Posterior acceleration as a mechanism of blunt traumatic injury of the aorta

Transcription

1 Journal of Biomechanics 41 (2008) Short communication Posterior acceleration as a mechanism of blunt traumatic injury of the aorta Jason Forman a,, Stephen Stacey a, Jay Evans b, Richard Kent a a University of Virginia, Center for Applied Biomechanics, 1011 Linden Avenue, Charlottesville, VA 22903, USA b University of Virginia Health System, Department of Pathology, Charlottesville, VA, USA Accepted 31 January Abstract Rupture of the thoracic aorta is a leading cause of rapid fatality in automobile crashes, but the exact mechanisms of this injury remain unidentified. One commonly postulated mechanism is a differential motion of the aortic arch relative to the heart and its neighboring vessels caused by high-magnitude acceleration of the thorax. This paper investigates acceleration as an aortic injury mechanism using nine impact-sled tests with human cadaver thoraces. The test system utilized generates very high posteriorly directed thoracic accelerations with minimal compression of the chest. The sled tests resulted in peak mid-spine accelerations of g (mean7standard deviation) with sustained mid-spine accelerations of up to 80 g for 20 ms in most cases. The tests resulted in maximum chest compressions of 773.1% of the total chest depth, and maximum recorded increases in intra-aortic, tracheal, and esophageal pressure of 177, 112, and 156 kpa, respectively. No macroscopic injuries to the thoracic aorta resulted from these tests, though other limited visceral injury was observed. The results suggest that posteriorly directed acceleration alone (up to the magnitudes studied here) is not sufficient to cause gross aortic injury. Furthermore, the observed transient increases in intra-aortic and extra-aortic pressure indicate that complex pressure distributions are present during dynamic thoracic deceleration events. This suggests that any attempt to model traumatic aortic injury should include consideration for both the intra-aortic fluid pressure and the extra-aortic, intra-thoracic pressure present during the event. r 2008 Elsevier Ltd. All rights reserved. Keywords: Aorta; Trauma; Acceleration; Injury mechanisms; Injury biomechanics 1. Introduction Rupture of the thoracic aorta is a leading cause of rapid fatality in automobile crashes (McGwin et al., 2003; Richens et al., 2003; Shkrum et al., 1999). Inertial loading of the aortic arch caused by high thoracic acceleration is a commonly postulated, but rarely investigated mechanism of aortic injury (Richens et al., 2002; Shkrum et al., 1999). Similarly, Lundevall (1964) theorized that acceleration of the fluid inside the aorta may result in an injurious increase in intra-aortic fluid pressure. Limited field data and experimental studies have provided some insight into thoracic tolerance to posteriorly Corresponding author. Tel.: ; fax: addresses: jlf3m@virginia.edu (J. Forman), steve.stacey@gmail.com (S. Stacey), jcevans11@gmail.com (J. Evans), rwk3c@virginia.edu (R. Kent). directed acceleration (Stapp, 1970; Mertz and Gadd, 1971; Melvin et al., 1998; Baque et al., 2006). None of those studies, however, explicitly attempted to isolate acceleration as a mechanical input into the body. The purpose of this study was to investigate the injury tolerance of the aorta to posteriorly directed thoracic acceleration isolated from other potentially causative factors (i.e., chest compression). 2. Methods Nine impact-sled tests were performed subjecting human cadaver thoraces (Table 1) to posteriorly directed, high-magnitude accelerations with minimal chest compression. A detailed description of the development of the test methods used here is reported in Forman et al. (2005). The cadavers used here were pre-screened for HIV, hepatitis A, B, and C, and pre-existing injury and pathologic anomalies (via computed tomography scan). The causes of death for these subjects were unlikely to /$ - see front matter r 2008 Elsevier Ltd. All rights reserved. doi: /j.jbiomech

2 1360 J. Forman et al. / Journal of Biomechanics 41 (2008) Table 1 Subject information Test number Cadaver ID Age at death/gender 47/M 62/M 42/F 57/F 53/M 74/F 52/M 65/M 74/M Mass (kg) Stature (cm) Chest depth (supine, a th rib, cm) Chest breadth (supine, 8th rib, cm) a Cause of death ETOH complications Cerebral vascular accident Drug overdose Cardiopulmonary arrest Respiratory failure Cardiovascular accident ETOH and vascular disease Pancre-atic cancer Intracranial bleed a Chest depth and chest breadth are not available for this subject. Chest depth and breadth estimated (from photographs) to be approximately equal to that of subject 206. Fig. 1. Schematic illustration of the sled test system. The subject is connected to the target sled with a series of brackets attached to its spine, and is mounted inside a rigid tank filled with stiff polypropylene beads. have degraded tissue significantly antemortem. The subjects were unembalmed and were preserved