WTA 2014 PLENARY PAPER
Blunt abdominal aortic injury: A Western Trauma Association multicenter study Sherene Shalhub, MD, MPH, Benjamin W. Starnes, MD, Megan L. Brenner, MD, MS, Walter L. Biffl, MD, Ali Azizzadeh, MD, Kenji Inaba, MD, Dimitra Skiada, MD, Ben Zarzaur, MD, MPH, Cayce Nawaf, MD, Evert A. Eriksson, MD, Samir M. Fakhry, MD, Jasmeet S. Paul, MD, Krista L. Kaups, MD, MSc, David J. Ciesla, MD, S. Rob Todd, MD, Mark J. Seamon, MD, Lisa M. Capano-Wehrle, MPH, Gregory J. Jurkovich, MD, and Rosemary A. Kozar, MD, Seattle, Washington
Blunt abdominal aortic injury (BAAI) is a rare injury. The objective of the current study was to examine the presentation and management of BAAI at a multi-institutional level. METHODS: The Western Trauma Association Multi-Center Trials conducted a study of BAAI from 1996 to 2011. Data collected included demographics, injury mechanism, associated injuries, interventions, and complications. RESULTS: Of 392,315 blunt trauma patients, 113 (0.03%) presented with BAAI at 12 major trauma centers (67% male; median age, 38 years; range, 6Y88; median Injury Severity Score [ISS], 34; range, 16Y75). The leading cause of injury was motor vehicle collisions (60%). Hypotension was documented in 47% of the cases. The most commonly associated injuries were spine fractures (44%) and pneumothorax/hemothorax (42%). Solid organ, small bowel, and large bowel injuries occurred in 38%, 35%, and 28% respectively. BAAI presented as free aortic rupture (32%), pseudoaneurysm (16%), and injuries without aortic external contour abnormality on computed tomography such as large intimal flaps (34%) or intimal tears (18%). Open and endovascular repairs were undertaken as first-choice therapy in 43% and 15% of cases, respectively. Choice of management varied by type of BAAI: 89% of intimal tears were managed nonoperatively, and 96% of aortic ruptures were treated with open repair. Overall mortality was 39%, the majority (68%) occurring in the first 24 hours because of hemorrhage or cardiac arrest. The highest mortality was associated with Zone II aortic ruptures (92%). Follow-up was documented in 38% of live discharges. CONCLUSION: This is the largest BAAI series reported to date. BAAI presents as a spectrum of injury ranging from minimal aortic injury to aortic rupture. Nonoperative management is successful in uncomplicated cases without external aortic contour abnormality on computed tomography. Highest mortality occurred in free aortic ruptures, suggesting that alternative measures of early noncompressible torso hemorrhage control are warranted. (J Trauma Acute Care Surg. 2014;77: 879Y885. Copyright * 2014 by Lippincott Williams & Wilkins) LEVEL OF EVIDENCE: Epidemiologic study, level III; therapeutic study, level IV. KEY WORDS: Blunt trauma; abdominal aorta; minimal aortic injury; aortic rupture; blunt aortic injury. BACKGROUND:
B
lunt abdominal aortic injury (BAAI) is rare, with a reported incidence of less than 1% of all blunt trauma injuries. It is one of the most challenging injuries because of the associated injuries and hemodynamic instability of those who survive to reach the hospital.1Y3 Most of the BAAI literature has been limited to case reports, cases series, and literature review,4Y8 with the largest review to date obtained
from the National Trauma Data Bank.9 A standardized radiologic abdominal aortic injury classification and management details at a multi-institutional level have not been fully delineated. The aim of this study was to examine the presentation, radiologic classification, and management of this rare injury at multi-institutional level in a relatively large cohort of patients.
Submitted: February 16, 2014, Revised: March 31, 2014, Accepted: April 9, 2014, Published online: Sepember 22, 2014. From the Division of Vascular Surgery (S.S., B.W.S.), Department of Surgery, University of Washington, Seattle, Washington; R. Adams Cowley Shock Trauma Center (M.L.B.), University of Maryland, Baltimore, Maryland; Department of Surgery (W.L.B., G.J.J.), Denver Health Medical Center and the University of Colorado School of Medicine, Denver, Colorado; Department of Cardiothoracic and Vascular Surgery (A.A.), University of Texas Medical School at Houston; and Department of Surgery (R.A.K.), University of Texas Health Science Center at Houston, Houston, Texas; Division of Trauma and Surgical Critical Care (K.I., D.S.), Department of Surgery, Keck School of Medicine, Los Angeles County + University of Southern California Medical Center, Los Angeles; and Community Regional Medical Center (K.L.K.), Department of Surgery, University of California, San Francisco-Fresno Campus, Fresno, California; Division of Trauma and Critical Care (B.Z., C.N.), Department of Surgery, University of Tennessee Health Science Center, Memphis, Tennessee; Division of General Surgery (E.A.E., S.M.F.), Department of Surgery, Medical University of South Carolina, Charleston, South Carolina; Division of Trauma and Critical Care (J.S.P.), Department of Surgery, Medical College of Wisconsin, Milwaukee, Wisconsin; Department of Surgery (D.J.C.), University of South Florida Colleges of Medicine, Tampa, Florida; Department of Surgery (S.R.T.), New York University Langone Medical Center, New York, New York; Division of Trauma and Surgical Critical Care (M.J.S., L.M.C.-W.), Department of Surgery, Cooper University Hospital, Camden, New Jersey. This study was presented at the 44th Annual Meeting of the Western Trauma Association, March 6, 2014, in Steamboat Springs, Colorado. Address for reprints: Sherene Shalhub, MD, MPH, Division of Vascular Surgery, Department of General Surgery, The University of Washington, 1959 NE Pacific St, Box 356410, Seattle, WA 98195; email: [email protected]. DOI: 10.1097/TA.0000000000000353
J Trauma Acute Care Surg Volume 77, Number 6
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
879
J Trauma Acute Care Surg Volume 77, Number 6
Shalhub et al.
PATIENTS AND METHODS The Western Trauma Association Multi-Center Trials Group performed a retrospective cohort study of patients presenting with BAAI from 1996 to 2011. Institutional review board approval was obtained at each of the contributing sites. Included in this series are data previously described as a single-center experience.7,8 BAAI was defined as an injury to the abdominal aorta from the diaphragmatic hiatus to the aortic bifurcation and identified by the International Classification of DiseasesV9th Rev. (ICD-9) code 902.0 (injury aorta, abdominal). Exclusion criteria included cases with blunt thoracic aortic injury with dissection extending to the abdominal aorta and cases diagnosed on autopsy alone but not recognized at the time of patient care. Aortic injuries classification was based on the presence of external aortic contour abnormality on computed tomography (CT) or the presence of free rupture found intraoperatively.7 In the intimal tears and large intimal flaps (LIFs), the normal external aortic contour is preserved on CT. Intimal tears were defined as intimal defect and/or thrombus of 10 mm in length or width. LIF were defined as intimal defect and/or thrombus of 10 mm in length or width. Intramural hematoma was included in this category. A pseudoaneurysm was thought of as a contained rupture and clearly associated with an external aortic contour abnormality on CT. Aortic ruptures were associated with external aortic contour abnormality with free contrast extravasation on CT or hemoperitoneum found upon laparotomy. Abdominal aorta zones of injury were classified based on possible endovascular surgical approaches as follows (Fig. 1): Zone I injuries occur from the diaphragmatic hiatus to above the superior mesenteric artery (SMA). Zone II injuries include the region from the SMA to and including the renal arteries. Zone III injuries are below the renal arteries and to the aortic bifurcation. Branch vessel avulsion injury was noted by type (celiac, SMA, inferior mesenteric artery, renal and lumbar arteries). Specific data collected included age, sex, mechanism of injury, initial clinical presentation, Injury Severity Score (ISS), presence of a seat belt sign, other associated injuries, diagnostic methods, management, modes of repair, procedural complications, mortality, subsequent radiographic imaging, and follow-up. Representative images from CT, CT angiogram (CTA), and angiograms were independently reviewed by a single author (S.S.) to confirm the classification of the injury when imaging was available to be retrieved from the archives (n = 41). The remainder of the injuries was classified based on operative and imaging reports. Data were analyzed using Microsoft Excel 2007 software (Microsoft, Redmond, WA) and SPSS version 19 for Windows (SPSS, Inc., Chicago, IL). Continuous data are presented as medians and ranges. Means of continuous data were compared using the Student’s t test. Categorical data were compared using the Fisher’s exact test or W2 analysis, where appropriate.
