- Email: [email protected]
Dynamic impact force and association with structural damage to the knee joint: An ex-vivo study
Annals of Anatomy 196 (2014) 456–463
Contents lists available at ScienceDirect
Annals of Anatomy journal homepage: www.elsevier.de/aanat
Research article
Dynamic impact force and association with structural damage to the knee joint: An ex-vivo study Richard Brill a , Walther A. Wohlgemuth b , Harald Hempfling c , Klaus Bohndorf d , Ursula Becker e , Ulrich Welsch f , Alexander Kamp g , Frank W. Roemer h,a,∗ a
Department of Radiology, Klinikum Augsburg, Augsburg, Germany Department of Radiology, University of Regensburg, Regensburg, Germany c Berufsgenossenschaftliche Unfallklinik, Murnau, Germany d MR Center of Excellence, Department of Radiology, University of Vienna, Vienna, Austria e F. Hoffmann-La Roche Ltd., Basel, Switzerland f Department of Cell Biology, Ludwigs-Maximilians University, Munich, Germany g Center for Radiology and Nuclear Medicine, Murnau, Germany h Department of Radiology, University of Erlangen-Nuremberg, Erlangen, Germany b
a r t i c l e
i n f o
Article history: Received 9 February 2014 Received in revised form 7 July 2014 Accepted 29 July 2014 Keywords: Dynamic impact Knee Osteochondral injury Histology
s u m m a r y No systematic, histologically confirmed data are available concerning the association between magnitude of direct dynamic impact caused by vertical impact trauma and the resulting injury to cartilage and subchondral bone. The aim of this study was to investigate the association between dynamic impact and the resulting patterns of osteochondral injury in an ex-vivo model. A mechanical apparatus was employed to perform ex-vivo controlled dynamic vertical impact experiments in 110 pig knees with the femur positioned in a holding fixture. A falling body with a thrust plate and photo sensor was applied. The direct impact to the trochlear articular surface was registered and the resulting osteochondral injuries macroscopically and histologically correlated and categorized. The relationship between magnitude of direct impact and injury severity could be classified as stage I injuries (impact <7.3 MPa): elastic deformation, no histological injury; stage II injuries (impact 7.3–9.6 MPa): viscoelastic imprint of the cartilaginous surface, subchondral microfractures; stage III injuries (impact 9.6–12.7 MPa): disrupted cartilage surface, chondral fissures and subchondral microfractures; stage IV injuries (impact >12.7 MPa): osteochondral impression, histologically imprint and osteochondral macrofractures. The impact ranges and histologic injury stages determined from this vertical dynamic impact experiment allowed for a biomechanical classification of direct, acute osteochondral injury. In contrast to static load commonly applied in ex-vivo experiments, dynamic impact more realistically represents actual trauma to the knee joint. © 2014 Elsevier GmbH. All rights reserved.
1. Introduction The zonal structure of hyaline articular cartilage and the adjacent subchondral bone has been a topic of histologic evaluation for more than 80 years, when Benninghoff described and differentiated four distinct cartilage zones (Benninghoff, 1925). The
∗ Corresponding author at: Associate Professor of Radiology Department of Radiology University of Erlangen-Nuremberg Maximiliansplatz 1 91054 Erlangen, Germany. Tel.: +49 9131 8536065; fax: +49 9131 8536068. E-mail addresses: [email protected], [email protected], [email protected] (F.W. Roemer). http://dx.doi.org/10.1016/j.aanat.2014.07.007 0940-9602/© 2014 Elsevier GmbH. All rights reserved.
advent of electron microscopic studies has subsequently markedly refined our knowledge of the cartilage ultrastructure (O’Connor et al., 1988). In addition, the biomechanical behavior of cartilage and bone as an osteochondral unit has been extensively explored (Radin et al., 1970; Spahn and Wittig, 2003; Räth et al., 2008; Lochmüller et al., 2008). By employing basic static material testing equipment these studies investigated the deformation caused by static load applied to cartilage and bone specimens. Through these experiments the elastic, viscoelastic and plastic properties of cartilage could be determined. The elasticity modulus, the rigidity, and resistance to fracture could be defined and the time dependence of viscoelastic deformation was established. In more recent in-vivo and ex-vivo animal experiments single severe and minor
R. Brill et al. / Annals of Anatomy 196 (2014) 456–463
457
repetitive impact trauma was applied to the knee joint. Consequent injuries were described and graded, and the healing process of the different osteochondral injuries was followed-up radiologically and clinically (Radin et al., 1984; Lahm et al., 2005; Ewers et al., 2002a,b; Newberry et al., 1998). Additional experiments have helped to classify injuries caused by radial and shear force including isolated and combined injuries (Tomatsu et al., 1992). With the advent of magnetic resonance imaging, classifications for traumatic osteochondral injuries have been proposed that were consequently refined (Deutsch and Mink, 1990; Mink and Deutsch, 1989; Bohndorf, 1996, 1999). In addition, ultrasound-based work on impact injuries to cartilage and the subchondral bone has been presented (Virén et al., 2009, 2011, 2012). In the present study we used a drop-tower apparatus for simulation of direct dynamic impact trauma (Kokkonen et al., 2011). As opposed to the static load employed in material testing, this device better simulates the reality of falls under natural conditions. Vertical dynamic impact was chosen in order to avoid shear forces. The aim of this ex-vivo study was to investigate whether direct acute dynamic impact trauma demonstrates a force-dependent association between direct impact to the cartilaginous articular surface and adjacent subchondral bone. We further wished to develop a grading system of injury caused by direct dynamic impact based on visual inspection and histological assessment. 2. Material and methods Due to its retrospective nature and use of abbatoir joints, this study was exempted from institutional review board procedures.
Fig. 1. Schematic diagram of the experimental set-up (1 ground plate, 2 holding fixture, 3 experimental specimen, 4 stand, 5 laser with cable and laser beam, 6 supporting rod, 7 falling tube, 8 falling body with cable, 9 reflector, 10 light barrier, 11 computer and monitor).