by freezing until the time of testing. The subjects upper and lower extremities were removed prior to testing. All cadaver testing and handling procedures were approved by the University of Virginia (UVA) Center for Applied Biomechanics Oversight Committee and the UVA Institutional Biosafety Committee. For each test, the subject was mounted on a lightweight target sled (initially stationary), which was then struck by a heavy bullet sled (traveling at approximately 60 km/h, Fig. 1). The subject s spine and pelvis were connected to the target sled with a series of vertebral brackets (eight to eleven U-shaped brackets per subject, attached to the posterior lamina with wood screws inserted through the lamina and pedicles and into the vertebral bodies). These connections were adjusted to attain a spinal curvature representative of a seated automobile occupant. A steel tank was then installed on the target sled around the subject, packed with stiff polypropylene beads, and sealed with a lid. This provided a loading path that was distributed over the entire thorax, and prevented the thorax from expanding upon the application of posteriorly directed loading to the spine. The subjects pulmonary and cardiovascular systems were pressurized (independently) to the typical in vivo mean arterial pressure (approximately 10 kpa externally) immediately prior to testing. Pulmonary pressurization was accomplished via a tracheostomy. Cardiovascular pressurization was accomplished via fluid (blood plasma replacement solution and ink) introduced through a cannula inserted into a carotid artery. In tests , the abdominal aorta was occluded just superior to the bifurcation with a Foley catheter inserted via a femoral artery. That method of occlusion caused artifactual, partial-thickness injury in the abdominal aorta (see below), and for subsequent tests the femoral arteries were ligated instead. For all tests, the remaining carotid arteries and exposed subclavian arteries were ligated to prevent excessive fluid loss. To mitigate fluid loss through the muscular tissue transected during the spinal hardware installation, for test 3.1 and 3.2 a capillary occlusion solution was injected into the cardiovascular system prior to pressurization (Ziperman et al., 1975). Nominally 390 ml of solution containing 950 ml saline and 250 g of polyethylene micro-spheres (5 50 mm particle diameter, Microthene s FN 501, Lyondell Chemical Company, Houston, TX) was injected into the cardiovascular system via a carotid artery, followed by an injection of normal saline to force the micro-sphere solution into the peripheral arterioles and capillaries. Acceleration of the test subject was measured by two tri-axial accelerometer arrays, one mounted on the sternum and one mounted on the seventh thoracic vertebra. The resulting accelerations were resolved into the subject reference frame (subject X-axis parallel to the direction of acceleration) using the initial orientations of the accelerometer arrays. Chest deflection was calculated by twice numerically integrating the differential, X-axis acceleration between the sternum and the spine. For tests 2.1 through 3.2, chest deflection was also measured with a string potentiometer mounted to the anterior of the sled tank and attached to the subject s sternum. Intra-aortic fluid pressure was measured using two ultra-miniature, catheter pressure transducers (model number SPR-524, Millar Instruments, Inc., Houston, TX) inserted into the aortic arch via the carotid arteries. Following the observation of large intra-aortic pressures in tests , subsequent tests included pressure measurement in the trachea (tests 2.3 through 3.2) and in the esophagus (tests 3.1 and 3.2). 3. Results The signal data are summarized in Table 2 (including calculated values of the thoracic soft tissue injury criterion,

3 J. Forman et al. / Journal of Biomechanics 41 (2008) Table 2 Data results summary Test # Impact speed (km/h) Peak target sled accel. (g) a (CFC 60) 3 ms clip peak target sled accel. (g) a Peak midspine accel. (g) b (CFC180) 3 ms clip mid-spine accel. (g) b Peak sternum accel. (g) b (CFC 180) 3 ms clip peak sternum accel. (g) b Chest deflection (% of chest depth at 8th rib) (max/min) c Peak viscous criterion, VC (m/s) d / / / / / / / Avg / Std / Dev. 3.2 e /NM e NM e a X-axis acceleration. Negative indicates posteriorly directed acceleration. CFC ¼ Channel filter class. b Peak resultant acceleration. c Chest deflection is the motion of the sternum relative to the spine. Tests and 3.2, chest deflection was calculated through double integration of the differential spine/sternum acceleration; tests , chest deflection was measured with a string potentiometer. Positive chest deflection denotes chest expansion. d The peak product of normalized chest compression and chest compression velocity (Lau and Viano, 1986). e The sled system failed during this test resulting in a different acceleration pulse. The results of this test are not included in the average or