Volume 39, Issue 2 , Pages 155-159, February 2010
An Analysis of 50 Surgically Managed Penetrating Subclavian Artery Injuries☆☆☆
Article Outline
- Abstract
- 1. Introduction
- 2. Patients and methods
- 3. Operative technique
- 4. Results
- 5. Discussion
- Acknowledgements
- References
- Copyright
Abstract
Objectives
The surgical management and outcome of penetrating subclavian artery (SCA) injuries is presented in this article.
Design
A retrospective chart review is used to detail the management and outcome of penetrating SCA injuries.
Patients and methods
Patients with penetrating SCA injuries presenting to the Groote Schuur Hospital from January 1997 to December 2007 were reviewed. Demographic data, mechanism of injury, associated injuries, angiographic findings, surgical treatment, hospital stay, complications and mortality were noted.
Results
Fifty patients with penetrating SCA injuries were identified from an operating trauma database. Stab and gunshot wounds accounted for 40 and 10 SCA injuries, respectively. The mean Revised Trauma Score (RTS) was 7.2. Angiography was obtained in 37 patients; false aneurysm (13) and total occlusion (nine) were the two most common findings. A median sternotomy was required in 25 (50%) patients and emergency room thoracotomy was performed in two patients (4%) for initial haemorrhage control. Primary repair of SCA injuries was possible in 52% of the patients. Three SCA injuries (6%) were ligated and one patient received an endovascular stent. Morbidity was restricted to associated brachial plexus injuries. The limb salvage rate was 100% and there were no deaths.
Conclusion
Preoperative angiography was useful in planning an operative approach. Primary repair was possible in the majority of the patients and ligation of SCA injuries was life-saving in critically ill patients.
Keywords: Vascular trauma, Penetrating, Subclavian artery injuries
1. Introduction
Penetrating subclavian artery (SCA) injuries occur infrequently and constitute less than 2% of all civilian vascular traumas.1, 2, 3, 4 Limited clinical experience, complex local anatomy and difficult exposure of proximal mediastinal vessels have rendered management of these injuries particularly difficult. The operative mortality rate ranges from 5% to 30% and is attributed to rapid exsanguination and a high incidence of concomitant injuries.5, 6 The purpose of this study was to review the surgical treatment and outcome of SCA injuries in an urban trauma centre with a high incidence of penetrating trauma.
2. Patients and methods
The records of all patients undergoing surgery for a penetrating SCA injury in the Trauma Center at the Groote Schuur Hospital during the 11-year period from January 1997 to December 2007 were retrieved from a prospective trauma operating theatre database and retrospectively reviewed. There were no cases of non-operated SCA injuries.
Standard demographic information, mechanism of injury, admission vital signs, peripheral neurological deficit, haemoglobin concentration, blood transfusion requirements and diagnostic methods were recorded. Operation notes documented the location of the SCA injury, method of repair and local associated injuries. Admissions to the intensive care unit (ICU), duration of ICU and hospital stays as well as complications were noted. Initial management and resuscitation were conducted along standard Advanced Trauma and Life Support (ATLS)® guidelines. Emergency room thoracotomy (ERT) was performed in patients with no signs of life or with imminent cardiac arrest and/or immediate drainage of 1500
ml of blood from tube thoracostomy. Patients presenting in shock with active bleeding, or with an ischaemic limb, were resuscitated and expediently taken to the operative room (OR) for emergency exploration. Haemodynamically stable patients and those who stabilised after simple resuscitation (less than 2l crystalloids) underwent further evaluation. Indications for angiography were: ipsilateral distal pulse discrepancy, absent pulse in the presence of a viable limb, large haematoma, thrill/bruit and mediastinal changes on plain chest radiograph. Transmediastinal tracts were investigated with a computed tomography of the chest and/or neck to determine trajectory, and angiography was then performed in the presence of mediastinal haematoma and/or an inconclusive CT finding, that is, the radiologist was not confident that an injury could be excluded.
