*Reprint req

Biomechanical T

rials Engineerin

E-mail addre

J Shoulder Elbow Surg (2013) 22, 1248-1255

1058-2746/$ - s

http://dx.doi.org

www.elsevier.com/locate/ymse

The effect of stem surface treatment and materialon pistoning of ulnar components in linked cementedelbow prostheses

Yara K. Hosein, BSca, Graham J.W. King, MD, MSc, FRCSCa,b,d,Cynthia E. Dunning, PhD, PEnga,b,c,d,*

aBiomedical Engineering Graduate Program, Western University, London, ON, CanadabDepartment of Surgery, Western University, London, ON, CanadacDepartment of Materials and Mechanical Engineering, Western University, London, ON, CanadadDepartment of Medical Biophysics, Western University, London, ON, Canada

Background: The ulnar component of a total elbow replacement can fail by ‘‘pistoning.’’ Stem surfacetreatments have improved stability at the stem-cement interface but with varied success. This study inves-tigated the role of surface treatment and stem substrate material on implant stability under axial loading.Materials and methods: Sixty circular stems (diameter, 8 mm) made of cobalt chrome (n ¼ 30) or tita-nium (n ¼ 30) had different surfaces: smooth, sintered beads, and plasma spray. The surface treatmentlength was either 10 mm or 20 mm. Stems were potted in bone cement, allowed to cure for 24 hours,and tested in a materials testing machine under a compressive staircase loading protocol. Failure wasdefined as 2 mm of push-out or completion of the protocol. Two-way analyses of variance comparedthe effects of surface treatment and substrate material on interface strength and motion.Results: Significant interactions were found between surface treatment and substrate material for bothinterface strength and motion (P < .05). For titanium, the 20-mm beaded stems had greater interfacestrength than all other stems (P < .05) and had less motion than the 10-mm plasma-spray and smoothstems (P < .05). For cobalt chrome, the 20-mm beaded stems showed greater interface strength(P < .05) and similar motion (P > .05) to the 20-mm plasma-spray stems (P < .05), which outperformedall other stems (P < .05). Mechanisms of catastrophic failure varied: smooth stems debonded at the stem-cement interface, beaded stems experienced debonding of the beads from the stem, and plasma-spray stemsshowed loss of frictional force between the surface treatment and cement.Discussion and conclusion: Stem surface treatment can enhance ulnar component stability but is depen-dent on substrate material.Level of evidence: Basic Science, Biomechanical Study.� 2013 Journal of Shoulder and Elbow Surgery Board of Trustees.

Keywords: Surface treatment; ulnar component; sintered beads; plasma spray; implant fixation; implant

stability

uests: Cynthia E. Dunning, PhD, PEng, Jack McBain

esting Laboratory, Department of Mechanical and Mate-

g, Western University, London, ON, Canada N6A 5B9.

ss: cdunning@uwo.ca (C.E. Dunning).

ee front matter � 2013 Journal of Shoulder and Elbow Surgery

/10.1016/j.jse.2013.03.007

Ulnar component pistoning has been described as one ofthe main failure mechanisms of total elbow prostheses,4

leading to loosening of the implant system. Joint reactionforces that result in tension/compression loads acting across

Board of Trustees.

Stem surface treatment and ulnar component stability 1249

the ulnohumeral joint occur when the elbow is in the flexedposition and can lead to this pistoning effect.1,8,11 Stemdesign factors, including the application of surface treat-ments, are expected to resist the effects of mechanicalloosening caused by these forces.6 With surface treatments,improved cement fixation can be hypothesized based on thepremise of a mechanical interlock formed between the bulkcement and treated stem surface. For ulnar componentsspecifically, a clinical study by Jeon et al10 showed thatvarious surface-treated stems had different rates of loos-ening; however, no known in vitro studies have comparedthe success of these surface treatments relative to oneanother.

Sintered beads and thermal plasma sprays are twocommon surface treatments used in ulnar stem designs. TheCoonrad-Morrey Total Elbow (Zimmer, Warsaw, IN, USA),in particular, has modified its titanium prosthesis designover the years based in part on varying these two surfacetreatments.10 In addition, the Latitude EV (Tornier SAS,Montbonnot Saint Martin, France) (cobalt chrome stem)and Discovery Elbow System (Biomet, Warsaw, IN, USA)(titanium stem) incorporate a plasma-spray surface treat-ment in their ulnar component designs. When one takesinto consideration the varying surface topographies of therespective surface treatments (ie, beaded and plasma spray),it can be hypothesized that the type of stem surface treat-ment may affect the strength of the mechanical interlockformed between the stem and cement. Comparisons of thismechanical interlock can offer insight into failure mecha-nisms associated with these surface-treated implants.

