Posttraumatic arthritis continues to be a major cause of long-term impairment after a displaced intraarticular fracture.28 The causes of these arthritic changes often are said to be multifactorial. However, of all the multiple factors involved, articular malreduction and cartilage impaction injury consistently have been implicated as having a major role in the development of posttraumatic arthritis.15,20
Displacement of the articular surface alters a mechanical environment of the joint. These loading alterations are thought to lead to cartilage degeneration.7,11,12 Cartilage impaction injury results in damage to the structural properties of the articular cartilage and a decrease in viability of chondrocytes.4,5,31 Unfortunately, very little is known about how these two variables interrelate. Basic science investigations to date have focused on osteotomies of a joint (simulated articular fracture) without impaction of the articular cartilage, or cartilage impaction without fracture of the underlying articular surface.4,5,16,17,31 Clinically, these two factors are inseparable, because fractures with more displacement and comminution are also the same fractures in which increased amounts of cartilage impaction injury can be anticipated.
This study represents the first step in beginning to approach this problem by developing an in vivo model of a displaced articular fracture. Previously, we investigated acute changes of the mechanical environment of the hip after acetabular fracture in cadavers.21–23 We have chosen to expand our work with the study of injury to the dorsal wall (equivalent to the posterior wall in humans) of the acetabulum of a goat for our model. This study reflects two distinct components of a development of this model: first, the development of a bench top model to measure intraarticular loading conditions; and second, it represents the first set of in vivo survival data of this animal model.
MATERIALS AND METHODS: DEVELOPMENT OF BENCH TOP MODEL
The entire pelvis and articulated femurs were harvested from five fresh adult female Nubian goat cadavers Both hips were tested in each specimen for a total of 10 specimens. The specimens were harvested with a layer of periosteum and muscle insertions and ligaments, and the capsules of both hips were intact. The specimens then were preserved at −20°C until the time of testing. The specimens were kept moist throughout testing with normal saline solution at room temperature.
Each specimen was attached to a loading jig that provided secure and reproducible positioning in an Instron TM 1122 materials-testing system (Instron, Canton, MA) (Fig 1). The distal portion of each femoral shaft was anchored in polymethylmethacrylate (PMMA). The femur and the hip being tested were mounted on a device, which allowed adjustable rotation in the coronal and sagittal planes. The entire femoral mount was on a translation table allowing unconstrained translation of the femur in the AP plane and the mediolateral plane relative to the pelvis, while holding rotation consistent. Two large AO (Synthes, Paoli, PA) external fixator rods were placed parallel through the sacral body. The sacrum, posterior portion of the ilium, and rods were potted in PMMA. This provided attachment of the pelvis to the upper portion of the loading jig.
The pelvis was mounted with the sacrum horizontal to simulate the normal posture of the goat at the midstance phase of gait. The position of the pelvis and femur during testing was based on the work of Page et al24 In the frontal plane, the femoral shaft was placed in 15% abduction relative to the pelvis; in the sagittal plane, the femoral shaft was placed in 20° flexion. The femurs were mounted in neutral rotation using the lesser trochanter as a landmark. The specimens initially were mounted with the joint capsule intact. The loading jig was marked to allow the specimen to be returned to the proper position after disarticulation. The entire capsule and its insertion into the labrum were removed. A continuous cable was used to simulate the major muscle groups surrounding the hip including the hamstrings, the hip abductors, and the hip flexors (Fig 1). At each of the muscle insertions about the pelvis, an eyebolt was placed. The cable was attached to a load cell positioned at the posterior aspect of the distal femur and routed to the ischial tuberosity, to the greater trochanter, to the anterior iliac spine. The cable was tensioned to 50 N before loading of the hip. This tension was sufficient to maintain joint stability throughout loading. The load cell allowed us to monitor and control tension in the cable during loading to ensure consistent loading from trial to trial.
Pressure patterns were recorded with the Fuji Prescale Film (C. Itoh Corporation, New York, NY). The specimens were kept moist throughout testing with normal saline solution at room temperature. The specimens were loaded at 0.4 mm per second to 200 N for each of three conditions: (1) with the acetabulum intact; (2) with a simulated fracture of the posterior wall (with the fragment removed), and (3) after reduction and internal fixation of the fragment. These three conditions subsequently will be referred to as intact, fractured, and repaired. Loads were maintained for 30 seconds. The load of 200 N corresponds to approximately 40% of body weight, which is the load through the hip of a quadruped during the stance phase of gait.13 Three loading cycles were done for each test condition. Pretensioning of the cable system to 50 N did not result in any recorded articular contact on the Fuji film.
