A biomechanics study about ligaments injury in anterior posterior compression type Ⅱ pelvic injury

Background Anterior posterior compression(APC)type II pelvis fracture is considered to be a typical one caused by the destruction of pelvic ligaments, while the mechanism of ligaments injury is still controversial. This study aims to explore ligaments injury in APC type II pelvic injury. Method: Fourteen human cadaveric pelvis samples with sacrospinous ligament (SPL), sacrotuberous ligament (SBL), anterior sacroiliac ligament (ASL) and partial bone retaining unilaterally were made for this study. They were divided into hemipelvis restricted group and unrestricted group randomly. Record the separation distance of pubic symphysis and anterior sacroiliac joint, external rotation angle and force when the ASL ruptured. Continuing external rotation violence, observing bone and pelvic ligaments change. all of anterior sacroiliac ligament, sacrotuberous ligament and sacrospinous ligament failed.

A biomechanics study about ligaments injury in anterior posterior compression type Ⅱ pelvic injury

Orthopedics
Orthopedic Surgery injured. Second, when ASL fails, pubic symphysis displacement has a high fluctuation. Third, ASL rupture does not mean SBL and SPL are injured inevitably.

Introduction
Fractures of the pelvic ring are frequently seen, especially in polytrauma patients or geriatric patients with poor bone quality [1][2][3]. The anterior and posterior compression (APC) pelvic injury, which is a classic type, also known as the external rotation injury, is considered to be a typical one caused by the destruction of pelvic ligaments. On basis of the anatomical and biomechanical research, the anterior structures, including the pubic symphysis and the pubic rami, contribute approximately 40% to the stability of the pelvis. The remaining 60% of the stability of the pelvis is support with the posterior structures, including the sacroiliac joint [4]. APC type Ⅱ fracture is the intermediate type between stable and unstable external rotation pelvic fracture, the mechanism of ligaments injury of which is still controversial [5,6]. Thus, the injury mechanism and classification of APC type II need to be further refined, which is significant and representative. There have been some biomechanical articles about APC pelvic fracture treatment [7,8]

Main experimental reagents
Embedded specimens of denture powder (PMMA, type II) and denture water (Shanghai coral chemical plant).The volume ratio of powder and water is 2:1, 30 minutes at room temperature.

Experiment specimens
Inclusion criteria: (1) specimens were more than 18 years old; (2) specimens were not associated with pelvic tumor, previous fracture or sacroiliac joint fusion; (3) By naked eye and X-ray plain film observation, specimens were bilateral symmetry. Consequently, fourteen fresh adult cadaver specimens without anticorrosive treatment were collected for this study (5 females and 9 males, all specimens are from the Department of human anatomy, Wenzhou Medical University). We amputated Waist 5 and symphysis pubis, leaving the complete pelvic part retaining, including bone and ligaments (mainly include pubic symphysis ligament, SPL, SBL, anterior and posterior sacroiliac ligaments and sacroiliac interosseous ligament), then they were numbered 1-14. The specimens were soaked in normal saline, sealed and wrapped in double-layer plastic bags to prevent drying and dehydration when they were not used. They were cryopreserved at − 20 ℃ and ablated at room temperature 12 hours before the experiment. (Table 1, Fig. 1)

Experiment Method
The exact mechanism of pelvic external rotation injury in vivo was not clear, and the specific steps of original Tile test was also not clear. Therefore, we randomly divided the specimens into two groups, and made two kinds of pelvic external rotation injury test models. One group was the hemi-pelvic restricted group (1,3,5,7,9,11,13), the other one was the unrestricted group (2,4,6,8,10,12,14).
In both of two groups, 3-4 steel K-wires with diameter of 1 cm were inserted into the pelvis from the middle of the fifth lumbar vertebrae to the center of the sacrum, leaving about 5 cm longer than the fifth lumbar, and then the pelvis was embedded with fixator, so that it could be fixed on the biomechanical machine, and the biomechanical machine on the test side could rotate the sacroiliac joint to back direction, simulating the external rotation of the lateral pelvis. For figuring out the separation distance of pubic symphysis when ASL ruptured, a 0.3 cm steel plate was inserted into pubic symphysis to maintain a 0.3 cm reference separation point and calculate for separation distance. The plate was fixed on the base of the test machine, in the restricted group, two screws were fixed on the pelvic acetabulum test side, which restricted other directions movement expect internal and external rotation. But the unrestricted group did not fix with screws, which allowed vertical and sagittal movements (Fig. 2). The former group provided with a pure external rotation mode, while the latter group provided with a composite model (external rotation with flexion and extension displacement). As it was not clear which model was closer to the typical injury, we tested both of two models. In order to exclude the difference between the left and right sides, we also randomly selected half of the specimens to external rotate on the right side and the rest on left side.
Each pelvic specimen was designed with one side of the sacroiliac joint was restricted and the other side was unrestricted. Then we simulated APC injury mode, set the sacrum as the center and 250 nm as the maximum possible, and performed a pure external rotation of 2°/s.  Statistical analysis was performed using SPSS version 22.0. We analysed the data using paired sample t tests. Differences were considered statistically significant if P < 0.05.