Titanium and cobalt chrome alloys are both used as thesubstrate material in elbow prosthesis stem designs.15 Stemsubstrate material may play a role in the success of theapplied surface treatment because the composition of thesubstrate can influence the strength of the stem-treatmentbond formed during the treatment process.5 As such, it isimportant to consider stem material when investigating therole of surface treatments in prosthesis loosening.

Therefore, the purpose of this in vitro study was toinvestigate the role of stem surface treatment and substratematerial on the stability of a simulated ulnar implant stemsubjected to cyclic compression loading.

Materials and methods

Sixty smooth, circular implant stems (diameter, 8 mm) werecustom machined from either titanium (n ¼ 30) or cobalt chrome(n ¼ 30) by Tornier SAS (Montbonnot Saint Martin, France). Foreach material, the stems were subdivided into 5 equal groups.Stems in the first group retained their full-length smooth surface(n ¼ 6), whereas the remaining 4 groups had standard commer-cially used stem surface treatments applied by Orchid Bio-Coat(Southfield, MI, USA). Two groups received a beaded surfacetreatment with a coverage length of 20 mm (n ¼ 6) or 10 mm (n ¼6), whereas two groups received a plasma-spray surface treatmentwith a coverage length of 20 mm (n ¼ 6) or 10 mm (n ¼ 6)

(Fig. 1). The beaded surfaces consisted of 1 layer of beads, witha bead diameter of approximately 500 mm. For the plasma-sprayedsurfaces, a titanium plasma spray was used with an averageroughness Ra value of 48.5 � 3.9 mm.

All stems were potted to a fixed depth of 20 mm in squarealuminum tubes by use of vacuum-mixed polymethyl methacry-late bone cement (Simplex P; Stryker, Kalamazoo, MI, USA) suchthat there was full coverage of the surface-treated regions. For the10-mm surface-treated stems, this potting method allowed forcement coverage of the 10-mm treated region, as well as anadditional 10 mm of proximal smooth stem surface (Fig. 1). Thepotted stems were maintained in air at 22�C for 24 hours duringcuring. Subsequently, they were secured in a materials testingmachine (Fig. 2) (Instron 8874; Instron, Norwood, MA, USA) byuse of fixturing to accommodate stem push-out under compres-sion. A Delrin stopper (Du Pont de Nemours International SA, LeGrand-Saconnex, Switzerland) was placed inside the aluminumtube at the base of the cement mantle to ensure push-out of thestem only, without slippage of the cement mantle within thealuminum tube. Loading was cycled at 1.5 Hz under compression,with loads being kept between 500 N and an upper limit.The upper limit started at 1000 N and increased in increments of1000 N every 100 cycles to a maximum 10,000 N, after which itcycled for a further 25,000 cycles to complete the testing protocol(ie, total of 25,900 cycles). Failure of the stem-cement interfacewas defined as 2 mm of push-out of the stem relative to thecement, termed ‘‘catastrophic failure,’’ or until completion of theloading protocol.

Motion at the stem-cement interfacewas quantifiedwith a customoptical tracking system (Basler Pilot GigE Camera [Basler,Ahrensburg, Germany]; Opto Engineering Telecentric Lens [OptoEngineering, Mantua, Italy]; Axial Diffuse Illuminator [AdvancedIllumination, Rochester, VT, USA]; and LabVIEW Vision Acqui-sition System [National Instruments, Austin, TX, USA]). Thissystem incorporated a color thresholding method to optically trackthe centroid of markers placed on the stem and cement (Fig. 2)to determine their relative distances throughout loading.

Immediately after testing, the cemented stems were placed inacetone (Caledon Laboratories, Georgetown, ON, Canada) for 24hours to allow dissolution of the surrounding bone cement. Thestems were subsequently cleaned and visually inspected todetermine the presence of any surface treatment damage associ-ated with testing.