standard deviation calculation. The chest compression and VC were not calculated due to failure of the instrumentation part way through the test. VC, Lau and Viano, 1986). Plots of subject resultant accelerations and chest deflection are included in Fig. 2. Maximum intra-aortic, tracheal, and esophageal pressures (where available) are included in Fig. 3. Differences were observed (qualitatively) in the spinal accelerations (intertest) and spine and sled accelerations (intra-test), likely due to vibrations in the sled (the sled accelerometer was a damped accelerometer, whereas the subject accelerometers were not). These differences are considerably reduced with increased filtering (Table 2), and the peak sled accelerations are comparable to the sustained-plateau values for spinal acceleration (Fig. 2). During test 3.2 the impact attenuator placed at the interface between the bullet and target sleds failed. This resulted in a subject acceleration of greater magnitude and shorter duration than the other tests, and precluded the measurement of chest compression for that test. The maximum acceleration and pressure measurements for that test have been excluded from the calculated averages and standard deviations in Table 2. Soft tissue injury results are included in Table 3. No test in this study produced identifiable injury to the thoracic aorta. The atherosclerotic condition of each subject s aorta was qualitatively graded at autopsy (Table 3). In test 1.2 it was not possible to either identify or exclude aortic injury due to the presence of severe atherosclerosis with calcification and ulceration (test 1.2 was thus removed from this analysis). In tests 1.1, 1.3, and 1.4, artifactual tears were observed in the abdominal aorta (caused by Foley catheter balloons used to occlude the abdominal aorta). These consisted of small, partial-thickness tears in the intima and media, and were not observed in subsequent tests ( ) where the abdominal aorta was not occluded. 4. Discussion This study did not generate any thoracic vascular trauma in human cadavers exposed to chest accelerations as high as 211 g (CFC, 180), though minor organ trauma consistent with an inertial mechanism was observed. The lacerations of the splenic capsule at the hilar attachment (test 1.3) and the small bowel mesentery (tests 1.2 and 1.3) occurred at points of connection that may have experienced stress from inertial movement of their respectively tethered organs. For each of these cases, it is possible that postmortem degradation of the viscera may have weakened the tissues, resulting in an underestimation of their injury tolerance. Considerable atherosclerosis in tests 1.4, 2.3, and 3.1 likely decreased the tolerance of those subjects to aortic injury (Viano et al., 1978), although those tests did not result in injury despite this probable increased susceptibility. Atherosclerosis may have also affected variations in the measured intra-aortic pressures (due to a change in the extensibility of the aorta), although the nature of such an effect is less clear (Fig. 4). In the study of injury in complex environments where multiple mechanical factors may contribute to, or result from, loading of the body, careful distinctions must be made between causes of injury and correlations with injury when interpreting experimental and epidemiological observations. In the study of aortic injury, thoracic acceleration is often suggested as the cause of injury

4 1362 J. Forman et al. / Journal of Biomechanics 41 (2008) Fig. 2. Chest acceleration and deflection time histories. All accelerations are resultant accelerations filtered to CFC 180. The chest deflection signals measure the motion of the sternum relative to the spine, and have been truncated to avoid integration error accumulation. Tests and 3.2: chest deflection was measured through double integration of the differential spine and sternum accelerations (each filtered at CFC 180). Tests : chest deflection was measured with a string potentiometer connecting the sternum to the anterior surface of the containment tank (filtered at CFC 600). This was corrected with double-integration of the differential sled and tank accelerations (CFC 60 and CFC 180, respectively). Positive deflection denotes chest expansion. secondary to either a perceived or a quantified correlation with injury (Cavanaugh et al., 2005; Shkrum et al., 1999). Such correlations, however, are not sufficient criteria for establishing injury causation, as acceleration may serve simply as a surrogate measure of the impact severity. The relationship between impact severity and injury outcome often changes when different methods of applying load to the body are employed. The subjects of this study were restrained to minimize thoracic deformation, resulting in no aortic injury despite sustaining very large accelerations. The realization that high thoracic accelerations may be tolerated if chest compression is minimized may prompt the development of vehicle restraints that minimize