RESULTS During the study period, there were 392,315 patients admitted for blunt trauma to 12 major trauma centers. Of 880
Figure 1. Classification of abdominal aorta zones of injury based on possible endovascular surgical approaches. Zone I injuries occur from the diaphragmatic above the SMA. Zone II injuries include the SMA to the renal arteries. Zone III injuries are between the renal arteries and the aortic bifurcation. Zone I and III injuries are amenable to endovascular repair.
those, 113 patients were identified with BAAI giving rise to an incidence of 0.03% of all blunt trauma injuries. The majority of the cases were male (n = 78, 67.3%) with a median age of 38 years (range, 6Y88 years). The median ISS was 34 (range, 16Y75). The leading cause of injury was motor vehicle collisions (n = 68, 60.2%) followed by pedestrian versus vehicle crashes (13, 11.5%). Nearly half of the cases presented with hypotension (systolic blood pressure [SBP] G 90 mm Hg) on arrival (n = 53, 46.9%). Seat belt sign was documented in 24 cases (35.3% of motor vehicle crash cases). Cardiac arrest in the prehospital setting or in the emergency department (ED) was documented in 23 cases (20.4%). ED thoracotomy was performed in 13 cases (11.5%). CT or CTA was used in 74.3% of the cases. The likelihood of obtaining a CT scan was highly related to the patient’s hemodynamic stability as 31% with reported SBP less than 90 mm Hg had a CT scan compared with 69% of those without hypotension (p G 0.000). Diagnostic peritoneal lavage was used in 2 (1.8%) and a diagnostic angiogram in 10 cases (8.8%). The most commonly associated injuries were spine fractures (n = 50, 44.2%), solid organ injury (n = 43, 38.1%), small bowel injury (n = 39, 34.5%), and large bowel injury (n = 32, 28.3%). Table 1 summarizes the demographics, * 2014 Lippincott Williams & Wilkins
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
J Trauma Acute Care Surg Volume 77, Number 6
Shalhub et al.
TABLE 1. Demographics, Mechanism of Injury, Clinical Presentation, Associated Injuries, Type and Zone of Injury in Patients With Blunt Abdominal Aortic Injuries
TABLE 1. (Continued)
n = 113 Male, n (%) Age, median (range), y ISS, median (range) Mechanism of injury, n (%) Motor vehicle crash Car vs. pedestrian Fall Motor cycle crash Crush injury Car vs. bicycle All-terrain vehicle crash Other Clinical presentation, n (%) SBP G 90 mm Hg on arrival Cardiac arrest before surgery ED thoracotomy Mortality in ED or on the way to surgery Seat belt sign documented Comorbidities, n (%) Hypertension Coronary artery disease Peripheral vascular disease Creatinine Q 1.5 Diagnostic studies, n (%) Trauma CT CTA Angiogram Diagnostic peritoneal lavage Associated injuries, n (%) Spine fracture Cervical spine fracture Thoracic spine fracture Lumbar spine fracture Pneumothorax/hemothorax, n (%) Hemoperitoneum, n (%) Solid organ injury, n (%) Liver Spleen Kidney Pancreas Rib fractures, n (%) Small bowel injury, n (%) Mesenteric hematoma, n (%) Traumatic brain injury, n (%) Colon injury, n (%) Pelvic fracture, n (%) Abdominal wall ecchymosis, n (%) Abdominal wall degloving, n (%) Inferior vena cava injury, n (%) Superior mesenteric/portal vein injury, n (%) BAAI type, n (%) Intimal tear LIF
76 (67.3) 38 (6Y88) 34 (16Y75) 68 (60.2) 13 (11.5) 10 (8.8) 7 (6.2) 6 (5.3) 2 (1.8) 2 (1.8) 5 (4.4) 53 (46.9) 23 (20.4) 13 (11.5) 5 (4.4) 24 (21.2) 9 (7.9) 5 (4.4) 4 (3.5) 5 (4.4) 53 (46.9) 31 (27.4) 10 (8.8) 2 (1.8) 50 (44.2) 7 (6.2) 14 (12.4) 33 (29.2) 47 (41.6) 45 (39.8) 43 (38.1) 24 (21.2) 20 (17.7) 8 (7.1) 6 (5.3) 39 (34.5) 39 (34.5) 38 (33.6) 35 (31.0) 32 (28.3) 28 (24.8) 27 (23.9) 10 (8.8) 10 (8.8) 1 (0.8) 20 (17.7) 38 (33.6)
n = 113 Pseudoaneurysm Rupture Injury to preexisting AAA Complicated LIFs, n (%) Aortic dissection Extension of dissection into the iliac artery Thrombosis Aortic rupture description, n (%) Aortic wall rupture Branch vessel avulsion Celiac artery avulsion Superior mesenteric artery avulsion Inferior mesenteric artery avulsion Right renal artery Left renal artery Lumbar arteries avulsion Iliac artery avulsion Aortoiliac junction rupture Zone of injury, n (%) Zone I Zone II Zone III
18 (15.9) 36 (31.9) 1 (0.9) 7 (6.2) 9 (8.0) 4 (3.5) 15 (13.3) 24 (21.2) 3 (2.7) 5 (4.4) 3 (2.7) 4 (3.5) 3 (2.7) 8 (7.1) 1 (0.9) 1 (0.9) 22 (19.5) 22 (19.5) 75 (66.4)
mechanism of injury, clinical presentation, and associated injuries in this cohort.
BAAI Management Choice of management varied by type of BAAI. BAAI presented as intimal tear (n = 20, 17.7%), LIF (n = 38, 36%), pseudoaneurysm (n = 18, 15.9%), and free rupture (n = 36, 31.9%). In one case, the patient presented with a thrombosis of a small infrarenal abdominal aortic aneurysm (AAA) without rupture. Traumatic aortic dissection with long intimal flaps was seen in seven cases (6.2%), and these were included in the LIF group. Table 1 details the location and type of BAAI seen in this cohort. The majority of intimal tears (89.5%) were managed nonoperatively, while all aortic ruptures were managed with open repair. In the cases managed operatively, the operative approach varied according to the zone of injury, with the majority of Zone II injuries managed by open repair (42%), while none of the Zone II injuries were suitable for endovascular repair. Table 2 details the management of BAAI by type and zone of injury. The next three sections describe BAAI management in detail.
Nonoperative Management A total of 40 cases (35.4%) were managed nonoperatively. The majority of the BAAI injuries managed nonoperatively were those presenting without aortic contour abnormalities (89.5% of intimal tears and 44.7% of LIFs). One third of the pseudoaneurysms were also managed nonoperatively.
* 2014 Lippincott Williams & Wilkins
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
881
J Trauma Acute Care Surg Volume 77, Number 6
Shalhub et al.
TABLE 2. BAAI Management by Type and Zone of Injury Nonoperative Management Management by injury type Intimal tear (n = 19) Large intimal flap (n = 38) Pseudoaneurysm (n = 18) Aortic rupture (n = 30) Management by injury zone Zone I (n = 18) Zone II (n = 14) Zone III (n = 69) Multiple zones (n = 5)
Open Repair
Endovascular Repair
17 (89.5%) 17 (44.7%) 6 (33.3%) 0
1 13 4 30
(5.3%) (34.2%) (22.2%) (100%)
1 (5.3%) 8 (21.1%) 8 (44.4%) 0
6 (33.3%) 3 (21.4%) 28 (40.6%) 3 (60%)
9 (50%) 11 (78.6%) 29 (42.0%) 0
3 (16.7%) 0 12 (17.4%) 2 (40%)
Table 3 summarizes the management and outcomes of the pseudoaneurysms. Failure of nonoperative management occurred in three cases: BAAI with aortic dissection along the length of the abdominal aorta presenting initially with temporary paraplegia. The patient became hemodynamically unstable 1 day after admission, and a large retroperitoneal hematoma was found intraoperatively. The patient died intraoperatively (Fig. 2A). Another case of failed nonoperative management was that of a BAAI associated with acute thrombosis of the right common iliac artery at the aortic bifurcation. The patient refused treatment and died of sepsis related to an ischemic lower extremity. A third case was that of a Zone II thrombus managed with anticoagulation initially; however, there was thrombus extension on follow-up imaging, which led to a hybrid repair with aortic debranching and a stent graft placement.
Open Surgical Management Forty-nine cases (43.4%) including 30 aortic ruptures required open exploration and repair. Of those, 32 (65%) presented with SBP less than 90 mm Hg, 14 (28.6%) experienced a cardiac arrest before arrival to the operating room. ED thoracotomy with aortic cross clamping was performed in 7 (14.3%) of these cases. Open repair was performed by interposition tube or bifurcated graft (n = 16, 32.7%), primary repair (n = 13, 26.5%), and branch ligation (n = 6, 12.2%). In eight cases (16.3%), an exploratory laparotomy was started, but no repairs were undertaken because of cardiac arrest from hemorrhagic shock. In three cases, no repair was documented, and the abdomen was packed, followed by a transfer to the angiography suite for arterial embolization (n = 3, 6.1%). Intraoperative findings were described as retroperitoneal hematoma (n = 29, 59.2%), contused aorta (n = 10, 20.4%), and aorta thrombosis (n = 8, 16.3%) (Fig. 2B) as well as one normal appearing aorta, which was associated with an intimal tear later seen on CTA. One case of LIF was treated with primary repair but required take back the next day for endovascular repair because of thrombosis of an iliac artery.
