Innominate artery trauma: A thirty-year experience

Transcription

1 Innominate artery trauma: A thirty-year experience Robert H. Johnston, Jr., MD, Matthew J. Wall, Jr., MD, and Kenneth L. Mattox, MD, Houston, Texas Purpose: Injury to the innominate artery may represent a zone I cervical, thoracic outlet, or intrathoracic vascular injury and poses diagnostic, exposure, and management problems for the general, vascular, and thoracic surgeon. This complex injury often becomes a new learning experience with each infrequent encounter. Methods: Between 1960 and patients with innominate artery injuries were treated. Penetrating injuries were from gunshot wounds in 25, stab wounds in 7, and shotgun wounds in 2 patients. Blunt injuries accounted for seven of the patients. In 28 patients there were multiple injuries, including multiple vascular injuries. Median sternotomy was used in all planned operations in the later part of the study, and bilateral transsternal anterolateral thoracotomy was used in patients undergoing urgent or emergency center thoracotomy. Results: Blunt injury resulted in tears near the aortic arch with intimal disruption. Bypass grafting without hypothermia, shunts, or systemic heparinization is now used in all patients. Thirty-two patients survived to leave the hospital with no new complications related to the procedure. Postoperative neurologic complications were associated with preexisting neurologic deficits. Conclusions: Patients with innominate artery injury and stable vital signs can usually be transported without difficulty and treated without complex perioperative adjuncts. These patients can undergo revascularization with simple vascular surgical techniques and should expect an uncomplicated postoperative course unless there has been associated central nervous system injury or related injuries leading to systemic infection. Synthetic conduits have been used with success and have not required systemic heparinization or complex temporary shunting. (J VAsc StrRG 1993;17: ) Injury to the great vessels at the thoracic outlet can be one of the most challenging surgical problems that will be seen by the surgeon. These injuries can be from blunt trauma or zone I cervical, thoracic outlet, or upper thoracic penetration. The patients are frequently hypovolemic, hypothermic, and acidotic when they arrive at the emergency center. The first report of traumatic innominate artery injury was by Wheeler 1 in 1932 when ligation was the recommended treatment. Later, Langley 2 reported the case of an American airman who sustained a penetrating innominate artery injury from a shell fragment in From The Cora and Webb Mading Department of Surgery, Baylor College of Medicine, and The Ben Taub General Hospital, Houston. Presented at the Forty-sixth Annual Meeting of the Society for Vascular Surgery, Chicago, Ill., June 8-9, Reprint requests: Robert H. Johnston, Jr., MD, Clinical Professor, Department of Surgery, Baylor College of Medicine, One Baylor Plaza, Houston, TX ]6]42299 Despite 60 years of managing this injury the mortality rate has continued to remain high and this injury has continued to be associated with complications and combined injury. MATERIAL AND METHODS These data were obtained from the medical records and the surgical department's weekly activity report of the Ben Taub General Hospital. This report covers a period when one or more of the authors were participants in the surgery department at the Ben Taub General Hospital. There were 43 patients treated with innominate artery injury between 1964 and Thirty-eight were men and five were women. The average age was 38 with the age ranging from 16 to 89. The etiology was penetrating in 79.1% (34), blunt in 16.3% (7), and other causes in two patients. The penetrating injuries were secondary to gunshot wounds in 24, stab wounds in 7, and shotgun wounds in 2 (Table I) /93/$