2.1. Experimental design To simulate a fall on a joint and to measure its impact, a dynamic material testing apparatus was developed. Using this apparatus, impact experiments were conducted in a series of 110 pig knees ex-vivo. Only one experiment was conducted for each knee. The femoral trochlea was chosen as the site of impact. A thrust plate diameter of 7 mm and the weight of the falling body (750 g) were kept constant with fall height being the only variable parameter. The ranges of impact examined extended from those inducing injury-free deformation to those causing frank osteochondral fractures. The induced injuries were initially examined macroscopically and then histologically. In pilot experiments, we first manually induced cartilage injury in 10 pig knees. For this, a metal rod was attached to the cartilage and force was exerted with a hammer onto the trochlear surface. This gave an initial impression of the force and the size of the plate required for the final experimental setup. The mechanical components of the custom made apparatus, which were composed of steel, included a falling tube, a falling body with thrust plate and a holding fixture for the specimen being examined (Fig. 1). The setup instrumentation included a sensor integrated in the falling body to measure impact, a laser-light barrier to start the measurement, and a multichannel computer. In the experiment, the falling body was released from a pre-determined clamped height to drop onto the articular trochlear surface of the fixed specimen. When the falling body crossed the light barrier, the measurement system was initiated and the computer calculated the force-time curve of the impact applied to the specimen in the holding fixture over a period of 150 ms. The sensitivity of the force sensor was 4 N and the monitoring frequency of the force was 6827 points per second. A total of 1024 measurements per experiment could be stored. The fall heights varied between 1 and 5 cm and were changed at 2 mm intervals. The energy of the impact may be determined indirectly based on the deceleration force (measured
in N) and the depth of the impact in mm. However, as the depth of impact is difficult to define and may only be approximated, we decided to measure the decelerating force directly and on-line with high frequency and then record the maximum decelerating force. While the decelerating force curves will show an identical shape for all measurements the maximum resulting force will be different depending on the height of the fall. The maximum force was defined as force in Newton (N). For the experiments, porcine specimens (German domestic pig) were used for reasons of biological similarity to humans, availability, and animal protection considerations. The animals were 5–6 months of age and weighed 80–100 kg. At the abattoir, the knee joints were removed within 1 h of the animals’ death and final preparation of the distal femoropatellar joint specimens was performed directly prior to the experiment within 3–4 h after death. The lateral femoral condyle was positioned exactly parallel to the thrust plate surface in order to avoid any shear forces. Altogether, 120 knee specimens were collected and prepared, including those for the pilot experiments. 2.2. Documentation of injuries The direct result of the impact experiment was the force-time curve for the impact from which the effective maximum force could be determined. Immediately after the experiment the specimen was analyzed macroscopically to document obtained injuries. The experiment number, maximum force and visual injury findings were registered and photographically documented. For histological investigation of the osteochondral injuries, a cube of 1.0 to 1.5 cm side-length that enveloped the impact site was excised. The cubes were then placed in 4% phosphate-buffered formaldehyde solution (Roti-Histofix® , Carl Roth GmbH & Co. KG, Karlsruhe, Germany), were decalcified for 6 weeks in nitric acid and embedded in paraffin (Paraplast® , Carl Roth GmbH & Co. KG,
458
R. Brill et al. / Annals of Anatomy 196 (2014) 456–463
Fig. 2. Frequencies of the force intervals measured for 110 pig knees.
Karlsruhe, Germany). The slice thickness for histological examination was 5 m. Masson trichrome and hematoxylin & eosin (HE) were used for staining. Articular surface damage, intrachondral injury and injuries of the subchondral bone were histologically assessed and qualitatively described. 2.3. Statistical analysis In addition to descriptive analyses, absolute and cumulative relative frequency distributions of the injury stages were determined for the applied ranges of impact. Impact forces were correlated with resulting injury pattern on the damaged knee specimens (Statistical package SPSS ® Version 17, Chicago, Illinois). 3. Results 3.1. Impact measurements The impact duration was comparable for all experiments at 7 ms (SD ± 1.2 s). The shape of the force-time curve for the impact was nearly identical for all experiments. The maximum force of the impact, in the following simply termed “impact:”, was between 102 N and 785 N. At a thrust plate diameter of 7 mm, this impact corresponded to a force of 2.6 to 20.4 MPa. As the magnitude of the impact of the falling body from the adjustable drop height could not be adequately estimated in advance, the frequency of impacts resulting from the different experiments was not evenly distributed (Fig. 2). 3.2. Macroscopic and microscopic examination By evaluating the changes induced by varying impact on the cartilage and subchondral bone in the 110 specimens both macroscopically and microscopically, a uniform pattern of injury emerged that could be classified into four grades. 3.2.1. Grade I: Elastic deformation No macroscopic changes of the osteochondral specimen are seen. Visually, a native specimen with an intact articular surface was observed. In physical terms, the impact was purely elastic with a simultaneous elastic deformation of the cartilage without causing
any permanent structural damage. Histologically, no changes were detected in the cartilage or subchondral bone (Fig. 3). 3.2.2. Grade II: Viscoelastic imprint A permanent imprint of the thrust plate was observed but no obvious cartilage disruption was seen. The imprint appeared to be reversible and disappeared completely within a 60 min period. In physical terms, the non-elastic force caused a viscoelastic, reversible deformation of the articular cartilage. In the subchondral bone, microfractures were observed microscopically (Fig. 4). 3.2.3. Grade III: Disrupted cartilage The impact of the thrust plate caused an imprint to the cartilage surface and a circular pattern of disrupted cartilage was observed at the perimeter of the imprint. The cartilage was irreversibly damaged and the impression of the thrust plate to the cartilage was at least partially irreversible. In physical terms, a non-elastic impact had been applied, causing a partial plastic, partial viscoelastic deformation of the cartilage. The histological section showed rupture of the cartilage and extensive microfractures in the subchondral bone (Fig. 5). 3.2.4. Grade IV: Osteochondral impression The injury to the osteochondral specimen is characterized macroscopically by bone fractures and continuously torn cartilage at the edges of the fissure. The imprint from the injury is irreversible. In physical terms, a non-elastic force was applied that caused a plastic deformation and osteochondral macrofracture. The histological section shows an osteochondral macrofracture with continuous chondral rupture down through zone IV of the cartilage and a macrofracture of the subchondral bone (Fig. 6). 3.3. Allocation of impact range to injury stage The impact force generated in the 110 impact experiments was subsequently allocated to the four qualitative injury stages according to the macroscopic and microscopic findings. Of the 110 impact experiments, 23 knees were classified as grade I injury, 20 knees as grade II injury, 33 knees as grade III injury, and 34 knees as grade IV injury according to the macro- and microscopic findings. The frequency distribution of the impacts measured for these four injury grades is presented in Fig. 7. Some overlap in the outer values of the
R. Brill et al. / Annals of Anatomy 196 (2014) 456–463
459
Impact range I: ≤280 N (≤7.3 MPa) Impact range II: >280–370 N (7.3–9.6 MPa) Impact range III: >370–490 N (7.3–9.6 MPa) Impact range IV: >490 N (>12.7 MPa) When allocating the impact ranges to the injury stages, 28 (25.5%) of the 110 injured knees were allocated to impact range I, 14 knees (12.7%) to impact range II, 41 knees (37.3%) to impact range III, and 27 knees (24.5%) to impact range IV (Table 1). By impact range classification, 62% of the 110 injured knees were correctly allocated to the right injury stage. Conversely by injury stage an average of 62% of the knees were likewise correctly allocated to the right impact range (Table 2). Deviation of impact range and injury grade was seen for 23 knees (20.9%) in grade I for 20 knees (18.2%), for 33 knees (30.0%) in grade III and for 34 knees (30.9%) in grade 4. A summary overview of impact, macroscopic and microscopic findings is presented in Table 2.
4. Discussion
Fig. 3. Stage I injury characterized by elastic deformation on the cartilage surface. (A). Schematic drawing. Arrow visualizes direction of impact. Cartilage shows reversible indentation. No changes are seen in the subchondral bone (oblique lines lower part of image). (B). Macroscopic image shows an unremarkable articular surface without imprint. (C). Histology shows unaffected cartilage (asterisk) and subchondral bone (double asterisks).
