- Email: [email protected]
In vitro assessment of the contact mechanics of reverse-engineered distal humeral hemiarthroplasty prostheses
Clinical Biomechanics 29 (2014) 990–996
Contents lists available at ScienceDirect
Clinical Biomechanics journal homepage: www.elsevier.com/locate/clinbiomech
In vitro assessment of the contact mechanics of reverse-engineered distal humeral hemiarthroplasty prostheses Ryan Willing a,b,⁎, Michael Lapner a,1, Graham J.W. King a,2,3, James A. Johnson a,3,4,5 a b
Bioengineering Research Laboratory, Roth | McFarlane Hand and Upper Limb Centre, St. Joseph's Health Care, The University of Western Ontario, London, ON, Canada Department of Mechanical Engineering, Binghamton University, Binghamton, NY, USA
a r t i c l e
i n f o
Article history: Received 28 April 2014 Accepted 13 August 2014 Keywords: Elbow Distal humerus Hemiarthroplasty Contact pattern Reverse-engineered
a b s t r a c t Background: Distal humeral hemiarthroplasty alters cartilage contact mechanics, which may predispose to osteoarthritis. Current prostheses do not replicate the native anatomy, and therefore contribute to these changes. We hypothesized that prostheses reverse-engineered from the native bone shape would provide similar contact patterns as the native articulation. Methods: Reverse-engineered hemiarthroplasty prostheses were manufactured for five cadaveric elbows based on CT images of the distal humerus. Passive flexion trials with constant muscle forces were performed with the native articulation intact while bone motions were recorded using a motion tracking system. Motion trials were then repeated after the distal humerus was replaced with a corresponding reverse-engineered prosthesis. Contact areas and patterns were reconstructed using computer models created from CT scan images combined with the motion tracker data. The total contact areas, as well as the contact area within smaller sub-regions of the ulna and radius, were analyzed for changes resulting from hemiarthroplasty using repeated-measures ANOVAs. Findings: Contact area at the ulna and radius decreased on average 42% (SD 19%, P = .008) and 41% (SD 42%, P = .096), respectively. Contact area decreases were not uniform throughout the different sub-regions, suggesting that contact patterns were also altered. Interpretation: Reverse-engineered prostheses did not reproduce the same contact pattern as the native joints, possibly because the thickness of the distal humerus cartilage layer was neglected when generating the prosthesis shapes or as a consequence of the increased stiffness of the metallic implants. Alternative design strategies and materials for hemiarthroplasty should be considered in future work. © 2014 Elsevier Ltd. All rights reserved.
1. Introduction Distal humeral hemiarthroplasty (DHH) is a treatment option for the management of comminuted distal humeral fractures (Burkhart et al., 2011; Hohman et al., 2014; Smith and Hughes, 2013). DHH places a metal prosthesis in direct contact with the native articular surfaces of the radial head and greater sigmoid notch of the ulna. The increased ⁎ Corresponding author at: Mechanical Engineering Department, Thomas J. Watson School of Engineering & Applied Science, Binghamton University — SUNY, P.O. Box 6000, Binghamton, NY 13902-6000, USA. E-mail address: [email protected] (R. Willing). 1 Division of Orthopedic Surgery, Sturgeon Hospital, University of Alberta, 201 Boudreau Rd, St. Albert, AB T8N 6C4, Canada. 2 Department of Surgery, The University of Western Ontario, 1151 Richmond St, London, ON N6A 3K7, Canada. 3 Roth | McFarlane Hand and Upper Limb Centre Bioengineering Laboratory, St. Joseph's Health Care, 268 Grosvenor St, London, ON N6A 4L6, Canada. 4 Biomedical Engineering Department, The University of Western Ontario, 1151 Richmond St, London, ON N6A 3K7, Canada. 5 Department of Mechanical and Materials Engineering, The University of Western Ontario, 1151 Richmond St, London, ON N6A 3K7, Canada.
http://dx.doi.org/10.1016/j.clinbiomech.2014.08.015 0268-0033/© 2014 Elsevier Ltd. All rights reserved.
stiffness of metal hemiarthroplasty devices relative to cartilage typically results in a decrease in contact area (Liew et al., 2003; Sabo et al., 2011). This may increase cartilage stresses due to the altered contact mechanics, resulting in cartilage wear and post-traumatic osteoarthritis (Dalldorf et al., 1995; Smith and Hughes, 2013). Reproducing the native contact patterns after DHH could reduce the likelihood of cartilage erosion. In an effort to recreate more natural cartilage contact mechanics, “anatomical” DHH designs have been developed and are commercially available (Hohman et al., 2014; Lapner et al., 2014). These implants attempt to mimic the native anatomy of the distal humerus; however, their shapes are simplified for general use. An in-vitro study by Lapner et al. (2014) showed that mean ulnohumeral and radiocapitellar contact areas both decreased as the result of DHH with a commercially available axisymmetric prosthesis (decreased by 44%, P = 0.03, and 4%, P = 0.07, respectively). The resulting contact pattern was different than that of the native joint, and varying prosthesis size did not have a significant effect on contact area or pattern. It is possible to create anatomically accurate reverse-engineered implants on a patient-specific basis using pre-operative 3-D imaging modalities. Patient specific hip (Leichtle et al., 2012; Reize and Wulker, 2007), knee (Fitz, 2009), spine (de Beer
R. Willing et al. / Clinical Biomechanics 29 (2014) 990–996
and Scheffer, 2012), and cranial (Sundseth and Berg-Johnsen, 2013) prosthesis components, as well as patient-specific cutting guides for total knee replacement (Cenni et al., 2014; Fitz, 2009; Slamin and Parsley, 2012) have been designed from pre-operative medical imaging. The efficacy of patient-specific reverse-engineered hemiarthroplasty prostheses for promoting more natural contact mechanics, particularly for DHH, has not been reported. While reductions in contact area can be expected due to the increased stiffness of these metal prostheses compared to the cartilage surfaces they are replacing, patient-specific reverse-engineered prostheses could result in contact patterns similar to those of the native joint. The objective of this study was to quantify the changes in cartilage contact which occur after DHH with a subject-specific reverseengineered prosthesis. We hypothesized that DHH with an anatomically shaped reverse-engineered prostheses would result in a reduction in contact area, but contact patterns similar in shape to those of a native joint. 2. Methods 2.1. Reverse-engineered DHH design CT images of 5 fresh-frozen intact left arm specimens (3 male, 2 female, average age 74 yrs, range: 66–84 yrs) with no radiographic evidence of joint pathology were obtained using a GE Discovery CT750 HD scanner (GE Health Care, Pewaukee, WI, USA) at 120 kV and 292 mAs with a slice thickness of 0.625 mm (in-plane pixel sizes ranging from 0.492 mm–0.586 mm). The CT data was imported into Mimics v14.12 (Materialise, Leuven, Belgium), and the distal humeral bone geometry was extracted using threshold based segmentation, which included any voxel with an attenuation value of 250 HU or greater (Lapner et al., 2014; Willing et al., 2013, 2014). Models were wrapped, exported in the stereolithography (STL) format, and remeshed using a radial basis function in Matlab (The Mathworks, Natick, MA, USA). The resulting models comprised uniformly sized triangles with approximately 0.4 mm edge lengths. A Boolean geometry subtract operation was performed using custom Blender script (The Blender Foundation, Amsterdam, NL), which cropped the model to the articular region and created interface geometry for attaching a custom stem component (Fig. 1, left). A sPro™ 125 direct metal selective laser melting (SLM) machine (3D Systems Corp., Rock Hill, SC, USA) was used to manufacture stainless steel prosthesis prototypes based on the computer models (Fig. 1, middle). The articular surfaces were sanded and polished to a mirror finish using a rotary hand tool (Fig. 1, right). A MicroScribe G2X digitizer (Immersion Corp., San Jose, CA, USA) was traced across the entire contact surface of the polished prosthesis prototype, and the surface geometry was saved as a 3D point cloud. The computer model was aligned to the 3D point cloud using an iterative closest point (ICP) algorithm (Besl, 1992), and the distance from each digitized point to the closest point on the computer model was measured to be, on average, 0.08 mm (SD 0.26 mm).