3. Operative technique
Surgical access to the subclavian artery depends on the clinical presentation and site of vascular injury. Patients with no signs of life or in imminent cardiac arrest require emergency room thoracotomy. Haemodynamically unstable patients are transferred to the operating room immediately. In the presence of active bleeding, patients are placed in the Trendelenburg position to minimise the risk of air embolism and control of external haemorrhage is achieved by simple digital or manual pressure. Foley-catheter balloon tamponade (FCBT) can be used in an attempt to control bleeding, especially from the retroclavicular region. Further, a ‘swab on a stick’ can be used to apply pressure to minimise bleeding during exposure.
In the OR, all patients are positioned supine with a bolster beneath the shoulders; the head is placed in a head ring and turned to the opposite side with slight extension of the neck. The ipsilateral arm is down at the side. All patients are prepped and draped as for a median sternotomy, which allows for supra-, cross- and infraclavicular extensions. The groin is also prepared for harvesting the saphenous vein. Patients bleeding actively from the supraclavicular region undergo median sternotomy with supraclavicular extension and proximal control of the subclavian vessels. Patients bleeding from the infraclavicular region undergo supraclavicular exposure and proximal control if possible, failing this, a median sternotomy is performed. Patients undergoing surgery following confirmation of an injury on angiography undergo median sternotomy initially if the injury is proximal to the vertebral artery and/or involving the first part of the subclavian artery; all other injuries are approached through a supraclavicular incision, failing which, a median sternotomy is rapidly performed. Distal control is mandatory because of the extensive collateral circulation in the neck and shoulder region and is usually achieved with an infraclavicular incision in the deltopectoral groove. The supra- and infraclavicular incisions can be joined over the clavicle, which can be divided when necessary. Damage control surgery is reserved for the critically ill exsanguinating patient approaching the ‘triad of death’ of hypothermia, coagulopathy and acidosis. An initial ‘bail out’ operation for penetrating SCA injuries will include temporary shunting with pieces of nasogastric tubes, suction catheters and intravenous lines or ligation of the artery with or without packing of the surgical dissection with swabs. The patient is then resuscitated in the ICU and definitive arterial repair performed 24–48h later on a stable, rewarmed patient with an acceptable coagulation profile.
4. Results
Fifty patients with penetrating SCA injuries comprised the study group. There were 49 men and one woman with a mean age of 27 (range: 15–54) years. The mechanism of injury was a stab wound in 40 (80%) patients and gunshot wound (GSW) in 10 patients. Twenty-eight patients (56%) sustained injuries to the left SCA and 22 (44%) had right-sided injuries. Forty-eight patients (96%) presented within 24h of their injuries and two patients (4%) had significant delays of 4 and 7days, respectively. The mean Revised Trauma Score (RTS) was 7.2 (range: 3.0–7.8). The mean finger-prick haemoglobin estimation on admission was 9.4 (range: 4–14) G%. Seven patients (14%) were hypotensive (systolic blood pressure <90mmHg) on arrival to the emergency room, whilst 43 patients (86%) were haemodynamically normal on presentation. A pulse deficit was present in 29 patients (58%). FCBT was used in four patients; in three patients bleeding was arrested and patients stabilised enough for formal angiography. The fourth patient had persistent bleeding, requiring immediate surgical exploration. Demographic data and admission clinical findings are listed in Table 1. Angiography was performed in 37 (74%) patients; the most common finding was a false aneurysm (Fig. 1). Table 2 lists the indications for angiography and their findings. A mean perioperative transfusion of 4.4 (range: 0–19) units of packed cells was administered. Thirteen patients (26%) required