impact ranges for the injury stages was observed. The cumulative relative frequency of the forces applied for the four injury stages is shown in Fig. 8. The following cut-offs for force intervals were defined in regard to classifying osteochondral injury stage according to magnitude of impact:
Based on direct traumatic impact we were able to define a force-dependent biomechanical classification for direct, acute osteochondral injuries in this ex-vivo pig model ranging from reversible viscoelastic deformation to osteochondral impression injury on histological assessment. The severity of mechanically induced changes to the cartilage and subchondral bone depends on the height and velocity of the force. Comparison of slow static loading with dynamic impact has shown that the rigidity of the cartilage surface as defined by increase in the compression stress-strain curve also increases in relation to both the deformation and the deformation velocity (Radin et al., 1970; Langelier and Buschmann, 2003). Similarly, cartilage absorption of dynamic peak loads or impact increases in relation to the deformation and strain velocity. Bone loading experiments have demonstrated mechanical behavior analogous to that of cartilage. However, bone is 10- to 15-fold more rigid and with respect to the layer thickness has a 10- to 15-fold lower absorbing effect than cartilage (Radin et al., 1970). In vivo, subchondral bone contributes to reducing peak loads to the same extent as joint cartilage does (Radin et al., 1970). In animal experiments Radin et al. (1984) found that as a result of repeated but physiologically tolerated impact to a joint over a longer period of time, structural changes first develop in the subchondral bone and that, only later, cartilaginous pathology is detected. In our experiment using high impact forces subchondral bone changes were also observed prior to irreversible cartilage alterations. The dynamic impact experiments carried out in the present study can be compared with the static loading experiments on osteochondral specimens from the knee joints of 1-year-old pigs that Spahn and Wittig (2003) conducted using material testing equipment. Static loading of up to 6 MPa caused elastic deformation, which is in line with injury stage I—elastic deformation in our study, which extends to 7.3 MPa. Span and Wittig also considered 6–10 MPa as the range causing plastic reversible deformation that they observed already 1 min after relieving the strain on the specimen. Indeed, this complies with “impact range II—viscoelastic imprint” of 7.3–9.6 MPa where the deformation was viscoelastic and resolved within an hour, as shown in our study. A plastic irreversible deformation did not develop in our study until “stage III—torn cartilage” in the impact range from 9.6 to 12.7 MPa. The comprehensive micromechanical, light-microscopic and electron microscopic study of O’Connor et al. (1988) on the structure and biomechanical behavior in the individual joint
460
R. Brill et al. / Annals of Anatomy 196 (2014) 456–463
cartilage zones can be applied to explain cartilaginous disruption in this stage. Hyaline cartilage is described as a fiber-reinforced material comprised of collagen fibers and resin from proteoglycan (PG). The surface zone (zone I) of the cartilage, which is the most elastic, consists of a high concentration of collagen fibers that run in parallel to the cartilage surface and contain a low concentration of PG. In contrast, the middle zone (zone II and zone III), which is the most rigid, has the lowest concentration of unordered collagen fibers and a high concentration of PG. The deep zone (zone IV), on the other hand, which is moderately rigid is comprised of a moderate concentration of collagen fibers that are oriented vertically to the cartilage surface and have a moderate concentration of PG. In all cartilage zones the rigidity is directly linear to the PG content. The higher the PG content the greater the rigidity (i.e. the lower the elasticity). By loading a large surface osteochondral specimen with a smaller, cylindrical thrust plate as in our study, tensile and shear forces are built up at the edge of the thrust plate surface—in addition to its direct force, pulling the cartilage into the fissure that develops. The tensile force running parallel to the cartilage surface is maintained in the collagen fibers in zones I and II, which likewise run parallel to the cartilage surface until they tear. Our impact apparatus is not the first to be able to create a reproducible impact to an articular surface. Vrahas et al. (1997) introduced an instrument allowing perpendicular impaction of the distal femoral articular surface by an impactor with the area of the impacted surface and the measured force allowing to calculate an estimate of the impact stress. The instrument designed by Borelli and co-workers applied impact through a metallic pendulum of different masses (0, 400 and 2400 g) with the force being transmitted to the articular surface through an aluminum impactor of 3 mm diameter (Borelli et al., 2002). The difference between the Borreli apparatus and ours is the indirect measurement of force extracted from the mass and the time to peak force with the apparatus only allowing for 3 different masses. The Borrelli instrument allowed for a model of post-traumatic osteoarthritis by a single high-energy impact load through disruption of the extracellular matrix and by causing a decrease in chondrocyte metabolism (Borelli et al., 2009). To estimate the force needed to cause cartilage disruption we assume that at the beginning of the impact half of the force flowing at the edge of the thrust plate exerts a tensile strain and the other half over the thrust plate surface a compressive strain on the cartilage and that the height of the cartilage surface zone subjected to the flow of the force is 0.5 mm. From this we can calculate a cylindrical surface around the edges of the thrust plate through which the tensile strain runs when tearing ensues, with a thrust plate diameter of 7 mm and a surface area of 11 mm2 . Assuming a total effective strain of 370 N at the beginning of a tear; i.e., from the lower end of impact range III, a tensile strain or tear force of 185 N/11 mm2 or 16.8 MPa prevails at the margins of the thrust plate. At a thickness of 0.75 mm of the cartilage zone subjected to the force, the tear force would be 185 N/16.5 mm2 or 11.2 MPa. In “stage IV—Osteochondral fracture” the cartilage continues to tear as a result of increasing force and penetration depth of the thrust plate into zone IV, slicing through the cartilage. After the cartilage is torn at the edge, the impact is still evenly distributed through the cartilage and the underlying subchondral bone, causing it to break. As a consequence, the force measured in stage IV constitutes the fracture stress for the subchondral bone. Considering the frequency of the impact measured in stage IV we assume 13.7 ± 1.0 MPa to be the average value for fracture stress. The fracture stress we measured for stage IV is considerably Fig. 4. Stage II injury characterized by viscoelastic imprint on the cartilage surface: (A). Schematic drawing shows direction of impact (arrow), indentation of the articular chondral surface and subchondral microfractures (dark shaded area in lower part of image). (B). Left part of image shows the acute but reversible imprint (arrows)
directly after impact. The right image part shows the macroscopic view 30 min later and complete resolution of impact zone. (C). The histological image is showing subchondral microfractures (arrows).
R. Brill et al. / Annals of Anatomy 196 (2014) 456–463
461
Fig. 6. Stage IV injury is characterized by osteochondral impression. (A). Schematic drawing exhibits direction of impact (arrow), cartilage tearing and subchondral macrofractures. (B). Macroscopic image shows osteochondral impression (arrows). (C). Histological image shows osteochondral macrofracture including surface disruption (black arrow) and subchondral fracture (white arrow).
Fig. 5. Stage III injury exhibits torn cartilage at the edges of the thrust plate. (A). Schematic drawing shows direction of impact (arrow), disrupted cartilage (light gray lines adjacent to probe) and subchondral microfractures (dark shaded area in lower part of image). (B). Macroscopic image shows permanent imprint and peripheral
cartilage fissuring (arrows). (C). Histological image shows torn cartilage (arrows) and subchondral microfractures (arrowheads).
462
R. Brill et al. / Annals of Anatomy 196 (2014) 456–463
Table 1 Frequency comparison of the injured knees in the actual injury stage and in the impact range of the classification. Impact range in classification Actual injury stage
I
II
III
IV
I II III IV Sum Probability that knee falls in injury stage x if impact range x was classified
22 5 1
1 7 4 2 14 50.0%
6 21 14 41 51.2%
2 7 18 27 66.7%
28 78.6%
Sum
Probability that knee will fall in impact range x if it represents actual injury stage x(%)
23 20 33 34 110
95.7 35.0 63.6 52.9
Table 2 Biomechanical distribution of direct, acute osteochondral injuries. Injury stage
Cartilage zone I–III
Cartilage zone IV
Subchondral bone
Force (MPa)
Stage I: elastic deformation Stage II: viscoelastic imprint Stage III: torn cartilage Stage IV: osteochondral impression
No findings No findings Tear Tear
No findings No findings No findings Tear
No findings Microfractures Microfractures Macrofracture
<7.3 7.3–9.6 9.6–12.7 >12.7
Fig. 7. Frequency distribution of force intervals for injury stages I–IV.