991
2.2. Specimen preparation Prior to testing, specimens were amputated mid-humerus and stored at room temperature for approximately 20 h to fully thaw. The radius and ulna were pinned together in neutral forearm rotation. A 10 cm segment of the humeral shaft was denuded of all soft tissues and was rigidly clamped to a mounting fixture. Optotrak Certus (Northern Digital Inc., Waterloo, Ontario, Canada) infrared marker triads were affixed to the shafts of the ulna and radius, and the humerus clamp. A capsulectomy was performed at the elbow joint, and the medial and lateral collateral ligaments (MCL and LCL) were released at their respective origins on the humerus, which allowed complete disarticulation of the elbow joint for exposure of the distal humeral articular surface. 2.3. Testing apparatus The arm was attached to an elbow testing apparatus (via the humerus clamp; Fig. 2). Pneumatic actuators (Bimba Co., University Park, IL, USA) provided simulated muscle tensions FBIC, FBRA and FTRI via stainless steel aircraft cables sutured to the biceps, brachialis and triceps tendons, respectively. The flexion–extension axis of the humerus was defined by connecting the center of a circle fit to the trochlear sulcus, to the center of a sphere fit of the capitellum (London, 1981; Shiba et al., 1988) using surface digitization data obtained by the optically tracked stylus tool. A 5 mm diameter hole was drilled medial-laterally through the distal humerus along the flexion–extension axis using computer navigation and a 5 mm diameter cylindrical brass ligament guide was installed (Willing et al., 2014). Heavy braided sutures were attached to the free ends of the LCL and MCL, routed over the ligament guide (Fig. 2, inset), and tensioned to 20 N. Constant ligament tension, rather than a constant ligament length, simulated perfect ligament balance and ensured that the joint reaction force passing through the articulation would be insensitive to prosthesis alignment. 2.4. Testing protocol Motion was recorded while the arm was guided through passive flexion motions (3 repetitions, 0°–120°) with ligament loads applied. Biceps, brachialis and triceps muscle tensions of 20 N, 20 N and 40 N were applied based on previously reported ratios (Morrey et al., 1991). The optical motion capture system tracked bone motion in 3D within 0.1 mm positional accuracy. Following the trials with the native articulation, four spherical 20 mm nylon fiducial markers were attached to the distal humerus. The fiducial locations were digitized with respect to the humerus motion tracker using an optically tracked stylus tool. The distal humerus, as well as the attached fiducial markers, was then excised from the humeral diaphysis using an oscillating saw and stored in normal saline. Using computer navigation, a custom distal humerus stem component was guided into place such that the prosthesis flexion axis aligned with the flexion axis of the native articulation (average error = 3.6°,
Fig. 1. Creation of reverse-engineered prostheses. Left: Computer model of reverse engineered hemiarthroplasty prosthesis created from CT imagery. Middle: Raw part manufactured using direct metal SLM. Right: Final polished prototype ready for testing.
992
R. Willing et al. / Clinical Biomechanics 29 (2014) 990–996
reported average surface alignment errors of 0.15 mm (maximum: 0.23 mm) using this registration technique (Willing et al., 2014). Displacements and rotations were applied to the corresponding 3D computer models using the motion tracker data recorded during the motion trials. Once models were positioned with respect to one another, contact surfaces would overlap (because the models represented the un-deformed cartilage layer, which would actually compress during contact) (Fig. 3). Contact was assumed to occur wherever cartilage– cartilage or prosthesis–cartilage overlap occurred. This contact measurement technique was validated in a previous study, where a good correlation between cast-measured and overlap-based contact area measurements was observed (ICC = 0.727), contact area error was typically less than 15%, and the changes in joint contact due to DHH could be reliably detected (Willing et al., 2014). Contact area was calculated in 15° increments from 0° to 120° for each motion (3 flexion trials) and joint configuration (native and hemi) at the ulna and radius. Results from repeated trials of the same motion were averaged. The areas were recorded in terms of percentage of the total available contact area in order to normalize for different specimen sizes.
2.6. Data analysis
Fig. 2. Elbow testing apparatus. Pneumatic actuators provided simulated muscle forces for the biceps, brachialis, and triceps. Ligament forces were provided by hanging weights from cable sutures to the ligaments, routed to pass over the ligament guide.
maximum error = 6.0°) and the prosthesis trochlear sulcus was in the same position as the native trochlear sulcus (average error = 3.1 mm, maximum error = 5.1 mm). The specimen-specific reverseengineered hemiarthroplasty prosthesis was installed onto the stem. The motion and loading protocol used for the native articulation were repeated with the prosthesis in place. Ligament lines of action were always directed towards the flexion axis, which was referenced to the corresponding native or prosthesis contact surfaces, so joint contact mechanics were insensitive to small errors in stem positioning. After testing, the ulna and radius were separated and denuded. The articular surfaces of the ulna, radius and prosthesis were digitized using the stylus tool. Four fiducial markers were attached to each of the proximal radius and ulna, and their locations digitized. The radius, ulna and distal humerus were CT scanned in air using the same scanning protocol as previously described in Section 2.1, but with in-plane pixel sizes now ranging from 0.283 mm to 0.402 mm.
2.5. Contact area calculation The post-experiment CT image data was imported into Mimics. Bone and cartilage surface models were created using threshold based segmentation. Bone models included all voxels with an attenuation value of 250 HU or greater. Cartilage models included all voxels with an attenuation value of −700 HU or greater (Lapner et al., 2014; Willing et al., 2014; Yeung et al., 2013). The attached fiducial markers were also segmented from the CT scan data, and a sphere fit algorithm was used to calculate the centroid of each fiducial in the CT scan data, and the fiducial digitization data. All of the models were registered to the experiment data using a fiducial-based registration method (Lalone et al., 2012; Lapner et al., 2014; Willing et al., 2013, 2014). Model positions were further refined via alignment to their corresponding surface digitizations using an ICP alignment procedure. In a previous study, we
The accumulated contact patterns were calculated at the proximal ulna and radius. Accumulated contact patterns were defined as all of the points on the cartilage surface which were in contact at any interval of an entire motion (Fig. 4). The contact area and patterns through flexion were further analyzed by evaluating how the amount of contact within quadrants of the ulna and radius changed (Fig. 5). Repeatedmeasures analyses of variance (ANOVAs) were used to determine the effect of joint condition (native versus DHH) on the contact area throughout flexion and the accumulated contact patterns within each quadrant.