Result
(1) When ASL ruptured (Fig. 3), between two groups, mean separation distance of pubic symphysis and anterior sacroiliac joint was 28.6 ± 8.4 mm to 23.6 ± 8.2 mm and 11.4 ± 3.8 mm to 9.7 ± 3.9 mm respectively, all P > 0.05. When we compared mean external rotation angle and force between two groups, mean data was 33.9 ± 5.5° to 48.9 ± 5.2° and 553.9 ± 82.6 N to 756.6 ± 41.4 respectively, all P < 0.01, there were significant differences, the unrestricted group showed more external rotation angle and force. Besides, in the restricted group, all samples associated with SPL and SBL injury, two samples had SBL and three specimens had SPL ruptured completely. In addition, two specimen suffered SPL and SBL ruptured completely at the same time. On the contrary, no distinct SBL or SPL injury was observed in unrestricted group. (Fig. 4,5, Table 3) Table 3 When anterior sacroiliac ligament rupture, comparison of sacrotuberous ligament and sacrospinous ligament between two groups Restricted Group Unrestricted Group 1 3 5 7 9 11 13 2 4 6 8 10 12 14 ligament is injured (including ligament is prolonged, torn, and the tension cannot be restored obviously); F = ligament ruptures completely (2) After ASL ruptured, we continued to external rotate externally. When samples underwent extremely external rotation force, there was still no SBL or SPL injury in unrestricted group. However, interosseous sacroiliac ligament, posterior sacroiliac ligament injury and slight sagittal rotation displacement of sacroiliac joint occurred (Fig. 6). When all of the three ligaments ruptured, the separation distance of pubic symphysis and sacroiliac joint was 42.0 ± 7.6 mm and 16.7 ± 4.2 mm respectively, which increased significantly when comparing to the distances ASL ruptured barely(P < 0.05). These data suggested that the pelvic ring was more unstable too. (Table 4, Fig. 7) Table 4 When all of the three ligaments fail, compared relevant mean distances to the mean distances when anterior sacroiliac ligament fails barely.