The simulated fracture of the posterior wall began at 10° posterior to the acetabular vertex along the rim of the acetabulum, and continued along the arc of the acetabular rim to 60° posterior to the vertex. The fragment extended medially to include 50% of the width of the retroacetabular surface, as measured from the posterior acetabular rim to the greater sciatic notch, and the entire width of the articular surface throughout the 50° arc was removed (Fig 2). The fracture was made with the use of holes drilled to the subchondral bone, but not through the articular cartilage. The drill holes were connected with an osteotome to create a realistic fracture in the articular surface.
The fracture subsequently was repaired by one of us (EAC), under ideal bench-top conditions with excellent access and visibility. A mini-fragment tubular plate with 2.7-mm screws was used in a buttress configuration for internal fixation of the posterior wall. Two bicortical screws were placed proximal and distal to the fragment. The articular surface was assessed visually and with digital palpation at the time of the reduction. All reductions were anatomic.
Low-range Fuji film, which is sensitive to pressures between 2.5 and 10.0 MPa, was used for intraarticular pressure measurement on the basis of a previous report of hip contact pressures.1 The film was cut into star-shaped patterns that were tacked to the surface of the femoral head with photographic mounting adhesive. Before the film was applied, the head of the femur was covered with a thin layer of latex (Trojan Latex Condom; Carter Wallace, New York, NY) to protect the film from the moisture on the articular surface. When placed over the femoral head, the star shape made a near perfect sphere for excellent coverage of the femoral head. A second layer of latex then was placed over the film. The resting thickness of the latex-film-latex layer was 300 μm. This thickness decreased to 250 μm under compressive joint load.
Before loading, 1-mm holes were drilled through the articular cartilage and the underlying bone of the acetabulum at two reproducible points on the anterior and posterior sides of the acetabular vertex, and midway between the acetabular fossa and rim. During loading, blunt probes inserted through these holes were used to mark the pressure sensitive film (low-range Fuji Prescale Film) to establish orientation of the film with respect to the acetabulum. The outline of the acetabular rim and the vertex of the acetabulum, along the rim, also were marked on the film. This method is similar to that used in tests of human hips.
After loading, the pressure-sensitive film was removed from the femoral head and aligned on paper, and laminated in plastic to facilitate handling. All films were analyzed with a three-stage process of digitizing, filtering, and measurement. Digitizing was done with a flatbed scanner (Abaton Scan 300/GS, Everex Systems, Fremont, CA) at a spatial resolution of 144 dpi and 256 levels of gray. Filtering and measurement were done with NIH Image software (Wayne Rasband, Research Services Branch, National Institute of Mental Health, Bethesda, MD) on a Power Macintosh computer (Apple Computer, Cupertino, CA). We used a calibration scale for Fuji Prescale film developed in our laboratory under similar environmental conditions as in this study.23
Three separate film patterns were recorded for each hip in each experimental condition. The repeat patterns were evaluated qualitatively for consistency at the time of the test; any test that had obvious disagreement was rejected, and the test was repeated. All measurements represent the mean of the three patterns. We chose these runs, based on our experience with Fuji Film in the hip in humans.
The images were divided into three regions on the basis of the reference marks established during testing: cranial, dorsal, and caudal. These regions correspond to the anterior wall, the superior aspect, and the posterior wall of the acetabulum in the human, respectively. This division separated the acetabulum approximately into thirds. For each region, contact area, mean pressure, and maximum pressure were measured. The load contained within each region was calculated by multiplying the area by the mean pressure. Parameters within each region of the acetabulum were analyzed by one-way repeated measures ANOVA (the factor was the condition and the levels were intact, fractured, and repaired). Probability values < 0.05 were considered significant. The results are given as the mean and standard deviation for each condition. The statistical analyses are based on data from the 10 hips.
The mean total contact area for the intact acetabulum was 137 (±38) mm2. For the intact and repaired conditions, area was concentrated in the caudal and cranial regions. The mean contact area decreased (p < 0.05) from 70 (±18) to 50 (±11) mm2 on the caudal region, and increased (p < 0.05) from 5 (±4) to 22 (±11) mm2 on the dorsal region of the acetabulum.