emergency exploration, the indications being active haemorrhage (10) and critical limb ischaemia (three). A variety of operative exposures were used; sternotomy provided adequate exposure in 25 patients (50%) (Table 3). A combination of supra- and infraclavicular incisions with or without dissection of the clavicle was required in 15 patients (30%). A supraclavicular incision was used in nine patients (18%). One patient (2%) with a false aneurysm, with a previous sternotomy and bilateral thoracotomy (all performed on previous separate admissions), had a prosthetic polytetrafluoroethylene (PTFE)-covered stent placed radiologically. Two patients (4%) with left SCA injuries needed an ERT for haemorrhage control in addition to supraclavicular exposure in the OR. Eleven patients had division of the clavicle: three were divided and plated, six were divided and not plated and two patients had medial resection of the clavicle. Lateral suture and end-to-end anastomosis following adequate vessel debridement were the most common methods of arterial repair used in 16 (32%) and 10 (20%) patients, respectively (Table 3). PTFE interposition grafting was performed in 11 patients (22%) because of size discrepancy between the native artery and autogenous vein graft. Arterial reconstruction with saphenous vein interposition grafting was used in nine patients (18%). Ligation of the SCA was performed in three patients (6%); two of these patients presented with significant haemorrhage and one patient had an infected false aneurysm with delayed presentation. Twelve (24%) patients sustained concomitant subclavian vein (SCV) injuries; nine were ligated and three were primarily repaired. Mild limb swelling was documented postoperatively in the patients who underwent SCV ligation. Brachial plexus injuries accounted for long-term morbidity and were seen in six (12%) patients. Transection of the brachial plexus was documented in two patients; primary repair was performed in one patient at the time of SCA injury surgery and delayed repair with a sural nerve graft was performed at 4months after injury in the other patient. Both these patients unfortunately experienced major neurological deficits following repair. The remaining four patients (three with brachial plexus incomplete lacerations and one with a contusion) were managed conservatively with physiotherapy and specialised splints, and recovered uneventfully. Postoperative complications were seen in nine (18%) patients. Three (6%) patients had chyle leaks – one patient with a high-output leak had ligation of the thoracic duct while the remaining two were managed conservatively with medium chain triglycerides. Two patients (4%) with PTFE grafts were presumed to have graft sepsis on the basis of pyrexia and superficial wound sepsis, and associated sternal sepsis, respectively. The latter patient required a pectoral flap with parenteral antibiotics while the other patient settled on parenteral antibiotics alone. Clotted haemothorax occurred as a complication in two cases (4%) and both were successfully managed with video-assisted thoracoscopic evacuation. One patient (2%) developed an empyema, which resolved after repeat ultrasound-guided tube thoracostomy. Their limb salvage rate was 100% and no deaths were recorded during our study.
Table 1. Demographics and clinical presentation of 50 patients with penetrating subclavian artery injuries.
| No. of patients (%) | |
|---|---|
| Gender | |
| Male | 49 (98) |
| Female | 1 (2) |
| Mechanism of injury | |
| Stab wounds | 40 (80) |
| Gunshot wounds | 10 (20) |
| Side | |
| Left | 28 (56) |
| Right | 22 (44) |
| Presentation | |
| <24h | 48 (96) |
| >24h | 2 (4)* |
| Haemodynamics on arrival | |
| Hypotensive (<90mmHg) | 7 (14) |
| Stable (>90mmHg) | 43 (86) |
| Clinical manifestation | |
| Pulse deficit | 29 (58) |
| Bleeding | 10 (20) |
| Bruit | 10 (20) |
| Large haematoma | 6 (12) |
| Pulsatile haematoma | 5 (10) |
| Limb ischaemia | 2 (4) |
| Mean haemoglobin | 9.4 (4–14) G% |
| Mean revised trauma score | 7.2 (3.0–7.8) |
∗Two patients presented with significant delay of 4 and 7 |
Figure 1
Left SCA aneurysm (block arrow) in patient with a stab wound.
