a biomechanical comparison of equine third metacarpal condylar bone fragment compression and screw...

Veterinary Surgery31:201-210, 2002

A Biomechanical Comparison of Equine Third MetacarpalCondylar Bone Fragment Compression and Screw PushoutStrength Between Headless Tapered Variable Pitch and AO

Cortical Bone Screws

LARRY D. GALUPPO, DVM, Diplomate ACVS, SUSAN M. STOVER, DVM, PhD, Diplomate ACVS, and DAVID G. JENSEN

Objectives—To compare bone fragment compression and the mechanical pushout strength andstiffness of 6.5-mm Acutrak Plus (AP) and 4.5-mm AO cortical (AO) bone screws after stabilizationof a simulated equine third metacarpal (MC3) bone complete lateral condylar fracture.Study Design—In vitro biomechanical paired study of screw insertion variables, bone fragmentcompression, and screw pushout tests using a bone screw stabilized simulated lateral condylarfracture model.Sample Population—Six pairs of cadaveric equine MC3s.Methods—Metacarpi were placed in a fixture and centered on a biaxial load cell in a materials testingsystem to measure torque, compressive force, and time for drilling, tapping, and screw insertion.Fragment compression was measured with a pressure-sensing device placed between the simulatedfracture fragments during screw insertion for fragment stabilization. Subsequently, screws werepushed out of the stabilized bone fragments in a single cycle to failure. A paired t test was used toassess differences between site preparation, screw insertion, fragment compression, and screwpushout variables, with significance set at P � .05.Results—Measured drilling variables were comparable for AO and AP specimens. However, the APtap had significantly greater insertion torque and force. Mean maximum screw insertion torque wassignificantly greater for AO screws. For fragment compression, AP screws generated 65% and 44%of the compressive pressure and force, respectively, of AO screws. AP screws tended to have higheroverall pushout strength. Pushout stiffness was similar between both screw types.Conclusion—The 6.5-mm tapered AP screw generated less interfragmentary compressive pressureand force but had similar pushout stiffness. Evaluation of failure patterns demonstrated that APscrews had greater pushout strength compared with 4.5-mm AO screws for fixation of a simulatedcomplete lateral condylar fracture.Clinical Relevance—The 6.5-mm tapered AP screw should provide ample holding strength butwould provide less interfragmentary compression than 4.5-mm AO screws for repair of completelateral condylar fractures in horses.© Copyright 2002 by The American College of Veterinary Surgeons

THE ACUTRAK PLUS (AP) bone screw(Acumed, Beaverton, OR) is a titanium alloy

(ASTM, F136), cannulated, headless, variable-pitch,

self-tapping, tapered screw designed to provide inter-fragmentary compression for treatment of variousorthopedic disorders in humans.1 The screw shaft is

From the J. D. Wheat Veterinary Orthopedic Research Laboratory, School of Veterinary Medicine, University of California, Davis, CA;and the Acumed Corporation, Beaverton, OR.

Supported in part by the Marcia MacDonald Rivas Research Fund, University of California-Davis.No reprints available.Address correspondence to Larry D. Galuppo, DVM, Department of Surgical and Radiological Sciences, School of Veterinary

Medicine, University of California, Davis, One Shields Avenue, Davis, CA 95616.© Copyright 2002 by The American College of Veterinary Surgeons0161-3499/02/3103-0004$35.00/0doi:10.1053/jvet.2002.32399 201

threaded along its entire length with thread pitch (thedistance between individual threads) becoming morenarrow as the threads progress from the apex to the base.The screw measures 6.5 mm in diameter at the base andtapers to 5.0 mm in diameter at the apex. Flutes are cut inthe apex and the base of the screw to collect bone debristhat may accumulate through the self-cutting action. Thefully-threaded, tapered, variable-pitch design creates alarge bone-screw interface by allowing the threads toengage bone on both sides of the fracture site whilegenerating compression along the entire screw shaft asthe screw is tightened.1 The headless design allows thescrew to be placed beneath bone surfaces, avoidinginterference of critical overlying soft tissue structures.This feature may be desirable with equine third metacar-pal bone (MC3) fracture stabilization where, historically,screw heads are positioned within the lateral collateralligament of the fetlock joint.