Anatomic reduction and fixation of the fragment of the posterior wall restored the total contact area and the distribution of contact within the acetabulum toward the intact state. The mean contact area for the total joint returned (p > 0.05) to 147 (±35) mm2, similar to the intact state. Likewise, the contact area for the three regions of the acetabulum showed no difference (p > 0.05) between the intact and repaired states.
There were significant changes in the mean and peak pressures in the dorsal region of the acetabulum after fracture of the posterior wall. The mean pressure increased (p < 0.05) from 1.9 (±1.5) to 3.7 (±0.9) MPa. The peak pressure increased (p < 0.05) from 2.8 (±2.3) to 6.1 (±1.9) MPa. The mean and peak pressures in the caudal region also showed changes in response to the fracture of the posterior wall. The mean pressure increased (p < 0.05) from 4.4 (±0.9) to 5.1 (±0.6) MPa. The peak pressure increased (p < 0.05) from 7.2 (±2.0) to 9.1 (±0.9) MPa. The mean and peak pressures in the cranial region showed no changes.
Reduction and fixation of the fracture fragment reversed changes in pressure caused by the simulated fracture. There were no significant differences in either mean or peak pressures in any of the regions of the acetabulum between the intact and repaired states. The mean load had a pattern that was similar to that of the pressure.
MATERIALS AND METHODS
Development of In Vivo Model
After completion of bench top testing, two groups of four adult female Nubian goats had a survival surgery to create a high-energy caudal wall acetabular fracture. Radiographs were obtained before surgery to ensure closure of epiphyseal plates, and that there were no hip abnormalities. For each goat, general anesthesia was induced in a veterinary operating suite with the animal placed in the lateral position. The right hip was approached through a posterior approach with tenotomy of the short external rotators to expose the caudal acetabular surface.6 A 1-cm2 template was placed on the caudal retroacetabular surface, 1-cm cranial to the inferior margin of the posterior wall. A 1.5-mm drill was used to outline the proposed fracture fragment, drilling only through the outer cortex. A small capsulotomy was made inferior to the intended fracture, and a small elevator was placed in the joint to protect the femoral head from inadvertent drill penetration. The capsular attachment to the fragment was preserved. An osteotome was used to complete the outer cortex scoring of the fracture fragment. The fracture was not completed through the articular surface at this time.
A 4-mm Schanz pin was placed in the proximal right femur perpendicular to the shaft. An accelerometer probe was firmly attached to the Schanz pin with external fixation clamps. The wound was covered with a sterile towel during the impaction. The animal then was turned supine and the operating table was placed in 15° of reverse Trendelenburg to best orient the direction of impaction to the prepared fracture fragment. The drop-tower impaction device then was placed on the distal femur in line with the femoral shaft with the hip positioned at 90° flexion so the femoral shaft was perpendicular to the dorsal wall. The drop-tower device consists of a bearing surface applied to the goat’s flexed knee with a gel pad to ensure a uniform impact. Ten kilograms of diving weights were mounted to slide down a 3 foot-½ inch diameter shaft, down to a base plate. A switch for the accelerometer was mounted to initiate recording when the weight was dropped. The impaction simulates a dashboard impact mechanism (Fig 3). The accelerometer and impactor device were removed, and the animal was returned to the lateral position. The hip then was inspected and, if necessary, the fracture was completed with an osteotome (this occurred in one hip). No hips dislocated and all were stable through a full ROM.
After creation of the fracture, four goats had anatomic reduction and internal fixation of the fracture. A six-hole, 2.7-mm reconstruction plate was contoured to provide a buttress for the wall fragment. The plate was secured with two, 2.7-mm bicortical screws above and below the fracture. Care was taken to ensure that all screws were extraarticular, and that the fixation was stable with an anatomic reduction. In the other four animals, the fragment was left displaced. In four goats, the posterior wall fragment was left in a displaced position. The wounds were closed with repair of the external rotators, and the animals were returned to their pens. Prophylactic antibiotics were used preoperatively. Analgesics were administered as needed. The goats were ambulating normally at approximately 2 weeks postoperatively, and were returned to pasture.
The animals were sacrificed at 90 days (range, 89–91 days). Intravenous doses of oxytetracycline (30 mg/kg) were administered 9 and 2 days before sacrifice to label actively remodeling bone. Anteroposterior radiographs were completed after harvest. The pelves and attached femurs were harvested with a layer of periosteum, ligaments, capsule, and muscle insertions intact. Mechanical testing was done on the day of harvest. The internal fixation group was tested before and after hardware removal. The specimens were kept moist throughout testing with normal saline at room temperature. Biomechanical testing was done with the methods described previously on bench-top testing.