Endovascular Management Endovascular management was used as a primary mode of therapy in 17 cases (15%). The majority of the cases were male (70.5%) with a median age of 35 years (range, 15Y62 years) and median ISS of 29 (range, 24Y50). In these cases, a variety of stents and stent grafts were used: wall stent (n = 2), Gore excluder (n = 4), Medtronic AneuRX cuff (n = 4), Atrium stent (n = 1), Gore Cuff (n = 1), Viabahn (n = 1), Gore TAG (n = 1), and other (n = 3). Use of intravascular ultrasound was documented in two cases. The BAAI injuries
TABLE 3. Management and Outcomes of BAAI Pseudoaneurysms Age/Sex
ISS
Injury Zone
Size, mm
Management
LOS, d
Died
Follow-up
42, Male 81 Female 41, Male 42, Female
29 50 50 34
III III I III
7 16 5 12 30 20 30
8 63 53 33
N Y, TBI N N
Y N/A Y N
21, Male 47, Male 77, Male 26, Male 36, Female 21, Male 24, Male 37, Male 6, Female 53, Female 20, Male 22, Female 43, Female 38, Female
43 NR 29 57 34 21 29 25 25 34 NR 26 25 NR
I III III III III II III I III III III III II I
37 21 27 10
Endovascular stent graft Nonoperative Endovascular stent graft Nonoperative (surgery for mesenteric bleeding, no comment on retroperitoneal findings) Open repair, interposition graft Endovascular stent graft Nonoperative (died before repair) Nonoperative Open repair, bifurcation graft Open repair, primary, hospital Day 2 Endovascular stent graft Endovascular stent graft Nonoperative Endovascular stent graft Endovascular stent graft Endovascular stent Nonoperative Open repair, interposition graft
0 31 13 31 95 0 5 6 29 14 33 9 114 10
Y, coagulopathy N Y, pneumonia N N Y N N N N N N Y, sepsis N
N/A Y N/A Y Y N/A Y Y Y N Y, endoleak Y 2nd cuff Y N/A Y
25 22 28 10 18 12 15 32 10 10 NR NR NR NR NR
N, no; N/A, not applicable; NR, not reported; TBI, traumatic brain injury; Y, yes.
882
* 2014 Lippincott Williams & Wilkins
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
J Trauma Acute Care Surg Volume 77, Number 6
Shalhub et al.
was transferred to the angiography suite for embolization of lumbar arteries (n = 2) or celiac artery embolization with stent graft placement in the Zone I (n = 1). In one case, an LIF was treated with open primary repair but required take back the next day for endovascular repair because of thrombosis with extension into an iliac artery.
Length of Stay and Follow-up The overall length of stay for the cohort was 9.5 days (range, 0Y113 days). Sixty-nine patients were discharged (63.8% male; median age, 31 years; range, 6Y62 years). Of interest, two of these cases required ED thoracotomy (15% survival post-ED thoracotomy). The median length of stay of survivors was 14 days (range, 2Y95 days). The predominant injury type in those discharged was intimal tear (n = 18, 26.1%) or LIF (n = 32, 46.4%), and the predominant zone of injury was Zone III (n = 52, 75.4%). Follow-up of at least one visit was documented in 39% of the cases (n = 27): 10 had open repairs, and 9 had endovascular repairs. Long-term follow-up data were not available, and no data regarding associated graft infections were available. One case of endovascular pseudoaneurysm repair demonstrated an endoleak on follow-up imaging and required placement of a second stent graft.
Mortality
Figure 2. A and B, Examples of complicated LIFs. A, Coronal and axial CTA imaging showing BAAI dissection extending along the length of the abdominal aorta in a patient presenting initially with temporary paraplegia and who developed a large periarotic hematoma the next day and died intraoperatively. B, Intraoperative imaging of BAAI presenting with normal contour abnormality on trauma CT with associated bilateral lower extremity limb ischemia. The torn intima can be seen superiorly (block arrow), and the thrombus is seen in the distal aorta (long arrow). C and D, Examples of Zone I and Zone III BAAI endovascular repair. C, Sagittal view of CTA imaging showing Zone I pseudoaneurysm and a three-dimensional reconstruction of follow-up CTA after stent graft placement with extension into the distal thoracic. D, Coronal view of CTA imaging showing Zone III pseudoaneurysm and three-dimensional reconstruction of follow-up CTA imaging after bifurcated stent graft placement.
treated were one intimal tear (5.9%), LIFs (n = 8, 47.1%), and pseudoaneurysms (n = 8, 47.1%). All these injuries were limited to the Zone I (n = 5, 29.4%) or Zone III (n = 12, 70.5%). Figure 2C and D illustrate examples of Zone I and Zone III BAAI endovascular repair. All patients in this group were discharged alive after a median length of stay of 12 days (range, 5Y53 days). Five additional cases were then managed by endovascular means as an adjunctive therapy. One case was a Zone II injury with thrombus that progressed on anticoagulation and was treated with hybrid repair. In three cases, an exploratory laparotomy was performed first, then the abdomen was packed, and the patient
Overall mortality was 38.9% (n = 44), with the majority (n = 30, 68.2%) occurring in the first 24 hours because of cardiac arrest from hemorrhagic shock with the exception of one death caused by severe traumatic brain injury. Those who died were older compared with survivors, with a median age of 48.5 years (range, 6Y88 years; p G 0.000), and had higher ISS, with a median ISS of 43 (range, 16Y74) versus 34 (range, 16Y59, p G 0.000). Mortality varied by the zone of injury and the type of injury, with the highest mortality occurring in Zone II aortic ruptures (n = 14) with a mortality of 92.9%. Table 4 details BAAI mortality by initial presentation as well as type and zone of injury. Overall mortality in the open repair group was 57.1% (n = 28) including 11 intraoperative deaths (22.4%). Other causes of death in this group included two complications related to aortic thrombosis and ischemia-reperfusion of the bilateral lower extremities, which required amputations, and two caused by traumatic brain injuries. There was 0% mortality in the group treated with endovascular repair as the primary modality of treatment and 100% mortality in the four cases managed by endovascular means after an open exploration. Late causes of mortality (hospital days 3Y114 days) included multiorgan system failure and sepsis (n = 11). One death was caused by SMA bypass thrombosis in a Zone II thrombosis hybrid repair and subsequent bowel ischemia.
DISCUSSION This is the largest BAAI series reported to date. BAAI is rare, with an incidence less than 0.03% with even the busiest trauma centers managing zero to three cases a year. With the use of high-resolution CTA in the trauma workup, an increasing number of intimal tears are expected to be diagnosed.10Y12 Disruption of the intima is the underlying pathologic lesion
* 2014 Lippincott Williams & Wilkins
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
883
J Trauma Acute Care Surg Volume 77, Number 6
Shalhub et al.