Insertion of the AP screw in cadaveric equine bonehas been described.2 With its unique self-compressingdesign; the screw is designed to be inserted after onlydrilling a single hole perpendicular to the fracture.Drill bits are marked in 5-mm increments, whichfacilitates drilling accuracy. Because the AP screw isunable to cut its own path through the dense epiphy-seal bone of equine MC3,2 a noncannulated taperedtap is used to create threads for screw purchase.

The main objective of this study was to comparebone fragment compression and pushout mechanicalproperties between 4.5-mm AO cortical bone (AO),(Synthes, Paoli, PA) and AP screws after stabilizingsimulated lateral condylar fractures with either screwtype. It was hypothesized that the AO screw insertedin lag fashion would have greater overall fragmentcompression and compressed surface area because ofincreased contact area of the screw head. However,because of a larger screw diameter, and the ability ofeach thread to purchase new bone along the entirelength of the AP screw,1 it was hypothesized that theAP screw would have superior pushout mechanicalperformance than AO screws. Although the perfor-mance of AO and AP insertion equipment has beencompared,2 screw insertion variables (drill, tap andscrew compressive force, torque, and time) werequantitated to assess the repeatability of the measure-ments obtained previously.

MATERIALS AND METHODS

Six MC3 pairs, with no gross evidence of metacarpopha-langeal joint osteoarthritis, were collected from 6 Thor-

oughbred racehorse cadavers (3 geldings, 2 males, 1 female;mean [� SD] age, 5.2 � 1.8 years; range, 2 to 7 years).Specimens were wrapped in towels soaked with normal(0.09%) saline, packaged in plastic, and stored at �25°Cuntil testing. The left or right metacarpus of each pair wasrandomly assigned to the AO screw group, and contralateralmetacarpi were assigned to the AP screw group.

All bones were thawed at room temperature in normalsaline (0.9% NaCl) solution before specimen preparationand mechanical testing. The bone was irrigated continu-ously with normal saline for all sawing, drilling, tapping,and screw insertion procedures. A bone saw and guidedevice were used to manually create an 8-cm-long sagittalplane osteotomy in the lateral condyle near the midsagittalridge (Fig 1). This saw cut was the first of a 2-stageprocedure that was designed to simulate a complete lateralcondylar fracture.

To measure and record the screw insertion variables,drill, tap, and screw insertion torques, compressive forces,and times, all specimens were placed in a fixture andcentered on an axial-torsional, servohydraulic materialstesting system (model 809; MTS Systems Corporation,MN) fitted with a 2,200-N biaxial load cell (MTS). Drilling,tapping, and screw insertion torques and compressive forceswere measured by the biaxial load cell, and all data,including time, were acquired at 250 Hz throughout eachtest by analog/digital conversion and stored as a data file.For the AO screw, insertion torque, compressive force, andtime were measured for the 4.5-mm and 3.2-mm drill bits,the 4.5-mm bone tap, and the 4.5-mm AO screw. For the APscrew, insertion torque, compressive force, and time wererecorded for the 6.5-mm tapered drill bit, tap, and screw.Mean and maximum torque and compressive force andmaximum time were reported for all drilling, tapping, andscrew insertion procedures. Mean drilling, tapping, andscrew insertion torque and compressive force were calcu-lated by averaging torques and compressive forces for theentire drilling, tapping, or screw insertion procedure. Max-imum drilling, tapping, and screw insertion torques andcompressive forces were derived from the force or torquemeasurements that corresponded to the highest valuesobtained during mechanical testing. A single operator per-formed all drilling, tapping, and screw insertions manually.

For insertion of all 4.5-mm diameter, 48-mm–long AOscrews (total thread length of 45 mm), a 4.5-mm bit with adrill stop, inserted in a pneumatic drill (Small Hand Drill;Synthes) was used to drill a hole from the lateral epicondyleto the simulated fracture. The hole was centered in theepicondylar fossa of the lateral condyle and was perpendic-ular to the saw cut and parallel with the articular surface.Compressed nitrogen gas set to the manufacturer’s recom-mended pressure (90 psi) was used to power the drill. Afterinsertion of a 3.2-mm insert sleeve, a 3.2-mm bit was usedto drill through the remaining epiphyseal bone (completelyexiting the medial epicondyle), and a 4.5-mm bone tap was

202 EQUINE THIRD METACARPAL CONDYLAR BONE FRAGMENT COMPRESSION

used to create threads for screw purchase. Countersinkingwas not performed.

