- Email: [email protected]
Does prior sustained compression make cartilage-on-bone more vulnerable to trauma?
Clinical Biomechanics 27 (2012) 637–645
Contents lists available at SciVerse ScienceDirect
Clinical Biomechanics journal homepage: www.elsevier.com/locate/clinbiomech
Does prior sustained compression make cartilage-on-bone more vulnerable to trauma? Woong Kim, Ashvin Thambyah, Neil Broom ⁎ Tissue Mechanics Laboratory, Department of Chemical and Materials Engineering, University of Auckland, New Zealand
a r t i c l e
i n f o
Article history: Received 1 December 2011 Accepted 22 March 2012 Keywords: Cartilage-on-bone Prior creep loading Mechanical response to impact Quantification of induced structural damage
a b s t r a c t Background: This study investigated how varying levels of prior creep deformation in cartilage-on-bone samples influences their mechanical response and vulnerability to structural damage following a single traumatic impact. Methods: Bovine patellae were subjected to varying intervals of prior creep loading at a constant stress of 4 MPa. Immediately following removal of this stress the samples were impacted with a pendulum indenter system at a fixed energy of 2.2 J. Findings: With increasing prior creep, the peak force on impact rose, the duration of impact and time to reach peak force both decreased, and both the energy dissipated during impact and the magnitude of impulse were both unchanged by the level of prior creep. With increasing prior creep, the severity of impact-induced osteochondral damage increased: articular cartilage cracks penetrated to a greater depth, extending to the calcified cartilage layer resulting in hairline fractures or articular cartilage delamination and associated secondary damage to the vascular channels in the subchondral bone. Interpretation: The study shows that exposure of the cartilage-on-bone system to prior creep can significantly influence its response to subsequent impact, namely force attenuation and severity of damage to the articular cartilage, calcified cartilage and vascular channel network in the subchondral bone. © 2012 Elsevier Ltd. All rights reserved.
1. Introduction Articular cartilage (AC) is a relatively compliant connective tissue layer covering the diarthrodial joint surfaces (Ghadially, 1983; Meachim and Stockwell, 1979). The deformability of AC facilitates an increase in joint contact area and thus acts in a stress-reducing role. It also allows the joint to articulate with minimal friction (Brower and Hsu, 1969; Ghadially, 1983; Mankin and Radin, 1993; Meachim and Stockwell, 1979). The functional unit of AC is a dense collagenous network within which the hydrophilic proteoglycans are constrained. In its fully hydrated state AC contains some 60–80% water. At low rates of loading the deformation of AC occurs via the exclusion of this water out through an ultra-low permeability matrix, a mode of behaviour that has been described as poro-visco-hyperelastic (Oloyede et al., 1992). At high rates of loading, fluid movement is restricted or even prevented and the cartilage matrix behaves as an instantaneously elastic material (Flachsmann et al., 2001). AC is also known to be vulnerable to mechanical trauma (Radin, 1999; Radin and Paul, 1971b; Radin et al., 1972) and this can induce a posttraumatic osteoarthritis (OA) (Felson, 2004; Mink and Deutsch,
⁎ Corresponding author. E-mail address: [email protected] (N. Broom). 0268-0033/$ – see front matter © 2012 Elsevier Ltd. All rights reserved. doi:10.1016/j.clinbiomech.2012.03.007
1989; Vellet et al., 1991; Walker, 1998). Both in vivo and in vitro studies have sought to determine the levels of threshold stress, and especially impact, for cartilage damage (see review by Scott et al. (Scott and Athanasiou, 2006)). However, no rigorous body of in vivo data exists especially in regard to the level of stress, the stress rate and the duration of stress. Aspen et al. (Aspden et al., 2002) suggest that a non-pathological stress rate in vivo is between 100–1000 MPa/s with a duration of 30–150 ms. This contrasts with a traumatic impact, such as in a knee-dashboard injury which may have a duration of around 1 ms (Repo and Finlay, 1977). Under in vitro conditions it has been found that an impact load (single or repeated) of 11 to 36 MPa can cause visible cartilage damage, sometimes together with chondrocyte necrosis and apoptosis (Atkinson et al., 1998; Burgin and Aspden, 2007; Flachsmann et al., 2005; Haut, 1989; Jeffrey and Aspden, 2006; Kerin et al., 2003; Obeid et al., 1994; Repo and Finlay, 1977; Torzilli et al., 1999; Verteramo and Seedhom, 2007; Wilson et al., 2003). Many of these studies employed either a free-flight mass, drop tower or pendulum impacting system to create stress durations of less than 30 ms, and where peak forces or the energy of the impact have been reported (Scott and Athanasiou, 2006). Magnetic resonance imaging has been used to measure in vivo changes in cartilage thickness in human subjects before and after different exercise regimes (Eckstein et al., 1995, 1998, 1999, 2001, 2005, 2006; Mosher et al., 2010). Patellar cartilage thickness was
638
W. Kim et al. / Clinical Biomechanics 27 (2012) 637–645
loaded but having no prior deformation? This paper therefore reports on a study investigating how varying levels of prior creep deformation influence the vulnerability of cartilage-on-bone samples to a defined intensity of impact. 2. Methods 2.1. Tissues and test groups
Fig. 1. Percentage bar graph showing the relationship between the components of prior creep strain (black), impact strain (white) and remaining cartilage thickness (grey) as a function of duration of creep loading. Note that all strain/thickness values are rounded for clarity.
found to be reduced ~ 6% after 30 knee bends, ~5% after running and cycling, and ~ 3% after walking (Eckstein et al., 2006). In a study of femoral and tibial cartilage a decrease in thickness of up to 12% was reported following running for 30 min (Mosher et al., 2010). These studies suggest that in vivo AC, in some habitually loaded regions of the joint, may operate in a near continuously deformed state (~ 90% of its fully hydrated thickness), with even higher levels of deformation during more extreme physical activities. An interesting question therefore presents itself: are these habitually deformed regions of the joint more or less susceptible to a superimposed rapidly applied load than those regions similarly
Patellae free of gross AC defects were collected from freshly slain prime bovine bulls (~2 years old) and stored at − 20 °C. Prior to experimentation each patella was thawed in cold running water for 20 min and then two cartilage–bone samples with en face dimensions 14 mm × 14 mm and ~ 10 mm in height were sawn from the distal– lateral quadrant (see sites A and B in Supplemental Fig. 1A). This region was chosen because of its relative flatness and thus suitability for mechanical testing. A total of 160 samples obtained from 91 patellae were employed in this study, and allocated into eight groups of 20 samples with each group subjected to varying durations of creep. All samples were tested within 48 hours of initial thawing. Each sample was embedded into the well of a custom-made stainless steel holder using dental cement so as to align the articular surface perpendicular to the path of a pendulum indenter (Supplemental Fig. 2) as previously described (Flachsmann et al., 2005). The stainless steel indenter (Supplemental Fig. 1A) had a semi-cylindrical profile of radius ~2 mm and length 10 mm and positioned such that its initial contact with the sample was a line parallel to the cylinder axis along the cylinder surface. The indenter ends were slightly rounded to minimise the risk of edge-induced damage (Kerin et al., 1998). A flat-faced rectangular stainless-steel cap of en face
Fig. 2. Scatter plots showing relationships between prior creep strain and (A) impact strain, (B) peak force, (C) impact duration, (D) time to reach peak force, (E) impulse and (F) energy loss.
W. Kim et al. / Clinical Biomechanics 27 (2012) 637–645
639
Fig. 3. (A) En face view of impacted site in a sample exhibiting articular surface disruption revealed by India ink staining. Samples were bisected along the vertical dotted line. (B) Scanned cross-sectional image of the bisected sample showing that the impact-induced damage has extended into the full cartilage depth as indicated by the ink staining penetrating into the underlying subchondral bone. (C) Microphotograph of (B) showing both articular cartilage delamination along the tidemark (see arrow) and ink residues in the vascular channels (see arrowheads). The indenter size and profile relative to the sample size is also shown. (D) shows an enlargement of the boxed region in (C) near the extremity of the delamination. The India ink has flowed into the tip of a vascular channel (see white arrowhead) and spread into other interconnected vascular channels (see black arrowhead). Note that the impact does not appear to have fractured the subchondral plate.
dimensions 6 × 14 mm was designed to fit snugly over the semicylindrical impacting indenter (Supplemental Fig. 1A, lower). With this indenter cap in position, and prior to the dental cement setting, a pre-load of ~0.005 MPa was applied to the pendulum via a weight and pulley system so as to align the articular surface of the sample with this planar indenter cap (Supplemental Fig. 1B). This ensured that the contact between the indenter and cartilage surface of the sample was optimised. Once the dental cement had set the holder, with its embedded sample, was temporarily removed from the target base and rehydrated for at least 2 hours in 0.15 M saline at 4 °C.