After biomechanical testing, both acetabulums were cut from the pelvis and placed in ethanol. The acetabulum was bisected in the coronal plane into anterior and posterior halves, dividing the acetabulum at the midportion of the fracture fragment. The articular surface was marked for orientation to map the Fuji film patterns to the histologic analysis. The posterior half of the acetabulum was decalcified, dehydrated, and embedded in paraffin. Seven-micron serial sections in the sagittal plane crossing the fracture lines were cut and stained with hematoxylin and eosin and safranin-O fast green.
Six sections of each acetabulum were evaluated: six stained with hematoxylin and eosin and three stained with safranin-O. The sections were rated using the Mankin Histologic Rating System18 (Table 1). Two independent reviewers determined one rating for each of the specimens. The experimental group was assigned two grades, one for each side of the fracture, intact and wall segments.
Histologic data were evaluated by ANOVA to compare the differences between groups. For comparison between experimental and control hips, the intact and dorsal wall fracture groups were averaged to compare with controls. For comparison between experimental groups, direct comparison between intact and wall scores was done.
In Vivo Model
After the drop tower impact loading, the fracture was complete in seven of eight hips. The fracture created was one piece of dorsal wall with capsular attachment material. The fracture was displaced easily once completed. Displaced fractures had a minimum of 5 mm of articular displacement.
All animals had normal gait before harvest. There were no postoperative complications. All fractures healed. The location of the fracture was consistently within the dorsal region, near the acetabular vertex. The intraarticular portion of the fractures was slightly smaller than the bench-top osteotomy.
The appearance of the acetabula treated with reduction and fixation showed anatomic position of the fracture fragment with no intraarticular step-off (Fig 4). The fracture lines were visible with what appeared to be a fibrocartilaginous line. The fragment articular cartilage appeared normal.
The hips with the fracture fragments left displaced healed in the displaced position. The articular surface filled in with what appeared to be fibrocartilage with no articular step-off. The remainder of the articular cartilage appeared grossly normal.
The radiographs from the internal fixation group showed normal appearing hips. In two goats, there was evidence of screw loosening, but no broken or intraarticular hardware. In the displaced group, the fracture fragment healed in the displaced position lateral to the normal rim of the acetabulum. There were radiographic signs of osteoarthrosis (joint space narrowing and sclerosis) in the displaced posterior wall hips. The accelerometer data were used to calculate force transmission data. The proximal femur recorded consistently. When done over the arc of the joint, the contact stress delivered was between 60 MPa (recorded constant on Fuji Film) and 40 MPa (estimated whole joint area).
The contact area for the normal goat acetabulum is concentrated in the cranial and caudal facets, with very little loading in the dorsal region (Table 2). The fractured, reduced, and plated side showed 63.8 mm2 in the cranial, 58 mm2 in the caudal, and 0.4 mm2 in the dorsal regions. After plate removal, the mean area was 65.9 mm2 in the cranial, 61.2 mm2 in the caudal, and 0.38 mm2 in the dorsal regions. Hardware removal did not substantially alter the contact area.
There was no difference in contact area between the repaired and displaced group for any region.
There was no major difference in mean contact pressure between the displaced and repaired conditions (Table 2). However, there was a nonsignificant trend toward increased mean pressure in the dorsal region of the no repaired group compared with the repaired group.
Peak pressures in reduced and repaired hips, nonreduced and nonrepaired hips, and control hips were similar (Table 2).
There was no consequential difference in maximum pressure between the displaced and repaired conditions. However, there was a trend toward increased peak pressure in the dorsal region of the displaced group compared with the repaired group.
Results of articular load (calculated as mean contact pressure × contact area) were similar to results of contact areas with no significant alterations.
There was a difference (p < 0.05) in the mean Mankin scores18 from the control hips to the experimental hips, regardless of treatment (Table 3). There was evidence of increased chondrocyte cloning and clefts in the cartilage matrix in the intact portion of the experimental hips, with or without reduction. There was no evidence of chondrocyte necrosis, or necrosis of any cartilaginous or fibrous tissue seen in any specimen 90 days after fracture. Comparison of histologic grades between experimental groups of reduced and displaced posterior wall segment was almost significant (p = 0.06) (Figs 5, 6). Displaced fractures had a markedly abnormal structure with near complete destruction of the cartilaginous surface in the displaced segment, and blood vessels crossing the tidemark (Figs 7, 8). There also was a nonsignificant trend toward decreased safranin-O staining in specimens that were not reduced.