3. Results 3.1. Effect of joint configuration on contact area throughout flexion Representative cartilage–cartilage (native joint) and prosthesis– cartilage (DHH joint) overlap distances for a single specimen at 30°, 60°, 90°, and 120° of flexion are shown in Fig. 3. The average contact area versus flexion angle data for the entire ulna, radius, and corresponding sub-regions, considering all specimens, are shown in Fig. 6. Contact area at the native ulnohumeral joint tended to increase with increasing flexion angle (Fig. 6). After hemiarthroplasty with reverseengineered prostheses, ulnohumeral contact area decreased by 42% (SD 19%). If expressed as a percentage of the entire areas of the articular surfaces, this decrease was 17% (SD 8%). This decrease was statistically significant (P = .008). The trend of increased contact area with increasing flexion was maintained. Within the four ulnohumeral quadrants, statistically significant decreases in contact area were also observed at the lateral olecranon (P = .004), medial olecranon (P = .049), and lateral coronoid (P = .015). Changes in the relationship between flexion angle and contact area were more prominent within the lateral aspect of the ulnohumeral joint. Contact area at the native radiocapitellar joint tended to decrease with increasing flexion angle (Fig. 6). After hemiarthroplasty, radiocapitellar contact area decreased by 41% (SD 42%), however this decrease was not statistically significant (P = .096). If expressed as a percentage of the entire areas of the articular surfaces, this decrease was 8% (SD 8%). The trend of decreasing radiocapitellar contact area with increasing flexion was not provided by the reverse-engineered hemiarthroplasty implants. Statistically significant decreases in average contact area were observed at the medial anterior region of the radius (P = .046).
R. Willing et al. / Clinical Biomechanics 29 (2014) 990–996
993
Fig. 3. Example overlap measurements. Representative contact surface overlap distances at 30°, 60°, 90°, and 120° for a single specimen in the native and DHH state. Darker grayscale values represent greater cartilage–cartilage or prosthesis–cartilage surface overlap. Overlap occurs because the rigid computational models do not account for real cartilage deformations which allow closer proximity between models. Contact can be assumed wherever overlap occurs.
3.2. Effect of joint configuration on accumulated contact pattern The areas of the accumulated contact patterns across the ulna and radius, and the sub-regions of each, are shown in Fig. 7. The mean overall ulnohumeral accumulated contact patch size decreased by 35% (SD 20%), and the mean overall radiocapitellar accumulated contact patch size decreased by 36% (SD 26%). Both of these decreases were statistically significant (P = .016 and P = .039, respectively), and when expressed as percentages of the entire areas of the articular surfaces, these mean decreases were 27% (SD 15%) and 36% (SD 26%), respectively. These decreases indicate that contact is isolated to a smaller region following DHH with the reverse-engineered prostheses. Considering the ulna and radius sub-regions, significant decreases in accumulated contact were detected at the lateral olecranon region of the ulna (P = .015), and the medial anterior (P = .013) and medial posterior (P = .033) regions of the radius. None of the sub-regions experienced
a significant increase in accumulated contact following DHH, and only subtle changes were observed in the coronoid regions of the ulna and lateral aspects of the radius. The non-uniform reduction in the accumulated contact patch sizes suggests that the contact shape or location also changed. 4. Discussion We hypothesized that DHH with reverse-engineered prostheses would cause a reduction in contact area due to the increased stiffness of the prostheses compared to the native cartilage, but that contact patterns would reflect those of the native joints. The results of this study confirm our first hypothesis; however, significant changes in the contact distribution patterns were also observed. While significant reductions in contact area were observed at the ulnohumeral joint when the entire flexion motion was considered, a
Fig. 4. Calculation of accumulated contact area. Example data depicting how the accumulated contact area represents the integration of contact patterns from multiple instances.
994
R. Willing et al. / Clinical Biomechanics 29 (2014) 990–996
Fig. 5. Contact sub-regions. Sub-regions of the ulna and radius contact surfaces used for analysis of contact area and accumulated contact patterns.
significant difference in radiocapitellar contact area throughout flexion was not observed. This likely was the result of variations in radiocapitellar contact patterns between specimens and the smaller size of the articulation. The overall decreases in contact area observed in this study were expected. The increased stiffness of the metallic hemiarthroplasty prostheses relative to native articular cartilage likely accounts for these differences. Similar decreases in contact area have been described for
other hemiarthroplasty prostheses (Liew et al., 2003; Sabo et al., 2011). More recently, an in vitro study employing similar techniques with a commercially available axisymmetric DHH implant demonstrated comparable decreases in elbow joint contact area as reported herein with reverse-engineered prostheses (Lapner et al., 2014). We anticipated that the contact distribution patterns after DHH would be similar to those of the native joint because the articular shape of the native distal humerus was being replicated. However, the results indicate a change in the contact distribution patterns, since large decreases occurred in some regions (e.g. the lateral olecranon region) but little change occurred in others (e.g. the medial coronoid region). Disrupting the natural contact pattern of the remaining cartilage surfaces after hemiarthroplasty is of concern, potentially leading to post-traumatic osteoarthritis (Dalldorf et al., 1995). The reasons for the change in ulnohumeral and radiocapitellar contact patterns are not entirely clear. The most likely explanations are related to the fact that the reverse-engineered DHH prostheses were designed based on distal humeral osseous anatomy and lacked inclusion of the cartilage thickness. The geometry of the distal humeral osseous anatomy can be readily obtained using clinical CT scan images, and we chose to limit ourselves to this widespread and accessible imaging modality. Segmentation of osseous anatomy from clinical CT scan images is relatively quick and can be automated, and we thought that it would be the most suitable approach for designing reverse-engineered prostheses. However, the resulting DHH prostheses were therefore slightly smaller than the original cartilage surface. Joint congruency may have decreased and changes in contact distribution would reflect a new equilibrium joint alignment. Previous studies have reported that the cartilage thickness distributions across the ulnohumeral and radiocapitellar joints are not uniform (Rafehi et al., 2012; Schenck et al., 1994; Tillmann, 1978; Willing et al., 2013, 2014). This means neglecting the cartilage
Fig. 6. Contact area versus flexion angle. Contact area at the ulna (left) and radius (right) as a function of flexion angle. Lower plots show the data for the individual sub-regions. Solid lines represent the average data of 5 specimens and shaded areas represent the standard deviation. Levels of significance (n = 5), testing for differences between native and hemi configuration data, are stated beside region titles.