Fig. 8. Cumulative, relative frequency of the range of forces measured for each injury stage (Stage 1—Elastic Deformation (purple line), Stage 2—Viscoelastic Imprint (light blue line), Stage 3—Disrupted Cartilage (dark blue line), Stage 4—Osteochondral Impression (maroon line).
lower than 25.8 ± 5.2 MPa as reported by Spahn and Wittig (2003). However, these authors also observed lower fracture stress values of 18–23 MPa using static loading, which are closer to our study results. One limitation of our experimental design was that despite aligning the articular surface as parallel as possible to the thrust plate even a minimal tilt of the thrust plate will exert an asymmetrical force on the cartilage surface. This in turn may produce heterogeneous loads and shear forces and possibly explains some of the overlap of the stages that we identified in the outer limits of the impact ranges in the proposed classification. Summarizing our results, we were able to suggest a forcedependent, histologically confirmed biomechanical classification of direct, acute osteochondral injury in an ex-vivo model, which is in line with radiological staging of traumatic osteochondral damage and suggests that the subchondral bone is affected by microfractures at dynamic impact forces that do not create permanent articular surface disruption, which explains reversibility of subchondral changes seen on MRI after indirect impact of two
R. Brill et al. / Annals of Anatomy 196 (2014) 456–463
opposing articular surfaces. Altogether four different stages could be determined that are observed for distinct ranges of dynamic impact. Authors contributions (1) All authors were involved in the conception and design of the study, or acquisition of data, or analysis and interpretation of data. (2) All authors contributed to drafting the article or revising it critically for important intellectual content. (3) All authors gave their final approval of the manuscript to be submitted. Additional contributions: • Analysis and interpretation of the data: RB, WAW, HH, KB, UB, UW, AK, FWR • Drafting of the article: RB, WAW, UB, FWR • Provision of study materials or patients: RB, HH • Statistical expertise: UB, FWR • Obtaining of funding: N/A • Collection and assembly of data: RB, HH, KB, UW, AK, FWR Responsibility for the integrity of the work as a whole, from inception to finished article, is taken by R. Brill, MD (first author; [email protected]) Funding and role of the funding source No funding was received. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.aanat. 2014.07.007. References Benninghoff, A., 1925. Form und Bau der Gelenkknorpel in ihren Beziehungen zur Funktion Die modellierenden und formerhaltenden Faktoren des Knorpelreliefs. Zeitschrift für Anatomie. 76, 43–63. Bohndorf, K., 1996. Injuries at the articulating surface of bone (chondral, osteochondrol and subchondral fractures and osteochondosis dissecans). Eur. J. Radiol. 22, 22–29. Bohndorf, K., 1999. Imaging of acute injuries of the articular surface (chondral, osteochondral and subchondral fractures). Skelet. Radiol. 28, 545–560. Borelli, J., Burns, M.E., Ricci, W.M., Silva, M.J., 2002. A method for delivering variable impact stresses on the articular cartilage of rabbit knees. J. Orthop. Res. 16, 182–188.
463
Borelli, J., Silva, M.J., Zaegel, M.A., Franz, C., Sandell, L.J., 2009. Single high-energy impact load causes posttraumatic OA in young rabbits via a decrease in cellular metabolism. J. Orthop. Res. 27, 347–352. Deutsch, A.L., Mink, J.H., 1990. Magnetic resonance imaging of injuries to bone and articular cartilage. Orthop. Rev. 19, 66–75. Ewers, B.J., Weaver, B.T., Haut, R.C., 2002a. Impact orientation can significantly affect the outcome of blunt impact to the rabbit patellofemoral joint. J. Biomech. 35, 1591–1598. Ewers, B.J., Weaver, B.T., Sevensma, E.T., Haut, R.C., 2002b. Chronic changes in rabbit retro-patellar cartilage and subchondral bone after blunt impact loading of the patellofemoral joint. J. Orthop. Res. 20, 545–550. Kokkonen, H.T., Jurvelin, J.S., Tiitu, V., Töyräs, J., 2011. Detection of mechanical injury of articular cartilage using contrast enhanced computed tomography. Osteoarthritis Cartilage 19, 295–301. Lahm, A., Uhl, M., Edlich, M., Erggelet, C., Haberstoh, J., Kreuz, P.C., 2005. An experimental canine model for subchondral lesions of the knee joint. Knee 12, 51–55. Langelier, E., Buschmann, M.D., 2003. Increasing strain and strain rate strengthen transient stiffness but weaken the response to subsequent compression for articular cartilage in unconfined compression. J. Biomech. 36, 853–859. Lochmüller, E.M., Kristin, J., Matsuura, M., Kuhn, V., Hudelmaier, M., Link, T.M., Eckstein, F., 2008. Measurement of trabecular bone microstructure does not improve prediction of mechanical failure loads at the distal radius compared with bone mass alone. Calcif. Tissue Int. 83, 293–299. Mink, J.H., Deutsch, A.L., 1989. Occult cartilage and bone injuries of the knee detection, classification and assessment with MR imaging. Radiology 170, 823–829. Newberry, W.N., Mackenzie, C.D., Haut, R.C., 1998. Blunt impact causes changes in bone and cartilage in a regularly exercised animal model. J. Orthop. Res. 16, 348–354. O’Connor, P., Orford, C.R., Gardner, D.L., 1988. Differential response to compressive loads of zones of canine hyaline articular cartilage: micromechanical, light and electron microscopic studies. Ann. Rheum. Dis. 47, 414–420. Radin, E.L., Paul, I.L., Lowy, M.A., 1970. Comparison of the dynamic force transmitting properties of subchondral bone and articular cartilage. J. Bone Joint Surg. Am. 52, 444–455. Radin, E.L., Martin, R.B., Burr, D.B., Caterson, B., Boyd, R.D., Doodwin, C., 1984. Effects of the mechanical loading on the tissues of the rabbit knee. J. Orthop. Res. 2, 221–234. Räth, C., Monetti, R., Bauer, J., Sidorenko, I., Müller, D., Matsuura, M., Lochmüller, E.-M., Zysset, P., Eckstein, F., 2008. Strength through structure: visualization and local assessment of the trabecular bone structure. New J. Phys. 10, 125010. Spahn, G., Wittig, R., 2003. Biomechanical properties (compressive strength and compressive pressure at break) of hyaline cartilage under axial load. Zentralbl Chir. 128, 78–82. Tomatsu, T., Imai, N., Takeuchi, N., Takahashi, K., Kimura, N., 1992. Experimentally produced fractures of articular cartilage and bone the effects of shear forces on the pig knee. J. Bone Joint Surg. Br. 74, 457–462. Virén, T., Saarakkala, S., Kaleva, E., Nieminen, H.J., Jurvelin, J.S., Töyräs, J., 2009. Minimally invasive ultrasound method for intra-articular diagnostics of cartilage degeneration. Ultrasound Med. Biol. 35, 1546–1554. Virén, T., Saarakkala, S., Tiitu, V., Puhakka, J., Kiviranta, I., Jurvelin, J., Töyräs, J., 2011. Ultrasound evaluation of mechanical injury of bovine knee articular cartilage under arthroscopic control. IEEE Trans. Ultrason. Ferroelectr. Freq. Control 58, 148–155. Virén, T., Timonen, M., Tyrväinen, H., Tiitu, V., Jurvelin, J.S., Töyräs, J., 2012. Ultrasonic evaluation of acute impact injury of articular cartilage in vitro. Osteoarthritis Cartilage 20, 719–726. Vrahas, M.S., Smith, G.A., Rosler, D.M., Baratta, R.V., 1997. Method to impact in vivo rabbit femoral cartilage with blows of quantifiable stress. J. Orthop. Res. 15, 314–317.