A power analysis done on the histologic data indicated that five animals per experimental group would be adequate to reproducibly show a significant difference in Mankin grades of four points or more between reduced and displaced posterior wall segments, and for histologic analysis.
The importance of articular reduction in preventing posttraumatic arthritis has been emphasized for many years.12,20 Advances have identified articular impaction as a possible independent factor in the development of posttraumatic arthritis.20 The development of animal models with realistic injury mechanisms is important in this area of investigation.
Our study is a first attempt to develop such a model. The Nubian goat seems to have several positive aspects as a model for study of the posterior wall acetabular fracture. The animals are homogenous and easily obtainable. The femoral head and acetabulum are large enough for application of pressure-sensitive film. Moreover, unlike dogs, the goats do not have a predisposition for hip dysplasia or other disorders of the hip, which may lead to arthritis.13 Bergmann et al3 reported a biomechanical comparison of forces and muscle actions around the hip in quadrupeds versus humans. They reported the quadruped was an adequate model for hip function. Bergmann et al3 also showed hand-limb mechanics were similar for dogs and goats. Page et al24 also reported that quadrupeds can be used to model total hip arthroplasty. The injury mechanism is more realistic involving articular impaction and fracture creation. The dorsal wall fracture created is one articular piece.
Although this model shows promise, there are numerous limitations. Displaced posterior wall acetabular fractures in the Nubian Goat resulted acutely in noteworthy changes in the intraarticular load distribution in the hip. Unlike the hip in humans, these changes were reversed with reduction of the articular fragment and internal fixation. Olson et al21–23 observed this normal loading pattern is lost after an acetabular fracture is created experimentally.
The explanation for the acute restoration of normal load distribution in the goat after fixation is unclear. There are several possible contributing factors. The stiffness of the fixation used may be noticeably greater relative to the forces applied to the hip of a goat. The size of the fixation implant also is relatively larger when compared with the size of the posterior wall fragment in the goat than in the human. The geometry of the pelvis of the goat also may prevent or inhibit bending deformation as postulated in the hip in humans. Which of these factors predominate is unclear.
Changes in intraarticular load transmission across the hip have been shown acutely in cadaveric models of acetabular fractures in the human hip and the goat.21–23 It has been presumed that these changes would continue to be detectable in a short-term survival. However, in Part II of our experiment, we did not observe significant differences in the load distribution in the hip between control hips, repaired fractures, or displaced fractures. Again, the exact explanation for this observation is unclear. The biomechanical testing was done in one position of the hip. Different pressure alterations may have been reported with varying positions of the hip. We observed that the experimental articular defect was slightly smaller than the 10 × 10-mm defect made in the cortical bone of the retroacetabular surface of the hip of the goats. The displaced fractures healed with a firm fibrous tissue. This tissue filled the articular defect and blended with the surrounding articular cartilage grossly. This may represent an intermediate step in an attempt at cartilage remodeling.
An advantage of pressure-sensitive film is that it can be used to measure joint-contact areas and pressures simultaneously.22 Other methods that have been used in investigations of the hip can be used to measure contact area but not pressure (dye studies and castings) or to measure pressure locally but not globally (piezoelectric transducers and mechanical transducer).1,6,11,12 Fuji film, however, is not free of technical problems. The dye transfer basis of the film, and the need to digitize and process image data, can lead to measurements that are more variable than those made with some transducer-based techniques. The latex-film-latex barrier increases the diameter of the femoral head, in turn altering the head-acetabulum geometric relationship. Pressures measured are in a range, not continuously, to a maximum limiting value. Accurate measurements also require application and maintenance of load during a time that is substantially longer than the stance phase of gait.
There is creep of the cartilage and subchondral bone during the test, although the exact extent is difficult to evaluate. The expected results of creep, however, do not weaken our conclusion about the changes associated with a fracture and repair. Creep of cartilage would act to more evenly distribute pressure within the joint than in the instantaneous-load state. Within the context of our analysis, this is a conservative error. It would tend to produce differences in contact area and pressure that are less, not more, than the instantaneous-load values. A more sophisticated method of measurement of pressure is needed to address this issue fully.