For insertion of all 6.5-mm-diameter, 45-mm-long APscrews, a noncannulated, tapered bit was used to drill(Synthes, PA; compressed nitrogen at 90 psi) a hole 50 mmdeep, and a 45-mm-long, tapered bone tap was used tocreate threads for screw purchase.

After drilling and tapping, the second stage of theostectomy was performed. A 4-cm-long segment of bonewas removed by making 2 saw cuts with the distal most cutplaced 4 cm proximal to the lateral metacarpal condyle (Fig1). Lateral condylar bone fragments were then removed andphotographed with a measurement scale to obtain surfacearea calculations of the entire bone fragment (Scion Imagesoftware; Scion Corporation, NY). A circular pressuresensor with a central hole to accommodate screws (Fig 2A,Tekscan Inc, Boston, MA), calibrated for a pressure rangefrom 0 to 10,000 kPa with the MTS (model 809; MTSSystems Corporation, MN), was placed between bonefragments. The fragments were then secured with therespective AP or AO screw type. Sensors were positionedbetween bone fragments to obtain pressure measurementfrom the epiphysis of MC3. All screws were tightenedmanually to a perceived maximum torque (determined byperforming preliminary tests, which assessed the maximumtorque required to strip the recessed screw driver insert ofthe AP screw and cause screw head failure of the AOscrew). Compressive pressure, force, and surface area undercompression were measured for all specimens. Compres-sion was reported as total contact pressure or force betweenbone fragments with a threshold set at �0 kPa or N,respectively. Compressed surface area was determined bythe area of the sensor that had detectable contact pressures(�0 kPa; Fig 2B). The pressure sensor was then removed,

and the simulated lateral condylar fracture was restabilizedwith the respective screw type. Each screw was tightenedmanually to the previously recorded maximum torque valueobtained during compression testing.

Stabilized specimens were prepared for pushout testingby drilling a 4.5-mm hole from the medial epicondyle to thedistal end of each screw. For AO screws, a 4.5-mm drill bit(Synthes) was used to enlarge the 3.2-mm hole that exitedthe medial epicondyle during the drilling and tappingprocedure. Over drilling, the 3.2-mm pilot hole allowedaccurate alignment of the 4.5-mm hole for insertion of a4.5-mm pushout device. Because the 50-mm-deep holedrilled for all AP stabilized specimens did not exit themedial epicondyle, accurate alignment of the 4.5-mm holewas accomplished by using a guide pin placed through theproximal end of the cannulated screw. A 1.6-mm guide pin(Acumed, Beaverton, OR) secured in a pneumatic drill(Synthes) was inserted through the epiphyseal bone to exitthe medial epicondyle. A cannulated 4.5-mm drill bit(Synthes) inserted over the guide pin was then used to drilla hole extending from the medial epicondyle to the distalend of the AP screw.

After final preparation, a dorsopalmar radiograph (80 kV,0.1 s, 15 mA, [MinXray; Northbrook, IL], with Lanex finedouble-screen system [Eastman Kodak, Rochester, NY])was taken of each MC3 construct.

Each MC3 construct was aligned in a materials testingmachine (MTS Systems Corporation, Minneapolis, MN)equipped with a 15,000-N biaxial load cell to allow align-ment of the pushout device with the distal end of the screwin each specimen (Fig 3A). A separate pushout shaft, eachwith an outer diameter of 4.5 mm, was used to accommo-date each screw type (Fig 3B). All MC3 constructs weretested in a single cycle to failure under displacement control

Fig 1. Diagrammatic representation of the distal epiphysis of an equine third metacarpal bone (MC3), (A) dorsal view and (B)sagittal view, showing the orientation of saw cuts that were used to create a simulated lateral condyle fracture. The distal 4-cmsegment of bone was affixed with either a 4.5-mm AO or 6.5-mm AP bone screw. The proximal 4-cm segment of bone was removedto enable accurate positioning of the pressure sensor to obtain interfragmentary compression measurements.