TABLE 4. BAAI Mortality by Initial Presentation, Type and Zone of Injury Lived Age, mean (range), y ISS, mean (range)
Died
33.6 (6Y62) 48.1 (6Y88) 33.3 (16Y59) 42.7 (16Y74) Mortality SBP G 90 mm Hg (n = 53) 38 (71.7%) Cardiac arrest before surgery (n = 23) 21 (91.3%) Cardiac arrest (n = 31) 29 (93.5) Mortality with ED thoracotomy (n = 13) 11 (84.6%) BAAI type Intimal tears 2 (10%) LIFs 6 (15.8%) Pseudoaneurysm 5 (27.8%) Aortic rupture 30 (83.3%) Occluded small AAA 1 (100%) BAAI zone Zone I 8 (42.1%) Zone II 15 (83.3%) Zone III 19 (26.8%) Multiple zones 2 (40%)
p 0.000 0.001 p
0.000
0.000
of BAAI, which presents as a spectrum ranging from minimal intimal disruption to aortic rupture. For this study, the classification of the BAAI was based on the presence or absence of aortic contour abnormality as seen by axial images on CT. Intimal tears that are the equivalent of minimal aortic injury13,14 and LIF have preserved external aortic contour by CT, while external aortic contour abnormality is seen in the case of pseudoaneurysms, aortic wall disruption, and branch avulsions. Management varied somewhat between centers, depending on available resources and regional practice variation. Based on the available data in this study and previous studies,4Y9 several recommendations can be made. & All patients with suspected BAAI should undergo rapid CTA if hemodynamically fit. & Intimal tears can be managed nonoperatively with antiplatelet therapy and A-blockers followed by interval CTA follow-up within 30 days. & Uncomplicated LIF can be managed nonoperatively with antiplatelet therapy and A-blockers followed by interval CTA within 48 hours to assess for progression. If there is evidence of progression (intraluminal thrombus, aneurysmal degeneration, and pseudoaneurysm formation), then LIF is complicated and can be managed, when possible, with endovascular repair. & When known or suspected BAAI patients are taken to the operating room, they should be placed on a fluoroscopycompatible table, and preparation should be wide and include both groin regions for femoral access should endovascular repair be contemplated. Aortic Zones I and III are amenable to endovascular repair with particular benefit in the setting of concurrent gross contamination. & When BAAI is identified in the setting of gross contamination or hollow viscus injury, strong consideration should be given to endovascular repair. 884
& The use of antiplatelet therapy for 6 weeks after injury should be considered if there are no contraindications. & Postdischarge follow-up is strongly recommended. In the cases managed nonoperatively, follow-up with CTA at 1 month, 6 months, 1 year, and every other year thereafter is recommended or until the injury has resolved. & The timing of repair is based on the hemodynamic stability and concurrent associated injuries. The repair can be open or via endovascular means such as stent graft placement and branch vessel embolization. Mortality remains high among those who survive to the hospital with a traumatic abdominal aortic rupture. A significant number of patients (52.8%) presented in extremis and cardiac arrest before arriving in the operating room with an overall mortality of 83%. This supports the need for improved modalities of early hemorrhage control. Given the poor outcomes of ED thoracotomy,15 evaluation of newer modalities for early control of noncompressible torso hemorrhage such as resuscitative endovascular balloon occlusion of the aorta is warranted.16Y19 The use of a hybrid operating room in the treatment of these cases would allow the surgeon the full breadth of available treatment modalities with the potential for improved survival; this however, remains to be assessed. This study was subject to the normal limitations seen with a retrospective chart review including inability to retrieve old images, data regarding use of antiplatelet therapies, or long-term follow-up data. In light of the rare occurrence of this injury, ongoing prospective data collection across multiple centers would provide additional data for analysis.
CONCLUSION This is the largest BAAI series reported to date. BAAI is rare and presents as a spectrum of injury, ranging from minimal aortic injury to aortic rupture. Nonoperative management is successful in uncomplicated cases with preserved external aortic contour on CT. The highest mortality occurred in those with free aortic rupture, suggesting that alternative measures of early hemorrhage control are needed. Ongoing prospective data collection is warranted.
AUTHORSHIP S.S., G.J.J., and R.A.K. contributed to the study conception and design. S.S., R.A.K., B.W.S., and W.L.B. performed analysis and interpretation. S.S., M.L.B., W.L.B., K.I., D.S., B.Z., C.N., E.A.E., S.M.F., J.S.P., K.L.K., D.J.C., S.R.T., M.J.S., L.M.C.-W., and R.A.K. conducted the data collection. S.S., R.A.K., and B.W.S. wrote the article. S.S., B.W.S., A.E.A., R.A.K., W.L.B., S.M.F., and K.I. provided critical revision.
ACKNOWLEDGMENT We thank the patients involved in this study. We also thank Chris Akers for his support with the medical illustrations.
DISCLOSURE The authors declare no conflicts of interest.
* 2014 Lippincott Williams & Wilkins
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
J Trauma Acute Care Surg Volume 77, Number 6
Shalhub et al.
REFERENCES 1. Deree J, Shenvi E, Fortlage D, et al. Patient factors and operating room resuscitation predict mortality in traumatic abdominal aortic injury: a 20year analysis. J Vasc Surg. 2007;45:493Y497. 2. Asensio JA, Chahwan S, Hanpeter D, et al. Operative management and outcome of 302 abdominal vascular injuries. Am J Surg. 2000;180: 528Y533. 3. Burkhart HM, Gomez GA, Jacobson LE, Pless JE, Broadie TA. Fatal blunt aortic injuries: a review of 242 autopsy cases. J Trauma. 2001;50:113Y115. 4. Lock JS, Huffman AD, Johnson RC. Blunt trauma to the abdominal aorta. J Trauma. 1987;27:674Y677. 5. Roth SM, Wheeler JR, Gregory RT, et al. Blunt injury of the abdominal aorta: a review. J Trauma. 1997;42:748Y755. 6. Reisman JD, Morgan AS. Analysis of 46 intra-abdominal aortic injuries from blunt trauma: case reports and literature review. J Trauma. 1990; 30:1294Y1297. 7. Shalhub S, Starnes BW, Tran NT, et al. Blunt abdominal aortic injury. J Vasc Surg. 2012;55:1277Y1285. 8. Harris DG, Drucker CB, Brenner ML, Sarkar R, Narayan M, Crawford RS. Patterns and management of blunt abdominal aortic injury. Ann Vasc Surg. 2013;27:1074Y1080. 9. de MC, Dueck AD, Gomez D, Haas B, Nathens AB. Associated injuries, management, and outcomes of blunt abdominal aortic injury. J Vasc Surg. 2012;56:656Y660. 10. Mellnick VM, McDowell C, Lubner M, Bhalla S, Menias CO. CT features of blunt abdominal aortic injury. Emerg Radiol. 2012;19:301Y307. 11. Nguyen D, Platon A, Shanmuganathan K, Mirvis SE, Becker CD, Poletti PA. Evaluation of a single-pass continuous whole-body 16-MDCT protocol for patients with polytrauma. AJR Am J Roentgenol. 2009;192:3Y10. 12. Paul JS, Webb TP, Aprahamian C, Weigelt JA. Intraabdominal vascular injury: are we getting any better? J Trauma. 2010;69:1393Y1397. 13. Malhotra AK, Fabian TC, Croce MA, Weiman DS, Gavant ML, Pate JW. Minimal aortic injury: a lesion associated with advancing diagnostic techniques. J Trauma. 2001;51:1042Y1048. 14. Aladham F, Sundaram B, Williams DM, Quint LE. Traumatic aortic injury: computerized tomographic findings at presentation and after conservative therapy. J Comput Assist Tomogr. 2010;34:388Y394. 15. Burlew CC, Moore EE, Moore FA, et al. Western Trauma Association critical decisions in trauma: resuscitative thoracotomy. J Trauma Acute Care Surg. 2012;73:1359Y1363. 16. Brenner ML, Moore LJ, DuBose JJ, et al. A clinical series of resuscitative endovascular balloon occlusion of the aorta for hemorrhage control and resuscitation. J Trauma Acute Care Surg. 2013;75:506Y511. 17. Stannard A, Eliason JL, Rasmussen TE. Resuscitative endovascular balloon occlusion of the aorta (REBOA) as an adjunct for hemorrhagic shock. J Trauma. 2011;71:1869Y1872.
18. Morrison JJ, Percival TJ, Markov NP, et al. Aortic balloon occlusion is effective in controlling pelvic hemorrhage. J Surg Res. 2012;177:341Y347. 19. Scott DJ, Eliason JL, Villamaria C, et al. A novel fluoroscopy-free, resuscitative endovascular aortic balloon occlusion system in a model of hemorrhagic shock. J Trauma Acute Care Surg. 2013;75:122Y128.