equilibrium strain is achieved in 60 min and steady-state by 180 min (Oloyede and Broom, 1994). Creep displacements were recorded using a linear variable differential transformer (LVDT-Model DFg 2.5, Solartron Metrology, West Sussex, England) temporarily engaged with the pendulum arm, together with an A/D data-acquisition system. In order to prevent dehydration of the sample during the creep loading period (i.e. up to 180 min), a saline-soaked sponge was wrapped around the sample. Just prior to impacting the sponge was removed. 2.3. Impact loading
2.2. Creep loading In order to induce varying amounts of creep deformation in the cartilage prior to impacting (i.e. a physiologically relevant compression of ~10% (Eckstein et al., 2006; Mosher et al., 2010) and above), the samples were subjected to creep loading by means of a pulley and dead-weight assembly attached to the pendulum arm. The flat-faced rectangular cap was temporarily fitted over the impacting indenter and the sample then statically compressed at 4 MPa (Supplemental Fig. 1C) for prescribed durations of creep ranging from 0 min (control) to 0.2, 1, 5, 15, 30, 60 and 180 min. This produced an en face area of creep-deformed articular cartilage that extended beyond the projected area of the subsequent impact footprint (Supplemental Fig. 1D). The maximum period of creep loading was based on previous consolidation studies conducted in our laboratory demonstrating that although the strain vs time response of cartilage under static compression is highly non-linear, a substantial fraction of the
The indenter assembly incorporated a dynamic piezoelectric force transducer (Model 9021A, Kistler AG, Schweiz, Germany) whose signal was fed to a charge amplifier (Model 5015A, Kistler AG) and recorded at a sampling rate of 40 kHz using a data acquisition system. The indenter pendulum was released from a fixed height of 80 mm which gave an impact mean velocity of 1.13 m/s (SE, 0.003) associated with a mean kinetic energy Ek of 2.2 J (SE, 0.01) as estimated from its second moment of mass (I), angular velocity of pendulum (ω), linear velocity of indenter at impact (v), and radius of the pendulum arm (r; Eq. (1)). Ek ¼
1 2 1 v2 Iω ¼ I 2 2 r
ð1Þ
A high speed video camera (2000 fps; Fastcam MC 2.1, Photron, San Diego, USA) fitted with a macro lens and a 1 kW theatre spot light
640
W. Kim et al. / Clinical Biomechanics 27 (2012) 637–645
Fig. 4. Schematics illustrating damage scoring system developed to describe and quantify the osteochondral damage observed. (A) Illustrations of the four common damage types I, II, III and IV. Only Types III and IV were associated with vascular channel (VC) damage arising from either fracture of the calcified cartilage or delamination of the articular cartilage (see magnified insets). (B) and (C) show how articular cartilage fissures were scored with respect to their depth of penetration and number of branches.
with heat deflection were used to capture the progression of the impact. The camera was mounted so as to provide a side view of the indenter-cartilage contact and generate a video file of the impact event. The video data were analysed (i-Speed 3.0.2.6. Olympus NDT Inc. Waltham, MA, USA and Image J 1.43u, National Institutes of Health, Bethesda, USA) to obtain cartilage thickness, impact and rebound velocities (vi, vf), and impact deformation (Supplemental Fig. 1D). The kinetic energy lost (%) during the impact was calculated using a modified rotational kinetic energy equation (Eq. (2)) based on the kinetic energy delivered Ek, measured impact and rebound velocities vi and vf, angular velocity ω.
Energy lost ð%Þ ¼
ΔEkði−f Þ EkðiÞ
¼
v2i −v2f v2i
It should be emphasised that the indenter geometry was not intended to simulate actual joint loading. Rather, this geometry and the dimensions of the tissue sample were chosen as a means of delivering an impact of defined energy in which the extent of obvious tissue disruption was confined to an area well within the en face dimensions of the sample surface. The aim was to obtain an accurate comparison of disruption at the microstructural level between the non-creep and creep loaded states. In effect our choice of indenter served to confine the extent of ‘injury’ to a manageable tissue volume.
2.4. Macroscopic assessment of damage ð2Þ
The force–time data provided by the stress transducer was used to determine the peak force, time to peak force, duration of impact and magnitude of impulse, the latter calculated by numerical integration of the force–time curve. Following impact the cartilage surface was wet-rubbed with India ink, washed with running water, and then any AS damage revealed by the ink staining recorded using macrophotography.
Samples were fixed in 10% formalin for approximately 24 hours then mildly decalcified using 10% formic acid for 3 to 5 days (Ippolito et al., 1981; Thambyah and Broom, 2007) and finally rinsed in water for 1 hour to remove residual chemicals. A cross-sectional cut was then made in order to bisect the cartilage–bone sample at a point approximately midway along the length of the impacted site (Supplemental Fig. 1E, F). The exposed cross-section was then scanned using a flatbed scanner (4800 dpi; Canoscan F8800, Canon, Tokyo, Japan) to obtain high resolution macro images of the damage morphologies (number of fissures, penetration depth and presence of
W. Kim et al. / Clinical Biomechanics 27 (2012) 637–645
641
Fig. 5. (A) Sample exhibiting all four damage types: Type I fissures marked with arrow, Type II cartilage chips (see shaded regions), Type III calcified cartilage fracture (white arrowhead) and Type IV delamination (black arrowhead). (B) Higher magnification view of LH boxed region in (A) showing Type III damage involved calcified cartilage CC fracture propagating to the tip of a vascular channel VC (see arrowheads) and leading to ink penetration into the SCB. (C) Magnified view of boxed area in (B). (D) Magnified view of RH boxed region in (A) showing Type IV cartilage delamination that has resulted in three VCs being damaged and leading to ink penetration of the VC network (see arrowheads). (E) Magnified image of a region of delamination showing its typical zigzag fracture morphology intercepting a VC (arrowhead).
delamination) in the hyaline cartilage and any India ink staining in the subchondral bone (SCB).
2.5. Microscopic assessment of damage Those samples exhibiting ink penetration into the SCB were then sectioned using a freezing microtome to obtain 20–30 osteochondral serial sections (~ 30 μm thick). At least 10 sections per sample were examined in their fully hydrated state using both bright field (Nikon AZ100, Nikon Instruments, Melville, NY, USA) and differential interference contrast optical microscopy (DIC, Nikon Eclipse 80i, Nikon Instruments) in order to assess the degree of impact-induced damage in the calcified cartilage (CC) and SCB.
2.6. Impact damage types and quantification system From a preliminary set of experiments a scoring system was developed to classify and quantify the modes and severity of damage observed in the AC, CC, osteochondral junction (OCJ) and SCB (see Section 3). 2.7. Statistical analysis Pearson's correlation tests (SPSS, IBM, Armonk, NY, USA) were used to investigate correlations between the mechanical parameters (level of creep strain, impact strain, peak force, time to peak force, impact duration, impulse and percentage energy lost) and their relationship to the damage severity and type.
642
W. Kim et al. / Clinical Biomechanics 27 (2012) 637–645
decreased from 0.8 to 0.6 ms (r = − 0.584, P b 0.05, Fig. 2D). Conversely, neither the impulse (r = 0.120, P = 0.128, Fig. 2E) nor percentage energy loss with impact (r = 0.203, P b 0.05, Fig. 2F) correlated with the level of prior creep strain. 3.2. Quantification of structural damage Fig. 3A shows an en face view of an impacted sample exhibiting AS damage along the contact line of the indenter's cylindrical axis as revealed by India ink staining. A scanned cross-section through this damage (Fig. 3B) indicates that fissures have propagated through the full thickness of the AC and into the calcified cartilage (CC). Higher magnification views of sections taken from this same sample are shown in Fig. 3C and D. The scoring system devised to quantify the impact damage (Fig. 4) took into consideration the depth of fissure penetration, the number of fissures found per sample and the type of damage observed. With reference to the set of higher magnification images shown in Fig. 5A–E the four common types of damage observed in the impacted samples were as follows:
Fig. 6. Venn diagram showing frequencies of the damage Types I, II, III and IV. Note that the shaded region represents the sample numbers associated with ink staining in the subchondral bone.