203GALUPPO, STOVER, AND JENSEN

at a rate of 1 mm/min to a maximum of 10 mm displace-ment. Displacement and load data were acquired at 4 Hzthroughout each test by analog/digital conversion and storedas a data file. Dorsopalmar radiographic projections wererepeated after testing on all specimens using the same filmscreen combination and radiographic technique.

Load-deformation curves were generated for all mechan-ical tests. The yield point was determined as the point wherethe curve first deviated from a linear region. Stiffness wasdefined as the slope of the linear region. The failure pointwas determined as the point of maximum load. Yield andfailure loads and displacements were determined by therespective values for the yield and failure points. Yield and

failure energies were derived from the area under the curvecorresponding to the yield and failure points, respectively.Failure site, described as bone-screw interface, screw, orboth were recorded for all specimens.

Mean values obtained for screw insertion variables (tapand screw torques, compressive forces, and times), bonefragment compression (force, pressure, and surface areameasurements), total surface area, and surface area undercompression and for mechanical testing variables (yieldload, displacement, and energy; failure load, displacement,and energy; and stiffness) were compared between AO andAP screws with a paired Student t test. A Bonferroniadjustment for multiple comparisons was used to compare

Fig 2. (A) The pressuresensor (Tekscan, MA) usedto obtain interfragmentarycompression measurementsfor 4.5-mm AO and 6.5-mmAP screws. Sensors wereplaced between bone frag-ments of the simulated lat-eral condylar fracture, andfragments were compressedby placing the respectivescrew type from the epicon-dyle of the lateral condylebone fragment through thecentral hole of the sensorinto the parent bone. (B)Pressure mapping of theinterfragmentary compres-sion generated by 4.5-mmAO and 6.5-mm AP screws.The pressure scale (kPa)correlates highest pressureswith the lightest shade andlowest with the darkestshade. The center of force(white arrows) estimatedfrom the program softwarewas located centrally forAO and was slightly distalto center for AP screws. Thesmall wedge section, with-out piezoelectric sensors(black arrows), was placedproximally so that inter-fragmentary compressioncould be measured on thedistal portion of the con-dyle.


drill torques, compressive forces, and times between the3.2-mm and 4.5-mm AO, and the 6.5-mm tapered AP drillbits. Significance for all analyses was set at P � .05. Therelationships between maximum screw torque and bonefragment compression and pushout strength were evaluatedfor both AO and AP fixations by regression analysis. Theamount of bone fragment surface area under compressioncompared with total surface area was also expressed as apercentage for both screw types.

RESULTS

There were no significant differences in mean(�SD) maximum and mean drilling torques betweenthe 3.2-mm and 4.5-mm AO and the noncannulated,tapered AP drill bits. Mean drilling force was similarfor the 3.2-mm and 4.5-mm AO bits, and the AP and4.5-mm AO bits, but the 3.2-mm bit had significantlygreater mean drilling force than the AP bit (P � .02).Maximum drilling force was greater for the taperedAP and the 3.2-mm AO drill bits than the 4.5-mm AObit (AP, P � .01, and 3.2-mm, P � .03) but was notdifferent between the two. The drilling time for the3.2-mm AO bit was similar to the 4.5-mm bit, and thecombined drilling time of the 3.2- and 4.5-mm bitswas not different from the AP bit.

The tapered AP bone tap had greater mean andmaximum tapping torques (P � .01) and mean andmaximum tapping forces (P � .04 and .03, respec-tively) than the 4.5-mm AO tap, but maximum tappingtimes were similar. The AP and AO screws had similarmean and maximum insertion forces, mean insertiontorque and total insertion time; however, the AO screwhad a greater maximum insertion torque (P � .006)compared with the AP screw (Table 1).

All 45-mm-long, 6.5-mm-diameter AP (4.1-mmaverage core diameter) and all 48-mm-long, 4.5-mm-diameter AO (3.1-mm core diameter) screws wereinserted in the distal epiphysis of MC3 without com-plications. AP screws had 45 mm of thread purchase,and AO screws had 25 mm of total thread purchase.Compression of the osteotomy appeared adequate inall specimens, as determined by an inability to observe