Stem-cement interface strength and stem motion were bothquantified. Interface strength was determined from the number ofcycles required to cause failure for each of the stems, as obtainedfrom the materials testing machine. Stem motion was determinedfrom the relative distances between the markers on the stem andcement before failure (termed ‘‘interface toggle’’) (Fig. 2). Forthose stems that reached 10,000 N in the loading protocol andsurvived beyond the first 100 cycles, an additional measure ofstem motion (termed ‘‘global stem motion’’) was used. Globalstem motion measured the change in stem motion from 901 cycles(ie, start of the 10,000 N load step) until failure and allowedrelative comparison of surface treatments at the constant 10,000 Nload level (Fig. 2).

Two-way analyses of variance (ANOVAs) with post hocStudent-Newman-Keuls tests (a ¼ .05) were used to examine therole of stem surface treatment and substrate material on cementedstem stability, based on the measures of interface strength andstem motion.

Figure 2 Schema of the cemented stem in a materials testing machine, showing the application of a compressive load. The camera wasused to track markers placed on the stem and cement throughout loading as shown in the inset. The relative distance between the stem andcement was used to determine interface toggle just before failure (I.T.). The change in global stem motion from 901 cycles (ie, start of10,000 N load step) until failure (G.S.M.) was used to compare axial motion of surface-treated stems at a constant load level.

Figure 1 Implant stems with various surface treatments: smooth (A), 20-mm-long beaded treatment (B), 10-mm-long beaded treatment(C), 20-mm-long plasma-spray treatment (D), and 10-mm-long plasma-spray treatment (E). Stems were cemented to a fixed 20-mm depth,as highlighted by the region between the parallel lines. For the 10-mm-long surface-treated stems, this allowed full cement coverage of thesurface-treated region, as well as an additional 10-mm length of proximal smooth stem surface.

1250 Y.K. Hosein et al.

Results

Post-testing visual inspection of the stems found thatthe 10-mm beaded titanium and 10-mm beaded cobaltchrome stems were the only surfaces to have mechanicaldamage at failure (Fig. 3). This was observed asdebonding of the beaded treatments from the stemsurfaces. All other stem surfaces remained intact, with novisual damage evident.

Survival curves showed that stem surface treatment didaffect stem stability under compression loading, where the20-mm-long beaded stems outlasted the other stem surfaces(Fig. 4). Data obtained from the optical tracking systemshowed that stem motion increased simultaneously with thecyclic staircase loading protocol (Fig. 5), and as such, allmotion data (ie, interface toggle and global stem motion)

were normalized to their respective loads and cycles toallow for relative comparison of stem surfaces.

With regard to cycles to failure, the 2-way ANOVAfound an overall effect of surface treatment (P < .05) andno overall effect of substrate material (P ¼ .25), but therewas a significant interaction between these factors (P ¼ .02)(Fig. 6, A). Therefore, 1-way ANOVAs were performedand showed that for titanium, the 20-mm beaded stemsoutlasted all other treatments (P < .05). For cobalt chrome,the 20-mm beaded stems outlasted all other treatments(P < .05), but the 20-mm plasma-spray stems alsoperformed better than the 10-mm beaded, 10-mm plasma-spray, and smooth stems (P < .05).

With regard to interface toggle, the 2-way ANOVA againfound an overall effect of surface treatment (P < .05), nooverall effect of substrate material (P ¼ .78), and a signifi-cant interaction between these factors (P ¼ .01) (Fig. 6, B).

Figure 4 Survival curves for titanium (left) and cobalt chrome (right). The 20-mm-long beaded stems had the greatest survival, whereasthe stems with a smooth surface had the shortest survival, as defined by the number of cycles required to cause failure.

Figure 3 Post-testing inspection of stem surface treatments: titanium (left) and cobalt chrome (right) 20-mm-long beaded stems (A),titanium (left) and cobalt chrome (right) 10-mm-long beaded stems (B), titanium (left) and cobalt chrome (right) 20-mm-long plasma-spraystems (C), and titanium (left) and cobalt chrome (right) 10-mm-long plasma-spray stems (D). The 10-mm-long beaded titanium and cobaltchrome stems were the only stems to have debonding of the treatment from the stem surface at failure (star).

Stem surface treatment and ulnar component stability 1251

One-way ANOVAs showed that for titanium, the 20-mmbeaded, 10-mm beaded, and 20-mm plasma-spray stemshad less toggle than the 10-mm plasma spray and smoothstems (P < .05), with the 10-mm plasma-spray stemsexhibiting less toggle than the smooth stems (P < .05). Forcobalt chrome, the 20-mm beaded and 20-mm plasma-spray stems had less toggle than the 10-mm-long surfacetreatments and smooth stems (P < .05).