3. Results The cohort mean thickness of the articular cartilage in the 160 original samples prepared for testing was 1.60 mm (SE 0.02 mm). 3.1. Mechanical response to impact Supplemental Fig. 2 shows a typical force vs time curve obtained from a sample subjected to 5 min of prior creep at 4 MPa. The asymmetry of the curve was a consistent feature of all tests and possibly arises from rig-specific characteristics, the inertia of the tissue, its embedded plaster matrix/holder, flexion in the pendulum arm and associated stored energy dissipation. Although this asymmetry in the force vs time response was not investigated any further in the study, its consistency throughout all of the impact tests reassured us that a comparison of responses between the various test groups was, in fact, legitimate. The video images in Supplemental Fig. 3 compare side views of the same sample as in Supplemental Fig. 2 at both zero and peak impact load. The non-planar contact geometry does not permit accurate determination of the impact stresses in the AC. However, the video images do provide an approximate profile of the AC/indenter contact at peak force. Averaged over the 160 samples tested the projected contact area was ~ 30 mm 2, yielding an average peak impact stress of ~ 42 MPa. The bar graph in Fig. 1 shows the averaged values of (a) induced creep strain, (b) subsequent impact strain and (c) remaining deformed AC thickness (normalised to its original un-deformed thickness) for the samples in each prior-creep deformed group (i.e. 0 to 180 min). With increasing duration of creep the creep strain increased progressively to 62%, while the subsequent impact strains decreased from 34% to 16% (SE, 1%). There was no significant change in the levels of either creep or impact strain from 60 to 180 min. Fig. 2 shows scatter plots for the various mechanical parameters as a function of creep strain. There was a strong inverse linear relationship between creep strain and impact strain (Pearson's coefficient r = −0.707, P b 0.05, Fig. 2A). With increasing creep strain the mean peak force increased from 1249 N to 1388 N (r = 0.353, P b 0.05, Fig. 2B), the duration of impact decreased from 4.1 ms to 3.4 ms (r = − 0.544, P b 0.05, Fig. 2C) and the time to reach peak force
Type I AC fissure damage: the formation of a single fissure or branching fissures in the AC (Figs. 4A, C, 5A) but without either AC loss or delamination. Type II AC chip damage: AC loss which occurred when two “fissures” intersected (Figs. 4A, 5A). Type III CC fracture: the penetration of a hairline crack into the CC layer and intercepting a single vascular channel (VC) (Figs. 4A, 5A–C), which allowed the ink stain to spread along the interconnected VC network and then focally stain the SCB (Fig. 3B). Type IV AC delamination: AC delaminating along the tidemark (Figs. 4A, 5A, D, E) resulting in damage to the tip of any VC intercepted by the propagating crack creating the delamination. Again, this led to ink staining of the SCB (Fig. 3B). The severity of damage was further quantified from the depth of fissure penetration relative to the full AC thickness (scored as less than 15%, 25%, 50%, 75% or 100%) (Fig. 4B), and the frequency of AC fissures per sample by summing the number of main fissures and fissure branches (Fig. 4C). The Venn diagram in Fig. 6 summarises the frequency of the four damage types. All 160 samples suffered at least Type I damage. Of these 160 samples 55 suffered, in addition, various combinations of Types II, III and IV damage as indicated. A total of 25 samples exhibited SCB ink staining and these all contained either CC fracture (i.e. Type III) or AC delamination (i.e. Type IV), or both. Further, none of these 25 samples exhibited any detectable damage in the SCB. The depth of fissure penetration in the AC was positively associated with the level of prior creep strain (Fig. 7A; r = 0.805). The frequency of Type III and IV damage (and the associated SCB staining) increased only when the prior creep strain reached 30% or higher (Fig. 7D–F). Conversely, there was no obvious correlation between the level of creep strain and the number of AC fissures found per sample (r = 0.266; Fig. 7B) or the frequency of AC chip damage (r = 0.007; Fig. 7C). 4. Discussion This study has shown that under the impact conditions employed the level of prior creep in the AC is a critical factor in influencing both the mechanical response and susceptibility to trauma of the osteochondral tissues. With increasing creep strain the peak impact force transmitted to the SCB increases whereas both the time to reach peak force and the duration of impact decrease (Fig. 2B, C, D). Further, with increasing creep strain the depth of penetration of the AC fissures increased (Fig. 7A). Creep strains of 30% or higher were
W. Kim et al. / Clinical Biomechanics 27 (2012) 637–645
643
Fig. 7. Percentage bar graphs indicating severity of damage (see A, B) and types of damage (see C–F) as a function of creep strain (expressed in ranges of 10%). The numbers of samples affected are indicated inside their respective bars. Percentage bar graph for Type I damage was omitted since all 160 samples were affected.
associated with an increasing frequency of both Type III and IV damage involving the CC and VCs (Fig. 7D, E, F). By contrast, the level of prior creep appeared to have little or no influence on the impulse or amount of energy lost during the impact (Fig. 2E, F), and this suggests cartilage itself plays a minimal role in energy or shock absorption. This finding is consistent with the much earlier postulate of Radin and Paul (1971a) and the whole joint studies of Hoshino and Wallace (1987) that the underlying bone is the main contributor to shock absorption in the joint. Also, there was no obvious association between the level of prior creep and either the number of fissures found per sample or the frequency of AC chip damage (Fig. 7B, C). This suggests that both Types I and II modes of damage were, in part, indenter-geometry dependent. Conversely, Types III and IV modes involving damage to the CC would have arisen from forces transmitted through the impacted AC rather than by the action of the indenter cutting directly into the CC layer. In fact, even at the near-equilibrium creep strain of 62% the additional strain of 16% from the subsequent impact still left a residual thickness of AC which was at least 22% of the original AC thickness (Fig. 1). It is the transmission of force through this residual AC thickness that would have inflicted the observed damage in the CC.
With the impacting energy employed in the present study (~2.2 J), disruption of the calcified cartilage and vascular channels (VC) commenced at a prior creep strain of around 30%, a value that considerably exceeds the physiological range of 3–12% measured in human knee AC in vivo (Eckstein et al., 2006; Mosher et al., 2010). Our data do, however, suggest that high stress impulsive loading of the joint tissues following a prolonged period of sustained loading may increase the risk of osteochondral damage (as shown by damage Types III and IV) and that this may in turn result in haemorrhaging into the SCB (Ahstrom, 1965; Armstrong and Mow, 1982; Bauer and Jackson, 1988; Brown and Cruess, 1982; Butler and Andrews, 1988; Gillquist et al., 1977; Noyes et al., 1980; Sandberg et al., 1986). Such defects have been also shown to affect the normal joint compliance, contact pressures and cartilage strains in both tibia and femur as demonstrated using finite element modelling (Shirazi and ShiraziAdl, 2009) and known to be associated with the development of a secondary osteoarthrosis (Oegema and Thompson, 1995; Thompson et al., 1991). A novel aspect of the present study was the effectiveness of ink staining in the SCB as an indicator of the extent to which CC fracture (Type III), AC delamination (Type IV) and VC disruption resulted from the impacting event (Fig. 3). Bauer and Jackson reported that
644
W. Kim et al. / Clinical Biomechanics 27 (2012) 637–645
osteochondral damage in the human knee, similar to that observed in the present investigation, can occur from rotational or direct impact trauma (Bauer and Jackson, 1988). They also noted that this type of subchondral damage can lead to a haemarthrosis or bleeding into the joint space (Ahstrom, 1965; Armstrong and Mow, 1982; Bauer and Jackson, 1988; Brown and Cruess, 1982; Butler and Andrews, 1988; Gillquist et al., 1977; Noyes et al., 1980; Sandberg et al., 1986). From a clinical perspective, hair-line fissures in the otherwise normal appearing articular cartilage would not be detected using routinely performed arthroscopy or radiographic diagnosis. And yet, as the present study demonstrates, such subtle damage can be associated with significant disruption to both the calcified cartilage and vascular channels originating in the subchondral bone. AC, being avascular, has a relatively passive response to injury. Conversely, trauma to the underlying vascularised subchondral bone can induce a complex inflammatory response that may lead to the eventual degeneration of the joint (Oegema and Thompson, 1995; Thompson et al., 1991). The average peak impact stress of ~ 42 MPa calculated from the force data and high speed video images is not dissimilar to levels of peak stress experienced in vivo during certain relatively common types of trauma where 25–30 MPa have been recorded in the human patellofemoral joint (Finlay and Repo, 1978; Haut, 1989; Repo and Finlay, 1977; Scott and Athanasiou, 2006). Interestingly Vener et al. (Vener et al., 1992), using a peak impact stress of 40 MPa, reported both CC and SCB fracture in canine metacarpophalangeal and metatarsophalangeal joints. This contrasts with our present findings in which no subchondral damage was observed despite the high peak forces employed. 5. Conclusions With the aim of developing a more complete understanding of the impact behaviour of the cartilage–bone system this study has demonstrated that the level of prior creep in articular cartilage does have a significant influence both on the mechanical response of the integrated cartilage–bone system and the modes and severity of damage induced in the articular cartilage, the calcified cartilage and vascular channels. Conflict of interest statement We authors (Drs. Woong Kim, Neil D. Broom and Ashvin Thambyah) of the manuscript disclose that we have no financial and personal relationships with other people or organisations that could inappropriately influence (bias) their work. Supplementary materials related to this article can be found online at doi:10.1016/j.clinbiomech.2012.03.007. Acknowledgements Funding for this research was generously provided by the Auckland Medical Research Foundation (AMRF). References Ahstrom Jr., J.P., 1965. Osteochondral fracture in the knee joint associated with hypermobility and dislocation of the patella. Report of eighteen cases. J. Bone Joint Surg. Am. 47, 1491–1502. Armstrong, C.G., Mow, V.C., 1982. Variations in the intrinsic mechanical properties of human articular cartilage with age, degeneration, and water content. J. Bone Joint Surg. Am. 64, 88–94. Aspden, R.M., Jeffrey, J.E., Burgin, L.V., 2002. Impact loading of articular cartilage. Osteoarthritis Cartilage 10, 588–589 discussion 590. Atkinson, T.S., Haut, R.C., Altiero, N.J., 1998. Impact-induced fissuring of articular cartilage: an investigation of failure criteria. J. Biomech. Eng. 120, 181–187. Bauer, M., Jackson, R.W., 1988. Chondral lesions of the femoral condyles: a system of arthroscopic classification. Arthroscopy 4, 97–102.