Contents lists available at ScienceDirect
Annals of Anatomy journal homepage: www.elsevier.de/aanat
Research article
Dynamic impact force and association with structural damage to the knee joint: An ex-vivo study Richard Brill a , Walther A. Wohlgemuth b , Harald Hempfling c , Klaus Bohndorf d , Ursula Becker e , Ulrich Welsch f , Alexander Kamp g , Frank W. Roemer h,a,∗ a
Department of Radiology, Klinikum Augsburg, Augsburg, Germany Department of Radiology, University of Regensburg, Regensburg, Germany c Berufsgenossenschaftliche Unfallklinik, Murnau, Germany d MR Center of Excellence, Department of Radiology, University of Vienna, Vienna, Austria e F. Hoffmann-La Roche Ltd., Basel, Switzerland f Department of Cell Biology, Ludwigs-Maximilians University, Munich, Germany g Center for Radiology and Nuclear Medicine, Murnau, Germany h Department of Radiology, University of Erlangen-Nuremberg, Erlangen, Germany b
a r t i c l e
i n f o
Article history: Received 9 February 2014 Received in revised form 7 July 2014 Accepted 29 July 2014 Keywords: Dynamic impact Knee Osteochondral injury Histology
s u m m a r y No systematic, histologically confirmed data are available concerning the association between magnitude of direct dynamic impact caused by vertical impact trauma and the resulting injury to cartilage and subchondral bone. The aim of this study was to investigate the association between dynamic impact and the resulting patterns of osteochondral injury in an ex-vivo model. A mechanical apparatus was employed to perform ex-vivo controlled dynamic vertical impact experiments in 110 pig knees with the femur positioned in a holding fixture. A falling body with a thrust plate and photo sensor was applied. The direct impact to the trochlear articular surface was registered and the resulting osteochondral injuries macroscopically and histologically correlated and categorized. The relationship between magnitude of direct impact and injury severity could be classified as stage I injuries (impact <7.3 MPa): elastic deformation, no histological injury; stage II injuries (impact 7.3–9.6 MPa): viscoelastic imprint of the cartilaginous surface, subchondral microfractures; stage III injuries (impact 9.6–12.7 MPa): disrupted cartilage surface, chondral fissures and subchondral microfractures; stage IV injuries (impact >12.7 MPa): osteochondral impression, histologically imprint and osteochondral macrofractures. The impact ranges and histologic injury stages determined from this vertical dynamic impact experiment allowed for a biomechanical classification of direct, acute osteochondral injury. In contrast to static load commonly applied in ex-vivo experiments, dynamic impact more realistically represents actual trauma to the knee joint. © 2014 Elsevier GmbH. All rights reserved.
1. Introduction The zonal structure of hyaline articular cartilage and the adjacent subchondral bone has been a topic of histologic evaluation for more than 80 years, when Benninghoff described and differentiated four distinct cartilage zones (Benninghoff, 1925). The
∗ Corresponding author at: Associate Professor of Radiology Department of Radiology University of Erlangen-Nuremberg Maximiliansplatz 1 91054 Erlangen, Germany. Tel.: +49 9131 8536065; fax: +49 9131 8536068. E-mail addresses: [email protected], [email protected], [email protected] (F.W. Roemer). http://dx.doi.org/10.1016/j.aanat.2014.07.007 0940-9602/© 2014 Elsevier GmbH. All rights reserved.
advent of electron microscopic studies has subsequently markedly refined our knowledge of the cartilage ultrastructure (O’Connor et al., 1988). In addition, the biomechanical behavior of cartilage and bone as an osteochondral unit has been extensively explored (Radin et al., 1970; Spahn and Wittig, 2003; Räth et al., 2008; Lochmüller et al., 2008). By employing basic static material testing equipment these studies investigated the deformation caused by static load applied to cartilage and bone specimens. Through these experiments the elastic, viscoelastic and plastic properties of cartilage could be determined. The elasticity modulus, the rigidity, and resistance to fracture could be defined and the time dependence of viscoelastic deformation was established. In more recent in-vivo and ex-vivo animal experiments single severe and minor
R. Brill et al. / Annals of Anatomy 196 (2014) 456–463
457
repetitive impact trauma was applied to the knee joint. Consequent injuries were described and graded, and the healing process of the different osteochondral injuries was followed-up radiologically and clinically (Radin et al., 1984; Lahm et al., 2005; Ewers et al., 2002a,b; Newberry et al., 1998). Additional experiments have helped to classify injuries caused by radial and shear force including isolated and combined injuries (Tomatsu et al., 1992). With the advent of magnetic resonance imaging, classifications for traumatic osteochondral injuries have been proposed that were consequently refined (Deutsch and Mink, 1990; Mink and Deutsch, 1989; Bohndorf, 1996, 1999). In addition, ultrasound-based work on impact injuries to cartilage and the subchondral bone has been presented (Virén et al., 2009, 2011, 2012). In the present study we used a drop-tower apparatus for simulation of direct dynamic impact trauma (Kokkonen et al., 2011). As opposed to the static load employed in material testing, this device better simulates the reality of falls under natural conditions. Vertical dynamic impact was chosen in order to avoid shear forces. The aim of this ex-vivo study was to investigate whether direct acute dynamic impact trauma demonstrates a force-dependent association between direct impact to the cartilaginous articular surface and adjacent subchondral bone. We further wished to develop a grading system of injury caused by direct dynamic impact based on visual inspection and histological assessment. 2. Material and methods Due to its retrospective nature and use of abbatoir joints, this study was exempted from institutional review board procedures.
Fig. 1. Schematic diagram of the experimental set-up (1 ground plate, 2 holding fixture, 3 experimental specimen, 4 stand, 5 laser with cable and laser beam, 6 supporting rod, 7 falling tube, 8 falling body with cable, 9 reflector, 10 light barrier, 11 computer and monitor).