thoracic compression by loading the relatively stiff structures of the pelvis and the clavicles. Other studies have investigated the cause of aortic injury by isolating other possible mechanisms of injury (e.g. intraaortic pressure Bass et al., 2001; distraction of the arch Hardy et al., 2006; chest compression Kroell and Schneider, 1974; Nusholtz et al., 1985; Viano et al., 1978; Voigt and Wilfert, 1969). Although none of those studies prove that their respective injury mechanisms are the cause of aortic injury in the field, they do show that those mechanisms can cause aortic injury in the absence of other factors. In contrast, this study shows that, in the absence of chest compression, posteriorly directed acceleration alone up to the magnitudes studied here is not sufficient to cause aortic injury in human cadavers. These results are consistent with the limited in vivo acceleration tolerance data available in the literature (Melvin et al., 1998; Mertz

5 and Gadd, 1971; Stapp, 1970). Aortic injury may occur, however, at higher magnitudes of acceleration. Thoracic acceleration may also contribute to aortic injury in combination with chest compression and fluid pressure. Preliminary computational studies have suggested that thoracic acceleration may non-linearly magnify maximum stresses in the aortic wall at the isthmus when combined with an intra-aortic pressure pulse and posterior displacement of the heart (Lee and Kent, 2007). Despite the lack of aortic injury in the cadaver tests presented here, internal pressure measurements exceeded the expected burst pressure of the aorta (approximately 101 kpa, Bass et al., 2001) in many cases. The pressureinduced stress in the aortic wall depends on the difference between the pressure inside the aorta and the pressure in J. Forman et al. / Journal of Biomechanics 41 (2008) Fig. 3. Maximum pressures measured inside the aorta, trachea, and esophagus compared to the aortic pressure corresponding to a 50% risk of injury as reported by Bass et al. (2001). The differential pressure is the maximum difference between the intra-aortic and tracheal pressures. Fig. 4. Examples of visceral injuries. Top Lacerations of the splenic capsule from test 1.3. Bottom Superficial laceration of the small bowel mesentery from test 2.3. Table 3 Summary of soft tissue injury results from autopsy Test no. Aortic injuries Atherosclerosis Other soft tissue injury Mild to moderate None 1.1 Two artifactual abdominal aortic tears just superior to the iliac bifurcation a 1.2 Not identifiable b Very severe Superficial laceration of the pericardium just deep to a rib fracture; multiple superficial lacerations of the left visceral pleura; superficial laceration of the small bowel mesentery 1.3 One artifactual abdominal aortic tear just superior to the iliac bifurcation a Moderate Hemorrhage of the posterior pericardium c, three lacerations of the splenic capsule at the hilar attachment 1.4 One artifactual abdominal aortic tear just Moderate to severe None superior to the iliac bifurcation a 2.1 None Moderate Multiple superficial lacerations of the visceral pleura in the right lower lobe 2.2 None Moderate Two superficial pleural lacerations on the posterior, left lower lobe 2.3 None Severe Superficial mesenteric laceration located 10 cm proximal to the ileocecal junction 3.1 None Moderate to severe None 3.2 None Mild to moderate None a These tears were caused artifactually by Foley catheters used to occlude the abdominal aorta. b The presence of severe atherosclerosis with calcification and ulceration precluded the identification of injury to the aorta. c This injury could not be identified on either the pre-test or post-test CT scans, and thus could not be ruled out as pre-existing. It is possible that this may have resulted from resuscitation efforts.

6 1364 J. Forman et al. / Journal of Biomechanics 41 (2008) the surrounding environment. These tests resulted in extraaortic pressures (tracheal and esophageal pressures) comparable to, or exceeding, the pressures measured inside the aorta (Fig. 3). Although it is unknown to what extent these pressures represent the pressure environment immediately surrounding aorta, if the tracheal pressure is comparable to the pressure in the thoracic cavity then the peak differential pressure across the aortic wall may have met the Bass et al. 50% risk of injury in test 3.2, but may have remained well below that threshold for tests 2.3 and 3.1. These results suggest that in a dynamic environment there are pressure gradients throughout the thorax. The characteristics of these gradients are likely complex due to the inhomogeneity of the thoracic structure and it s contents (as evidenced by the large variation in pressures measured in this study), and the extra-aortic pressure within the thorax may be as large in magnitude, or larger, than the pressure inside the aorta. This suggests that considering intra-aortic pressure alone is not sufficient to determine the state of pressureinduced stress in the walls of the aorta. Instead, thoracic computational models and future experimental studies that include intra-aortic pressure as an aortic injury mechanism should also consider extra-aortic, thoracic pressure. 