Brower, T.D., Hsu, W.Y., 1969. Normal articular cartilage. Clin. Orthop. Relat. Res. 64, 9–17. Brown, K.L., Cruess, R.L., 1982. Bone and cartilage transplantation in orthopaedic surgery. A review. J. Bone Joint Surg. Am. 64, 270–279. Burgin, L.V., Aspden, R.M., 2007. A drop tower for controlled impact testing of biological tissues. Med. Eng. Phys. 29, 525–530. Butler, J.C., Andrews, J.R., 1988. The role of arthroscopic surgery in the evaluation of acute traumatic hemarthrosis of the knee. Clin. Orthop. Relat. Res. 150–152. Eckstein, F., Sittek, H., Milz, S., Schulte, E., Kiefer, B., Reiser, M., et al., 1995. The potential of magnetic resonance imaging (MRI) for quantifying articular cartilage thickness —a methodological study. Clin. Biomech. 10, 434–440. Eckstein, F., Tieschky, M., Faber, S.C., Haubner, M., Kolem, H., Englmeier, K.H., et al., 1998. Effect of physical exercise on cartilage volume and thickness in vivo: MR imaging study. Radiology 207, 243–248. Eckstein, F., Tieschky, M., Faber, S., Englmeier, K.H., Reiser, M., 1999. Functional analysis of articular cartilage deformation, recovery, and fluid flow following dynamic exercise in vivo. Anat. Embryol. (Berl) 200, 419–424. Eckstein, F., Reiser, M., Englmeier, K.H., Putz, R., 2001. In vivo morphometry and functional analysis of human articular cartilage with quantitative magnetic resonance imaging—from image to data, from data to theory. Anat. Embryol. (Berl) 203, 147–173. Eckstein, F., Lemberger, B., Gratzke, C., Hudelmaier, M., Glaser, C., Englmeier, K.H., et al., 2005. In vivo cartilage deformation after different types of activity and its dependence on physical training status. Ann. Rheum. Dis. 64, 291–295. Eckstein, F., Hudelmaier, M., Putz, R., 2006. The effects of exercise on human articular cartilage. J. Anat. 208, 491–512. Felson, D.T., 2004. Risk factors for osteoarthritis: understanding joint vulnerability. Clin. Orthop. Relat. Res. S16–S21. Finlay, J.B., Repo, R.U., 1978. Instrumentation and procedure for the controlled impact of articular cartilage. IEEE Trans. Biomed. Eng. 25, 34–39. Flachsmann, R., Broom, N.D., Hardy, A.E., 2001. Deformation and rupture of the articular surface under dynamic and static compression. J. Orthop. Res. 19, 1131–1139. Flachsmann, R., Kim, W., Broom, N., 2005. Vulnerability to rupture of the intact articular surface with respect to age and proximity to site of fibrillation: a dynamic and static-investigation. Connect. Tissue Res. 46, 159–169. Ghadially, F.N., 1983. Fine structure of synovial joints : a text and atlas of the ultrastructure of normal and pathological articular tissues. Butterworths, London, Boston. Gillquist, J., Hagberg, G., Oretorp, N., 1977. Arthroscopy in acute injuries of the knee joint. Acta Orthop. Scand. 48, 190–196. Haut, R.C., 1989. Contact pressures in the patellofemoral joint during impact loading on the human flexed knee. J. Orthop. Res. 7, 272–280. Hoshino, A., Wallace, W.A., 1987. Impact-absorbing properties of the human knee. J. Bone Joint Surg. Br. 69, 807–811. Ippolito, E., LaVelle, S., Pedrini, V., 1981. The effect of various decalcifying agents on cartilage proteoglycans. Stain Technol. 56, 367–373. Jeffrey, J.E., Aspden, R.M., 2006. The biophysical effects of a single impact load on human and bovine articular cartilage. Proc. Inst. Mech. Eng. H 220, 677–686. Kerin, A.J., Wisnom, M.R., Adams, M.A., 1998. The compressive strength of articular cartilage. Proc. Inst. Mech. Eng. H 212, 273–280. Kerin, A.J., Coleman, A., Wisnom, M.R., Adams, M.A., 2003. Propagation of surface fissures in articular cartilage in response to cyclic loading in vitro. Clin. Biomech. 18, 960–968. Mankin, H.J., Radin, E.L., 1993. Structure and function of joints, Arthritis and Allied Conditions, 12th Ed. , pp. 181–197. Meachim, G., Stockwell, R., 1979. The matrix. In: Freeman, M.A.R. (Ed.), Adult Articular Cartilage. Pitman Medical, New York. 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. Mosher, T.J., Liu, Y., Torok, C.M., 2010. Functional cartilage MRI T2 mapping: evaluating the effect of age and training on knee cartilage response to running. Osteoarthritis Cartilage 18, 358–364. Noyes, F.R., Bassett, R.W., Grood, E.S., Butler, D.L., 1980. Arthroscopy in acute traumatic hemarthrosis of the knee. Incidence of anterior cruciate tears and other injuries. J. Bone Joint Surg. Am. 62 (687–95), 757. Obeid, E.M., Adams, M.A., Newman, J.H., 1994. Mechanical properties of articular cartilage in knees with unicompartmental osteoarthritis. J. Bone Joint Surg. Br. 76, 315–319. Oegema, T.R., Thompson, R.C., 1995. Histopathology and pathobiochemistry of the cartilage–bone interface in osteoarthritis. Osteoarthritic Disord. 205–217. Oloyede, A., Broom, N.D., 1994. Complex nature of stress inside loaded articular cartilage. Clin. Biomech. 4, 149–156. Oloyede, A., Flachsmann, R., Broom, N.D., 1992. The dramatic influence of loading velocity on the compressive response of articular cartilage. Connect. Tissue Res. 27, 211–224. Radin, E.L., 1999. Subchondral bone changes and cartilage damage. Equine Vet. J. 31, 94–95. Radin, E.L., Paul, I.L., 1971a. Importance of bone in sparing articular cartilage from impact. Clin. Orthop. Relat. Res. 78, 342–344. Radin, E.L., Paul, I.L., 1971b. Response of joints to impact loading. I. In vitro wear. Arthritis Rheum. 14, 356–362. Radin, E.L., Paul, I.L., Rose, R.M., 1972. Role of mechanical factors in pathogenesis of primary osteoarthritis. Lancet 1, 519–522. Repo, R.U., Finlay, J.B., 1977. Survival of articular cartilage after controlled impact. J. Bone Joint Surg. Am. 59, 1068–1076. Sandberg, R., Balkfors, B., Henricson, A., Westlin, N., 1986. Traumatic hemarthrosis in stable knees. Acta Orthop. Scand. 57, 516–517.
W. Kim et al. / Clinical Biomechanics 27 (2012) 637–645 Scott, C.C., Athanasiou, K.A., 2006. Mechanical impact and articular cartilage. Crit. Rev. Biomed. Eng. 34, 347–378. Shirazi, R., Shirazi-Adl, A., 2009. Computational biomechanics of articular cartilage of human knee joint: effect of osteochondral defects. J. Biomech. 42, 2458–2465. Thambyah, A., Broom, N., 2007. On how degeneration influences load-bearing in the cartilage–bone system: a microstructural and micromechanical study. Osteoarthritis Cartilage 15, 1410–1423. Thompson Jr., R.C., Oegema Jr., T.R., Lewis, J.L., Wallace, L., 1991. Osteoarthrotic changes after acute transarticular load. An animal model. J. Bone Joint Surg. Am. 73, 990–1001. Torzilli, P.A., Grigiene, R., Borrelli Jr., J., Helfet, D.L., 1999. Effect of impact load on articular cartilage: cell metabolism and viability, and matrix water content. J. Biomech. Eng. 121, 433–441.
645
Vellet, A.D., Marks, P.H., Fowler, P.J., Munro, T.G., 1991. Occult posttraumatic osteochondral lesions of the knee: prevalence, classification, and short-term sequelae evaluated with MR imaging. Radiology 178, 271–276. Vener, M.J., Thompson Jr., R.C., Lewis, J.L., Oegema Jr., T.R., 1992. Subchondral damage after acute transarticular loading: an in vitro model of joint injury. J. Orthop. Res. 10, 759–765. Verteramo, A., Seedhom, B.B., 2007. Effect of a single impact loading on the structure and mechanical properties of articular cartilage. J. Biomech. 40, 3580–3589. Walker, J.M., 1998. Pathomechanics and classification of cartilage lesions, facilitation of repair. J. Orthop. Sports Phys. Ther. 28, 216–231. Wilson, W., van Rietbergen, B., van Donkelaar, C.C., Huiskes, R., 2003. Pathways of loadinduced cartilage damage causing cartilage degeneration in the knee after meniscectomy. J. Biomech. 36, 845–851.