EDITORIAL CRITIQUE Shalhub and colleagues have added to the growing literature addressing the rare entity of blunt aortic injury. They point out that the mortality is high, even when treated at some of the best trauma centers in the country. Mortality, as in all cases of trauma, was related to patient age, severity of injury and admitting physiology; most of the deaths occurred within 24 hours and were due to hemorrhage. This brings into sharp relief the authors emphasis on early diagnosis and early appropriate management. Not mentioned as a potential diagnostic modality, particularly in the hypotensive patient, is FAST. The abdominal aorta, and surrounding hematoma when present, is easily visualized in both sagittal and axial views. Useful are the authors’ recommendations, although not all are directly supported by their data. They have clearly provided us with fodder for a prospective, comparative effectiveness, multicenter trial evaluating endoluminal therapy, beta-blockade, anti-platelet agents and nonoperative management. However, in this era of financial stewardship, the comment that a ‘‘hybrid operating room’’ might improve survival is exaggeration in the extreme. A high resolution C-arm in a standard operating room is quite satisfactory for endoluminal management of the cases described. With that caveat, Shalhub and coworkers are to be congratulated for presenting extensive data on a rare injury and for providing us with clinically sound recommendations for diagnosis and management. Steven R. Shackford, MD Trauma Services Scripps Mercy Medical Center San Diego, California
* 2014 Lippincott Williams & Wilkins
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
885
Blunt abdominal aortic injury: A Western Trauma Association multicenter study Sherene Shalhub, MD, MPH, Benjamin W. Starnes, MD, Megan L. Brenner, MD, MS, Walter L. Biffl, MD, Ali Azizzadeh, MD, Kenji Inaba, MD, Dimitra Skiada, MD, Ben Zarzaur, MD, MPH, Cayce Nawaf, MD, Evert A. Eriksson, MD, Samir M. Fakhry, MD, Jasmeet S. Paul, MD, Krista L. Kaups, MD, MSc, David J. Ciesla, MD, S. Rob Todd, MD, Mark J. Seamon, MD, Lisa M. Capano-Wehrle, MPH, Gregory J. Jurkovich, MD, and Rosemary A. Kozar, MD, Seattle, Washington
Blunt abdominal aortic injury (BAAI) is a rare injury. The objective of the current study was to examine the presentation and management of BAAI at a multi-institutional level. METHODS: The Western Trauma Association Multi-Center Trials conducted a study of BAAI from 1996 to 2011. Data collected included demographics, injury mechanism, associated injuries, interventions, and complications. RESULTS: Of 392,315 blunt trauma patients, 113 (0.03%) presented with BAAI at 12 major trauma centers (67% male; median age, 38 years; range, 6Y88; median Injury Severity Score [ISS], 34; range, 16Y75). The leading cause of injury was motor vehicle collisions (60%). Hypotension was documented in 47% of the cases. The most commonly associated injuries were spine fractures (44%) and pneumothorax/hemothorax (42%). Solid organ, small bowel, and large bowel injuries occurred in 38%, 35%, and 28% respectively. BAAI presented as free aortic rupture (32%), pseudoaneurysm (16%), and injuries without aortic external contour abnormality on computed tomography such as large intimal flaps (34%) or intimal tears (18%). Open and endovascular repairs were undertaken as first-choice therapy in 43% and 15% of cases, respectively. Choice of management varied by type of BAAI: 89% of intimal tears were managed nonoperatively, and 96% of aortic ruptures were treated with open repair. Overall mortality was 39%, the majority (68%) occurring in the first 24 hours because of hemorrhage or cardiac arrest. The highest mortality was associated with Zone II aortic ruptures (92%). Follow-up was documented in 38% of live discharges. CONCLUSION: This is the largest BAAI series reported to date. BAAI presents as a spectrum of injury ranging from minimal aortic injury to aortic rupture. Nonoperative management is successful in uncomplicated cases without external aortic contour abnormality on computed tomography. Highest mortality occurred in free aortic ruptures, suggesting that alternative measures of early noncompressible torso hemorrhage control are warranted. (J Trauma Acute Care Surg. 2014;77: 879Y885. Copyright * 2014 by Lippincott Williams & Wilkins) LEVEL OF EVIDENCE: Epidemiologic study, level III; therapeutic study, level IV. KEY WORDS: Blunt trauma; abdominal aorta; minimal aortic injury; aortic rupture; blunt aortic injury. BACKGROUND:
B
lunt abdominal aortic injury (BAAI) is rare, with a reported incidence of less than 1% of all blunt trauma injuries. It is one of the most challenging injuries because of the associated injuries and hemodynamic instability of those who survive to reach the hospital.1Y3 Most of the BAAI literature has been limited to case reports, cases series, and literature review,4Y8 with the largest review to date obtained
from the National Trauma Data Bank.9 A standardized radiologic abdominal aortic injury classification and management details at a multi-institutional level have not been fully delineated. The aim of this study was to examine the presentation, radiologic classification, and management of this rare injury at multi-institutional level in a relatively large cohort of patients.
Submitted: February 16, 2014, Revised: March 31, 2014, Accepted: April 9, 2014, Published online: Sepember 22, 2014. From the Division of Vascular Surgery (S.S., B.W.S.), Department of Surgery, University of Washington, Seattle, Washington; R. Adams Cowley Shock Trauma Center (M.L.B.), University of Maryland, Baltimore, Maryland; Department of Surgery (W.L.B., G.J.J.), Denver Health Medical Center and the University of Colorado School of Medicine, Denver, Colorado; Department of Cardiothoracic and Vascular Surgery (A.A.), University of Texas Medical School at Houston; and Department of Surgery (R.A.K.), University of Texas Health Science Center at Houston, Houston, Texas; Division of Trauma and Surgical Critical Care (K.I., D.S.), Department of Surgery, Keck School of Medicine, Los Angeles County + University of Southern California Medical Center, Los Angeles; and Community Regional Medical Center (K.L.K.), Department of Surgery, University of California, San Francisco-Fresno Campus, Fresno, California; Division of Trauma and Critical Care (B.Z., C.N.), Department of Surgery, University of Tennessee Health Science Center, Memphis, Tennessee; Division of General Surgery (E.A.E., S.M.F.), Department of Surgery, Medical University of South Carolina, Charleston, South Carolina; Division of Trauma and Critical Care (J.S.P.), Department of Surgery, Medical College of Wisconsin, Milwaukee, Wisconsin; Department of Surgery (D.J.C.), University of South Florida Colleges of Medicine, Tampa, Florida; Department of Surgery (S.R.T.), New York University Langone Medical Center, New York, New York; Division of Trauma and Surgical Critical Care (M.J.S., L.M.C.-W.), Department of Surgery, Cooper University Hospital, Camden, New Jersey. This study was presented at the 44th Annual Meeting of the Western Trauma Association, March 6, 2014, in Steamboat Springs, Colorado. Address for reprints: Sherene Shalhub, MD, MPH, Division of Vascular Surgery, Department of General Surgery, The University of Washington, 1959 NE Pacific St, Box 356410, Seattle, WA 98195; email: [email protected]. DOI: 10.1097/TA.0000000000000353
J Trauma Acute Care Surg Volume 77, Number 6
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
879
J Trauma Acute Care Surg Volume 77, Number 6
Shalhub et al.
PATIENTS AND METHODS The Western Trauma Association Multi-Center Trials Group performed a retrospective cohort study of patients presenting with BAAI from 1996 to 2011. Institutional review board approval was obtained at each of the contributing sites. Included in this series are data previously described as a single-center experience.7,8 BAAI was defined as an injury to the abdominal aorta from the diaphragmatic hiatus to the aortic bifurcation and identified by the International Classification of DiseasesV9th Rev. (ICD-9) code 902.0 (injury aorta, abdominal). Exclusion criteria included cases with blunt thoracic aortic injury with dissection extending to the abdominal aorta and cases diagnosed on autopsy alone but not recognized at the time of patient care. Aortic injuries classification was based on the presence of external aortic contour abnormality on computed tomography (CT) or the presence of free rupture found intraoperatively.7 In the intimal tears and large intimal flaps (LIFs), the normal external aortic contour is preserved on CT. Intimal tears were defined as intimal defect and/or thrombus of 10 mm in length or width. LIF were defined as intimal defect and/or thrombus of 10 mm in length or width. Intramural hematoma was included in this category. A pseudoaneurysm was thought of as a contained rupture and clearly associated with an external aortic contour abnormality on CT. Aortic ruptures were associated with external aortic contour abnormality with free contrast extravasation on CT or hemoperitoneum found upon laparotomy. Abdominal aorta zones of injury were classified based on possible endovascular surgical approaches as follows (Fig. 1): Zone I injuries occur from the diaphragmatic hiatus to above the superior mesenteric artery (SMA). Zone II injuries include the region from the SMA to and including the renal arteries. Zone III injuries are below the renal arteries and to the aortic bifurcation. Branch vessel avulsion injury was noted by type (celiac, SMA, inferior mesenteric artery, renal and lumbar arteries). Specific data collected included age, sex, mechanism of injury, initial clinical presentation, Injury Severity Score (ISS), presence of a seat belt sign, other associated injuries, diagnostic methods, management, modes of repair, procedural complications, mortality, subsequent radiographic imaging, and follow-up. Representative images from CT, CT angiogram (CTA), and angiograms were independently reviewed by a single author (S.S.) to confirm the classification of the injury when imaging was available to be retrieved from the archives (n = 41). The remainder of the injuries was classified based on operative and imaging reports. Data were analyzed using Microsoft Excel 2007 software (Microsoft, Redmond, WA) and SPSS version 19 for Windows (SPSS, Inc., Chicago, IL). Continuous data are presented as medians and ranges. Means of continuous data were compared using the Student’s t test. Categorical data were compared using the Fisher’s exact test or W2 analysis, where appropriate.