2.1. Experimental design To simulate a fall on a joint and to measure its impact, a dynamic material testing apparatus was developed. Using this apparatus, impact experiments were conducted in a series of 110 pig knees ex-vivo. Only one experiment was conducted for each knee. The femoral trochlea was chosen as the site of impact. A thrust plate diameter of 7 mm and the weight of the falling body (750 g) were kept constant with fall height being the only variable parameter. The ranges of impact examined extended from those inducing injury-free deformation to those causing frank osteochondral fractures. The induced injuries were initially examined macroscopically and then histologically. In pilot experiments, we first manually induced cartilage injury in 10 pig knees. For this, a metal rod was attached to the cartilage and force was exerted with a hammer onto the trochlear surface. This gave an initial impression of the force and the size of the plate required for the final experimental setup. The mechanical components of the custom made apparatus, which were composed of steel, included a falling tube, a falling body with thrust plate and a holding fixture for the specimen being examined (Fig. 1). The setup instrumentation included a sensor integrated in the falling body to measure impact, a laser-light barrier to start the measurement, and a multichannel computer. In the experiment, the falling body was released from a pre-determined clamped height to drop onto the articular trochlear surface of the fixed specimen. When the falling body crossed the light barrier, the measurement system was initiated and the computer calculated the force-time curve of the impact applied to the specimen in the holding fixture over a period of 150 ms. The sensitivity of the force sensor was 4 N and the monitoring frequency of the force was 6827 points per second. A total of 1024 measurements per experiment could be stored. The fall heights varied between 1 and 5 cm and were changed at 2 mm intervals. The energy of the impact may be determined indirectly based on the deceleration force (measured
in N) and the depth of the impact in mm. However, as the depth of impact is difficult to define and may only be approximated, we decided to measure the decelerating force directly and on-line with high frequency and then record the maximum decelerating force. While the decelerating force curves will show an identical shape for all measurements the maximum resulting force will be different depending on the height of the fall. The maximum force was defined as force in Newton (N). For the experiments, porcine specimens (German domestic pig) were used for reasons of biological similarity to humans, availability, and animal protection considerations. The animals were 5–6 months of age and weighed 80–100 kg. At the abattoir, the knee joints were removed within 1 h of the animals’ death and final preparation of the distal femoropatellar joint specimens was performed directly prior to the experiment within 3–4 h after death. The lateral femoral condyle was positioned exactly parallel to the thrust plate surface in order to avoid any shear forces. Altogether, 120 knee specimens were collected and prepared, including those for the pilot experiments. 2.2. Documentation of injuries The direct result of the impact experiment was the force-time curve for the impact from which the effective maximum force could be determined. Immediately after the experiment the specimen was analyzed macroscopically to document obtained injuries. The experiment number, maximum force and visual injury findings were registered and photographically documented. For histological investigation of the osteochondral injuries, a cube of 1.0 to 1.5 cm side-length that enveloped the impact site was excised. The cubes were then placed in 4% phosphate-buffered formaldehyde solution (Roti-Histofix® , Carl Roth GmbH & Co. KG, Karlsruhe, Germany), were decalcified for 6 weeks in nitric acid and embedded in paraffin (Paraplast® , Carl Roth GmbH & Co. KG,
458
R. Brill et al. / Annals of Anatomy 196 (2014) 456–463
Fig. 2. Frequencies of the force intervals measured for 110 pig knees.
Karlsruhe, Germany). The slice thickness for histological examination was 5 m. Masson trichrome and hematoxylin & eosin (HE) were used for staining. Articular surface damage, intrachondral injury and injuries of the subchondral bone were histologically assessed and qualitatively described. 2.3. Statistical analysis In addition to descriptive analyses, absolute and cumulative relative frequency distributions of the injury stages were determined for the applied ranges of impact. Impact forces were correlated with resulting injury pattern on the damaged knee specimens (Statistical package SPSS ® Version 17, Chicago, Illinois). 3. Results 3.1. Impact measurements The impact duration was comparable for all experiments at 7 ms (SD ± 1.2 s). The shape of the force-time curve for the impact was nearly identical for all experiments. The maximum force of the impact, in the following simply termed “impact:”, was between 102 N and 785 N. At a thrust plate diameter of 7 mm, this impact corresponded to a force of 2.6 to 20.4 MPa. As the magnitude of the impact of the falling body from the adjustable drop height could not be adequately estimated in advance, the frequency of impacts resulting from the different experiments was not evenly distributed (Fig. 2). 3.2. Macroscopic and microscopic examination By evaluating the changes induced by varying impact on the cartilage and subchondral bone in the 110 specimens both macroscopically and microscopically, a uniform pattern of injury emerged that could be classified into four grades. 3.2.1. Grade I: Elastic deformation No macroscopic changes of the osteochondral specimen are seen. Visually, a native specimen with an intact articular surface was observed. In physical terms, the impact was purely elastic with a simultaneous elastic deformation of the cartilage without causing
any permanent structural damage. Histologically, no changes were detected in the cartilage or subchondral bone (Fig. 3). 3.2.2. Grade II: Viscoelastic imprint A permanent imprint of the thrust plate was observed but no obvious cartilage disruption was seen. The imprint appeared to be reversible and disappeared completely within a 60 min period. In physical terms, the non-elastic force caused a viscoelastic, reversible deformation of the articular cartilage. In the subchondral bone, microfractures were observed microscopically (Fig. 4). 3.2.3. Grade III: Disrupted cartilage The impact of the thrust plate caused an imprint to the cartilage surface and a circular pattern of disrupted cartilage was observed at the perimeter of the imprint. The cartilage was irreversibly damaged and the impression of the thrust plate to the cartilage was at least partially irreversible. In physical terms, a non-elastic impact had been applied, causing a partial plastic, partial viscoelastic deformation of the cartilage. The histological section showed rupture of the cartilage and extensive microfractures in the subchondral bone (Fig. 5). 3.2.4. Grade IV: Osteochondral impression The injury to the osteochondral specimen is characterized macroscopically by bone fractures and continuously torn cartilage at the edges of the fissure. The imprint from the injury is irreversible. In physical terms, a non-elastic force was applied that caused a plastic deformation and osteochondral macrofracture. The histological section shows an osteochondral macrofracture with continuous chondral rupture down through zone IV of the cartilage and a macrofracture of the subchondral bone (Fig. 6). 3.3. Allocation of impact range to injury stage The impact force generated in the 110 impact experiments was subsequently allocated to the four qualitative injury stages according to the macroscopic and microscopic findings. Of the 110 impact experiments, 23 knees were classified as grade I injury, 20 knees as grade II injury, 33 knees as grade III injury, and 34 knees as grade IV injury according to the macro- and microscopic findings. The frequency distribution of the impacts measured for these four injury grades is presented in Fig. 7. Some overlap in the outer values of the
R. Brill et al. / Annals of Anatomy 196 (2014) 456–463
459
Impact range I: ≤280 N (≤7.3 MPa) Impact range II: >280–370 N (7.3–9.6 MPa) Impact range III: >370–490 N (7.3–9.6 MPa) Impact range IV: >490 N (>12.7 MPa) When allocating the impact ranges to the injury stages, 28 (25.5%) of the 110 injured knees were allocated to impact range I, 14 knees (12.7%) to impact range II, 41 knees (37.3%) to impact range III, and 27 knees (24.5%) to impact range IV (Table 1). By impact range classification, 62% of the 110 injured knees were correctly allocated to the right injury stage. Conversely by injury stage an average of 62% of the knees were likewise correctly allocated to the right impact range (Table 2). Deviation of impact range and injury grade was seen for 23 knees (20.9%) in grade I for 20 knees (18.2%), for 33 knees (30.0%) in grade III and for 34 knees (30.9%) in grade 4. A summary overview of impact, macroscopic and microscopic findings is presented in Table 2.
4. Discussion
Fig. 3. Stage I injury characterized by elastic deformation on the cartilage surface. (A). Schematic drawing. Arrow visualizes direction of impact. Cartilage shows reversible indentation. No changes are seen in the subchondral bone (oblique lines lower part of image). (B). Macroscopic image shows an unremarkable articular surface without imprint. (C). Histology shows unaffected cartilage (asterisk) and subchondral bone (double asterisks).