Discussion
Many researches about pelvic anatomy, morphology and biomechanics have been reported previously, which provides a good foundation for conducting pelvic ligaments study. The ASL covers the front of the sacroiliac joint, which is a wide and thin fiber bundle. As the ligament structure is relatively weak, which has little impact on maintaining the sacroiliac joint stable. The interosseous sacroiliac ligament, which is composed of lots of short and strong fiber bundle, filling much irregular joint space at the upper back of sacroiliac joint. The posterior sacroiliac ligament, also known as the dorsal sacroiliac ligament, dividing into shallow and deep layers. The posterior sacroiliac ligament and the interosseous ligament constitute the sacroiliac ligament complex, which forms the main mechanical resistance at the back of the sacroiliac joint [6][7][8][9][10][11][12][13][14][15]. Tile had conducted mechanics experiment, which confirmed that if the posterior sacroiliac ligament complex remained intact, even if other pelvic ligaments ruptured, there still would be no backward and up-down displacement of the hemipelvis [13]. But in that experiment, the control of rotation force of the posterior sacroiliac ligament complex was poor. In our study, we design two different models to support with better control. The SBL, which starts from the posterior inferior iliac spine, the external edge of the lower part of the sacrum and the upper part of the tailbone, and ends at the ischial node; the SPL is triangular, starting from the external edge of the sacrum and ending at the ischial spine [6,13]. which also had no effect on sacroiliac joint movement [17].
The APC pelvic injury is considered as a typical pelvic ligaments injury, which is classified into three types. Type I, one or both sides of pubic branches fracture or separation distance of pubic symphysis is less than 2.5 cm, and / or slight separation of anterior sacroiliac joint, but the anterior and posterior ligaments are intact; type II, the pubic symphysis separates more than 2.5 cm, all of the ASL, SPL and SBL rupture, and the sacroiliac joint separates slightly; type III, the hemipelvis separates completely, but there is no longitudinal displacement. The anterior and posterior ligaments are injured at the same time, and the sacroiliac joint is separate [6,17]. Among three types, APC type II is the intermediate type, which is between stable and unstable type, still with a lot of controversy about the ligaments injury at present. Thus, it is necessary to have further study about the mechanism of APC pelvic ligament injury.
For APC type I fracture, ASL is considered to be intact and could be taken conservative treatment; While for APC type II fracture, ASL is considered to be injured, operative treatment would be performed [18]. Besides, it is accepted that when the ASL ruptures, the separation distance of pubic symphysis must be more than 2.5 cm. Of note, the 2.5 cm distance is used in the Tile classification system to distinguish the chart type of book type B1 pelvic fracture as well [17]. According to our experiment outcome, when APC type II fracture associated with ASL failure, the mean separation distance is 2.38 cm, which is close to 2.5 cm. However, the distance of each specimen shows great differences, ranging from 1.4 cm to 4.0 cm. This result indicates that the ASL is likely to rupture when the displacement is more than 4.0 cm, and it may not happen between 1.4-4.0 cm.
Besides, we find that the sacrospinous and SBLs not always rupture completely in APC type II fracture.
In unrestricted group, all the pelvic floor ligaments are intact, while in the restricted group, among the seven samples, only two samples completely rupture. Slocumn and Terry reported that the sacrum was easier to rotate and bend after removal of SPL and SBL [19]. EM et al reported that the sacrum was unstable when the two ligaments were torn [20]. In unrestricted group, our result shows that rotational force enables the pelvis to rotate around the sacroiliac joint axis without damaging the two ligaments necessarily. Just as we find in the restricted group, the deformation force requires a vertical component, which is related to the axis of sacroiliac joint. Thus, our research provides much supplementary content to APC pelvic injury. The two ligaments can make the hemipelvis open by external rotation force, guiding hemipelvis to move downward. APC type II injury with intact SPL and SBL is a relatively stable pelvic injury, which can be cured without surgical intervention. Of course, we need more clinical outcomes to prove this viewpoint.
In summary, we mainly have the following findings: (1) pelvic external injury can divide into two situations: pelvis is restricted and pelvis is unrestricted. It is not certain which situation is closer to the real situation, but the results of which are different distinctly. When ASL ruptures, the angle and force of external rotation between the restricted group and the unrestricted group are significantly different, and the latter group is bigger. In the unrestricted group, when the ASL ruptures, no SPL or SBL damage is observed, while the outcome in the restricted group is opposite. We predict that the main reason is that in unrestricted group, the pelvis can rotate around the axis of sacroiliac joint without damaging the two ligaments while the models in restricted group eliminate the rotation and vertical displacement, which resulting in the two ligaments being injured easily. (2) When ASL fails, mean pubic symphysis distance is 2.38 cm, which is close to 2.5 cm, but the specific data of each samples is various, ranging from 1.4 to 4.0 cm. Thus, we think the 2.5 cm is not the critical standard to distinguish the APC type I from type II pelvic injury. We suggest that clinicians should be cautious to make clinical treatment decisions when the anterior sacroiliac and pelvic ring ligaments are damaged and the pubic symphysis displacement is more than 2.5 cm. Although we think that static and dynamic pelvic imaging can provide useful evidence, we can not make absolute clinical judgment on the ASL completely based on the separation distance of pubic symphysis, unless the distance is Our study also has a number of limitations. First, it is typically limited by cadaveric biomechanics, our cadaveric samples are a bit older when comparing to the pelvis of patients, which may lead to different characteristics; Our load-bearing model is not enough to represent this kind of injury accurately. However, when we try to relieve this impact through the restricted and unrestricted models, we find similar results. Second, when pelvis is under high-energy injury, the loading rate will be slower than expected. Considering the viscoelasticity of biological tissue, the ligament in particular, the load rate data in our injury model are various. Although these data is controversial in higher tension at present, it is generally accepted by the public that with the tension increasing, the ultimate load-bearing, stiffness and energy absorption of ligaments also increase, and the ligaments failure will be more likely to occur [21,22]. However, there is no literature about the relation between ligament elongation failure and stress. Although the strain rate of our acute pelvic injury model is not as big as we expected, we believe that the anatomical relationship between pelvic ligament and bone can provide useful information about APC injury [21][22][23]. The recorded force is not intended to reflect the actual force in pelvic injury. In fact, the distances to the specific ligament failure is the goal of the investigation, which is less affected by viscoelasticity. Third, the quantity of our sample is small, we need more cadaver specimens to support our theory.

Conclusion
First, pelvic external rotation injury can divide into two situations: hemipelvis is restricted and unrestricted, which results into two different outcomes. When ASL rupture, the unrestricted group external rotation angle and force is bigger and no obvious SBL or SPL appear, but in the restricted group, both of two ligaments are injured. Second, when ASL fails, pubic symphysis displacement distance has a high fluctuation, ranging from 1.4 to 4.0 cm. Third, when ASL is destroyed, we does not observe the inevitable destruction of the pelvic floor ligaments (SPL and SBL). A. X-ray image of the pelvis specimen before the experiment; B C Restricted hemipelvis model and unrestricted hemipelvis model  A-C In restricted group, ASL and SPL rupture.

Figure 5
A-C In the restricted group, all of ASL and the SPL and SBL rupture. Figure 6 A B In the unrestricted group, sample is extreme external rotation force, no obvious SPL or the SBL injury is observed, but the posterior sacroiliac ligament injury is seen.

Figure 7
A-C. In the restricted group, the X-ray images are taken when ASL, SPL and SBL rupture in turn, all of which shows that the separation distance of pubic symphysis increases apparently.