 1988
Operative procedure
Nonpenetrating subclavian artery trauma
Pathologic features
Supraclavicular incision (L) carotidsubclavian bypass (vein) No operation
Intimal disruption and thrombosis Shredded brachial plexus
Sternotomy Interposition graft (Dacron) Supraclavicular approach - vessel ligated
Intimal tcar and thrombosis
Sternotomy Interposition graft (Dacron)
--
Total disruption artery Shredded brachial plexus trunks Intimal tear and thrombosis
Result 6 mo Pulses returned Anesthetic flail limb 3 mo Anesthetic flail limb Cansalgia 3 mo Pulse returned Normal function 6 mo Causalgia Above-elbow amputation 12 mo Pulse returnedfull function
Sternotomy Interposition graft (vein)
Total disruption artery Shredded brachial plexus trunks
1 wk Gross sepsis Above-elbow amputation
Sternotomy Interposition graft (Dacron) (L) Thoracotomy Balloon catheter Thrombectomy
Total disruption artery
3 mo Pulse returned Full fimction 6 wk No return of pulse--severe claudication of arm Carotid-subclavian bypass 6 mo Pulse returned-full fimction 2 wk Pulse restored Died--brain death 6 mo Pulse returned Flail limb
Intimal tear and thrombosis
Sternotomy Interposition graft (Dacron)
Total disruption artery
Supraclavicular approach Interposition graft
Total disruption artery Shredded brachial plexus
(vein) No operation
3 mo
Flail limb Causalgia 5 mo Flail limb
No operation
No operation
6 'mo
Flail limb Supraclavicular approach End-to-end anastomosis subclavian Sternotomy Interposition graft (vein)
Intimal tear and thrombosis
Total disruption artery Shredded brachial plexus
3 mo
Pulse returned Full function 3 mo
Pulse returned Flail limb
73
patients, the distal segment (third part) in five, and the proximal segment close to its origin (first part) in three patients. There was no predilection for any particular side, with eight involving the right. No patient had an isolated vascular injury. Multiple rib fractures were encountered in 10 patients and were the most common associated injuries, including the first rib in three; one additional patient had an isolated fracture of the first rib. Other bone and joint injuries included fractures of the clavicle (five patients), humerus (three), scapula (one), and facial bones (one). Dislocation of the sternoclavicular joint was encountered in one patient. Nine patients sustained total brachial plexus injuries. One of these (patient 4) had an obvious associated Homer's syndrome, suggesting avulsion of the nerves at root level. These patients had an anesthetic, totally paralyzed arm, four of the nine manifested critically ischemic changes. Four patients had a flail segment of the chest wall with associated lung contusion, whereas two patients had a significant head injury with an impaired level of consciousness. Clinically, a radial pulse was not palpable in all patients, with a discernible temperature difference compared with the opposite upper limb. Eight patients showed signs of critical ischemia with a cold, blue, mottled hand and forearm.