Contents lists available at SciVerse ScienceDirect
Clinical Biomechanics journal homepage: www.elsevier.com/locate/clinbiomech
Does prior sustained compression make cartilage-on-bone more vulnerable to trauma? Woong Kim, Ashvin Thambyah, Neil Broom ⁎ Tissue Mechanics Laboratory, Department of Chemical and Materials Engineering, University of Auckland, New Zealand
a r t i c l e
i n f o
Article history: Received 1 December 2011 Accepted 22 March 2012 Keywords: Cartilage-on-bone Prior creep loading Mechanical response to impact Quantification of induced structural damage
a b s t r a c t Background: This study investigated how varying levels of prior creep deformation in cartilage-on-bone samples influences their mechanical response and vulnerability to structural damage following a single traumatic impact. Methods: Bovine patellae were subjected to varying intervals of prior creep loading at a constant stress of 4 MPa. Immediately following removal of this stress the samples were impacted with a pendulum indenter system at a fixed energy of 2.2 J. Findings: With increasing prior creep, the peak force on impact rose, the duration of impact and time to reach peak force both decreased, and both the energy dissipated during impact and the magnitude of impulse were both unchanged by the level of prior creep. With increasing prior creep, the severity of impact-induced osteochondral damage increased: articular cartilage cracks penetrated to a greater depth, extending to the calcified cartilage layer resulting in hairline fractures or articular cartilage delamination and associated secondary damage to the vascular channels in the subchondral bone. Interpretation: The study shows that exposure of the cartilage-on-bone system to prior creep can significantly influence its response to subsequent impact, namely force attenuation and severity of damage to the articular cartilage, calcified cartilage and vascular channel network in the subchondral bone. © 2012 Elsevier Ltd. All rights reserved.
1. Introduction Articular cartilage (AC) is a relatively compliant connective tissue layer covering the diarthrodial joint surfaces (Ghadially, 1983; Meachim and Stockwell, 1979). The deformability of AC facilitates an increase in joint contact area and thus acts in a stress-reducing role. It also allows the joint to articulate with minimal friction (Brower and Hsu, 1969; Ghadially, 1983; Mankin and Radin, 1993; Meachim and Stockwell, 1979). The functional unit of AC is a dense collagenous network within which the hydrophilic proteoglycans are constrained. In its fully hydrated state AC contains some 60–80% water. At low rates of loading the deformation of AC occurs via the exclusion of this water out through an ultra-low permeability matrix, a mode of behaviour that has been described as poro-visco-hyperelastic (Oloyede et al., 1992). At high rates of loading, fluid movement is restricted or even prevented and the cartilage matrix behaves as an instantaneously elastic material (Flachsmann et al., 2001). AC is also known to be vulnerable to mechanical trauma (Radin, 1999; Radin and Paul, 1971b; Radin et al., 1972) and this can induce a posttraumatic osteoarthritis (OA) (Felson, 2004; Mink and Deutsch,
⁎ Corresponding author. E-mail address: [email protected] (N. Broom). 0268-0033/$ – see front matter © 2012 Elsevier Ltd. All rights reserved. doi:10.1016/j.clinbiomech.2012.03.007
1989; Vellet et al., 1991; Walker, 1998). Both in vivo and in vitro studies have sought to determine the levels of threshold stress, and especially impact, for cartilage damage (see review by Scott et al. (Scott and Athanasiou, 2006)). However, no rigorous body of in vivo data exists especially in regard to the level of stress, the stress rate and the duration of stress. Aspen et al. (Aspden et al., 2002) suggest that a non-pathological stress rate in vivo is between 100–1000 MPa/s with a duration of 30–150 ms. This contrasts with a traumatic impact, such as in a knee-dashboard injury which may have a duration of around 1 ms (Repo and Finlay, 1977). Under in vitro conditions it has been found that an impact load (single or repeated) of 11 to 36 MPa can cause visible cartilage damage, sometimes together with chondrocyte necrosis and apoptosis (Atkinson et al., 1998; Burgin and Aspden, 2007; Flachsmann et al., 2005; Haut, 1989; Jeffrey and Aspden, 2006; Kerin et al., 2003; Obeid et al., 1994; Repo and Finlay, 1977; Torzilli et al., 1999; Verteramo and Seedhom, 2007; Wilson et al., 2003). Many of these studies employed either a free-flight mass, drop tower or pendulum impacting system to create stress durations of less than 30 ms, and where peak forces or the energy of the impact have been reported (Scott and Athanasiou, 2006). Magnetic resonance imaging has been used to measure in vivo changes in cartilage thickness in human subjects before and after different exercise regimes (Eckstein et al., 1995, 1998, 1999, 2001, 2005, 2006; Mosher et al., 2010). Patellar cartilage thickness was
638
W. Kim et al. / Clinical Biomechanics 27 (2012) 637–645
loaded but having no prior deformation? This paper therefore reports on a study investigating how varying levels of prior creep deformation influence the vulnerability of cartilage-on-bone samples to a defined intensity of impact. 2. Methods 2.1. Tissues and test groups
Fig. 1. Percentage bar graph showing the relationship between the components of prior creep strain (black), impact strain (white) and remaining cartilage thickness (grey) as a function of duration of creep loading. Note that all strain/thickness values are rounded for clarity.
found to be reduced ~ 6% after 30 knee bends, ~5% after running and cycling, and ~ 3% after walking (Eckstein et al., 2006). In a study of femoral and tibial cartilage a decrease in thickness of up to 12% was reported following running for 30 min (Mosher et al., 2010). These studies suggest that in vivo AC, in some habitually loaded regions of the joint, may operate in a near continuously deformed state (~ 90% of its fully hydrated thickness), with even higher levels of deformation during more extreme physical activities. An interesting question therefore presents itself: are these habitually deformed regions of the joint more or less susceptible to a superimposed rapidly applied load than those regions similarly
Patellae free of gross AC defects were collected from freshly slain prime bovine bulls (~2 years old) and stored at − 20 °C. Prior to experimentation each patella was thawed in cold running water for 20 min and then two cartilage–bone samples with en face dimensions 14 mm × 14 mm and ~ 10 mm in height were sawn from the distal– lateral quadrant (see sites A and B in Supplemental Fig. 1A). This region was chosen because of its relative flatness and thus suitability for mechanical testing. A total of 160 samples obtained from 91 patellae were employed in this study, and allocated into eight groups of 20 samples with each group subjected to varying durations of creep. All samples were tested within 48 hours of initial thawing. Each sample was embedded into the well of a custom-made stainless steel holder using dental cement so as to align the articular surface perpendicular to the path of a pendulum indenter (Supplemental Fig. 2) as previously described (Flachsmann et al., 2005). The stainless steel indenter (Supplemental Fig. 1A) had a semi-cylindrical profile of radius ~2 mm and length 10 mm and positioned such that its initial contact with the sample was a line parallel to the cylinder axis along the cylinder surface. The indenter ends were slightly rounded to minimise the risk of edge-induced damage (Kerin et al., 1998). A flat-faced rectangular stainless-steel cap of en face
Fig. 2. Scatter plots showing relationships between prior creep strain and (A) impact strain, (B) peak force, (C) impact duration, (D) time to reach peak force, (E) impulse and (F) energy loss.
W. Kim et al. / Clinical Biomechanics 27 (2012) 637–645
639
Fig. 3. (A) En face view of impacted site in a sample exhibiting articular surface disruption revealed by India ink staining. Samples were bisected along the vertical dotted line. (B) Scanned cross-sectional image of the bisected sample showing that the impact-induced damage has extended into the full cartilage depth as indicated by the ink staining penetrating into the underlying subchondral bone. (C) Microphotograph of (B) showing both articular cartilage delamination along the tidemark (see arrow) and ink residues in the vascular channels (see arrowheads). The indenter size and profile relative to the sample size is also shown. (D) shows an enlargement of the boxed region in (C) near the extremity of the delamination. The India ink has flowed into the tip of a vascular channel (see white arrowhead) and spread into other interconnected vascular channels (see black arrowhead). Note that the impact does not appear to have fractured the subchondral plate.
dimensions 6 × 14 mm was designed to fit snugly over the semicylindrical impacting indenter (Supplemental Fig. 1A, lower). With this indenter cap in position, and prior to the dental cement setting, a pre-load of ~0.005 MPa was applied to the pendulum via a weight and pulley system so as to align the articular surface of the sample with this planar indenter cap (Supplemental Fig. 1B). This ensured that the contact between the indenter and cartilage surface of the sample was optimised. Once the dental cement had set the holder, with its embedded sample, was temporarily removed from the target base and rehydrated for at least 2 hours in 0.15 M saline at 4 °C.