impact ranges for the injury stages was observed. The cumulative relative frequency of the forces applied for the four injury stages is shown in Fig. 8. The following cut-offs for force intervals were defined in regard to classifying osteochondral injury stage according to magnitude of impact:
Based on direct traumatic impact we were able to define a force-dependent biomechanical classification for direct, acute osteochondral injuries in this ex-vivo pig model ranging from reversible viscoelastic deformation to osteochondral impression injury on histological assessment. The severity of mechanically induced changes to the cartilage and subchondral bone depends on the height and velocity of the force. Comparison of slow static loading with dynamic impact has shown that the rigidity of the cartilage surface as defined by increase in the compression stress-strain curve also increases in relation to both the deformation and the deformation velocity (Radin et al., 1970; Langelier and Buschmann, 2003). Similarly, cartilage absorption of dynamic peak loads or impact increases in relation to the deformation and strain velocity. Bone loading experiments have demonstrated mechanical behavior analogous to that of cartilage. However, bone is 10- to 15-fold more rigid and with respect to the layer thickness has a 10- to 15-fold lower absorbing effect than cartilage (Radin et al., 1970). In vivo, subchondral bone contributes to reducing peak loads to the same extent as joint cartilage does (Radin et al., 1970). In animal experiments Radin et al. (1984) found that as a result of repeated but physiologically tolerated impact to a joint over a longer period of time, structural changes first develop in the subchondral bone and that, only later, cartilaginous pathology is detected. In our experiment using high impact forces subchondral bone changes were also observed prior to irreversible cartilage alterations. The dynamic impact experiments carried out in the present study can be compared with the static loading experiments on osteochondral specimens from the knee joints of 1-year-old pigs that Spahn and Wittig (2003) conducted using material testing equipment. Static loading of up to 6 MPa caused elastic deformation, which is in line with injury stage I—elastic deformation in our study, which extends to 7.3 MPa. Span and Wittig also considered 6–10 MPa as the range causing plastic reversible deformation that they observed already 1 min after relieving the strain on the specimen. Indeed, this complies with “impact range II—viscoelastic imprint” of 7.3–9.6 MPa where the deformation was viscoelastic and resolved within an hour, as shown in our study. A plastic irreversible deformation did not develop in our study until “stage III—torn cartilage” in the impact range from 9.6 to 12.7 MPa. The comprehensive micromechanical, light-microscopic and electron microscopic study of O’Connor et al. (1988) on the structure and biomechanical behavior in the individual joint
460
R. Brill et al. / Annals of Anatomy 196 (2014) 456–463
cartilage zones can be applied to explain cartilaginous disruption in this stage. Hyaline cartilage is described as a fiber-reinforced material comprised of collagen fibers and resin from proteoglycan (PG). The surface zone (zone I) of the cartilage, which is the most elastic, consists of a high concentration of collagen fibers that run in parallel to the cartilage surface and contain a low concentration of PG. In contrast, the middle zone (zone II and zone III), which is the most rigid, has the lowest concentration of unordered collagen fibers and a high concentration of PG. The deep zone (zone IV), on the other hand, which is moderately rigid is comprised of a moderate concentration of collagen fibers that are oriented vertically to the cartilage surface and have a moderate concentration of PG. In all cartilage zones the rigidity is directly linear to the PG content. The higher the PG content the greater the rigidity (i.e. the lower the elasticity). By loading a large surface osteochondral specimen with a smaller, cylindrical thrust plate as in our study, tensile and shear forces are built up at the edge of the thrust plate surface—in addition to its direct force, pulling the cartilage into the fissure that develops. The tensile force running parallel to the cartilage surface is maintained in the collagen fibers in zones I and II, which likewise run parallel to the cartilage surface until they tear. Our impact apparatus is not the first to be able to create a reproducible impact to an articular surface. Vrahas et al. (1997) introduced an instrument allowing perpendicular impaction of the distal femoral articular surface by an impactor with the area of the impacted surface and the measured force allowing to calculate an estimate of the impact stress. The instrument designed by Borelli and co-workers applied impact through a metallic pendulum of different masses (0, 400 and 2400 g) with the force being transmitted to the articular surface through an aluminum impactor of 3 mm diameter (Borelli et al., 2002). The difference between the Borreli apparatus and ours is the indirect measurement of force extracted from the mass and the time to peak force with the apparatus only allowing for 3 different masses. The Borrelli instrument allowed for a model of post-traumatic osteoarthritis by a single high-energy impact load through disruption of the extracellular matrix and by causing a decrease in chondrocyte metabolism (Borelli et al., 2009). To estimate the force needed to cause cartilage disruption we assume that at the beginning of the impact half of the force flowing at the edge of the thrust plate exerts a tensile strain and the other half over the thrust plate surface a compressive strain on the cartilage and that the height of the cartilage surface zone subjected to the flow of the force is 0.5 mm. From this we can calculate a cylindrical surface around the edges of the thrust plate through which the tensile strain runs when tearing ensues, with a thrust plate diameter of 7 mm and a surface area of 11 mm2 . Assuming a total effective strain of 370 N at the beginning of a tear; i.e., from the lower end of impact range III, a tensile strain or tear force of 185 N/11 mm2 or 16.8 MPa prevails at the margins of the thrust plate. At a thickness of 0.75 mm of the cartilage zone subjected to the force, the tear force would be 185 N/16.5 mm2 or 11.2 MPa. In “stage IV—Osteochondral fracture” the cartilage continues to tear as a result of increasing force and penetration depth of the thrust plate into zone IV, slicing through the cartilage. After the cartilage is torn at the edge, the impact is still evenly distributed through the cartilage and the underlying subchondral bone, causing it to break. As a consequence, the force measured in stage IV constitutes the fracture stress for the subchondral bone. Considering the frequency of the impact measured in stage IV we assume 13.7 ± 1.0 MPa to be the average value for fracture stress. The fracture stress we measured for stage IV is considerably Fig. 4. Stage II injury characterized by viscoelastic imprint on the cartilage surface: (A). Schematic drawing shows direction of impact (arrow), indentation of the articular chondral surface and subchondral microfractures (dark shaded area in lower part of image). (B). Left part of image shows the acute but reversible imprint (arrows)
directly after impact. The right image part shows the macroscopic view 30 min later and complete resolution of impact zone. (C). The histological image is showing subchondral microfractures (arrows).
R. Brill et al. / Annals of Anatomy 196 (2014) 456–463
461
Fig. 6. Stage IV injury is characterized by osteochondral impression. (A). Schematic drawing exhibits direction of impact (arrow), cartilage tearing and subchondral macrofractures. (B). Macroscopic image shows osteochondral impression (arrows). (C). Histological image shows osteochondral macrofracture including surface disruption (black arrow) and subchondral fracture (white arrow).
Fig. 5. Stage III injury exhibits torn cartilage at the edges of the thrust plate. (A). Schematic drawing shows direction of impact (arrow), disrupted cartilage (light gray lines adjacent to probe) and subchondral microfractures (dark shaded area in lower part of image). (B). Macroscopic image shows permanent imprint and peripheral
cartilage fissuring (arrows). (C). Histological image shows torn cartilage (arrows) and subchondral microfractures (arrowheads).
462
R. Brill et al. / Annals of Anatomy 196 (2014) 456–463
Table 1 Frequency comparison of the injured knees in the actual injury stage and in the impact range of the classification. Impact range in classification Actual injury stage
I
II
III
IV
I II III IV Sum Probability that knee falls in injury stage x if impact range x was classified
22 5 1
1 7 4 2 14 50.0%
6 21 14 41 51.2%
2 7 18 27 66.7%
28 78.6%
Sum
Probability that knee will fall in impact range x if it represents actual injury stage x(%)
23 20 33 34 110
95.7 35.0 63.6 52.9
Table 2 Biomechanical distribution of direct, acute osteochondral injuries. Injury stage
Cartilage zone I–III
Cartilage zone IV
Subchondral bone
Force (MPa)
Stage I: elastic deformation Stage II: viscoelastic imprint Stage III: torn cartilage Stage IV: osteochondral impression
No findings No findings Tear Tear
No findings No findings No findings Tear
No findings Microfractures Microfractures Macrofracture
<7.3 7.3–9.6 9.6–12.7 >12.7
Fig. 7. Frequency distribution of force intervals for injury stages I–IV.