MANAGEMENT Four patients were treated nonoperatively. All had total brachial plexus lesions, a viable limb, and severe associated chest wall injuries that required positive pressure ventilation. With the above four exceptions all patients in whom operative treatment was contemplated were subjected to arch angiograplay via the retrograde femoral route. Exploration of the neurovascular bundle was done in 11 patients. In six patients the proximal subclavian artery was approached via a limited median sternotomy and in the remainder with more distal injuries, a supraclavicular approach with division of the clavicle in mid-shaft: was used. These approaches have been described in detailY In one patient the proximal left subclavian artery was approached via a fourth intercostal space thoracotorny. In four patients the artery was found to be totally disrupted and in the remainder, intimal disruption with superimposed thrombosis was found. Of five patients with brachial plexus lesions subjected to exploration, four of whom had severe peripheral ischemia, all were found to have total disruption of the brachial plexus with shredding of the trunks in the supraclavicular fossa. In no patient was any attempt made at primary nerve repair and the nerve ends were marked with nonabsorbable suture material as best
74
Costaand Robb
possible to facilitate subsequent surgical exploration if required. To restore arterial continuity, seven patients required an interposition graft, with saphenous vein used in five and Dacron in two. On one occasion it was possible to achieve an end-to-end anastomosis without tension. In one patient (patient 4) in whom the brachial plexus was shredded, the vessel was ligated. Carotid-subclavian bypass on the left side was performed primarily in one patient (patient 1). One patient (patient 15) came to an outlying hospital after a motor vehicle accident, with severe chest injuries and an ischemic hand. Via a left thoracotomy a catheter thrombectomy of the left subclavian artery was attempted. The hand ischemia improved, although peripheral pulses were not restored. The patient was referred to the Vascular Service 6 weeks later with severe claudication of the hand and forearm. Angiography revealed total occlusion of the proximal left subclavian artery. Via a supraclavicular approach, carotid-subclavian bypass with reversed saphenous vein was performed; pulses were restored. All surviving patients were reassessed in the Vascuiar Clinic at monthly intervals for a period ranging between 3 and 12 months. All those with nerve injuries were referred to the peripheral nerve injury clinic, under the aegis of the Orthopaedic Department, at the time of discharge from hospital for longterm assessment and management. Details of this aspect of patient management are beyond the scope of this article. RESULTS One patient died (patient 9) as a result of severe diffuse brain injury. She was declared brain dead after 2 weeks and life support was discontinued. Two patients required amputation. One (patient 6) had severe soft tissue trauma and a total brachial plexus lesion. Gross infection mandated amputation within a week of operation at the subcapital level of the humerus despite a patent vascular repair. The other patient had severe cansalgia in an otherwise flail anesthetic limb and requested amputation 6 months after injury. This was also performed just below shoulder level. O f the remaining eight patients who had arterial reconstruction, all regained peripheral pulses; five have regained full function of the upper limb and three had associated brachial plexus lesions. The four patients who did not have initial exploratory surgery had viable limbs but with no neurologic function. To summarize the short-term outcome in the nine patients with brachial plexus disruption, two
Journal of VASCULAR SURGERY
had amputations and the remainder had a flail anesthetic upper limb 3 to 6 months later. In addition, two of these patients had intermittent episodes of unpleasant spontaneous phantom sensations (cansalgia). DISCUSSION The relative rarity of injury to the subclavian vessels is well documented in civilian and military surgical practice) s which probably reflects their anatomic inaccessibility. Nonpenetrating trauma resulting in damage to these vessels is even rarer. In the present series, most of the injuries resulted from motor vehicle accidents. In eight patients automobile passenger-restraining devices could be incriminated. In each case a shoulder-harness seat belt with a lap strap (three-point fixation) had been used. On close questioning of the individual patients, the diagonal component (shoulder strap) had been loosefitting, allowing considerable play between the chest wall and the belt. This had been a consequence of either faulty adjustment of the older (noninertia reel) belts, which require manual tightening, or the person had taken up the slack on the inertia reel by means of a peg or clip to relieve pressure across the shoulder. In two cases women were attempting to avoid creasing their dresses in this way. It is thought that at the moment of impact the forward acceleration of the torso, pivoting on the lap strap component, is suddenly stopped by the shoulder strap across the shoulder girdle, which being loose allows the torso to continue its movement anteriorly with the shoulder as a fixed point. The resultant distraction and shearing forces result in disruption of the neurovascular bundle and bony structures around the shoulder girdle. Similar stretchtraction-rotation stress mechanisms may be involved in injuries resulting from falls, such as from a moving vehicle. Previous reports have highlighted the skeletal complications of the use of automobile restraining devices6 and only isolated reports of arterial injuries consequent to their use have appeared. 