R. Willing et al. / Clinical Biomechanics 29 (2014) 990–996
995
Fig. 7. Changes in accumulated contact patterns. Measured areas of the accumulated contact patterns across the surface of the ulna, radius, and corresponding sub-regions of each. Error bars are 1 standard deviation. * denotes statistically significant differences (P b 0.05, n = 5).
thickness would have a varied result throughout the different contact regions. Patient-specific cartilage measurements obtained using MR imaging could help design hemiarthroplasty prostheses which account for the cartilage thickness. It is important to note, however, that the cartilage thickness distributions were measured after the in-vitro experiments were completed, and have been accounted for (where relevant) when calculating the overlap-based contact patches. Some limitations present in the current study must be acknowledged. One limitation was the small number of specimens. Further experiments with a larger sample size would likely strengthen the statistical significance of some findings. Another limitation is the accuracy of using motion capture and reconstructed computer models to calculate contact based on cartilage–cartilage or prosthesis–cartilage surface overlap. We have previously validated our contact measurement technique and found that contact area error was typically less than 15% (Willing et al., 2014). More importantly, however, our previous work demonstrated that the technique was reliable for measuring changes in contact area resulting from DHH, which justifies the use of this technique in the current study. We opted to use this technique over interpositional sensors or casting, which may potentially alter contact mechanics. Future studies should consider reverse-engineered
hemiarthroplasty prosthesis designs which account for the native cartilage thickness and more compliant prosthesis materials in order to more closely resemble the native distal humerus. 5. Conclusions In summary, this study on the efficacy of a new concept in DHH based on revere-engineering prostheses from native anatomy CT scans found that these designs still caused a decrease in contact area at the elbow joint, and that the contact distribution pattern was also affected. This suggests that the reverse-engineered prostheses tested in this study may provide little benefit over existing commercially available designs. Future work should focus on other design strategies such as accounting for the non-uniform cartilage thickness distribution, shape optimization, or more compliant materials for DHH in order to promote natural cartilage contact mechanics and improve clinical outcomes. Conflict of interest statement Graham King receives royalty and consulting payments from Tornier Inc., but not for the prosthesis design investigated in this study. None of the other authors have any conflict of interest, including any financial
996
R. Willing et al. / Clinical Biomechanics 29 (2014) 990–996
and personal relationships with other people or organizations that could inappropriately influence their work. Acknowledgments Funding was received from the Canadian Institute for Health Research (CIHR, 210521) in support of the data collection for this study. The first author was supported in part by the Joint Motion Program — A CIHR Training Program in Musculoskeletal Joint Research and Leadership. References Besl, P.J., 1992. A method for registration of 3-D shapes. IEEE Trans. Pattern Anal. Mach. Intell. 14, 239–256. Burkhart, K.J., Nijs, S., Mattyasovszky, S.G., Wouters, R., Gruszka, D., Nowak, T.E., et al., 2011. Distal humerus hemiarthroplasty of the elbow for comminuted distal humeral fractures in the elderly patient. J. Trauma 71, 635–642. Cenni, F., Timoncini, A., Ensini, A., Tamarri, S., Belvedere, C., D'Angeli, V., et al., 2014. Threedimensional implant position and orientation after total knee replacement performed with patient-specific instrumentation systems. J. Orthop. Res. 32, 331–337. Dalldorf, P.G., Banas, M.P., Hicks, D.G., Pellegrini Jr., V.D., 1995. Rate of degeneration of human acetabular cartilage after hemiarthroplasty. J. Bone Joint Surg. Am. 77, 877–882. de Beer, N., Scheffer, C., 2012. Reducing subsidence risk by using rapid manufactured patient-specific intervertebral disc implants. Spine J. 12, 1060–1066. Fitz, W., 2009. Unicompartmental knee arthroplasty with use of novel patient-specific resurfacing implants and personalized jigs. J. Bone Joint Surg. Am. 91 (Suppl. 1), 69–76. Hohman, D.W., Nodzo, S.R., Qvick, L.M., Duquin, T.R., Paterson, P.P., 2014. Hemiarthroplasty of the distal humerus for acute and chronic complex intraarticular injuries. J. Shoulder Elbow Surg. 23, 265–272. Lalone, E.A., McDonald, C.P., Ferreira, L.M., Peters, T.M., King, G.W., Johnson, J.A., 2012. Development of an image-based technique to examine joint congruency at the elbow. Comput. Methods Biomech. Biomed. Engin. 16, 280–290. Lapner, M., Willing, R., Johnson, J.A., King, G.J.W., 2014. The effect of distal humeral hemiarthroplasty on articular contact of the elbow. Clin. Biomech. 29, 537–544.
Leichtle, U.G., Leichtle, C.I., Martini, F., 2012. Microsection analysis of cortical form-fit of a custom femoral component in total hip arthroplasty: an in vitro study. Orthop. Rev. (Pavia) 4, e31. Liew, V.S., Cooper, I.C., Ferreira, L.M., Johnson, J.A., King, G.J., 2003. The effect of metallic radial head arthroplasty on radiocapitellar joint contact area. Clin. Biomech. 18, 115–118. London, J.T., 1981. Kinematics of the elbow. J. Bone Joint Surg. Am. 63, 529–535. Morrey, B.F., Tanaka, S., An, K.N., 1991. Valgus stability of the elbow. A definition of primary and secondary constraints. Clin. Orthop. Relat. Res. 265, 187–195. Rafehi, S., Lalone, E., Johnson, M., King, G.J., Athwal, G.S., 2012. An anatomic study of coronoid cartilage thickness with special reference to fractures. J. Shoulder Elbow Surg. 21, 961–968. Reize, P.K., Wulker, N., 2007. The adaptiva custom-made stem—our reasons for not using it anymore. Int. Orthop. 31, 217–222. Sabo, M.T., Shannon, H., Ng, J., Ferreira, L.M., Johnson, J.A., King, G.J., 2011. The impact of capitellar arthroplasty on elbow contact mechanics: implications for implant design. Clin. Biomech. 26, 458–463. Schenck Jr., R.C., Athanasiou, K.A., Constantinides, G., Gomez, E., 1994. A biomechanical analysis of articular cartilage of the human elbow and a potential relationship to osteochondritis dissecans. Clin. Orthop. Relat. Res. 299, 305–312. Shiba, R., Sorbie, C., Siu, D.W., Bryant, J.T., Cooke, T.D., Wevers, H.W., 1988. Geometry of the humeroulnar joint. J. Orthop. Res. 6, 897–906. Slamin, J., Parsley, B., 2012. Evolution of customization design for total knee arthroplasty. Curr. Rev. Musculoskelet. Med. 5, 290–295. Smith, G.C., Hughes, J.S., 2013. Unreconstructable acute distal humeral fractures and their sequelae treated with distal humeral hemiarthroplasty: a two-year to eleven-year follow-up. J. Shoulder Elbow Surg. 22, 1710–1723. Sundseth, J., Berg-Johnsen, J., 2013. Prefabricated patient-matched cranial implants for reconstruction of large skull defects. J. Cent. Nerv. Syst. Dis. 5, 19–24. Tillmann, B., 1978. A contribution to the functional morphology of articular surfaces. Norm. Pathol. Anat. (Stuttg.) 34, 1–50. Willing, R.T., Lalone, E.A., Shannon, H., Johnson, J.A., King, G.J., 2013. Validation of a finite element model of the human elbow for determining cartilage contact mechanics. J. Biomech. 46, 1767–1771. Willing, R., Lapner, M., Lalone, E.A., King, G.J., Johnson, J.A., 2014. Development of a computational technique to measure cartilage contact area. J. Biomech. 47, 1193–1197. Yeung, C., Willing, R., Shannon, H., Lalone, E.A., Johnson, M., King, G.J., 2013. Comparing CT to Surface-Digitization Based Measurements of Cartilage Thickness. in: Proceedings of the Orthopaedic Research Society Annual Meeting San Antonio, TX.