equilibrium strain is achieved in 60 min and steady-state by 180 min (Oloyede and Broom, 1994). Creep displacements were recorded using a linear variable differential transformer (LVDT-Model DFg 2.5, Solartron Metrology, West Sussex, England) temporarily engaged with the pendulum arm, together with an A/D data-acquisition system. In order to prevent dehydration of the sample during the creep loading period (i.e. up to 180 min), a saline-soaked sponge was wrapped around the sample. Just prior to impacting the sponge was removed. 2.3. Impact loading
2.2. Creep loading In order to induce varying amounts of creep deformation in the cartilage prior to impacting (i.e. a physiologically relevant compression of ~10% (Eckstein et al., 2006; Mosher et al., 2010) and above), the samples were subjected to creep loading by means of a pulley and dead-weight assembly attached to the pendulum arm. The flat-faced rectangular cap was temporarily fitted over the impacting indenter and the sample then statically compressed at 4 MPa (Supplemental Fig. 1C) for prescribed durations of creep ranging from 0 min (control) to 0.2, 1, 5, 15, 30, 60 and 180 min. This produced an en face area of creep-deformed articular cartilage that extended beyond the projected area of the subsequent impact footprint (Supplemental Fig. 1D). The maximum period of creep loading was based on previous consolidation studies conducted in our laboratory demonstrating that although the strain vs time response of cartilage under static compression is highly non-linear, a substantial fraction of the
The indenter assembly incorporated a dynamic piezoelectric force transducer (Model 9021A, Kistler AG, Schweiz, Germany) whose signal was fed to a charge amplifier (Model 5015A, Kistler AG) and recorded at a sampling rate of 40 kHz using a data acquisition system. The indenter pendulum was released from a fixed height of 80 mm which gave an impact mean velocity of 1.13 m/s (SE, 0.003) associated with a mean kinetic energy Ek of 2.2 J (SE, 0.01) as estimated from its second moment of mass (I), angular velocity of pendulum (ω), linear velocity of indenter at impact (v), and radius of the pendulum arm (r; Eq. (1)). Ek ¼
1 2 1 v2 Iω ¼ I 2 2 r
ð1Þ
A high speed video camera (2000 fps; Fastcam MC 2.1, Photron, San Diego, USA) fitted with a macro lens and a 1 kW theatre spot light
640
W. Kim et al. / Clinical Biomechanics 27 (2012) 637–645
Fig. 4. Schematics illustrating damage scoring system developed to describe and quantify the osteochondral damage observed. (A) Illustrations of the four common damage types I, II, III and IV. Only Types III and IV were associated with vascular channel (VC) damage arising from either fracture of the calcified cartilage or delamination of the articular cartilage (see magnified insets). (B) and (C) show how articular cartilage fissures were scored with respect to their depth of penetration and number of branches.
with heat deflection were used to capture the progression of the impact. The camera was mounted so as to provide a side view of the indenter-cartilage contact and generate a video file of the impact event. The video data were analysed (i-Speed 3.0.2.6. Olympus NDT Inc. Waltham, MA, USA and Image J 1.43u, National Institutes of Health, Bethesda, USA) to obtain cartilage thickness, impact and rebound velocities (vi, vf), and impact deformation (Supplemental Fig. 1D). The kinetic energy lost (%) during the impact was calculated using a modified rotational kinetic energy equation (Eq. (2)) based on the kinetic energy delivered Ek, measured impact and rebound velocities vi and vf, angular velocity ω.
Energy lost ð%Þ ¼
ΔEkði−f Þ EkðiÞ
¼
v2i −v2f v2i
It should be emphasised that the indenter geometry was not intended to simulate actual joint loading. Rather, this geometry and the dimensions of the tissue sample were chosen as a means of delivering an impact of defined energy in which the extent of obvious tissue disruption was confined to an area well within the en face dimensions of the sample surface. The aim was to obtain an accurate comparison of disruption at the microstructural level between the non-creep and creep loaded states. In effect our choice of indenter served to confine the extent of ‘injury’ to a manageable tissue volume.
2.4. Macroscopic assessment of damage ð2Þ
The force–time data provided by the stress transducer was used to determine the peak force, time to peak force, duration of impact and magnitude of impulse, the latter calculated by numerical integration of the force–time curve. Following impact the cartilage surface was wet-rubbed with India ink, washed with running water, and then any AS damage revealed by the ink staining recorded using macrophotography.
Samples were fixed in 10% formalin for approximately 24 hours then mildly decalcified using 10% formic acid for 3 to 5 days (Ippolito et al., 1981; Thambyah and Broom, 2007) and finally rinsed in water for 1 hour to remove residual chemicals. A cross-sectional cut was then made in order to bisect the cartilage–bone sample at a point approximately midway along the length of the impacted site (Supplemental Fig. 1E, F). The exposed cross-section was then scanned using a flatbed scanner (4800 dpi; Canoscan F8800, Canon, Tokyo, Japan) to obtain high resolution macro images of the damage morphologies (number of fissures, penetration depth and presence of
W. Kim et al. / Clinical Biomechanics 27 (2012) 637–645
641
Fig. 5. (A) Sample exhibiting all four damage types: Type I fissures marked with arrow, Type II cartilage chips (see shaded regions), Type III calcified cartilage fracture (white arrowhead) and Type IV delamination (black arrowhead). (B) Higher magnification view of LH boxed region in (A) showing Type III damage involved calcified cartilage CC fracture propagating to the tip of a vascular channel VC (see arrowheads) and leading to ink penetration into the SCB. (C) Magnified view of boxed area in (B). (D) Magnified view of RH boxed region in (A) showing Type IV cartilage delamination that has resulted in three VCs being damaged and leading to ink penetration of the VC network (see arrowheads). (E) Magnified image of a region of delamination showing its typical zigzag fracture morphology intercepting a VC (arrowhead).
delamination) in the hyaline cartilage and any India ink staining in the subchondral bone (SCB).
2.5. Microscopic assessment of damage Those samples exhibiting ink penetration into the SCB were then sectioned using a freezing microtome to obtain 20–30 osteochondral serial sections (~ 30 μm thick). At least 10 sections per sample were examined in their fully hydrated state using both bright field (Nikon AZ100, Nikon Instruments, Melville, NY, USA) and differential interference contrast optical microscopy (DIC, Nikon Eclipse 80i, Nikon Instruments) in order to assess the degree of impact-induced damage in the calcified cartilage (CC) and SCB.
2.6. Impact damage types and quantification system From a preliminary set of experiments a scoring system was developed to classify and quantify the modes and severity of damage observed in the AC, CC, osteochondral junction (OCJ) and SCB (see Section 3). 2.7. Statistical analysis Pearson's correlation tests (SPSS, IBM, Armonk, NY, USA) were used to investigate correlations between the mechanical parameters (level of creep strain, impact strain, peak force, time to peak force, impact duration, impulse and percentage energy lost) and their relationship to the damage severity and type.
642
W. Kim et al. / Clinical Biomechanics 27 (2012) 637–645
decreased from 0.8 to 0.6 ms (r = − 0.584, P b 0.05, Fig. 2D). Conversely, neither the impulse (r = 0.120, P = 0.128, Fig. 2E) nor percentage energy loss with impact (r = 0.203, P b 0.05, Fig. 2F) correlated with the level of prior creep strain. 3.2. Quantification of structural damage Fig. 3A shows an en face view of an impacted sample exhibiting AS damage along the contact line of the indenter's cylindrical axis as revealed by India ink staining. A scanned cross-section through this damage (Fig. 3B) indicates that fissures have propagated through the full thickness of the AC and into the calcified cartilage (CC). Higher magnification views of sections taken from this same sample are shown in Fig. 3C and D. The scoring system devised to quantify the impact damage (Fig. 4) took into consideration the depth of fissure penetration, the number of fissures found per sample and the type of damage observed. With reference to the set of higher magnification images shown in Fig. 5A–E the four common types of damage observed in the impacted samples were as follows:
Fig. 6. Venn diagram showing frequencies of the damage Types I, II, III and IV. Note that the shaded region represents the sample numbers associated with ink staining in the subchondral bone.