Table 2. Indication for angiography and their findings in 37 patients.
| Number (%) | |
|---|---|
| Indications for angiogram | |
| Pulse discrepancy | 19 (38) |
| Bruit | 10 (20) |
| Large haematoma | 1 (2) |
| Moderate haematoma | 4 (8) |
| Pulsatile haematoma | 1 |
| Widened mediastinum on chest radiograph | 1 |
| Delayed presentation of brachial plexus deficit | 1 |
| Angiographic finding | |
| False aneurysm (FA) | 13 (26) |
| Occlusion/transection | 13 (26) |
| Arteriovenous fistula (AVF) | 5 (10) |
| Intimal flap | 4 (8) |
| FA and AVF | 2 (4) |
Table 3. Surgical approach and management of subclavian artery injuries.
| 1st | 2nd | 3rd | |
|---|---|---|---|
| Emergency | |||
| Emergency room thoracotomy | – | 2 | – |
| Sternotomy | 1 | – | – |
| Sternotomy with neck extension (ext.) | 1 | 1 | – |
| Sternotomy, supra- and infraclavicular ext. with clavicle division | 1 | 1 | – |
| Sternotomy with supraclavicular ext. | 1 | 1 | 1 |
| Supraclavicular and clavicle division | – | 2 | – |
| Supraclavicular and infraclavicular with clavicle division | – | 1 | – |
| Elective | |||
| Sternotomy | 2 | – | – |
| Sternotomy with supraclavicular ext. | 8 | 3 | – |
| Sternotomy, supraclavicular ext. with clavicle division | 1 | 2 | – |
| Sternotomy with supra- and infraclavicular ext. | 1 | – | – |
| Sternotomy, supra- and infraclavicular ext. with clavicle division | 1 | – | – |
| Supraclavicular | – | 3 | 2 |
| Supra- and infraclavicular | – | 6 | 5 |
| Supra- and infraclavicular and clavicle division | – | 1 | 1 |
| Management | N (%) | ||
| Ateriorraphy (lateral repair) | 16 (32) | ||
| Resection and primary anastomosis | 10 (20) | ||
| PTFE | 11 (22) | ||
| Reverse saphenous vein graft | 9 (18) | ||
| Endovascular stent | 1 (2) | ||
| Ligation | 3 (6) | ||
5. Discussion
Penetrating SCA injuries are particularly devastating and continue to pose a major surgical challenge. Surgical stress combined with a high incidence of concomitant injuries contributes to significant morbidity and mortality. Pre-hospital mortality is about 75% and long-term morbidity is mainly attributed to brachial plexus injuries.7, 8 Trauma severity is directly proportional to missile velocity and gunshot injuries tend to be more severe than stab wounds.5
The majority of our patients (96%) presented to the trauma unit within 24h of the injury and 58% of patients had a pulse deficit on admission. However, delayed presentation is not uncommon because hard vascular signs and a pulse deficit are only seen in about 20% of patients;8 extensive collateral circulation in the upper limb prevents ischaemia even in the presence of proximal injuries. The incidence of critical limb ischaemia on presentation was 6% and expedient exploration of these patients resulted in only one fasciotomy with delayed skin graft and no limb loss. Formal angiography was performed in 37 haemodynamically stable patients (74%). This diagnostic modality is extremely helpful in planning the surgical approach and screening for occult and associated vascular injuries. It can also help determine which patients can benefit from endovascular intervention.
A wide range of surgical approaches have been described for the management of penetrating SCA injuries.6, 7, 8, 9 In our experience, a combination of median sternotomy and clavicular incisions provide an excellent exposure for both left- and right-sided injuries. ERT for haemorrhage control followed by supraclavicular incision in the operating room can be useful in patients with left-sided injuries. When a long segment of damaged SCA was encountered at surgery, the clavicle was divided (11 patients, 22%) to enhance operative exposure. Amongst this group of patients, we reconstructed the clavicle in three cases and performed medial resection in two others. Complication rates of up to 50% have been observed with clavicle reconstruction and many advocate the use of medial resection, although it should be emphasised that the latter technique can significantly affect functional outcome in the long run.7 At a 2-week follow-up, no patients had complications related to clavicle division.