RESULTS During the study period, there were 392,315 patients admitted for blunt trauma to 12 major trauma centers. Of 880
Figure 1. Classification of abdominal aorta zones of injury based on possible endovascular surgical approaches. Zone I injuries occur from the diaphragmatic above the SMA. Zone II injuries include the SMA to the renal arteries. Zone III injuries are between the renal arteries and the aortic bifurcation. Zone I and III injuries are amenable to endovascular repair.
those, 113 patients were identified with BAAI giving rise to an incidence of 0.03% of all blunt trauma injuries. The majority of the cases were male (n = 78, 67.3%) with a median age of 38 years (range, 6Y88 years). The median ISS was 34 (range, 16Y75). The leading cause of injury was motor vehicle collisions (n = 68, 60.2%) followed by pedestrian versus vehicle crashes (13, 11.5%). Nearly half of the cases presented with hypotension (systolic blood pressure [SBP] G 90 mm Hg) on arrival (n = 53, 46.9%). Seat belt sign was documented in 24 cases (35.3% of motor vehicle crash cases). Cardiac arrest in the prehospital setting or in the emergency department (ED) was documented in 23 cases (20.4%). ED thoracotomy was performed in 13 cases (11.5%). CT or CTA was used in 74.3% of the cases. The likelihood of obtaining a CT scan was highly related to the patient’s hemodynamic stability as 31% with reported SBP less than 90 mm Hg had a CT scan compared with 69% of those without hypotension (p G 0.000). Diagnostic peritoneal lavage was used in 2 (1.8%) and a diagnostic angiogram in 10 cases (8.8%). The most commonly associated injuries were spine fractures (n = 50, 44.2%), solid organ injury (n = 43, 38.1%), small bowel injury (n = 39, 34.5%), and large bowel injury (n = 32, 28.3%). Table 1 summarizes the demographics, * 2014 Lippincott Williams & Wilkins
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
J Trauma Acute Care Surg Volume 77, Number 6
Shalhub et al.
TABLE 1. Demographics, Mechanism of Injury, Clinical Presentation, Associated Injuries, Type and Zone of Injury in Patients With Blunt Abdominal Aortic Injuries
TABLE 1. (Continued)
n = 113 Male, n (%) Age, median (range), y ISS, median (range) Mechanism of injury, n (%) Motor vehicle crash Car vs. pedestrian Fall Motor cycle crash Crush injury Car vs. bicycle All-terrain vehicle crash Other Clinical presentation, n (%) SBP G 90 mm Hg on arrival Cardiac arrest before surgery ED thoracotomy Mortality in ED or on the way to surgery Seat belt sign documented Comorbidities, n (%) Hypertension Coronary artery disease Peripheral vascular disease Creatinine Q 1.5 Diagnostic studies, n (%) Trauma CT CTA Angiogram Diagnostic peritoneal lavage Associated injuries, n (%) Spine fracture Cervical spine fracture Thoracic spine fracture Lumbar spine fracture Pneumothorax/hemothorax, n (%) Hemoperitoneum, n (%) Solid organ injury, n (%) Liver Spleen Kidney Pancreas Rib fractures, n (%) Small bowel injury, n (%) Mesenteric hematoma, n (%) Traumatic brain injury, n (%) Colon injury, n (%) Pelvic fracture, n (%) Abdominal wall ecchymosis, n (%) Abdominal wall degloving, n (%) Inferior vena cava injury, n (%) Superior mesenteric/portal vein injury, n (%) BAAI type, n (%) Intimal tear LIF
76 (67.3) 38 (6Y88) 34 (16Y75) 68 (60.2) 13 (11.5) 10 (8.8) 7 (6.2) 6 (5.3) 2 (1.8) 2 (1.8) 5 (4.4) 53 (46.9) 23 (20.4) 13 (11.5) 5 (4.4) 24 (21.2) 9 (7.9) 5 (4.4) 4 (3.5) 5 (4.4) 53 (46.9) 31 (27.4) 10 (8.8) 2 (1.8) 50 (44.2) 7 (6.2) 14 (12.4) 33 (29.2) 47 (41.6) 45 (39.8) 43 (38.1) 24 (21.2) 20 (17.7) 8 (7.1) 6 (5.3) 39 (34.5) 39 (34.5) 38 (33.6) 35 (31.0) 32 (28.3) 28 (24.8) 27 (23.9) 10 (8.8) 10 (8.8) 1 (0.8) 20 (17.7) 38 (33.6)
n = 113 Pseudoaneurysm Rupture Injury to preexisting AAA Complicated LIFs, n (%) Aortic dissection Extension of dissection into the iliac artery Thrombosis Aortic rupture description, n (%) Aortic wall rupture Branch vessel avulsion Celiac artery avulsion Superior mesenteric artery avulsion Inferior mesenteric artery avulsion Right renal artery Left renal artery Lumbar arteries avulsion Iliac artery avulsion Aortoiliac junction rupture Zone of injury, n (%) Zone I Zone II Zone III
18 (15.9) 36 (31.9) 1 (0.9) 7 (6.2) 9 (8.0) 4 (3.5) 15 (13.3) 24 (21.2) 3 (2.7) 5 (4.4) 3 (2.7) 4 (3.5) 3 (2.7) 8 (7.1) 1 (0.9) 1 (0.9) 22 (19.5) 22 (19.5) 75 (66.4)
mechanism of injury, clinical presentation, and associated injuries in this cohort.
BAAI Management Choice of management varied by type of BAAI. BAAI presented as intimal tear (n = 20, 17.7%), LIF (n = 38, 36%), pseudoaneurysm (n = 18, 15.9%), and free rupture (n = 36, 31.9%). In one case, the patient presented with a thrombosis of a small infrarenal abdominal aortic aneurysm (AAA) without rupture. Traumatic aortic dissection with long intimal flaps was seen in seven cases (6.2%), and these were included in the LIF group. Table 1 details the location and type of BAAI seen in this cohort. The majority of intimal tears (89.5%) were managed nonoperatively, while all aortic ruptures were managed with open repair. In the cases managed operatively, the operative approach varied according to the zone of injury, with the majority of Zone II injuries managed by open repair (42%), while none of the Zone II injuries were suitable for endovascular repair. Table 2 details the management of BAAI by type and zone of injury. The next three sections describe BAAI management in detail.
Nonoperative Management A total of 40 cases (35.4%) were managed nonoperatively. The majority of the BAAI injuries managed nonoperatively were those presenting without aortic contour abnormalities (89.5% of intimal tears and 44.7% of LIFs). One third of the pseudoaneurysms were also managed nonoperatively.
* 2014 Lippincott Williams & Wilkins
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
881
J Trauma Acute Care Surg Volume 77, Number 6
Shalhub et al.
TABLE 2. BAAI Management by Type and Zone of Injury Nonoperative Management Management by injury type Intimal tear (n = 19) Large intimal flap (n = 38) Pseudoaneurysm (n = 18) Aortic rupture (n = 30) Management by injury zone Zone I (n = 18) Zone II (n = 14) Zone III (n = 69) Multiple zones (n = 5)
Open Repair
Endovascular Repair
17 (89.5%) 17 (44.7%) 6 (33.3%) 0
1 13 4 30
(5.3%) (34.2%) (22.2%) (100%)
1 (5.3%) 8 (21.1%) 8 (44.4%) 0
6 (33.3%) 3 (21.4%) 28 (40.6%) 3 (60%)
9 (50%) 11 (78.6%) 29 (42.0%) 0
3 (16.7%) 0 12 (17.4%) 2 (40%)
Table 3 summarizes the management and outcomes of the pseudoaneurysms. Failure of nonoperative management occurred in three cases: BAAI with aortic dissection along the length of the abdominal aorta presenting initially with temporary paraplegia. The patient became hemodynamically unstable 1 day after admission, and a large retroperitoneal hematoma was found intraoperatively. The patient died intraoperatively (Fig. 2A). Another case of failed nonoperative management was that of a BAAI associated with acute thrombosis of the right common iliac artery at the aortic bifurcation. The patient refused treatment and died of sepsis related to an ischemic lower extremity. A third case was that of a Zone II thrombus managed with anticoagulation initially; however, there was thrombus extension on follow-up imaging, which led to a hybrid repair with aortic debranching and a stent graft placement.
Open Surgical Management Forty-nine cases (43.4%) including 30 aortic ruptures required open exploration and repair. Of those, 32 (65%) presented with SBP less than 90 mm Hg, 14 (28.6%) experienced a cardiac arrest before arrival to the operating room. ED thoracotomy with aortic cross clamping was performed in 7 (14.3%) of these cases. Open repair was performed by interposition tube or bifurcated graft (n = 16, 32.7%), primary repair (n = 13, 26.5%), and branch ligation (n = 6, 12.2%). In eight cases (16.3%), an exploratory laparotomy was started, but no repairs were undertaken because of cardiac arrest from hemorrhagic shock. In three cases, no repair was documented, and the abdomen was packed, followed by a transfer to the angiography suite for arterial embolization (n = 3, 6.1%). Intraoperative findings were described as retroperitoneal hematoma (n = 29, 59.2%), contused aorta (n = 10, 20.4%), and aorta thrombosis (n = 8, 16.3%) (Fig. 2B) as well as one normal appearing aorta, which was associated with an intimal tear later seen on CTA. One case of LIF was treated with primary repair but required take back the next day for endovascular repair because of thrombosis of an iliac artery.