Histologic analysis showed that the reduced fractures healed the bony injury; however, they were left with a residual cleft of unhealed articular surface at the fracture site. This cleft was similar to the healing response reported with a full-thickness chondral injury in an unfractured bone.2,19 There was evidence of increased chondrocyte cloning and surface fissures in the articular surface in these reduced specimens.
Displaced specimens showed gross destruction of the articular surface with a significantly higher Mankin score, indicating an aggressive degenerative process of the articular surface. Interestingly, the aggressive degenerative changes occurred predominantly on the depressed portion of the articular surface. This portion of the articular surface would be expected to be unloaded relative to the residual intact portion of the acetabulum. The defect of the displaced articular segment was filled with an aggressive fibrous healing response. There was no evidence of cartilage or fibrous tissue necrosis seen in either the reduced or displaced specimens in this study. We did not specifically assess for chondrocyte viability. We did not observe cartilage remodeling similar to that reported by Llinas et al17 and Lefkoe et al16 in this experimental model.
We were able to generate an impact in an attempt to simulate a dashboard impact. Considering the total contact area (as measured with Fuji film) of the acetabulum of the goats, we calculated the impact pressure to be approximately 60 MPa. This is a very large impact, greater than used in many cartilage impact studies.
The blunt impact to articular cartilage has been discussed as an etiology to posttraumatic arthrosis in acetabular fractures. Borelli et al4,5 and Torzilli et al31 showed cartilage death at loads of 25 MPa in a bovine cartilage model. Repo and Finlay27 observed chondrocyte death and articular surface fissuring at pressures of 25 N/mm2, or 30% strain. Donohue et al9 evaluated impact loads on canine cartilage, observing decreased proteoglycan, increased water content, and increased collagen fiber width. Thompson et al29,30 applied loads of 200 N to patellofemoral joints in canines. They showed fractures in the zone of calcified cartilage and displacement in subchondral bone, but no change in the articular surface. However, electron microscopy revealed fissures and clefts at the surface. At 1 year, the fissures remained but the calcified zone and subchondral bone had healed. Many other authors also have reported deleterious effects of impaction.2,7,8,10,14,25,26
One of the major problems for clinical researchers is the separation of confounding factors. Fractures with greater comminution of the articular surface are more likely to result in articular malreduction. Presumably, the comminution resulted from a more severe injury. This suggests that the articular impaction injury probably is greater. Comminuted fractures in general tend to have a poor outcome clinically. How can we separate the effects of articular impaction from the effects of articular comminution articular reduction? This is one of the most challenging questions facing researchers today. This study was a preliminary attempt at developing a model to begin to address some of these complex issues. Based on our preliminary data, extensive work is needed to validate this model as a method of reproducibly developing posttraumatic arthritis. Once validated, this model and others like it could be used to investigate this important area.
1. Afoke NYP, Byers PD, Hutton WC: Contact pressures in the human hip joint. J Bone Joint Surg 69B:536–541, 1987.
2. Atkinson PJ, Haut RC: Subfracture insult to the human cadaver patellofemoral joint produces occult injury. J Orthop Res 13:936–943, 1995.
3. Bergmann G, Siraky J, Rohlmann A, Koelbel R: A comparison of hip joint forces in sheep, dog and man. J Biomech 7:907–921, 1984.
4. Borrelli J, Burns ME, Ricci WM, Silva MJ: A method for delivering variable impact stresses to the articular cartilage of rabbit knees. J Orthop Trauma 16:182–188, 2002.
5. Borrelli J, Torzilli PA, Griglene R, Helfet DL: Effect of impact load on articular cartilage: Development of an intra-articular fracture model. J Orthop Trauma 1:319–326, 1997.
6. Brown TD, Shaw DT: In vitro contact stress distribution in the natural human hip. J Biomech 18:317–384, 1983.
7. Chrisman OD, Ladenbauer-Bellis IM, Panjabi M, Goeltz S: The relationship of mechanical trauma and the early biochemical reactions of osteoarthritic cartilage. Clin Orthop 161:275–284, 1981.
8. Cremer MA, Roslonier EF, Kang AH: The cartilage collagens: A review of their structure, organization, and role in the pathogenesis of experimental arthritis in animals and in human rheumatic disease. J Mol Med 76:275–277, 1998.