7 Another possible mechanism of injury is the result of direct trauma with a fractured clavicle or first rib compressing the neurovascular bundle by a scissor mechanism. The degree of violence necessary to cause this type of injury is reflected in the fact that no patient had an isolated vascular injury. All patients had an absent radial pulse and more than half had an obviously ischernic limb, which reflects the degree of disruption of the collateral circulation. This is at variance with others who report the presence of pulses in most of their patients. ~'9All
Volume 8 Number 1 July 1988
patients with seat belt injuries had a characteristic pattern of abrasions over the shoulder region and across the torso. As previously described,2'3 we believe that all clinically stable patients in whom exploration of the superior mediastinal or shoulder girdle arteries is contemplated should have routine angiography to plan the operative approach. This assumes particular importance in the presence of severe blunt trauma to ensure that the aortic isthmus is intact; this approach is supported by others? ,9 There is little doubt that subclavian injuries resulting from penetrating trauma should be routinely explored and repaired in view of the complications that may follow the subsequent development of a false aneurysm, in particular brachial plexus compression injuries. 2,3,1°12 Similarly, in the situation in which there is critical ischemia of the extremity, coupled with possible neurologic injury, arterial repair should be performed. This would allow for a full assessment of the extent of the neurologic deficit subsequently. There is room for debate in the patient with a wellperfused flail limb and obvious total brachial plexus disruption. In all five of those with a total flail limb who had exploratory surgery, the brachial plexus was shredded beyond repair. In one of these patients clasSic Homer's syndrome was noted, which indicates avulsion of the nerve roots from the spinal cord. This sign is subtle and perhaps a more diligent assessment in these patients may reveal this as a frequent finding, which could be regarded as a contraindication to exploration. Detailed evaluation of the long-term neurologic status in these patients is beyond the scope of this article. However, brachial plexus repair at trunk level is generally unrewarding, although some success has been reported with penetrating injury? 3 It appears that the lesions found in this group of patients cannot be treated by reconstruction. Surgical approaches to the artery followed standard lines and have been fully described in earlier publications from this unit. 2,3The results for standard
Nonpenetratmg subclavian artery trauma 75
arterial repair techniques are rewarding as borne out in the present series, although the prospects are bleak for restoration of nel-ce function at the brachial plexus. The merits and demerits of automobile restraining devices have been debated at great l e n g t h / A t present they are a legal requirement in the Republic of South Africa and our general impression is that in terms of survival and patterns of injury in automobile accidents, the advantages outweigh the hazards. However, it must be emphasized that these seat belts should be correctly and snugly applied to avoid the catastrophic injuries described herein. REFERENCES 1. Rich NM, Hobson RW, Jarsffer BS, et al. Subclavian artery trauma. J Trauma 1973;13:485-96. 2. Robbs IV~ Baker LW, Human RR, et al. Cervico-mediastinal arterial injuries: a surgical challenge. Arch Surg 1981; 116:663-8. 3. Robbs JV, Baker LW. Cardiovascular trauma. Curr Probl Surg, Chicago: Year Book Press, April 1984. 4. Graham JM, Feliciano DV, Mattox KL, et al. Management of subclavian vascular injuries. I Trauma 1980;20:537-44. 5. DeBakey ME, Simeone FA. Battle injuries of the arteries in World War H. An analysis of 2471 cases. Ann Surg 1946; 123:534-79. 6. Williams JS, Kirkpatrick JR. The nature of seat belt injuries. J Trauma 1971;11:207-18. 7. Woelfel GF, Moore EE, GoghiU TH, et al. Severe thoracic and abdominal injuries associated with lap-harness seatbeks. J Trauma 1984;24:166-7. 8. Zelenock GB~ Kazmers A, Graham LM, et al. Nonpenetrating subclavian artery injuries. Arch Surg 1985;120:685-91. 9. Lira LT, Saletra ~D, Flanigan DP, et al. Subclavian and innominate artery trauma. Surgery 1979;86:890-6. 10. Andersen SK, Lorentzen JE, Rohr N. Arterial injuries of the upper extremities. Acta Chir Scand 1983;149:473-7. 11. Tomaszek DE. Combined subdavian artery and brachial plexus injuries from blunt upper extremity trauma. J Trauma 1984;24:161-3. 12. Robbs JV, Naidoo KS. Nerve compression injuries due to traumatic false aneurysm. Ann Surg 1984;200:80-2. 13. Sarkin TL. The diagnosis and management ofbrachial plexus lesions. S Afr I Surg t975;13:107-11.