2 Volume 17 Number 1 January 1993 Innominate artery trauma 135 Table I. Iunominate artery injury: Cause of injury No. of No. of Mortality patients deaths rate (%) Penetrating Gunshot wound Stab wound Shotgun wound Blunt trauma Other ,0 Total Associated injuries were frequently seen. Anatomic structures most commonly injured with the innominate artery were the superior vena cava, the innominate vein, and the carotid artery. These structures accounted for seven associated injuries each. Other structures injured in association with the innominate artery were the aorta, lung, and trachea. An associated condition of great significance was present in the patients who had a central nervous system deficit on presentation. There were four patients who had this problem when they were first treated, and all four died. Two patients had arteriovenous fistulas between the aorta or the innominate artery and the innominate vein. Fourteen patients had an isolated injury to the innominate artery (Table II). Shock was commonly present in these patients on admission to the emergency center. There were 17 patients who were in shock, and cardiopulmonary resuscitation was in progress on two of these on admission. Neither of these patients survived. Aortography was used in 13 of the 43 patients. Although this is the gold standard for diagnosis, shock precluded use of this test in many patients. Among the patients studied by aortography the most common wounds were by blunt trauma (six patients) and by gunshot wound (six patients). There was one study done on a patient who sustained a shotgun wound in which the only false-negative result in this series was obtained. One patient who was in shock after a gunshot wound was able to be resuscitated and underwent aortography, which demonstrated an injury to the innominate artery. The surgical approach to the management of this injury changed during the time interval under study. During the 1960s the approach was to repair this defect primarily or by inserting an interposition graft. This was accomplished through an assortment of incisions. During the 1970s there was a marked increase in the number of patients, and primary repair Table II. Innominate artery injury: Associated injury (n = 28) Associated injury No. Innominate vein 7 Superior vena cava 7 Carotid artery 7 Aorta 6 Lung 5 Trachea 4 Central nervous system 4 Subclavian artery and vein 3 Abdominal injury 3 Arteriovenous fistulas 2 Esophagus 2 Pulmonary artery and vein 1 continued to be the technique of choice. Early in the 1970s one of us (K.L.M.) began using the bypass principle for correction of this injury. The incision used was the median sternotomy, with only two patients having a "book" or "trap-door" incision and two additional patients having bilateral anterior thoracotomy. Since that time this method has become our exclusive approach to correction of this defect. Once the bypass from the ascending aorta to distal innominate artery is complete, a pledgeted suture line is placed in the aortic arch to close the origin of the innominate artery (Table III). Twenty-four patients were treated by primary arterial repair, four by interposition graft, and 13 by the insertion of a bypass graft from the ascending aorta to the distal innominate artery (Fig. 1). The bypass graft material was Dacron in nine and polytetrafluoroethylene (Gore-Tex*) in four. RESULTS The survival rate for the series was 74.4%. However, when analyzed by decade the evolution of approach is very clear. During the decade of the sixties there were four patients treated with two deaths; during the seventies 19 patients were treated for this injury and there were seven deaths. In the eighties there were 16 patients treated with two deaths, and no deaths have yet occurred in patient treated in the nineties. This calculates to a crude mortality rate of 50% during the sixties, 36.8% during the seventies, 11.8% during the eighties, and 0.0% so far during the nineties (Table IV). The clearest predictor of bad outcome was preoperative central nervous system deficit. Patients in which this was present had hemiparesis associated *Gore-Tex is a registered trademark of W.L. Gore & Associates, Flagstaff, Ariz.