Fig. 8. Cumulative, relative frequency of the range of forces measured for each injury stage (Stage 1—Elastic Deformation (purple line), Stage 2—Viscoelastic Imprint (light blue line), Stage 3—Disrupted Cartilage (dark blue line), Stage 4—Osteochondral Impression (maroon line).
lower than 25.8 ± 5.2 MPa as reported by Spahn and Wittig (2003). However, these authors also observed lower fracture stress values of 18–23 MPa using static loading, which are closer to our study results. One limitation of our experimental design was that despite aligning the articular surface as parallel as possible to the thrust plate even a minimal tilt of the thrust plate will exert an asymmetrical force on the cartilage surface. This in turn may produce heterogeneous loads and shear forces and possibly explains some of the overlap of the stages that we identified in the outer limits of the impact ranges in the proposed classification. Summarizing our results, we were able to suggest a forcedependent, histologically confirmed biomechanical classification of direct, acute osteochondral injury in an ex-vivo model, which is in line with radiological staging of traumatic osteochondral damage and suggests that the subchondral bone is affected by microfractures at dynamic impact forces that do not create permanent articular surface disruption, which explains reversibility of subchondral changes seen on MRI after indirect impact of two
R. Brill et al. / Annals of Anatomy 196 (2014) 456–463
opposing articular surfaces. Altogether four different stages could be determined that are observed for distinct ranges of dynamic impact. Authors contributions (1) All authors were involved in the conception and design of the study, or acquisition of data, or analysis and interpretation of data. (2) All authors contributed to drafting the article or revising it critically for important intellectual content. (3) All authors gave their final approval of the manuscript to be submitted. Additional contributions: • Analysis and interpretation of the data: RB, WAW, HH, KB, UB, UW, AK, FWR • Drafting of the article: RB, WAW, UB, FWR • Provision of study materials or patients: RB, HH • Statistical expertise: UB, FWR • Obtaining of funding: N/A • Collection and assembly of data: RB, HH, KB, UW, AK, FWR Responsibility for the integrity of the work as a whole, from inception to finished article, is taken by R. Brill, MD (first author; [email protected]) Funding and role of the funding source No funding was received. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.aanat. 2014.07.007. References Benninghoff, A., 1925. Form und Bau der Gelenkknorpel in ihren Beziehungen zur Funktion Die modellierenden und formerhaltenden Faktoren des Knorpelreliefs. Zeitschrift für Anatomie. 76, 43–63. Bohndorf, K., 1996. Injuries at the articulating surface of bone (chondral, osteochondrol and subchondral fractures and osteochondosis dissecans). Eur. J. Radiol. 22, 22–29. Bohndorf, K., 1999. Imaging of acute injuries of the articular surface (chondral, osteochondral and subchondral fractures). Skelet. Radiol. 28, 545–560. Borelli, J., Burns, M.E., Ricci, W.M., Silva, M.J., 2002. A method for delivering variable impact stresses on the articular cartilage of rabbit knees. J. Orthop. Res. 16, 182–188.
463
Borelli, J., Silva, M.J., Zaegel, M.A., Franz, C., Sandell, L.J., 2009. Single high-energy impact load causes posttraumatic OA in young rabbits via a decrease in cellular metabolism. J. Orthop. Res. 27, 347–352. Deutsch, A.L., Mink, J.H., 1990. Magnetic resonance imaging of injuries to bone and articular cartilage. Orthop. Rev. 19, 66–75. Ewers, B.J., Weaver, B.T., Haut, R.C., 2002a. Impact orientation can significantly affect the outcome of blunt impact to the rabbit patellofemoral joint. J. Biomech. 35, 1591–1598. Ewers, B.J., Weaver, B.T., Sevensma, E.T., Haut, R.C., 2002b. Chronic changes in rabbit retro-patellar cartilage and subchondral bone after blunt impact loading of the patellofemoral joint. J. Orthop. Res. 20, 545–550. Kokkonen, H.T., Jurvelin, J.S., Tiitu, V., Töyräs, J., 2011. Detection of mechanical injury of articular cartilage using contrast enhanced computed tomography. Osteoarthritis Cartilage 19, 295–301. Lahm, A., Uhl, M., Edlich, M., Erggelet, C., Haberstoh, J., Kreuz, P.C., 2005. An experimental canine model for subchondral lesions of the knee joint. Knee 12, 51–55. Langelier, E., Buschmann, M.D., 2003. Increasing strain and strain rate strengthen transient stiffness but weaken the response to subsequent compression for articular cartilage in unconfined compression. J. Biomech. 36, 853–859. Lochmüller, E.M., Kristin, J., Matsuura, M., Kuhn, V., Hudelmaier, M., Link, T.M., Eckstein, F., 2008. Measurement of trabecular bone microstructure does not improve prediction of mechanical failure loads at the distal radius compared with bone mass alone. Calcif. Tissue Int. 83, 293–299. Mink, J.H., Deutsch, A.L., 1989. Occult cartilage and bone injuries of the knee detection, classification and assessment with MR imaging. Radiology 170, 823–829. Newberry, W.N., Mackenzie, C.D., Haut, R.C., 1998. Blunt impact causes changes in bone and cartilage in a regularly exercised animal model. J. Orthop. Res. 16, 348–354. O’Connor, P., Orford, C.R., Gardner, D.L., 1988. Differential response to compressive loads of zones of canine hyaline articular cartilage: micromechanical, light and electron microscopic studies. Ann. Rheum. Dis. 47, 414–420. Radin, E.L., Paul, I.L., Lowy, M.A., 1970. Comparison of the dynamic force transmitting properties of subchondral bone and articular cartilage. J. Bone Joint Surg. Am. 52, 444–455. Radin, E.L., Martin, R.B., Burr, D.B., Caterson, B., Boyd, R.D., Doodwin, C., 1984. Effects of the mechanical loading on the tissues of the rabbit knee. J. Orthop. Res. 2, 221–234. Räth, C., Monetti, R., Bauer, J., Sidorenko, I., Müller, D., Matsuura, M., Lochmüller, E.-M., Zysset, P., Eckstein, F., 2008. Strength through structure: visualization and local assessment of the trabecular bone structure. New J. Phys. 10, 125010. Spahn, G., Wittig, R., 2003. Biomechanical properties (compressive strength and compressive pressure at break) of hyaline cartilage under axial load. Zentralbl Chir. 128, 78–82. Tomatsu, T., Imai, N., Takeuchi, N., Takahashi, K., Kimura, N., 1992. Experimentally produced fractures of articular cartilage and bone the effects of shear forces on the pig knee. J. Bone Joint Surg. Br. 74, 457–462. Virén, T., Saarakkala, S., Kaleva, E., Nieminen, H.J., Jurvelin, J.S., Töyräs, J., 2009. Minimally invasive ultrasound method for intra-articular diagnostics of cartilage degeneration. Ultrasound Med. Biol. 35, 1546–1554. Virén, T., Saarakkala, S., Tiitu, V., Puhakka, J., Kiviranta, I., Jurvelin, J., Töyräs, J., 2011. Ultrasound evaluation of mechanical injury of bovine knee articular cartilage under arthroscopic control. IEEE Trans. Ultrason. Ferroelectr. Freq. Control 58, 148–155. Virén, T., Timonen, M., Tyrväinen, H., Tiitu, V., Jurvelin, J.S., Töyräs, J., 2012. Ultrasonic evaluation of acute impact injury of articular cartilage in vitro. Osteoarthritis Cartilage 20, 719–726. Vrahas, M.S., Smith, G.A., Rosler, D.M., Baratta, R.V., 1997. Method to impact in vivo rabbit femoral cartilage with blows of quantifiable stress. J. Orthop. Res. 15, 314–317.