Contents lists available at ScienceDirect
Clinical Biomechanics journal homepage: www.elsevier.com/locate/clinbiomech
In vitro assessment of the contact mechanics of reverse-engineered distal humeral hemiarthroplasty prostheses Ryan Willing a,b,⁎, Michael Lapner a,1, Graham J.W. King a,2,3, James A. Johnson a,3,4,5 a b
Bioengineering Research Laboratory, Roth | McFarlane Hand and Upper Limb Centre, St. Joseph's Health Care, The University of Western Ontario, London, ON, Canada Department of Mechanical Engineering, Binghamton University, Binghamton, NY, USA
a r t i c l e
i n f o
Article history: Received 28 April 2014 Accepted 13 August 2014 Keywords: Elbow Distal humerus Hemiarthroplasty Contact pattern Reverse-engineered
a b s t r a c t Background: Distal humeral hemiarthroplasty alters cartilage contact mechanics, which may predispose to osteoarthritis. Current prostheses do not replicate the native anatomy, and therefore contribute to these changes. We hypothesized that prostheses reverse-engineered from the native bone shape would provide similar contact patterns as the native articulation. Methods: Reverse-engineered hemiarthroplasty prostheses were manufactured for five cadaveric elbows based on CT images of the distal humerus. Passive flexion trials with constant muscle forces were performed with the native articulation intact while bone motions were recorded using a motion tracking system. Motion trials were then repeated after the distal humerus was replaced with a corresponding reverse-engineered prosthesis. Contact areas and patterns were reconstructed using computer models created from CT scan images combined with the motion tracker data. The total contact areas, as well as the contact area within smaller sub-regions of the ulna and radius, were analyzed for changes resulting from hemiarthroplasty using repeated-measures ANOVAs. Findings: Contact area at the ulna and radius decreased on average 42% (SD 19%, P = .008) and 41% (SD 42%, P = .096), respectively. Contact area decreases were not uniform throughout the different sub-regions, suggesting that contact patterns were also altered. Interpretation: Reverse-engineered prostheses did not reproduce the same contact pattern as the native joints, possibly because the thickness of the distal humerus cartilage layer was neglected when generating the prosthesis shapes or as a consequence of the increased stiffness of the metallic implants. Alternative design strategies and materials for hemiarthroplasty should be considered in future work. © 2014 Elsevier Ltd. All rights reserved.
1. Introduction Distal humeral hemiarthroplasty (DHH) is a treatment option for the management of comminuted distal humeral fractures (Burkhart et al., 2011; Hohman et al., 2014; Smith and Hughes, 2013). DHH places a metal prosthesis in direct contact with the native articular surfaces of the radial head and greater sigmoid notch of the ulna. The increased ⁎ Corresponding author at: Mechanical Engineering Department, Thomas J. Watson School of Engineering & Applied Science, Binghamton University — SUNY, P.O. Box 6000, Binghamton, NY 13902-6000, USA. E-mail address: [email protected] (R. Willing). 1 Division of Orthopedic Surgery, Sturgeon Hospital, University of Alberta, 201 Boudreau Rd, St. Albert, AB T8N 6C4, Canada. 2 Department of Surgery, The University of Western Ontario, 1151 Richmond St, London, ON N6A 3K7, Canada. 3 Roth | McFarlane Hand and Upper Limb Centre Bioengineering Laboratory, St. Joseph's Health Care, 268 Grosvenor St, London, ON N6A 4L6, Canada. 4 Biomedical Engineering Department, The University of Western Ontario, 1151 Richmond St, London, ON N6A 3K7, Canada. 5 Department of Mechanical and Materials Engineering, The University of Western Ontario, 1151 Richmond St, London, ON N6A 3K7, Canada.
http://dx.doi.org/10.1016/j.clinbiomech.2014.08.015 0268-0033/© 2014 Elsevier Ltd. All rights reserved.
stiffness of metal hemiarthroplasty devices relative to cartilage typically results in a decrease in contact area (Liew et al., 2003; Sabo et al., 2011). This may increase cartilage stresses due to the altered contact mechanics, resulting in cartilage wear and post-traumatic osteoarthritis (Dalldorf et al., 1995; Smith and Hughes, 2013). Reproducing the native contact patterns after DHH could reduce the likelihood of cartilage erosion. In an effort to recreate more natural cartilage contact mechanics, “anatomical” DHH designs have been developed and are commercially available (Hohman et al., 2014; Lapner et al., 2014). These implants attempt to mimic the native anatomy of the distal humerus; however, their shapes are simplified for general use. An in-vitro study by Lapner et al. (2014) showed that mean ulnohumeral and radiocapitellar contact areas both decreased as the result of DHH with a commercially available axisymmetric prosthesis (decreased by 44%, P = 0.03, and 4%, P = 0.07, respectively). The resulting contact pattern was different than that of the native joint, and varying prosthesis size did not have a significant effect on contact area or pattern. It is possible to create anatomically accurate reverse-engineered implants on a patient-specific basis using pre-operative 3-D imaging modalities. Patient specific hip (Leichtle et al., 2012; Reize and Wulker, 2007), knee (Fitz, 2009), spine (de Beer
R. Willing et al. / Clinical Biomechanics 29 (2014) 990–996
and Scheffer, 2012), and cranial (Sundseth and Berg-Johnsen, 2013) prosthesis components, as well as patient-specific cutting guides for total knee replacement (Cenni et al., 2014; Fitz, 2009; Slamin and Parsley, 2012) have been designed from pre-operative medical imaging. The efficacy of patient-specific reverse-engineered hemiarthroplasty prostheses for promoting more natural contact mechanics, particularly for DHH, has not been reported. While reductions in contact area can be expected due to the increased stiffness of these metal prostheses compared to the cartilage surfaces they are replacing, patient-specific reverse-engineered prostheses could result in contact patterns similar to those of the native joint. The objective of this study was to quantify the changes in cartilage contact which occur after DHH with a subject-specific reverseengineered prosthesis. We hypothesized that DHH with an anatomically shaped reverse-engineered prostheses would result in a reduction in contact area, but contact patterns similar in shape to those of a native joint. 2. Methods 2.1. Reverse-engineered DHH design CT images of 5 fresh-frozen intact left arm specimens (3 male, 2 female, average age 74 yrs, range: 66–84 yrs) with no radiographic evidence of joint pathology were obtained using a GE Discovery CT750 HD scanner (GE Health Care, Pewaukee, WI, USA) at 120 kV and 292 mAs with a slice thickness of 0.625 mm (in-plane pixel sizes ranging from 0.492 mm–0.586 mm). The CT data was imported into Mimics v14.12 (Materialise, Leuven, Belgium), and the distal humeral bone geometry was extracted using threshold based segmentation, which included any voxel with an attenuation value of 250 HU or greater (Lapner et al., 2014; Willing et al., 2013, 2014). Models were wrapped, exported in the stereolithography (STL) format, and remeshed using a radial basis function in Matlab (The Mathworks, Natick, MA, USA). The resulting models comprised uniformly sized triangles with approximately 0.4 mm edge lengths. A Boolean geometry subtract operation was performed using custom Blender script (The Blender Foundation, Amsterdam, NL), which cropped the model to the articular region and created interface geometry for attaching a custom stem component (Fig. 1, left). A sPro™ 125 direct metal selective laser melting (SLM) machine (3D Systems Corp., Rock Hill, SC, USA) was used to manufacture stainless steel prosthesis prototypes based on the computer models (Fig. 1, middle). The articular surfaces were sanded and polished to a mirror finish using a rotary hand tool (Fig. 1, right). A MicroScribe G2X digitizer (Immersion Corp., San Jose, CA, USA) was traced across the entire contact surface of the polished prosthesis prototype, and the surface geometry was saved as a 3D point cloud. The computer model was aligned to the 3D point cloud using an iterative closest point (ICP) algorithm (Besl, 1992), and the distance from each digitized point to the closest point on the computer model was measured to be, on average, 0.08 mm (SD 0.26 mm).