3 136 Johnston, Wall, and Mattox Journal of VASCULAR SURGERY A 0 Fig. 1. Drawing depicts bypass technique used for patients with innominate artery injury. Table III. Innominate artery injury: Type of operative procedure, by decade Decade Procedure Total Bypass Primary repair Interposition Other with the injury on admission to the emergency center. Three of these patients were in shock on admission and one was normotensive but with paraplegia. All of the patients who had hemiplegia on presentation eventually died. All patients who required cardiopulmonary resuscitation before arrival at the emergency center died. Seventeen patients were in shock on admission and 59% survived (Table V). The patients who sustained innominate artery injury secondary to bltmt trauma all had proximal injury, that is, at the junction of the innominate artery and the aorta. The locations of the penetrating injuries were more distal, however, they were found throughout the length of the innominate artery. Associated injury was present in 30 of these patients (69.7%). The most common associated injury was to the innominate vein, superior vena cava, and carotid arteries. Preadmission central nervous system deficits correlated with mortality but did not correlate with carotid artery injury. There were seven carotid artery injuries present in this series. Of the patients with these injuries, one had hemiplegia on

4 Volume 17 Number 1 January 1993 Innominate artery trauma 137 Table IV. Innominate artery injury: Mortality rate and complications, by decade Decade Total No. of patients No. of complications No. of deaths Mortality rate 50.0% 36.8% 11.8% 0.0% 43 (100%) 14 (33%) 11 (25.6%) Table V. Innominate artery injury: Mortality rate, by presentation No. No. of Mortality patients deaths rate (%) Cardiopulmonary resusci tation Shock Central nervous system injury presentation and one had a seizure after operation but no residual neurologic deficit. Five patients with carotid artery injury had no neurologic deficits. When the patients with or without associated injuries are compared, 88% of the patients who were in shock on admission had associated injuries, thus a significant mortality rate of 41.2% was found. The group with the worst outcome by etiology was that with stab wounds in which three of seven died, while in the larger group of patients with gunshot wounds six of 25 died. In the group of patients with blunt trauma only one of seven patients died, and this was associated with a head injury (Table I), which caused the death. Median sternotomy was the most common incision used in this group of patients and was found to have the lowest associated mortality. This may be because of its more common use in the later part of the study and its use in combination with the bypass principle. Overall 37.2% of our patients had some complication after operation, and the overall mortality rate was 26.2% (Tables VI and VII). DISCUSSION The successful management ofvascnlar injuries of the thoracic outlet is one of the most challenging, yet rewarding feats in surgery. It has been reported that 20.1% of vascular injuries are in the chest, 3 and of that group 71% of the patients with major vascular injuries of the thoracic outlet or base of the neck die before arrival at the hospital. 4 With the development of the Emergency Medical Service system in Hous- Table VI. Innominate artery injury: Postoperative complications Complication No. Wound infection 3 Pulmonary failure : 2 Gastrointestinal bleeding 2 Mediastinitis 1 Anastomosis disruption 1 Sternal dehiscence 1 Pneumonia 1 Coagulopathy 1 Meningitis 1 Systemic air embolus 1 Sepsis 1 ton, the aggressive use of emergency center thoracotomy and arteriography, and the increase in illegal drug activities, the surgical service at Ben Taub General Hospital is seeing more patients who have sustained cardiovascular injury, s In 1991 two groups reviewed innominate artery injury and reported on a total composite of 22 patients from the literature who sustained innominate artery injury and their managementfi '6 We reviewed our experience with this injury and found that 43 patients had been treated during a 30-year period (Table VIII). Of this group the most common etiologic factor was penetrating injury, which accounted for most of the mortality. In different geographic areas there is a difference in the etiology of the penetrating wound, with some areas of the United States reporting a higher incidence of stab wound compared with the 57% of patients in this series who sustained gunshot wound. Patients with innominate vascular injury are commonly hypovolemic, hypothermic, and acidotic on arrival at the emergency center, 7 which necessitates resuscitation of the patient by surgical personnel in the emergency center before operative intervention. Emergency center thoracotomy is aggressively used in our emergency center, however, in patients with innominate artery injury only one patient was treated in this manner and did not survive. Once