When comparing the global motion of surface-treatedstems at a constant 10,000 N load level, we found thatsurface-treated stems showed various magnitudes of stemmotion (Fig. 7). Overall, the 20-mm and 10-mm beadedtreatments had the least stem motion (P < .05), with the20-mm plasma-spray treatment also showing less stemmotion than the 10-mm plasma-spray treatment (P < .05).Cobalt chrome stems showed less motion than titanium(P ¼ .02).

Discussion

Surface treatments have been added to cemented ulnarcomponents of total elbow arthroplasties with the aim ofimproving stem fixation and long-term stability.10 Bonecement is considered a connecting agent, much like a groutmaterial. When used with a surface-treated component, itis expected to fill the spaces of the treated surface, initiatinga mechanical interlock between the stem and cement. Thestrength of the interlock formed may subsequently improveimmediate and long-term stem fixation.

The evolution of ulnar component designs over the lastdecade has incorporated different stem surface treatments,specifically sintered beads and plasma spray, with thegoal of improving the strength of the stem-cement inter-lock formed while preserving the mechanical integrity (ie,

Figure 5 Representative graphs of stem motion relative to cement (top row) and global stem motion (bottom row) for titanium and cobaltchrome stems. The 20-mm-long beaded stems did not have catastrophic failure, as defined by 2 mm of stem push-out, for 11 of the 12 stemstested. (Note: Interface toggle was extracted from the graph of relative stem motion, in the region just before failure.)

Figure 6 Interface stability offered by surface treatments. (A) Cycles at failure. For titanium, the 20-mm beaded treatment hadthe most cycles compared with all stem surfaces (P < .05 [U]). For cobalt chrome, the 20-mm beaded stem outlasted the other treatments(P < .05 [)]), with the 20-mm-long plasma-spray treatment also showing greater cycles to failure than the other stem surfaces (P < .05 [y]).(B) Interface toggle before failure. For titanium, the 20-mm beaded, 10-mm beaded, and 20-mm plasma-spray stems showed the smallestmagnitudes of interface toggle (P < .05 [z]), with the 10-mm plasma-spray stems having less toggle than the smooth stems (P < .05 [x]). Forcobalt chrome, the 20-mm beaded and 20-mm plasma-spray treatments showed reduced interface toggle compared with all other stems(P < .05 [¤]).

1252 Y.K. Hosein et al.

fracture properties) of the substrate stem material.10

Although clinical studies have assessed the success ofvarious commercially available elbow prosthesis designsincorporating these surface treatments,7,14,15 there are noknown in vitro studies that have compared their role incomponent stability or provided information on the failuremechanisms associated with these stem surface treatments.

Pistoning of the ulnar component of linked total elbowreplacements is one source of implant loosening and canoccur from axial forces acting at the implant-cementinterface. These axial forces result from dynamic loadingof the ulnohumeral joint in routine daily activities1,8,11

or from impingement of coronoid processes or protrud-ing cement during elbow hyperflexion, causing distraction

Figure 7 Global stem motion of surface-treated stems from901 cycles (ie, start of the 10,000 N load step) until failure.The 20-mm and 10-mm beaded stems showed the smallestmagnitudes of stem motion (P < .05 [infinity sign]), with the20-mm plasma-spray treatment also showing less stem motionthan the 10-mm plasma-spray treatment (P < .05 [paragraphsign]). Overall, cobalt chrome stems showed less motion thantitanium (P ¼ .02).

Stem surface treatment and ulnar component stability 1253

forces to occur on the ulnar component.4 Therefore, tocompare the effects of different surface treatments oncomponent stability, axial forces that may contribute tostem pistoning should be considered. Pistoning forces thatoccur in vivo typically act in tension at the proximal end ofthe ulnar component; however, the shear forces produced atthe stem-cement interface are the same as those expected ifthe stems were exposed to similar compression loads. Assuch, the purpose of this study was to investigate the role ofstem surface treatment on the stability of titanium andcobalt chrome implant stems using a study design incor-porating cyclic compression loading to mimic ulnar pis-toning at failure.