9. Donohue JM, Buss D, Oegema Jr TR, Thompson Jr RM: The effects of indirect blunt trauma on adult canine articular cartilage. J Bone Joint Surg 65A:948–957, 1983.
10. Ewers BJ, Weaver BT, Sevensma ET, Haut RC: Chronic changes in rabbit retropatellar cartilage and subchondral bone after blunt impact loading of the patellofemoral joint. J Orthop Res 20:545–550, 2002.
11. Hadley NA, Brown TD, Weinstein SL: The effects of contact pressure elevations and aseptic necrosis on the long-term outcome of congenital hip dislocation. J Orthop Res 8:504–513, 1990.
12. Huber-Betzer H, Brown TD, Mattheck C: Some effects of global joint morphology on local stress aberrations near imprecisely reduced intraarticular fractures. J Biomech 23:811–822, 1990.
13. Hulse DA, Root CR: Management of acetabular fractures: Long-term evaluation. Compend Contin Educ Pract Vet 2:189–197, 1980.
14. Jeffrey JE, Gregory DW, Aspden RM: Matrix damage and chondrocyte viability following a single impact load on articular cartilage. Arch Biochem Biophys 322:87–96, 1995.
15. Kellam JF: Articular fracture reconstruction. J Orthop Trauma 11:318, 1997.
16. Lefkoe TP, Walsh WR, Anastasatos J, Ehrlich MG, Barrach HS: Remodeling of articular step-offs. Clin Orthop 314:253–265, 1995.
17. Llinas A, McKellop HA, Marshall GJ, et al: Healing and remodeling of articular incongruities in a rabbit fracture model. J Bone Joint Surg 75A:1508–1523, 1993.
18. Mankin HJ: Current concepts review: The response of articular cartilage to mechanical injury. J Bone Joint Surg 64A:460–466, 1982.
19. Mankin HJ, Mow VC, Buckwalter JA, Iannotti JP, Ratcliffe A: Articular Cartilage Structure, Composition and Function. In Buckwalter JA (ed). Orthopaedic Basic Science. Biology and Biomechanics of the Musculoskeletal System. Rosemont, IL, American Academy of Orthopaedic Surgery 443, 1999.
20. Matta JM: Fractures of the acetabulum: Accuracy of reduction and clinical results in patients managed operatively within three weeks after the injury. J Bone Joint Surg 78A:1632–1645, 1996.
21. Olson SA, Bay BK, Chapman MW, Sharkey NA: Biomechanical consequences of fracture and repair of the posterior wall of the acetabulum. J Bone Joint Surg 77A:1184–1192, 1995.
22. Olson SA, Bay BK, Hamel A: Biomechanics of the hip joint and the effects of fracture of the acetabulum. Clin Orthop 339:96–104, 1997.
23. Olson SA, Bay BK, Pollak AN, Sharkey NA, Lee T: The effect of variable size posterior wall acetabular fractures on contact characteristics of the hip joint. J Orthop Trauma 10:395–402, 1996.
24. Page AE, Allen C, Jasty C, et al: Determination of loading parameters in the canine hip in vivo. J Biomech 26:571–579, 1993.
25. Quinn TM, Grodzinsky AJ, Hunziker EB, Sandy JD: Effects of injurious compression on matrix turnover around individual cells in calf articular cartilage explants. J Orthop Res 16:490–499, 1998.
26. Radin EL, Paul IL: Response of joints to impact loading: In vitro wear. Arthritis Rheum 14:356–362, 1971.
27. Repo RU, Finlay JB: Survival of articular cartilage after controlled impact. J Bone Joint Surg 59A:1068–1076, 1977.
28. Swiontkowski MF, Chapman JR: Cost and effectiveness issues in care of injured patients. Clin Orthop 318:17–24, 1995.
29. Thompson RC, Oegema TR, Lewis JL, Wallace L: Osteoarthritic changes after acute transarticular load. J Bone Joint Surg 73A:990–1001, 1991.
30. Thompson RC, Verner MJ, Griffiths HC, et al: Scanning electron microscopic and magnetic resonance imaging studies of injuries to the patellofemoral joint after acute transarticular loading. J Bone Joint Surg 75A:703–713, 1993.
© 2004 Lippincott Williams & Wilkins, Inc.
31. Torzilli PA, Grigiene R, Borelli Jr J, Helfet DL: Effect of impact load on articular cartilage: Cell metabolism and viability, and matrix water content. J Biomech 121:433–441, 1999.