5 Journal of VASCULAR 138 Johnston, Wall, and Mattox SURGERY Table VII. Innominate artery injury: Deaths Date Cause Complication Repair Presentation 4/64 Gunshot wound 9/69 Gunshot wound 5/70 Shotgun wound 12/70 Gunshot wound 10/71 Gunshot wound 12/72 Tracheostomy erosion 1/74 Gunshot wound 7/75 Blunt trattma 12/76 Shotgun wound 5/81 Shotgun wound 8/87 Gunshot wound Anastomotic breakdown Primary Unknown Pneumothorax, bronchopleural fistulas Interposition graft Shock Hypothermia, coagulopathy Primary Shock Central nervous system injury Primary Normotensive Central nervous system injury Ligation Shock Meningitis Graft interposition - Central nervous system injury Primary Unknown Central nervous system injury Bypass Shock Cardiopulmonary resuscitation Primary Shock Air embolus Bypass Shock Cardiopulmonary resuscitation Primary Shock resuscitated the patients who were in shock were transported rapidly to the operating room where definitive repair of the injury was undertaken. In the group who were stable on admission to the emergency center, aortography was done. There were 13 patients in this group who underwent aortography. In the 13 patients who underwent arteriography one false-negative result was obtained, and this was in a patient with a shotgun injury. As Richardson et al. 8 have pointed out, up to 23% of arteriograms may give false-negative results for penetrating trauma. Every effort is made to do arteriography in thoracic outlet injury. However, in a large number of patients this is precluded by the condition of the patient, which necessitates rapid mediastinal exploration. The overall mortality rate in this study was 25.6%. However, when the patient, treatment, and outcome are reviewed by decade an interesting observation is made. During a 30-year period, a fall in mortality rate from 50% to 36.8% to 11.8% is seen. During the study period there was an evolution of the incision used and the method of operative repair. Early on, all patients were treated by primary repair or graft interposition through an assortment of incisions. It was not until the seventies and eighties that the median sternotomy emerged as the incision of choice for repair of this injury, and now this is the standard approach in our trauma center. In the eighties the bypass principle became the method of choice for patients seen with this injury at the Ben Taub General Hospital. A total of 70% of our patients had associated injuries. When an associated injury existed it significantly increased the mortality rate, from 8.3% to 33.3%. As the operative management of these patients has evolved, a median sternotomy incision with a Table VIII. Innominate artery injury: McLean literature review, No. of No. Years Author Year ceases survived reviewed Lim Pate and Casini Graham Bladergroen Weaver Johnston li 30 right neck extension for better visualization of the innominate bifurcation is now preferred. Bleeding is controlled by pressure or clamp as is needed. Before this injury is dealt with a partial occluding clamp is placed on the ascending aorta and a graft is sutured in place. The innominate is then divided just before its bifilrcation and end-to-end anastomosis is accomplished. A partial occluding clamp is placed on the aorta at its junction with the innominate artery. The aorta is then closed with pledgeted sutures and fmally a second running suture line. No shunt, heparin, or cardiopulrnonary bypass is used for this repair. In the earlier bypass repairs Dacron fabric grafts were used and more recently 8 mm polytetrafluoroethylene (Gore-Tex) grafts have been the graft of choice. No problems with graft infection or thrombosis have been encountered. There were several arteriovenous fistulas present in this series. The most recent patient came to us for treatment several months after a high-velocity gunshot wound to the aortic arch-innominate junction with fismlization to the innominate vein. Symbas 9 suggests that the presence of arteriovenous fistulas may provide some protection for the patient with these injuries because the bleeding returns to the

6 Volume 17 Number 1 January 1993 Innominate artery trauma 139 vascular system. Another interesting phenomenon is that patients with a left carotid artery origin from the innominate artery seem to have more of a propensity for innominate disruption from blunt trauma.i This anomaly is present in 11% of the population, but is seen in a disproportionate number of patients who have innominate disruption when first seen (29%).I This may suggest an increased susceptibility of this anatomic variation to traumatic avulsion. The group of patients with blunt trauma tends to be more stable and transfer can be accomplished without significant fear of rupture. 