Stem surface treatment had a major effect on overallstem stability. Looking at interface strength and stemmotion, the 20-mm beaded treatment showed the greatestsurvival and least toggle when compared with the otherstem surfaces for both titanium and cobalt chrome stems(Fig. 6). Beaded treatments have greater surface deviationswith regard to their surface topography when comparedwith other stem surfaces. On the basis of the diameter of thesintered beads applied to the stem surface, this may allowfor greater infiltration of bone cement onto the stemsurface. Failure of the stem-cement interface would there-fore be dependent on the amount of shearing force neededto overcome each interlock along the stem’s surface.Greater cement infiltration, in terms of depth and number ofinterlocking sites, would therefore require greater cyclicloads to cause stem instability, resulting in interface failureand stem push-out.

For stems that failed catastrophically (ie, had 2 mm ofpush-out), significant interaction was observed betweenstem surface treatment and substrate material for measuresof interface strength and toggle. The 20-mm plasma-spraycobalt chrome stem performed better than both of the

10-mm-long surface treatments and smooth cobalt chromestems (Fig. 6). For titanium stems, however, the 20-mmplasma-spray treatment performed similar to the 10-mm-long surface treatments and smooth surfaces. Therefore,although stem surface treatment contributed to improvedinterface stability before catastrophic failure, this wasdependent on the stem substrate material.

When comparing the interface mechanics for the indi-vidual stem-cement interfaces at catastrophic failure, thesmooth stem surfaces appeared to offer no resistance to theshearing force at the interface, and as such, once the stem-cement bond was broken, the stem failed rapidly. Incomparison, for cobalt chrome stems, the roughenedsurface of the 20-mm plasma-spray stem provided fric-tional resistance to the shearing force at the stem-cementinterface, which was greater than that provided by the10-mm-long plasma-spray and 10-mm-long beaded treat-ments. Once the shearing force at the interface becamelarger than the frictional force, the stems had push-out.

Post-testing inspection of the stem surfaces found thatthe 10-mm beaded stem was the only surface to havemechanical damage at failure (Fig. 3). This suggests thatthe number of cycles to failure for the 10-mm beadedtreatment may not be representative of the stem-cementinterface strength but is more likely a measure of thestrength of the bond between the beads and stem surface.The strength of this bond is directly related to the carboncontent of the base substrate material, where a lower carboncontent can affect the success of the sintering process,causing a weaker bond between the stem substrate materialand attached beads.5 Although beaded treatments are likelyto contribute to satisfactory interface strengths, as observedfrom the 20-mm beaded treatments, there may be vari-ability in their performance based on the success of thesintering treatment during the fabrication process. As such,stringent standards should be placed on the fabricationprocess of beaded stem designs, ensuring adequate materialcomposition of the base substrate metals before applicationof the beaded treatments.

For stems that surpassed the staircase region of theloading protocol and reached 1000 cycles without failure,global stem motion was used to compare the contributionof stem surface treatments to stem motion at the constant10,000 N level. Global stem motion included motionof the stem with cement and offered information on thefailure patterns associated with the different stem surfacetreatments. Overall, beaded stems had less stem motioncompared with the plasma-spray surfaces. As mentionedpreviously, the roughened surface of the plasma-spraytreatment provided frictional resistance to the shearingforce at the stem-cement interface, and this resistancewas represented by the gradual increase in stem motionbefore catastrophic failure (ie, gradual interface failure). Incomparison, catastrophic failure of 10-mm-long beadedstems was influenced by the bond broken between thebeads and the stem surface, resulting in a stable interface

1254 Y.K. Hosein et al.

with little stem motion before stem push-out (ie, rapidinterface failure). When analyzing stem motion for the 20-mm-long beaded stems that did not have catastrophicfailure, the values observed were comparable to the 10-mm-long beaded surfaces, even at the end of the loadingprotocol (ie, 10,000 N and 25,900 cycles). This suggeststhat the 20-mm-long beaded treatment contributed toa stable stem-cement interface.