991
2.2. Specimen preparation Prior to testing, specimens were amputated mid-humerus and stored at room temperature for approximately 20 h to fully thaw. The radius and ulna were pinned together in neutral forearm rotation. A 10 cm segment of the humeral shaft was denuded of all soft tissues and was rigidly clamped to a mounting fixture. Optotrak Certus (Northern Digital Inc., Waterloo, Ontario, Canada) infrared marker triads were affixed to the shafts of the ulna and radius, and the humerus clamp. A capsulectomy was performed at the elbow joint, and the medial and lateral collateral ligaments (MCL and LCL) were released at their respective origins on the humerus, which allowed complete disarticulation of the elbow joint for exposure of the distal humeral articular surface. 2.3. Testing apparatus The arm was attached to an elbow testing apparatus (via the humerus clamp; Fig. 2). Pneumatic actuators (Bimba Co., University Park, IL, USA) provided simulated muscle tensions FBIC, FBRA and FTRI via stainless steel aircraft cables sutured to the biceps, brachialis and triceps tendons, respectively. The flexion–extension axis of the humerus was defined by connecting the center of a circle fit to the trochlear sulcus, to the center of a sphere fit of the capitellum (London, 1981; Shiba et al., 1988) using surface digitization data obtained by the optically tracked stylus tool. A 5 mm diameter hole was drilled medial-laterally through the distal humerus along the flexion–extension axis using computer navigation and a 5 mm diameter cylindrical brass ligament guide was installed (Willing et al., 2014). Heavy braided sutures were attached to the free ends of the LCL and MCL, routed over the ligament guide (Fig. 2, inset), and tensioned to 20 N. Constant ligament tension, rather than a constant ligament length, simulated perfect ligament balance and ensured that the joint reaction force passing through the articulation would be insensitive to prosthesis alignment. 2.4. Testing protocol Motion was recorded while the arm was guided through passive flexion motions (3 repetitions, 0°–120°) with ligament loads applied. Biceps, brachialis and triceps muscle tensions of 20 N, 20 N and 40 N were applied based on previously reported ratios (Morrey et al., 1991). The optical motion capture system tracked bone motion in 3D within 0.1 mm positional accuracy. Following the trials with the native articulation, four spherical 20 mm nylon fiducial markers were attached to the distal humerus. The fiducial locations were digitized with respect to the humerus motion tracker using an optically tracked stylus tool. The distal humerus, as well as the attached fiducial markers, was then excised from the humeral diaphysis using an oscillating saw and stored in normal saline. Using computer navigation, a custom distal humerus stem component was guided into place such that the prosthesis flexion axis aligned with the flexion axis of the native articulation (average error = 3.6°,
Fig. 1. Creation of reverse-engineered prostheses. Left: Computer model of reverse engineered hemiarthroplasty prosthesis created from CT imagery. Middle: Raw part manufactured using direct metal SLM. Right: Final polished prototype ready for testing.
992
R. Willing et al. / Clinical Biomechanics 29 (2014) 990–996
reported average surface alignment errors of 0.15 mm (maximum: 0.23 mm) using this registration technique (Willing et al., 2014). Displacements and rotations were applied to the corresponding 3D computer models using the motion tracker data recorded during the motion trials. Once models were positioned with respect to one another, contact surfaces would overlap (because the models represented the un-deformed cartilage layer, which would actually compress during contact) (Fig. 3). Contact was assumed to occur wherever cartilage– cartilage or prosthesis–cartilage overlap occurred. This contact measurement technique was validated in a previous study, where a good correlation between cast-measured and overlap-based contact area measurements was observed (ICC = 0.727), contact area error was typically less than 15%, and the changes in joint contact due to DHH could be reliably detected (Willing et al., 2014). Contact area was calculated in 15° increments from 0° to 120° for each motion (3 flexion trials) and joint configuration (native and hemi) at the ulna and radius. Results from repeated trials of the same motion were averaged. The areas were recorded in terms of percentage of the total available contact area in order to normalize for different specimen sizes.
2.6. Data analysis
Fig. 2. Elbow testing apparatus. Pneumatic actuators provided simulated muscle forces for the biceps, brachialis, and triceps. Ligament forces were provided by hanging weights from cable sutures to the ligaments, routed to pass over the ligament guide.
maximum error = 6.0°) and the prosthesis trochlear sulcus was in the same position as the native trochlear sulcus (average error = 3.1 mm, maximum error = 5.1 mm). The specimen-specific reverseengineered hemiarthroplasty prosthesis was installed onto the stem. The motion and loading protocol used for the native articulation were repeated with the prosthesis in place. Ligament lines of action were always directed towards the flexion axis, which was referenced to the corresponding native or prosthesis contact surfaces, so joint contact mechanics were insensitive to small errors in stem positioning. After testing, the ulna and radius were separated and denuded. The articular surfaces of the ulna, radius and prosthesis were digitized using the stylus tool. Four fiducial markers were attached to each of the proximal radius and ulna, and their locations digitized. The radius, ulna and distal humerus were CT scanned in air using the same scanning protocol as previously described in Section 2.1, but with in-plane pixel sizes now ranging from 0.283 mm to 0.402 mm.
2.5. Contact area calculation The post-experiment CT image data was imported into Mimics. Bone and cartilage surface models were created using threshold based segmentation. Bone models included all voxels with an attenuation value of 250 HU or greater. Cartilage models included all voxels with an attenuation value of −700 HU or greater (Lapner et al., 2014; Willing et al., 2014; Yeung et al., 2013). The attached fiducial markers were also segmented from the CT scan data, and a sphere fit algorithm was used to calculate the centroid of each fiducial in the CT scan data, and the fiducial digitization data. All of the models were registered to the experiment data using a fiducial-based registration method (Lalone et al., 2012; Lapner et al., 2014; Willing et al., 2013, 2014). Model positions were further refined via alignment to their corresponding surface digitizations using an ICP alignment procedure. In a previous study, we
The accumulated contact patterns were calculated at the proximal ulna and radius. Accumulated contact patterns were defined as all of the points on the cartilage surface which were in contact at any interval of an entire motion (Fig. 4). The contact area and patterns through flexion were further analyzed by evaluating how the amount of contact within quadrants of the ulna and radius changed (Fig. 5). Repeatedmeasures analyses of variance (ANOVAs) were used to determine the effect of joint condition (native versus DHH) on the contact area throughout flexion and the accumulated contact patterns within each quadrant.