Concomitant SCA and SCV injuries were seen in 24% of all patients and most venous injuries were ligated. Postoperative limb oedema was transient and resolved with elevation. Primary repair of the SCA was possible in more than half of our patients. Interposition grafting was more frequently used in patients with GSWs, possibly because of the blast effect;5 70% of patients with GSWs required interposition grafts whilst only 35% of stab wound victims received grafts. We used almost equal numbers of prosthetic grafts and autogenous veins. As with previous studies,5, 8 graft sepsis was limited to synthetic grafts and resolved with parenteral antibiotics. Ligation was reserved for cases with delayed presentation and sepsis and critically ill haemodynamically unstable patients needing damage control surgery. Innovations in interventional radiology and new endovascular techniques will probably drastically change the management of SCA trauma in the future. Only one patient from our study group benefitted from an endovascular stent. It is important to bear in mind that in the South African context, most patients have multiple severe injuries on admission and many present with haemodynamic instability, often in a resource-limited setting. Endovascular intervention can be beneficial for a small number of carefully selected patients. The current experience is mostly limited to small case series but early results are promising.10, 11, 12
Associated brachial plexus injuries were encountered in 12% of cases and neurological deficit remains the leading cause of morbidity; all brachial plexus repairs were performed by a hand surgeon. In some cases, the neurological deficit on presentation results from compression of the nerves by a haematoma or an aneurysm and resolves after expedient surgical intervention.
Although this study is limited by its retrospective nature and focuses only on the operative outcome of patients with SCA injuries, we had excellent results with no deaths and a limb salvage rate of 100%. Planning the surgical approach using preoperative angiography, rapid exposure to gain vascular control along with reconstruction to re-establish perfusion and ligation in carefully selected cases have contributed to the successful outcome of patients in this study.
6. Conflict of Interest
No conflicts of interest to be declared.
Acknowledgements
No sources of financial and material support to be declared.
References
- . Penetrating thoracic vascular injuries. Surg Clin North Am. 1996;76:749–761
- . Subclavian vascular injuries. Br J Surg. 1987;74:1001–1003
- . Injuries to branches of the aortic arch. Aust NZ J Surg. 1998;58:217–219
- . Penetrating injuries of the aortic arch and its branches. Ann Thorac Surg. 1993;55:585–592
- . Six-year experience with management of subclavian artery injuries. Am Surg. 2000;66:927–931
- . Penetrating injuries to the subclavian and axillary vessels. J Am Coll Surg. 1999;188:290–295
- . Management of proximal axillary and subclavian artery injuries. Br J Surg. 2000;87:79–85
- . Traumatic injuries to the subclavian and axillary arteries: a 13-year review. Surg Today. 2005;35:561–565
- Penetrating injuries of the subclavian artery. Am J Surg. 2003;185:580–584
- Endovascular repair of traumatic injuries of the subclavian and axillary arteries. Injury, Int J Care Injured. 2005;36:778–782
- . Long-term results of stent graft treatment of subclavian artery injuries: management of choice for stable patients?. J Vasc Surg. 2007;47(4):739–743
- . Endovascular graft repair of penetrating subclavian artery injuries. J Endovasc Surg. 1996;3:382–388
☆ Presented at the 36th Annual Meeting of the Surgical Research Society of Southern Africa, 3–4 July 2008, University of Cape Town, Cape Town, Republic of South Africa.
☆☆ Published in abstract form from the above meeting: Sanju Sobnach, A Nicol, H Nathire, D Kahn, S Edu, P Navsaria. The management of penetrating subclavian artery injuries. S Afr J Surg 2009 Feb;47(1):30.
PII: S1078-5884(09)00550-4
doi:10.1016/j.ejvs.2009.10.013
© 2009 European Society for Vascular Surgery. Published by Elsevier Inc. All rights reserved.
Volume 39, Issue 2 , Pages 155-159, February 2010