Titanium and cobalt chrome are two common stemmetals used in ulnar component designs.15 Each offers itsown advantages to implant systems in terms of biocom-patibility, wear resistance, and mechanical properties.9,13

However, application of surface treatments to thesubstrate metal can influence the success of the implantsystem. Clinical studies have reported that beaded treat-ments of titanium stems can cause increased rates ofcomponent fracture, believed to be caused by weakening ofthe metal during the bead-sintering process.3,10 Our studydid not evaluate the strength of the metal stemmedcomponents but found that the success of the surfacetreatments was directly related to stem material type. Asmentioned before, variations in the metal composition,specifically the carbon content of the substrate metal, candirectly affect the strength of the bond formed between thebeads and stem during the sintering process. Therefore,future studies should look into investigating the effect ofstem substrate material composition on the bondingstrength of beaded treatments, as well as the effect of thebead-sintering process on the fracture properties ofdifferent stem substrate materials commonly used withelbow prostheses.

During elbow flexion, the ulnohumeral joint is undercompression/tension, where resultant forces act upwardonto the distal end of the humerus,1 and the magnitude ofthese forces varies depending on the angle of elbow flexion.These joint forces can act on linked total elbow prosthesesin a similar manner, where resultant forces act upward ontothe humeral component, causing pullout of the connectedulnar component. Pullout of the ulnar component may alsooccur from distraction forces produced when the elbow ishyperflexed past a limit set by an impinging structure (ie,flange, cement, or bone), creating a fulcrum loadingscenario,10 or from carrying a heavy object during elbowextension, causing forces at the trochlea of up to 20 timesthe external load at the hand.2 Both axial loading examplescreate shear forces along the length of the stem-cementinterface, which can cause interface debonding and resul-tant pistoning of the ulnar component. Our study incorpo-rated a cyclic compressive load to mimic dynamic shearforces that may cause pistoning of the implant stem underaxial loading, similar to that experienced by the ulnarcomponent. As such, the loads used for compression testing(ie, 500-10,000 N) were intentionally chosen to comparethe effect of the different surface treatments in a cyclicpistoning scenario, similar to that caused by resultant jointforces or distraction forces at the ulnohumeral joint.

From the measures of stem motion as detected from theoptical tracking system, we observed that the cementdirectly surrounding the stem contributed to overall stemmotion. This was seen in measures of relative and globalstem motion, where the stem had greater motion in theglobal frame when compared relative to the cement(Fig. 5). This may be explained by the creep propertiesexhibited by bone cement under dynamic loading.12

This contribution of cement creep may also explain thevariability seen in our motion results for smooth stems.The stability of smooth stems was influenced by the stem-cement bond formed at the interface, and as such, anymotion detected before failure may have been solelydependent on the viscoelastic nature of the cement onindividual testing days. Bone cement properties could havealso affected the minimal stem motion observed for the20-mm beaded stems that completed the testing protocol(ie, 25,900 cycles) because previous work showed that bonecement becomes stiffer with increasing loading cycles.16

The testing setup used for our study incorporatedaluminum tubes instead of cadaveric specimens forcementing of the implant stems. Although this does notrepresent a clinically relevant scenario, it allowed us toinvestigate the effect of stem surface treatment on stem-cement interface stability, excluding any contributions fromvaried bone quality. In addition, all stems were pottedwithin standard-size aluminum tubes, which controlledcement mantle thickness and volume.

Conclusion

This in vitro study, to our knowledge, is the first tocompare the effects of stem surface treatment andsubstrate material on pistoning of an implant stem underaxial load. The study was successful at showing thecontribution of beaded and plasma-spray surface treat-ments to stem-cement interface stability and was able toprovide information on the failure mechanisms associ-ated with these surface-treated stems. Overall, the20-mm beaded stems offered the greatest stabilityamong all stem surfaces, and for cobalt chrome stemsonly, the 20-mm plasma-spray stems contributed toimproved stability as well. When we compared mecha-nisms of catastrophic failure, smooth stems failed bydebonding at the stem-cement interface, beaded stemsfailed by debonding of the beaded surface treatmentfrom the stem surface, and plasma-spray stems failed byloss of frictional force between the plasma-spray treat-ment and bone cement. It is expected that the resultsfrom this biomechanical analysis will help us understandthe contribution of surface treatments in componentpistoning and provide information about the failuremechanisms associated with similar clinical stemdesigns.

Stem surface treatment and ulnar component stability 1255

Acknowledgments

The authors thank Tornier SAS for donating the stemsused in this study; however, no financial benefits werereceived.