Endovascular Management Endovascular management was used as a primary mode of therapy in 17 cases (15%). The majority of the cases were male (70.5%) with a median age of 35 years (range, 15Y62 years) and median ISS of 29 (range, 24Y50). In these cases, a variety of stents and stent grafts were used: wall stent (n = 2), Gore excluder (n = 4), Medtronic AneuRX cuff (n = 4), Atrium stent (n = 1), Gore Cuff (n = 1), Viabahn (n = 1), Gore TAG (n = 1), and other (n = 3). Use of intravascular ultrasound was documented in two cases. The BAAI injuries
TABLE 3. Management and Outcomes of BAAI Pseudoaneurysms Age/Sex
ISS
Injury Zone
Size, mm
Management
LOS, d
Died
Follow-up
42, Male 81 Female 41, Male 42, Female
29 50 50 34
III III I III
7 16 5 12 30 20 30
8 63 53 33
N Y, TBI N N
Y N/A Y N
21, Male 47, Male 77, Male 26, Male 36, Female 21, Male 24, Male 37, Male 6, Female 53, Female 20, Male 22, Female 43, Female 38, Female
43 NR 29 57 34 21 29 25 25 34 NR 26 25 NR
I III III III III II III I III III III III II I
37 21 27 10
Endovascular stent graft Nonoperative Endovascular stent graft Nonoperative (surgery for mesenteric bleeding, no comment on retroperitoneal findings) Open repair, interposition graft Endovascular stent graft Nonoperative (died before repair) Nonoperative Open repair, bifurcation graft Open repair, primary, hospital Day 2 Endovascular stent graft Endovascular stent graft Nonoperative Endovascular stent graft Endovascular stent graft Endovascular stent Nonoperative Open repair, interposition graft
0 31 13 31 95 0 5 6 29 14 33 9 114 10
Y, coagulopathy N Y, pneumonia N N Y N N N N N N Y, sepsis N
N/A Y N/A Y Y N/A Y Y Y N Y, endoleak Y 2nd cuff Y N/A Y
25 22 28 10 18 12 15 32 10 10 NR NR NR NR NR
N, no; N/A, not applicable; NR, not reported; TBI, traumatic brain injury; Y, yes.
882
* 2014 Lippincott Williams & Wilkins
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
J Trauma Acute Care Surg Volume 77, Number 6
Shalhub et al.
was transferred to the angiography suite for embolization of lumbar arteries (n = 2) or celiac artery embolization with stent graft placement in the Zone I (n = 1). In one case, an LIF was treated with open primary repair but required take back the next day for endovascular repair because of thrombosis with extension into an iliac artery.
Length of Stay and Follow-up The overall length of stay for the cohort was 9.5 days (range, 0Y113 days). Sixty-nine patients were discharged (63.8% male; median age, 31 years; range, 6Y62 years). Of interest, two of these cases required ED thoracotomy (15% survival post-ED thoracotomy). The median length of stay of survivors was 14 days (range, 2Y95 days). The predominant injury type in those discharged was intimal tear (n = 18, 26.1%) or LIF (n = 32, 46.4%), and the predominant zone of injury was Zone III (n = 52, 75.4%). Follow-up of at least one visit was documented in 39% of the cases (n = 27): 10 had open repairs, and 9 had endovascular repairs. Long-term follow-up data were not available, and no data regarding associated graft infections were available. One case of endovascular pseudoaneurysm repair demonstrated an endoleak on follow-up imaging and required placement of a second stent graft.
Mortality
Figure 2. A and B, Examples of complicated LIFs. A, Coronal and axial CTA imaging showing BAAI dissection extending along the length of the abdominal aorta in a patient presenting initially with temporary paraplegia and who developed a large periarotic hematoma the next day and died intraoperatively. B, Intraoperative imaging of BAAI presenting with normal contour abnormality on trauma CT with associated bilateral lower extremity limb ischemia. The torn intima can be seen superiorly (block arrow), and the thrombus is seen in the distal aorta (long arrow). C and D, Examples of Zone I and Zone III BAAI endovascular repair. C, Sagittal view of CTA imaging showing Zone I pseudoaneurysm and a three-dimensional reconstruction of follow-up CTA after stent graft placement with extension into the distal thoracic. D, Coronal view of CTA imaging showing Zone III pseudoaneurysm and three-dimensional reconstruction of follow-up CTA imaging after bifurcated stent graft placement.
treated were one intimal tear (5.9%), LIFs (n = 8, 47.1%), and pseudoaneurysms (n = 8, 47.1%). All these injuries were limited to the Zone I (n = 5, 29.4%) or Zone III (n = 12, 70.5%). Figure 2C and D illustrate examples of Zone I and Zone III BAAI endovascular repair. All patients in this group were discharged alive after a median length of stay of 12 days (range, 5Y53 days). Five additional cases were then managed by endovascular means as an adjunctive therapy. One case was a Zone II injury with thrombus that progressed on anticoagulation and was treated with hybrid repair. In three cases, an exploratory laparotomy was performed first, then the abdomen was packed, and the patient
Overall mortality was 38.9% (n = 44), with the majority (n = 30, 68.2%) occurring in the first 24 hours because of cardiac arrest from hemorrhagic shock with the exception of one death caused by severe traumatic brain injury. Those who died were older compared with survivors, with a median age of 48.5 years (range, 6Y88 years; p G 0.000), and had higher ISS, with a median ISS of 43 (range, 16Y74) versus 34 (range, 16Y59, p G 0.000). Mortality varied by the zone of injury and the type of injury, with the highest mortality occurring in Zone II aortic ruptures (n = 14) with a mortality of 92.9%. Table 4 details BAAI mortality by initial presentation as well as type and zone of injury. Overall mortality in the open repair group was 57.1% (n = 28) including 11 intraoperative deaths (22.4%). Other causes of death in this group included two complications related to aortic thrombosis and ischemia-reperfusion of the bilateral lower extremities, which required amputations, and two caused by traumatic brain injuries. There was 0% mortality in the group treated with endovascular repair as the primary modality of treatment and 100% mortality in the four cases managed by endovascular means after an open exploration. Late causes of mortality (hospital days 3Y114 days) included multiorgan system failure and sepsis (n = 11). One death was caused by SMA bypass thrombosis in a Zone II thrombosis hybrid repair and subsequent bowel ischemia.
DISCUSSION This is the largest BAAI series reported to date. BAAI is rare, with an incidence less than 0.03% with even the busiest trauma centers managing zero to three cases a year. With the use of high-resolution CTA in the trauma workup, an increasing number of intimal tears are expected to be diagnosed.10Y12 Disruption of the intima is the underlying pathologic lesion
* 2014 Lippincott Williams & Wilkins
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
883
J Trauma Acute Care Surg Volume 77, Number 6
Shalhub et al.