5. Conclusions Nine high-acceleration cadaver sled tests were performed resulting in rearward X-axis sled accelerations up to 98 g, spinal acceleration as high as 211 g, and thoracic deflections less than 11% with no observed injury to the thoracic aorta. During half of these tests, the intra-aortic fluid pressure exceeded the expected injury tolerance (Bass et al., 2001), but the differential pressure across the aortic wall may not have exceeded the injury tolerance (in most cases) due to simultaneous increases in extra-aortic pressure. The determination of the true tolerance of the aorta to thoracic acceleration requires further investigation at higher accelerations, however the accelerations imposed here were not sufficient to cause aortic injury in the cadavers tested. Conflict of interest statement Jason Forman, Stephen Stacey, Jay Evans and Richard Kent have no conflicts of interest to disclose. Acknowledgments The authors would like to acknowledge the assistance of Jason Mattice, Michelle Oyen, and Peter Matthews Rurak. This study was supported by a grant from The Alliance of Automobile Manufacturers. The opinions, findings, and conclusions expressed in this publication are those solely of the authors and are not necessarily those of The Alliance or its members. References Baque, P., Serre, T., Cheynel, N., Arnoux, P.J., Thollon, L., Behr, M., Masson, C., Delotte, J., Berdah, S.V., Brunet, C., An experimental cadaveric study for a better understanding of blunt traumatic aortic rupture. Journal of Trauma 61, Bass, C.R., Darvish, K., Bush, B., Crandall, J.R., Srinivasan, S.C.M., Tribble, C., Fiser, L., Tourret, L., Evans, J.C., Patrie, J., Wang, C., Material properties for modeling traumatic aortic rupture. Stapp Car Crash Journal 46. Cavanaugh, J., Koh, S.-W., Kalehonkar, S.L., Hardy, W.N., An Analysis of Traumatic Rupture of the Aorta in Side Impact Sled Tests. Society of Automotive Engineers, Warrendale, PA Paper No Forman, J.L., Kent, R.W., Bolton, J.R., Evans, J.C., A Method for the Experimental Investigation of Acceleration as a Mechanism for Aortic Injury. Society of Automotive Engineers, Warrendale, PA Paper No Hardy, W.N., Shah, C.S., Kopacz, J.M., Yang, K.H., Van Ee, C.A., Morgan, R., Digges, K., Study of potential mechanisms of traumatic rupture of the aorta using in situ experiments. Stapp Car Crash Journal 50, Kroell, C.K., Schneider, D.C., Impact tolerance and response of the human thorax II. In: 25th Proceedings, Stapp Car Crash Conference. Lau, I.V., Viano, D.C., Viscous criterion bases and applications of an injury severity index for soft tissues. In: 30th Proceedings, Stapp Car Crash Conference. Lee, S.-H., Kent, R.W., Blood flow and fluid-structure interactions in the human aorta during traumatic rupture conditions. Stapp Car Crash Journal 51, Lundevall, J., Traumatic rupture of the aorta, with special reference to road accidents. Acta Pathologica et Microbiologica Scandinavica 62, McGwin Jr., G., Metzger, J., Moran, S.G., Rue III, L.W., Occupant- and collision-related risk factors for blunt thoracic aorta injury. Journal of Trauma 54, Melvin, J.W., Baron, K.J., Little, W.C., Gideon, T.W., Biomechanical analysis of indy race car crashes. Stapp Car Crash Journal 42, Mertz, H., Gadd, C., Thoracic tolerance to whole-body deceleration. In: 15th Proceedings, Stapp Car Crash Conference. Nusholtz, G.S., Kaiker, P.S., Bosio, A.C., Kirsh, M., Thoracic response to frontal impact. In: 29th Proceedings, Stapp Car Crash Conference. Richens, D., Field, M., Neale, M., Oakley, C., The mechanism of injury in blunt traumatic rupture of the aorta. European Journal of Cardio-Thoracic Surgery 21, Richens, D., Kotidis, K., Neale, M., Oakley, C., Fails, A., Rupture of the aorta following road traffic accidents in the United Kingdom , The results of the co-operative crash injury study. European Journal of Cardio-Thoracic Surgery 23, Shkrum, M.J., McClafferty, K.J., Green, R.N., Nowak, E.S., Young, J.G., Mechanisms of aortic injury in fatalities occurring in motor vehicle collisions. Journal of Forensic Sciences 44, Stapp, J.P., Voluntary human tolerance levels. In: Gurjan, E.S., Lange, W.A., Patrick, L.M., Thomas, L.M. (Eds.), Impact Injury and Crash Protection. C.C. Thomas, Springfield, IL. Viano, D., Haut, R., Golocovsky, M., Absolon, K., Factors influencing biomechaical respose and closed chest trauma in experimental thoracic impacts. In: 22nd Proceedings, Association for the Advancement of Automotive Medicine. Voigt, G.E., Wilfert, K., Mechanisms of injuries to unrestrained drivers in head-on collisions. In: 13th Proceedings, Stapp Car Crash Conference. Ziperman, H.H., Peel, H., Cromack, J.R., Arterial pressurization in cadavers in impact studies. In: 19th Proceedings, Association for the Advancement of Automotive Medicine.