3. Results The cohort mean thickness of the articular cartilage in the 160 original samples prepared for testing was 1.60 mm (SE 0.02 mm). 3.1. Mechanical response to impact Supplemental Fig. 2 shows a typical force vs time curve obtained from a sample subjected to 5 min of prior creep at 4 MPa. The asymmetry of the curve was a consistent feature of all tests and possibly arises from rig-specific characteristics, the inertia of the tissue, its embedded plaster matrix/holder, flexion in the pendulum arm and associated stored energy dissipation. Although this asymmetry in the force vs time response was not investigated any further in the study, its consistency throughout all of the impact tests reassured us that a comparison of responses between the various test groups was, in fact, legitimate. The video images in Supplemental Fig. 3 compare side views of the same sample as in Supplemental Fig. 2 at both zero and peak impact load. The non-planar contact geometry does not permit accurate determination of the impact stresses in the AC. However, the video images do provide an approximate profile of the AC/indenter contact at peak force. Averaged over the 160 samples tested the projected contact area was ~ 30 mm 2, yielding an average peak impact stress of ~ 42 MPa. The bar graph in Fig. 1 shows the averaged values of (a) induced creep strain, (b) subsequent impact strain and (c) remaining deformed AC thickness (normalised to its original un-deformed thickness) for the samples in each prior-creep deformed group (i.e. 0 to 180 min). With increasing duration of creep the creep strain increased progressively to 62%, while the subsequent impact strains decreased from 34% to 16% (SE, 1%). There was no significant change in the levels of either creep or impact strain from 60 to 180 min. Fig. 2 shows scatter plots for the various mechanical parameters as a function of creep strain. There was a strong inverse linear relationship between creep strain and impact strain (Pearson's coefficient r = −0.707, P b 0.05, Fig. 2A). With increasing creep strain the mean peak force increased from 1249 N to 1388 N (r = 0.353, P b 0.05, Fig. 2B), the duration of impact decreased from 4.1 ms to 3.4 ms (r = − 0.544, P b 0.05, Fig. 2C) and the time to reach peak force
Type I AC fissure damage: the formation of a single fissure or branching fissures in the AC (Figs. 4A, C, 5A) but without either AC loss or delamination. Type II AC chip damage: AC loss which occurred when two “fissures” intersected (Figs. 4A, 5A). Type III CC fracture: the penetration of a hairline crack into the CC layer and intercepting a single vascular channel (VC) (Figs. 4A, 5A–C), which allowed the ink stain to spread along the interconnected VC network and then focally stain the SCB (Fig. 3B). Type IV AC delamination: AC delaminating along the tidemark (Figs. 4A, 5A, D, E) resulting in damage to the tip of any VC intercepted by the propagating crack creating the delamination. Again, this led to ink staining of the SCB (Fig. 3B). The severity of damage was further quantified from the depth of fissure penetration relative to the full AC thickness (scored as less than 15%, 25%, 50%, 75% or 100%) (Fig. 4B), and the frequency of AC fissures per sample by summing the number of main fissures and fissure branches (Fig. 4C). The Venn diagram in Fig. 6 summarises the frequency of the four damage types. All 160 samples suffered at least Type I damage. Of these 160 samples 55 suffered, in addition, various combinations of Types II, III and IV damage as indicated. A total of 25 samples exhibited SCB ink staining and these all contained either CC fracture (i.e. Type III) or AC delamination (i.e. Type IV), or both. Further, none of these 25 samples exhibited any detectable damage in the SCB. The depth of fissure penetration in the AC was positively associated with the level of prior creep strain (Fig. 7A; r = 0.805). The frequency of Type III and IV damage (and the associated SCB staining) increased only when the prior creep strain reached 30% or higher (Fig. 7D–F). Conversely, there was no obvious correlation between the level of creep strain and the number of AC fissures found per sample (r = 0.266; Fig. 7B) or the frequency of AC chip damage (r = 0.007; Fig. 7C). 4. Discussion This study has shown that under the impact conditions employed the level of prior creep in the AC is a critical factor in influencing both the mechanical response and susceptibility to trauma of the osteochondral tissues. With increasing creep strain the peak impact force transmitted to the SCB increases whereas both the time to reach peak force and the duration of impact decrease (Fig. 2B, C, D). Further, with increasing creep strain the depth of penetration of the AC fissures increased (Fig. 7A). Creep strains of 30% or higher were
W. Kim et al. / Clinical Biomechanics 27 (2012) 637–645
643
Fig. 7. Percentage bar graphs indicating severity of damage (see A, B) and types of damage (see C–F) as a function of creep strain (expressed in ranges of 10%). The numbers of samples affected are indicated inside their respective bars. Percentage bar graph for Type I damage was omitted since all 160 samples were affected.
associated with an increasing frequency of both Type III and IV damage involving the CC and VCs (Fig. 7D, E, F). By contrast, the level of prior creep appeared to have little or no influence on the impulse or amount of energy lost during the impact (Fig. 2E, F), and this suggests cartilage itself plays a minimal role in energy or shock absorption. This finding is consistent with the much earlier postulate of Radin and Paul (1971a) and the whole joint studies of Hoshino and Wallace (1987) that the underlying bone is the main contributor to shock absorption in the joint. Also, there was no obvious association between the level of prior creep and either the number of fissures found per sample or the frequency of AC chip damage (Fig. 7B, C). This suggests that both Types I and II modes of damage were, in part, indenter-geometry dependent. Conversely, Types III and IV modes involving damage to the CC would have arisen from forces transmitted through the impacted AC rather than by the action of the indenter cutting directly into the CC layer. In fact, even at the near-equilibrium creep strain of 62% the additional strain of 16% from the subsequent impact still left a residual thickness of AC which was at least 22% of the original AC thickness (Fig. 1). It is the transmission of force through this residual AC thickness that would have inflicted the observed damage in the CC.
With the impacting energy employed in the present study (~2.2 J), disruption of the calcified cartilage and vascular channels (VC) commenced at a prior creep strain of around 30%, a value that considerably exceeds the physiological range of 3–12% measured in human knee AC in vivo (Eckstein et al., 2006; Mosher et al., 2010). Our data do, however, suggest that high stress impulsive loading of the joint tissues following a prolonged period of sustained loading may increase the risk of osteochondral damage (as shown by damage Types III and IV) and that this may in turn result in haemorrhaging into the SCB (Ahstrom, 1965; Armstrong and Mow, 1982; Bauer and Jackson, 1988; Brown and Cruess, 1982; Butler and Andrews, 1988; Gillquist et al., 1977; Noyes et al., 1980; Sandberg et al., 1986). Such defects have been also shown to affect the normal joint compliance, contact pressures and cartilage strains in both tibia and femur as demonstrated using finite element modelling (Shirazi and ShiraziAdl, 2009) and known to be associated with the development of a secondary osteoarthrosis (Oegema and Thompson, 1995; Thompson et al., 1991). A novel aspect of the present study was the effectiveness of ink staining in the SCB as an indicator of the extent to which CC fracture (Type III), AC delamination (Type IV) and VC disruption resulted from the impacting event (Fig. 3). Bauer and Jackson reported that
644
W. Kim et al. / Clinical Biomechanics 27 (2012) 637–645
osteochondral damage in the human knee, similar to that observed in the present investigation, can occur from rotational or direct impact trauma (Bauer and Jackson, 1988). They also noted that this type of subchondral damage can lead to a haemarthrosis or bleeding into the joint space (Ahstrom, 1965; Armstrong and Mow, 1982; Bauer and Jackson, 1988; Brown and Cruess, 1982; Butler and Andrews, 1988; Gillquist et al., 1977; Noyes et al., 1980; Sandberg et al., 1986). From a clinical perspective, hair-line fissures in the otherwise normal appearing articular cartilage would not be detected using routinely performed arthroscopy or radiographic diagnosis. And yet, as the present study demonstrates, such subtle damage can be associated with significant disruption to both the calcified cartilage and vascular channels originating in the subchondral bone. AC, being avascular, has a relatively passive response to injury. Conversely, trauma to the underlying vascularised subchondral bone can induce a complex inflammatory response that may lead to the eventual degeneration of the joint (Oegema and Thompson, 1995; Thompson et al., 1991). The average peak impact stress of ~ 42 MPa calculated from the force data and high speed video images is not dissimilar to levels of peak stress experienced in vivo during certain relatively common types of trauma where 25–30 MPa have been recorded in the human patellofemoral joint (Finlay and Repo, 1978; Haut, 1989; Repo and Finlay, 1977; Scott and Athanasiou, 2006). Interestingly Vener et al. (Vener et al., 1992), using a peak impact stress of 40 MPa, reported both CC and SCB fracture in canine metacarpophalangeal and metatarsophalangeal joints. This contrasts with our present findings in which no subchondral damage was observed despite the high peak forces employed. 5. Conclusions With the aim of developing a more complete understanding of the impact behaviour of the cartilage–bone system this study has demonstrated that the level of prior creep in articular cartilage does have a significant influence both on the mechanical response of the integrated cartilage–bone system and the modes and severity of damage induced in the articular cartilage, the calcified cartilage and vascular channels. Conflict of interest statement We authors (Drs. Woong Kim, Neil D. Broom and Ashvin Thambyah) of the manuscript disclose that we have no financial and personal relationships with other people or organisations that could inappropriately influence (bias) their work. Supplementary materials related to this article can be found online at doi:10.1016/j.clinbiomech.2012.03.007. Acknowledgements Funding for this research was generously provided by the Auckland Medical Research Foundation (AMRF). References Ahstrom Jr., J.P., 1965. Osteochondral fracture in the knee joint associated with hypermobility and dislocation of the patella. Report of eighteen cases. J. Bone Joint Surg. Am. 47, 1491–1502. Armstrong, C.G., Mow, V.C., 1982. Variations in the intrinsic mechanical properties of human articular cartilage with age, degeneration, and water content. J. Bone Joint Surg. Am. 64, 88–94. Aspden, R.M., Jeffrey, J.E., Burgin, L.V., 2002. Impact loading of articular cartilage. Osteoarthritis Cartilage 10, 588–589 discussion 590. Atkinson, T.S., Haut, R.C., Altiero, N.J., 1998. Impact-induced fissuring of articular cartilage: an investigation of failure criteria. J. Biomech. Eng. 120, 181–187. Bauer, M., Jackson, R.W., 1988. Chondral lesions of the femoral condyles: a system of arthroscopic classification. Arthroscopy 4, 97–102.