TABLE 4. BAAI Mortality by Initial Presentation, Type and Zone of Injury Lived Age, mean (range), y ISS, mean (range)
Died
33.6 (6Y62) 48.1 (6Y88) 33.3 (16Y59) 42.7 (16Y74) Mortality SBP G 90 mm Hg (n = 53) 38 (71.7%) Cardiac arrest before surgery (n = 23) 21 (91.3%) Cardiac arrest (n = 31) 29 (93.5) Mortality with ED thoracotomy (n = 13) 11 (84.6%) BAAI type Intimal tears 2 (10%) LIFs 6 (15.8%) Pseudoaneurysm 5 (27.8%) Aortic rupture 30 (83.3%) Occluded small AAA 1 (100%) BAAI zone Zone I 8 (42.1%) Zone II 15 (83.3%) Zone III 19 (26.8%) Multiple zones 2 (40%)
p 0.000 0.001 p
0.000
0.000
of BAAI, which presents as a spectrum ranging from minimal intimal disruption to aortic rupture. For this study, the classification of the BAAI was based on the presence or absence of aortic contour abnormality as seen by axial images on CT. Intimal tears that are the equivalent of minimal aortic injury13,14 and LIF have preserved external aortic contour by CT, while external aortic contour abnormality is seen in the case of pseudoaneurysms, aortic wall disruption, and branch avulsions. Management varied somewhat between centers, depending on available resources and regional practice variation. Based on the available data in this study and previous studies,4Y9 several recommendations can be made. & All patients with suspected BAAI should undergo rapid CTA if hemodynamically fit. & Intimal tears can be managed nonoperatively with antiplatelet therapy and A-blockers followed by interval CTA follow-up within 30 days. & Uncomplicated LIF can be managed nonoperatively with antiplatelet therapy and A-blockers followed by interval CTA within 48 hours to assess for progression. If there is evidence of progression (intraluminal thrombus, aneurysmal degeneration, and pseudoaneurysm formation), then LIF is complicated and can be managed, when possible, with endovascular repair. & When known or suspected BAAI patients are taken to the operating room, they should be placed on a fluoroscopycompatible table, and preparation should be wide and include both groin regions for femoral access should endovascular repair be contemplated. Aortic Zones I and III are amenable to endovascular repair with particular benefit in the setting of concurrent gross contamination. & When BAAI is identified in the setting of gross contamination or hollow viscus injury, strong consideration should be given to endovascular repair. 884
& The use of antiplatelet therapy for 6 weeks after injury should be considered if there are no contraindications. & Postdischarge follow-up is strongly recommended. In the cases managed nonoperatively, follow-up with CTA at 1 month, 6 months, 1 year, and every other year thereafter is recommended or until the injury has resolved. & The timing of repair is based on the hemodynamic stability and concurrent associated injuries. The repair can be open or via endovascular means such as stent graft placement and branch vessel embolization. Mortality remains high among those who survive to the hospital with a traumatic abdominal aortic rupture. A significant number of patients (52.8%) presented in extremis and cardiac arrest before arriving in the operating room with an overall mortality of 83%. This supports the need for improved modalities of early hemorrhage control. Given the poor outcomes of ED thoracotomy,15 evaluation of newer modalities for early control of noncompressible torso hemorrhage such as resuscitative endovascular balloon occlusion of the aorta is warranted.16Y19 The use of a hybrid operating room in the treatment of these cases would allow the surgeon the full breadth of available treatment modalities with the potential for improved survival; this however, remains to be assessed. This study was subject to the normal limitations seen with a retrospective chart review including inability to retrieve old images, data regarding use of antiplatelet therapies, or long-term follow-up data. In light of the rare occurrence of this injury, ongoing prospective data collection across multiple centers would provide additional data for analysis.
CONCLUSION This is the largest BAAI series reported to date. BAAI is rare and presents as a spectrum of injury, ranging from minimal aortic injury to aortic rupture. Nonoperative management is successful in uncomplicated cases with preserved external aortic contour on CT. The highest mortality occurred in those with free aortic rupture, suggesting that alternative measures of early hemorrhage control are needed. Ongoing prospective data collection is warranted.
AUTHORSHIP S.S., G.J.J., and R.A.K. contributed to the study conception and design. S.S., R.A.K., B.W.S., and W.L.B. performed analysis and interpretation. S.S., M.L.B., W.L.B., K.I., D.S., B.Z., C.N., E.A.E., S.M.F., J.S.P., K.L.K., D.J.C., S.R.T., M.J.S., L.M.C.-W., and R.A.K. conducted the data collection. S.S., R.A.K., and B.W.S. wrote the article. S.S., B.W.S., A.E.A., R.A.K., W.L.B., S.M.F., and K.I. provided critical revision.
ACKNOWLEDGMENT We thank the patients involved in this study. We also thank Chris Akers for his support with the medical illustrations.
DISCLOSURE The authors declare no conflicts of interest.
* 2014 Lippincott Williams & Wilkins
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
J Trauma Acute Care Surg Volume 77, Number 6
Shalhub et al.
REFERENCES 1. Deree J, Shenvi E, Fortlage D, et al. Patient factors and operating room resuscitation predict mortality in traumatic abdominal aortic injury: a 20year analysis. J Vasc Surg. 2007;45:493Y497. 2. Asensio JA, Chahwan S, Hanpeter D, et al. Operative management and outcome of 302 abdominal vascular injuries. Am J Surg. 2000;180: 528Y533. 3. Burkhart HM, Gomez GA, Jacobson LE, Pless JE, Broadie TA. Fatal blunt aortic injuries: a review of 242 autopsy cases. J Trauma. 2001;50:113Y115. 4. Lock JS, Huffman AD, Johnson RC. Blunt trauma to the abdominal aorta. J Trauma. 1987;27:674Y677. 5. Roth SM, Wheeler JR, Gregory RT, et al. Blunt injury of the abdominal aorta: a review. J Trauma. 1997;42:748Y755. 6. Reisman JD, Morgan AS. Analysis of 46 intra-abdominal aortic injuries from blunt trauma: case reports and literature review. J Trauma. 1990; 30:1294Y1297. 7. Shalhub S, Starnes BW, Tran NT, et al. Blunt abdominal aortic injury. J Vasc Surg. 2012;55:1277Y1285. 8. Harris DG, Drucker CB, Brenner ML, Sarkar R, Narayan M, Crawford RS. Patterns and management of blunt abdominal aortic injury. Ann Vasc Surg. 2013;27:1074Y1080. 9. de MC, Dueck AD, Gomez D, Haas B, Nathens AB. Associated injuries, management, and outcomes of blunt abdominal aortic injury. J Vasc Surg. 2012;56:656Y660. 10. Mellnick VM, McDowell C, Lubner M, Bhalla S, Menias CO. CT features of blunt abdominal aortic injury. Emerg Radiol. 2012;19:301Y307. 11. Nguyen D, Platon A, Shanmuganathan K, Mirvis SE, Becker CD, Poletti PA. Evaluation of a single-pass continuous whole-body 16-MDCT protocol for patients with polytrauma. AJR Am J Roentgenol. 2009;192:3Y10. 12. Paul JS, Webb TP, Aprahamian C, Weigelt JA. Intraabdominal vascular injury: are we getting any better? J Trauma. 2010;69:1393Y1397. 13. Malhotra AK, Fabian TC, Croce MA, Weiman DS, Gavant ML, Pate JW. Minimal aortic injury: a lesion associated with advancing diagnostic techniques. J Trauma. 2001;51:1042Y1048. 14. Aladham F, Sundaram B, Williams DM, Quint LE. Traumatic aortic injury: computerized tomographic findings at presentation and after conservative therapy. J Comput Assist Tomogr. 2010;34:388Y394. 15. Burlew CC, Moore EE, Moore FA, et al. Western Trauma Association critical decisions in trauma: resuscitative thoracotomy. J Trauma Acute Care Surg. 2012;73:1359Y1363. 16. Brenner ML, Moore LJ, DuBose JJ, et al. A clinical series of resuscitative endovascular balloon occlusion of the aorta for hemorrhage control and resuscitation. J Trauma Acute Care Surg. 2013;75:506Y511. 17. Stannard A, Eliason JL, Rasmussen TE. Resuscitative endovascular balloon occlusion of the aorta (REBOA) as an adjunct for hemorrhagic shock. J Trauma. 2011;71:1869Y1872.
18. Morrison JJ, Percival TJ, Markov NP, et al. Aortic balloon occlusion is effective in controlling pelvic hemorrhage. J Surg Res. 2012;177:341Y347. 19. Scott DJ, Eliason JL, Villamaria C, et al. A novel fluoroscopy-free, resuscitative endovascular aortic balloon occlusion system in a model of hemorrhagic shock. J Trauma Acute Care Surg. 2013;75:122Y128.
EDITORIAL CRITIQUE Shalhub and colleagues have added to the growing literature addressing the rare entity of blunt aortic injury. They point out that the mortality is high, even when treated at some of the best trauma centers in the country. Mortality, as in all cases of trauma, was related to patient age, severity of injury and admitting physiology; most of the deaths occurred within 24 hours and were due to hemorrhage. This brings into sharp relief the authors emphasis on early diagnosis and early appropriate management. Not mentioned as a potential diagnostic modality, particularly in the hypotensive patient, is FAST. The abdominal aorta, and surrounding hematoma when present, is easily visualized in both sagittal and axial views. Useful are the authors’ recommendations, although not all are directly supported by their data. They have clearly provided us with fodder for a prospective, comparative effectiveness, multicenter trial evaluating endoluminal therapy, beta-blockade, anti-platelet agents and nonoperative management. However, in this era of financial stewardship, the comment that a ‘‘hybrid operating room’’ might improve survival is exaggeration in the extreme. A high resolution C-arm in a standard operating room is quite satisfactory for endoluminal management of the cases described. With that caveat, Shalhub and coworkers are to be congratulated for presenting extensive data on a rare injury and for providing us with clinically sound recommendations for diagnosis and management. Steven R. Shackford, MD Trauma Services Scripps Mercy Medical Center San Diego, California
* 2014 Lippincott Williams & Wilkins
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
885