3. Results 3.1. Effect of joint configuration on contact area throughout flexion Representative cartilage–cartilage (native joint) and prosthesis– cartilage (DHH joint) overlap distances for a single specimen at 30°, 60°, 90°, and 120° of flexion are shown in Fig. 3. The average contact area versus flexion angle data for the entire ulna, radius, and corresponding sub-regions, considering all specimens, are shown in Fig. 6. Contact area at the native ulnohumeral joint tended to increase with increasing flexion angle (Fig. 6). After hemiarthroplasty with reverseengineered prostheses, ulnohumeral contact area decreased by 42% (SD 19%). If expressed as a percentage of the entire areas of the articular surfaces, this decrease was 17% (SD 8%). This decrease was statistically significant (P = .008). The trend of increased contact area with increasing flexion was maintained. Within the four ulnohumeral quadrants, statistically significant decreases in contact area were also observed at the lateral olecranon (P = .004), medial olecranon (P = .049), and lateral coronoid (P = .015). Changes in the relationship between flexion angle and contact area were more prominent within the lateral aspect of the ulnohumeral joint. Contact area at the native radiocapitellar joint tended to decrease with increasing flexion angle (Fig. 6). After hemiarthroplasty, radiocapitellar contact area decreased by 41% (SD 42%), however this decrease was not statistically significant (P = .096). If expressed as a percentage of the entire areas of the articular surfaces, this decrease was 8% (SD 8%). The trend of decreasing radiocapitellar contact area with increasing flexion was not provided by the reverse-engineered hemiarthroplasty implants. Statistically significant decreases in average contact area were observed at the medial anterior region of the radius (P = .046).
R. Willing et al. / Clinical Biomechanics 29 (2014) 990–996
993
Fig. 3. Example overlap measurements. Representative contact surface overlap distances at 30°, 60°, 90°, and 120° for a single specimen in the native and DHH state. Darker grayscale values represent greater cartilage–cartilage or prosthesis–cartilage surface overlap. Overlap occurs because the rigid computational models do not account for real cartilage deformations which allow closer proximity between models. Contact can be assumed wherever overlap occurs.
3.2. Effect of joint configuration on accumulated contact pattern The areas of the accumulated contact patterns across the ulna and radius, and the sub-regions of each, are shown in Fig. 7. The mean overall ulnohumeral accumulated contact patch size decreased by 35% (SD 20%), and the mean overall radiocapitellar accumulated contact patch size decreased by 36% (SD 26%). Both of these decreases were statistically significant (P = .016 and P = .039, respectively), and when expressed as percentages of the entire areas of the articular surfaces, these mean decreases were 27% (SD 15%) and 36% (SD 26%), respectively. These decreases indicate that contact is isolated to a smaller region following DHH with the reverse-engineered prostheses. Considering the ulna and radius sub-regions, significant decreases in accumulated contact were detected at the lateral olecranon region of the ulna (P = .015), and the medial anterior (P = .013) and medial posterior (P = .033) regions of the radius. None of the sub-regions experienced
a significant increase in accumulated contact following DHH, and only subtle changes were observed in the coronoid regions of the ulna and lateral aspects of the radius. The non-uniform reduction in the accumulated contact patch sizes suggests that the contact shape or location also changed. 4. Discussion We hypothesized that DHH with reverse-engineered prostheses would cause a reduction in contact area due to the increased stiffness of the prostheses compared to the native cartilage, but that contact patterns would reflect those of the native joints. The results of this study confirm our first hypothesis; however, significant changes in the contact distribution patterns were also observed. While significant reductions in contact area were observed at the ulnohumeral joint when the entire flexion motion was considered, a
Fig. 4. Calculation of accumulated contact area. Example data depicting how the accumulated contact area represents the integration of contact patterns from multiple instances.
994
R. Willing et al. / Clinical Biomechanics 29 (2014) 990–996
Fig. 5. Contact sub-regions. Sub-regions of the ulna and radius contact surfaces used for analysis of contact area and accumulated contact patterns.
significant difference in radiocapitellar contact area throughout flexion was not observed. This likely was the result of variations in radiocapitellar contact patterns between specimens and the smaller size of the articulation. The overall decreases in contact area observed in this study were expected. The increased stiffness of the metallic hemiarthroplasty prostheses relative to native articular cartilage likely accounts for these differences. Similar decreases in contact area have been described for
other hemiarthroplasty prostheses (Liew et al., 2003; Sabo et al., 2011). More recently, an in vitro study employing similar techniques with a commercially available axisymmetric DHH implant demonstrated comparable decreases in elbow joint contact area as reported herein with reverse-engineered prostheses (Lapner et al., 2014). We anticipated that the contact distribution patterns after DHH would be similar to those of the native joint because the articular shape of the native distal humerus was being replicated. However, the results indicate a change in the contact distribution patterns, since large decreases occurred in some regions (e.g. the lateral olecranon region) but little change occurred in others (e.g. the medial coronoid region). Disrupting the natural contact pattern of the remaining cartilage surfaces after hemiarthroplasty is of concern, potentially leading to post-traumatic osteoarthritis (Dalldorf et al., 1995). The reasons for the change in ulnohumeral and radiocapitellar contact patterns are not entirely clear. The most likely explanations are related to the fact that the reverse-engineered DHH prostheses were designed based on distal humeral osseous anatomy and lacked inclusion of the cartilage thickness. The geometry of the distal humeral osseous anatomy can be readily obtained using clinical CT scan images, and we chose to limit ourselves to this widespread and accessible imaging modality. Segmentation of osseous anatomy from clinical CT scan images is relatively quick and can be automated, and we thought that it would be the most suitable approach for designing reverse-engineered prostheses. However, the resulting DHH prostheses were therefore slightly smaller than the original cartilage surface. Joint congruency may have decreased and changes in contact distribution would reflect a new equilibrium joint alignment. Previous studies have reported that the cartilage thickness distributions across the ulnohumeral and radiocapitellar joints are not uniform (Rafehi et al., 2012; Schenck et al., 1994; Tillmann, 1978; Willing et al., 2013, 2014). This means neglecting the cartilage
Fig. 6. Contact area versus flexion angle. Contact area at the ulna (left) and radius (right) as a function of flexion angle. Lower plots show the data for the individual sub-regions. Solid lines represent the average data of 5 specimens and shaded areas represent the standard deviation. Levels of significance (n = 5), testing for differences between native and hemi configuration data, are stated beside region titles.