11 With the exclusion of patients who are admitted in shock or have concomitant cerebral or multisystem trauma, the survival rate exceeds 90%, and in this series all survived who were not in shock or had no associated injury. The reported mortality rate of innominate artery injury approaches 30% and the reported likelihood of complications may exceed 40%.12 This current series reports a mortality rate of 25.6% and a complication rate of 37.2%, which compare favorably with the literature. Innominate artery injury is a formidable entity that requires organization and planning to achieve a desirable result for the injured patient. REFERENCES 1. Wheeler WI de C. Traumatic subclavian aneurysm. Irish J Med Sci 1942;6th s: Langley GF. Gunshot wound of the innominate artery. BM[[ 1943;Dec 4: Mattox K, Feliciano D, Butch J, Beall A, Jordan G, DeBakey M. Five thousand seven hundred sixty cardiovascular injuries in 4459 patients. Ann Surg 1989;209: Kemmerer W, Eckert W, Gathright J, et al. Patterns of thoracic injuries in fatal traffic accidents. J Trauma 1961;1: McLean TR, McManus PR. Penetrating trauma involving the innominate artery. Ann Thorac Surg 1991;51: Gubler K, Wisner D, Blaisdell FW. Multiple vessel injury to branches of the aortic arch: case report. J Trauma 1991;31: Mattox K. Symposium: new approach in vascular trauma. [1 VAsc SURG 1988;7: Richardson J, Smith M, Grover F, Arom K, Trinkle JK. Management of subclavian and innominate artery injuries. Am 1 Surg 1977;134: Symbas PN. Cardiothoracic trauma. Philadelphia: W-B Sanders; 1989: Graham J, Feliciano D, Mattox K, BeaU A. Innominate vascular injury. J Trauma 1982;22: Mattox K. Thoracic great vessel injury. Surg Clin North Am 1988;68: Submitted June 12, 1992; accepted Sept. 1, DISCUSSION Dr. Peter C. Pairolero (Rochester, Minn.). I would like to thank Dr. Johnston for an excellent presentation. Unquestionably, injuries to the great vessels at the level of the thoracic inlet have, throughout the years, been most challenging. In the past, the mortality associated with these injuries has been one of the highest for any type of arterial trauma. Fortunately, however, improvement has occurred, and the authors are to be congratulated on their excellent survival rate of nearly 90%, achieved during the past decade, for isolated innominate artery injury. Nonetheless, controversy still continues to exist regarding the best method of establishing the diagnosis of great vessel injury and of surgical management. The diagnosis of major arterial injury is obvious in the presence of bleeding, expanding hematoma, absent pulses, and circulatory collapse. But what if these are not present? In only 17 of Dr. Johnston's 42 patients was shock present. Clearly a high index of suspicion is required and, undoubtedly, this is more likely to occur if the injury is penetrating. However, approximately 15% of their patients had blunt trauma. Should great vessel injury then be suspected in all patients who have significant trauma to the anterior chest and base of the neck? As Dr. Johnston has mentioned, most surgeons would agree that if the patient is stable, arteriography is the gold standard to define the extent and location of injury. Yet arteriography is not infallible, as demonstrated in the Houston series, which had a false-negative rate of approximately 10%. If we cannot safely observe the patient with a negative arteriography result, is there any role for computed tomography after arteriography? Or, if the index of suspicion remains high, should the patient simply undergo exploratory operation? I would like Dr. Johnston to expand on the algorithm used in establishing the diagnosis of great vessel injury. The goal of surgical intervention is simple: early operation, arterial control, and vascular repair. The decision regarding exposure, however, is more complex. Is the injury in the neck, thoracic outlet, or thorax? It is probably best to consider all injuries in this region as involving the

7 140 Johnston, Wall, and Mattox Journal of VASCULAR. SURGERY thoracic outlet, and as the authors have amply demonstrated, exposure via the median sternotomy route is quite adequate. The incision can be extended into the neck or into either thorax, if indicated. Moreover, exposure of all brachiocephalic vessels, including the left subclavian, is possible, and if necessary cardiopulmonary bypass can be initiated in the patient. Finally, I agree with the authors' choice of arterial bypass when reconstruction is needed. In our experience with chronic innominate artery disease, bypass was the most successful long-term reconstructive procedure. Dr. Robert H. Johnston, Jr. Concerning our diag- nostic approach to these patients, if we are afforded the opportunity of treating a patient who is more or less stable, aortography is the gold standard for diagnosis. To explore these patients without it is like flying blind without instruments, because you do not know the extent of the injuries that could be there. Aortography is still the gold standard. However, most of the patients who are in extremis when brought to the emergency center are either operated on in the emergency center or taken directly to the operating room where exploration is undertaken.