Brower, T.D., Hsu, W.Y., 1969. Normal articular cartilage. Clin. Orthop. Relat. Res. 64, 9–17. Brown, K.L., Cruess, R.L., 1982. Bone and cartilage transplantation in orthopaedic surgery. A review. J. Bone Joint Surg. Am. 64, 270–279. Burgin, L.V., Aspden, R.M., 2007. A drop tower for controlled impact testing of biological tissues. Med. Eng. Phys. 29, 525–530. Butler, J.C., Andrews, J.R., 1988. The role of arthroscopic surgery in the evaluation of acute traumatic hemarthrosis of the knee. Clin. Orthop. Relat. Res. 150–152. Eckstein, F., Sittek, H., Milz, S., Schulte, E., Kiefer, B., Reiser, M., et al., 1995. The potential of magnetic resonance imaging (MRI) for quantifying articular cartilage thickness —a methodological study. Clin. Biomech. 10, 434–440. Eckstein, F., Tieschky, M., Faber, S.C., Haubner, M., Kolem, H., Englmeier, K.H., et al., 1998. Effect of physical exercise on cartilage volume and thickness in vivo: MR imaging study. Radiology 207, 243–248. Eckstein, F., Tieschky, M., Faber, S., Englmeier, K.H., Reiser, M., 1999. Functional analysis of articular cartilage deformation, recovery, and fluid flow following dynamic exercise in vivo. Anat. Embryol. (Berl) 200, 419–424. Eckstein, F., Reiser, M., Englmeier, K.H., Putz, R., 2001. In vivo morphometry and functional analysis of human articular cartilage with quantitative magnetic resonance imaging—from image to data, from data to theory. Anat. Embryol. (Berl) 203, 147–173. Eckstein, F., Lemberger, B., Gratzke, C., Hudelmaier, M., Glaser, C., Englmeier, K.H., et al., 2005. In vivo cartilage deformation after different types of activity and its dependence on physical training status. Ann. Rheum. Dis. 64, 291–295. Eckstein, F., Hudelmaier, M., Putz, R., 2006. The effects of exercise on human articular cartilage. J. Anat. 208, 491–512. Felson, D.T., 2004. Risk factors for osteoarthritis: understanding joint vulnerability. Clin. Orthop. Relat. Res. S16–S21. Finlay, J.B., Repo, R.U., 1978. Instrumentation and procedure for the controlled impact of articular cartilage. IEEE Trans. Biomed. Eng. 25, 34–39. Flachsmann, R., Broom, N.D., Hardy, A.E., 2001. Deformation and rupture of the articular surface under dynamic and static compression. J. Orthop. Res. 19, 1131–1139. Flachsmann, R., Kim, W., Broom, N., 2005. Vulnerability to rupture of the intact articular surface with respect to age and proximity to site of fibrillation: a dynamic and static-investigation. Connect. Tissue Res. 46, 159–169. Ghadially, F.N., 1983. Fine structure of synovial joints : a text and atlas of the ultrastructure of normal and pathological articular tissues. Butterworths, London, Boston. Gillquist, J., Hagberg, G., Oretorp, N., 1977. Arthroscopy in acute injuries of the knee joint. Acta Orthop. Scand. 48, 190–196. Haut, R.C., 1989. Contact pressures in the patellofemoral joint during impact loading on the human flexed knee. J. Orthop. Res. 7, 272–280. Hoshino, A., Wallace, W.A., 1987. Impact-absorbing properties of the human knee. J. Bone Joint Surg. Br. 69, 807–811. Ippolito, E., LaVelle, S., Pedrini, V., 1981. The effect of various decalcifying agents on cartilage proteoglycans. Stain Technol. 56, 367–373. Jeffrey, J.E., Aspden, R.M., 2006. The biophysical effects of a single impact load on human and bovine articular cartilage. Proc. Inst. Mech. Eng. H 220, 677–686. Kerin, A.J., Wisnom, M.R., Adams, M.A., 1998. The compressive strength of articular cartilage. Proc. Inst. Mech. Eng. H 212, 273–280. Kerin, A.J., Coleman, A., Wisnom, M.R., Adams, M.A., 2003. Propagation of surface fissures in articular cartilage in response to cyclic loading in vitro. Clin. Biomech. 18, 960–968. Mankin, H.J., Radin, E.L., 1993. Structure and function of joints, Arthritis and Allied Conditions, 12th Ed. , pp. 181–197. Meachim, G., Stockwell, R., 1979. The matrix. In: Freeman, M.A.R. (Ed.), Adult Articular Cartilage. Pitman Medical, New York. 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. Mosher, T.J., Liu, Y., Torok, C.M., 2010. Functional cartilage MRI T2 mapping: evaluating the effect of age and training on knee cartilage response to running. Osteoarthritis Cartilage 18, 358–364. Noyes, F.R., Bassett, R.W., Grood, E.S., Butler, D.L., 1980. Arthroscopy in acute traumatic hemarthrosis of the knee. Incidence of anterior cruciate tears and other injuries. J. Bone Joint Surg. Am. 62 (687–95), 757. Obeid, E.M., Adams, M.A., Newman, J.H., 1994. Mechanical properties of articular cartilage in knees with unicompartmental osteoarthritis. J. Bone Joint Surg. Br. 76, 315–319. Oegema, T.R., Thompson, R.C., 1995. Histopathology and pathobiochemistry of the cartilage–bone interface in osteoarthritis. Osteoarthritic Disord. 205–217. Oloyede, A., Broom, N.D., 1994. Complex nature of stress inside loaded articular cartilage. Clin. Biomech. 4, 149–156. Oloyede, A., Flachsmann, R., Broom, N.D., 1992. The dramatic influence of loading velocity on the compressive response of articular cartilage. Connect. Tissue Res. 27, 211–224. Radin, E.L., 1999. Subchondral bone changes and cartilage damage. Equine Vet. J. 31, 94–95. Radin, E.L., Paul, I.L., 1971a. Importance of bone in sparing articular cartilage from impact. Clin. Orthop. Relat. Res. 78, 342–344. Radin, E.L., Paul, I.L., 1971b. Response of joints to impact loading. I. In vitro wear. Arthritis Rheum. 14, 356–362. Radin, E.L., Paul, I.L., Rose, R.M., 1972. Role of mechanical factors in pathogenesis of primary osteoarthritis. Lancet 1, 519–522. Repo, R.U., Finlay, J.B., 1977. Survival of articular cartilage after controlled impact. J. Bone Joint Surg. Am. 59, 1068–1076. Sandberg, R., Balkfors, B., Henricson, A., Westlin, N., 1986. Traumatic hemarthrosis in stable knees. Acta Orthop. Scand. 57, 516–517.
W. Kim et al. / Clinical Biomechanics 27 (2012) 637–645 Scott, C.C., Athanasiou, K.A., 2006. Mechanical impact and articular cartilage. Crit. Rev. Biomed. Eng. 34, 347–378. Shirazi, R., Shirazi-Adl, A., 2009. Computational biomechanics of articular cartilage of human knee joint: effect of osteochondral defects. J. Biomech. 42, 2458–2465. Thambyah, A., Broom, N., 2007. On how degeneration influences load-bearing in the cartilage–bone system: a microstructural and micromechanical study. Osteoarthritis Cartilage 15, 1410–1423. Thompson Jr., R.C., Oegema Jr., T.R., Lewis, J.L., Wallace, L., 1991. Osteoarthrotic changes after acute transarticular load. An animal model. J. Bone Joint Surg. Am. 73, 990–1001. Torzilli, P.A., Grigiene, R., Borrelli Jr., J., Helfet, D.L., 1999. Effect of impact load on articular cartilage: cell metabolism and viability, and matrix water content. J. Biomech. Eng. 121, 433–441.
645
Vellet, A.D., Marks, P.H., Fowler, P.J., Munro, T.G., 1991. Occult posttraumatic osteochondral lesions of the knee: prevalence, classification, and short-term sequelae evaluated with MR imaging. Radiology 178, 271–276. Vener, M.J., Thompson Jr., R.C., Lewis, J.L., Oegema Jr., T.R., 1992. Subchondral damage after acute transarticular loading: an in vitro model of joint injury. J. Orthop. Res. 10, 759–765. Verteramo, A., Seedhom, B.B., 2007. Effect of a single impact loading on the structure and mechanical properties of articular cartilage. J. Biomech. 40, 3580–3589. Walker, J.M., 1998. Pathomechanics and classification of cartilage lesions, facilitation of repair. J. Orthop. Sports Phys. Ther. 28, 216–231. Wilson, W., van Rietbergen, B., van Donkelaar, C.C., Huiskes, R., 2003. Pathways of loadinduced cartilage damage causing cartilage degeneration in the knee after meniscectomy. J. Biomech. 36, 845–851.