R. Willing et al. / Clinical Biomechanics 29 (2014) 990–996
995
Fig. 7. Changes in accumulated contact patterns. Measured areas of the accumulated contact patterns across the surface of the ulna, radius, and corresponding sub-regions of each. Error bars are 1 standard deviation. * denotes statistically significant differences (P b 0.05, n = 5).
thickness would have a varied result throughout the different contact regions. Patient-specific cartilage measurements obtained using MR imaging could help design hemiarthroplasty prostheses which account for the cartilage thickness. It is important to note, however, that the cartilage thickness distributions were measured after the in-vitro experiments were completed, and have been accounted for (where relevant) when calculating the overlap-based contact patches. Some limitations present in the current study must be acknowledged. One limitation was the small number of specimens. Further experiments with a larger sample size would likely strengthen the statistical significance of some findings. Another limitation is the accuracy of using motion capture and reconstructed computer models to calculate contact based on cartilage–cartilage or prosthesis–cartilage surface overlap. We have previously validated our contact measurement technique and found that contact area error was typically less than 15% (Willing et al., 2014). More importantly, however, our previous work demonstrated that the technique was reliable for measuring changes in contact area resulting from DHH, which justifies the use of this technique in the current study. We opted to use this technique over interpositional sensors or casting, which may potentially alter contact mechanics. Future studies should consider reverse-engineered
hemiarthroplasty prosthesis designs which account for the native cartilage thickness and more compliant prosthesis materials in order to more closely resemble the native distal humerus. 5. Conclusions In summary, this study on the efficacy of a new concept in DHH based on revere-engineering prostheses from native anatomy CT scans found that these designs still caused a decrease in contact area at the elbow joint, and that the contact distribution pattern was also affected. This suggests that the reverse-engineered prostheses tested in this study may provide little benefit over existing commercially available designs. Future work should focus on other design strategies such as accounting for the non-uniform cartilage thickness distribution, shape optimization, or more compliant materials for DHH in order to promote natural cartilage contact mechanics and improve clinical outcomes. Conflict of interest statement Graham King receives royalty and consulting payments from Tornier Inc., but not for the prosthesis design investigated in this study. None of the other authors have any conflict of interest, including any financial
996
R. Willing et al. / Clinical Biomechanics 29 (2014) 990–996
and personal relationships with other people or organizations that could inappropriately influence their work. Acknowledgments Funding was received from the Canadian Institute for Health Research (CIHR, 210521) in support of the data collection for this study. The first author was supported in part by the Joint Motion Program — A CIHR Training Program in Musculoskeletal Joint Research and Leadership. References Besl, P.J., 1992. A method for registration of 3-D shapes. IEEE Trans. Pattern Anal. Mach. Intell. 14, 239–256. Burkhart, K.J., Nijs, S., Mattyasovszky, S.G., Wouters, R., Gruszka, D., Nowak, T.E., et al., 2011. Distal humerus hemiarthroplasty of the elbow for comminuted distal humeral fractures in the elderly patient. J. Trauma 71, 635–642. Cenni, F., Timoncini, A., Ensini, A., Tamarri, S., Belvedere, C., D'Angeli, V., et al., 2014. Threedimensional implant position and orientation after total knee replacement performed with patient-specific instrumentation systems. J. Orthop. Res. 32, 331–337. Dalldorf, P.G., Banas, M.P., Hicks, D.G., Pellegrini Jr., V.D., 1995. Rate of degeneration of human acetabular cartilage after hemiarthroplasty. J. Bone Joint Surg. Am. 77, 877–882. de Beer, N., Scheffer, C., 2012. Reducing subsidence risk by using rapid manufactured patient-specific intervertebral disc implants. Spine J. 12, 1060–1066. Fitz, W., 2009. Unicompartmental knee arthroplasty with use of novel patient-specific resurfacing implants and personalized jigs. J. Bone Joint Surg. Am. 91 (Suppl. 1), 69–76. Hohman, D.W., Nodzo, S.R., Qvick, L.M., Duquin, T.R., Paterson, P.P., 2014. Hemiarthroplasty of the distal humerus for acute and chronic complex intraarticular injuries. J. Shoulder Elbow Surg. 23, 265–272. Lalone, E.A., McDonald, C.P., Ferreira, L.M., Peters, T.M., King, G.W., Johnson, J.A., 2012. Development of an image-based technique to examine joint congruency at the elbow. Comput. Methods Biomech. Biomed. Engin. 16, 280–290. Lapner, M., Willing, R., Johnson, J.A., King, G.J.W., 2014. The effect of distal humeral hemiarthroplasty on articular contact of the elbow. Clin. Biomech. 29, 537–544.
Leichtle, U.G., Leichtle, C.I., Martini, F., 2012. Microsection analysis of cortical form-fit of a custom femoral component in total hip arthroplasty: an in vitro study. Orthop. Rev. (Pavia) 4, e31. Liew, V.S., Cooper, I.C., Ferreira, L.M., Johnson, J.A., King, G.J., 2003. The effect of metallic radial head arthroplasty on radiocapitellar joint contact area. Clin. Biomech. 18, 115–118. London, J.T., 1981. Kinematics of the elbow. J. Bone Joint Surg. Am. 63, 529–535. Morrey, B.F., Tanaka, S., An, K.N., 1991. Valgus stability of the elbow. A definition of primary and secondary constraints. Clin. Orthop. Relat. Res. 265, 187–195. Rafehi, S., Lalone, E., Johnson, M., King, G.J., Athwal, G.S., 2012. An anatomic study of coronoid cartilage thickness with special reference to fractures. J. Shoulder Elbow Surg. 21, 961–968. Reize, P.K., Wulker, N., 2007. The adaptiva custom-made stem—our reasons for not using it anymore. Int. Orthop. 31, 217–222. Sabo, M.T., Shannon, H., Ng, J., Ferreira, L.M., Johnson, J.A., King, G.J., 2011. The impact of capitellar arthroplasty on elbow contact mechanics: implications for implant design. Clin. Biomech. 26, 458–463. Schenck Jr., R.C., Athanasiou, K.A., Constantinides, G., Gomez, E., 1994. A biomechanical analysis of articular cartilage of the human elbow and a potential relationship to osteochondritis dissecans. Clin. Orthop. Relat. Res. 299, 305–312. Shiba, R., Sorbie, C., Siu, D.W., Bryant, J.T., Cooke, T.D., Wevers, H.W., 1988. Geometry of the humeroulnar joint. J. Orthop. Res. 6, 897–906. Slamin, J., Parsley, B., 2012. Evolution of customization design for total knee arthroplasty. Curr. Rev. Musculoskelet. Med. 5, 290–295. Smith, G.C., Hughes, J.S., 2013. Unreconstructable acute distal humeral fractures and their sequelae treated with distal humeral hemiarthroplasty: a two-year to eleven-year follow-up. J. Shoulder Elbow Surg. 22, 1710–1723. Sundseth, J., Berg-Johnsen, J., 2013. Prefabricated patient-matched cranial implants for reconstruction of large skull defects. J. Cent. Nerv. Syst. Dis. 5, 19–24. Tillmann, B., 1978. A contribution to the functional morphology of articular surfaces. Norm. Pathol. Anat. (Stuttg.) 34, 1–50. Willing, R.T., Lalone, E.A., Shannon, H., Johnson, J.A., King, G.J., 2013. Validation of a finite element model of the human elbow for determining cartilage contact mechanics. J. Biomech. 46, 1767–1771. Willing, R., Lapner, M., Lalone, E.A., King, G.J., Johnson, J.A., 2014. Development of a computational technique to measure cartilage contact area. J. Biomech. 47, 1193–1197. Yeung, C., Willing, R., Shannon, H., Lalone, E.A., Johnson, M., King, G.J., 2013. Comparing CT to Surface-Digitization Based Measurements of Cartilage Thickness. in: Proceedings of the Orthopaedic Research Society Annual Meeting San Antonio, TX.