Disclaimer

This study received funding from the following sour-ces: the Natural Sciences and Engineering ResearchCouncil of Canada Discovery Grant (grant No. RGPIN/251354-2005); The Ontario Ministry of Research andInnovation Early Researcher Award (grant No. ER06-02-341); the Canadian Foundation for Innovation(project No. 9356); and the Joint Motion ProgramdACanadian Institute of Health Research TrainingProgram in Musculoskeletal Health Research andLeadership.

Graham J.W. King reports that he is a surgeondesigner and receives royalties for the Latitude TotalElbow Arthroplasty System (Tornier, Bloomington, MN,USA). All the other authors, their immediate families,and any research foundations with which they are affil-iated have not received any financial payments or otherbenefits from any commercial entity related to thesubject of this article.

References

1. Amis AA. Biomechanics of the elbow. In: Stanley D, Trail I, editors.

Operative elbow surgery. Edinburgh: Churchill Livingstone Elsevier;

2012. p. 29-44.

2. Amis AA, Dowson D, Wright V. Elbow joint force predictions for

some strenuous isometric actions. J Biomech 1980;13:765-75.

3. Athwal GS, Morrey BF. Revision total elbow arthroplasty for pros-

thetic fractures. J Bone Joint Surg Am 2006;88:2017-26. http://dx.doi.

org/10.2106/JBJS.E.00878

4. Cheung EV, O’Driscoll SW. Total elbow prosthesis loosening

caused by ulnar component pistoning. J Bone Joint Surg Am 2007;89:

1269-74. http://dx.doi.org/10.2106/JBJS.F.00376

5. Davis JR. Coatings. In: Handbook of materials for medical devices.

Materials Park (OH): ASM International; 2003. p. 179-94.

6. Evans BG, Daniels AU, Serbousek JC, Mann RJ. A comparison of

the mechanical designs of articulating total elbow prostheses.

Clin Mater 1988;3:235-48.

7. Fevang BS, Lie SA, Havelin LI, Skredderstuen A, Furnes O. Results

after 562 total elbow replacements: a report from the Norwegian

Arthroplasty Register. J Shoulder Elbow Surg 2009;18:449-56. http://

dx.doi.org/10.1016/j.jse.2009.02.020

8. Goldberg VM, Figgie HE III, Inglis AE, Figgie MP. Total elbow

arthroplasty. J Bone Joint Surg Am 1988;70:778-83.

9. Hallab NJ, Urban RM, Jacobs JJ. Corrosion and biocom-

patibility of orthopedic implants. In: Yaszemski MJ, Trantolo DJ,

Lewandrowski K, Hasirci V, Altobelli DE, Wise DL, editors.

Biomaterials in orthopedics. New York: M. Dekker; 2004. p.

63-91.

10. Jeon I, Morrey BF, Sanchez-Sotelo J. Ulnar component surface finish

influenced the outcome of primary Coonrad-Morrey total elbow

arthroplasty. J Shoulder Elbow Surg 2012;21:1229-35. http://dx.doi.

org/10.1016/j.jse.2011.08.062

11. Johnson JA, King GJW. Anatomy and biomechanics of the elbow. In:

Williams GR, Yamaguchi K, Ramsey ML, Galatz LM, editors.

Shoulder and elbow arthroplasty. Philadelphia: Lippincott Williams &

Wilkins; 2005. p. 279-96.

12. Lewis G. Viscoelastic properties of injectable bone cements

for orthopaedic applications: state-of-the-art review. J Biomed Mater

Res B Appl Biomater 2011;98:171-91. http://dx.doi.org/10.1002/

jbm.b.31835

13. Navarro M, Michiardi A, Castano O, Planell JA. Biomaterials in

orthopaedics. J R Soc Interface 2008;5:1137-58. http://dx.doi.org/

10.1098/rsif.2008.0151

14. Skytta ET, Eskelinen A, Paavolainen P, Ikavalko M, Remes V. Total

elbow arthroplasty in rheumatoid arthritis: a population-based study

from the Finnish Arthroplasty Register. Acta Orthop 2009;80:472-7.

http://dx.doi.org/10.3109/17453670903110642

15. van der Lugt JC, Rozing PM. Systematic review of primary total

elbow prostheses used for the rheumatoid elbow. Clin Rheumatol

2004;23:291-8. http://dx.doi.org/10.1007/s10067-004-0884-9

16. Verdonschot N, Huiskes R. Dynamic creep behavior of acrylic bone

cement. J Biomed Mater Res 1995;29:575-81.