4

Fig 3. (A) Diagrammatic representation of pushout mechan-ical testing of a 6.5-mm AP screw stabilizing a simulatedlateral condylar fracture of a third metacarpal bone (MC3).MC3 constructs were secured in a materials testing machine(MTS Systems Corporation, MN) equipped with a 15,000-Nbiaxial load cell to allow alignment of the pushout device withthe distal end of the screw. A recessed portion of the MC3support allowed for screw displacement (arrow). (B) The pushrods used for pushout mechanical testing of 4.5-mm AO (a)and 6.5-mm AP (b) screws from stabilized simulated lateralcondylar fractures. The cupped portion of rod “a” was de-signed to fit over the distal end of the 4.5-mm AO screw. Themetallic extension of rod “b” was inserted into the distal end ofthe cannulated 6.5-mm AP screw, which allowed accuratepushout rod alignment. Both pushout rods had an outerdiameter of 4.5 mm to prevent interference with the sides ofthe 4.5-mm guide hole (drilled into the medial epicondyle)during pushout testing.


the osteotomy gap on radiographic images. Becausethe self-tapping threads of the AP screw would noteasily cut the dense bone of the distal epiphysis, it wasnecessary to tap the entire 50-mm-deep hole with the45-mm-long bone tap. Turning the screw one half toone full revolution clockwise and one to two revolu-tions counterclockwise (as if tapping threads) alsofacilitated inserting the screw to the appropriate depth.

Mean (� SD) condylar fragment surface area andsurface area under compression was similar for APand AO screws. However, AO screws produced sig-nificantly greater total contact compressive force andpressure (Table 2).

When load-deformation curves were evaluated forpushout testing, the AO screw had significantly greatermean yield load (P � .02), and displacement (P �.02), but similar yield energy (P � .08) and stiffness(P � .7). The AP and AO group had similar overallpushout strength (failure load), but AP screws hadsignificantly greater failure displacement (P � .02)and energy (P � .02; Table 3).

All AO constructs failed by bone failure at thebone-screw interface. AP constructs failed initially bydeformation of the screw at the pushout rod interfacefollowed by progressive screw deformation at thescrew body (3 constructs), screw deformation fol-lowed by shearing of the bone at the bone-screwinterface along the entire length of the screw (1),primary bone failure at the bone-screw interface (1),and by deformation of the screw at the apex only (1),(Figs 4 and 5).

On regression analysis of the relationship betweenmaximum screw torque and compressive pressure,there was a nonsignificant inverse relationship for AO(y � �1,278.5x � 11,059, r2 � .0656, P � .28) andAP (y � �1,157.6x � 7,744.4, r2 � .287, P � .054)screws. There was a nonsignificant positive relation-ship between maximum screw torque and pushoutstrength for AO (y � 3.1815x � 1.5885, r2 � .4834,

P � .80) and AP (y � 3.9619x � 0.0866, r2 � .3574,P � .99) screws. For both analyses, torque valueswere plotted on the x axis, and compressive pressureor pushout strength was plotted on the y axis.

DISCUSSION

Our results suggest that a 6.5-mm tapered AP screwshould provide equivalent or greater holding powerthan an AO screw; however, it may not provideequivalent compressive force between fracture frag-ments. AO screws generated significantly more meaninterfragmentary compressive force and pressure buthad a similar amount of surface area under compres-sion compared with AP screws. Considering the com-parable interfragmentary shear strength and stiffnessof AP and 4.5-mm AO screw constructs tested in asimilar in vitro model,2 further investigation is neces-sary to determine the minimum amount of compres-sive force required to maintain stability and promotehealing of complete condylar fractures in horses.

Drilling, tapping and screw insertion torque, force,and time were measured as possible indicators ofinstrument or implant failure, excessive heat genera-

Table 2. Mean � SD Bone Fragment Surface Area and CompressionMeasurements for AO and AP Screws

FixationMethod

TotalSurface

Area (cm2)

CompressedSurface

Area (cm2)Compressive

Force (N)Compressive

Pressure (kPa)

AO 12.4 � 1.5 6.5 � 1.6 4121.7 � 1718.7a 6456.7 � 2546.0a

*(52%)AP 12.1 � 1.1 5.4 � 0.8 1817.2 � 650.0b 4175.0 � 1480.9b

*(44%) (44%)† (65%)†

For a given variable, values with different superscripts are significantly(P � .05) different from one another.

*Values in parenthesis are mean % of total surface area under compres-sion.

†Values in parenthesis are mean % of the compressive force andpressure generated by AAP as compared with AO screws.

Table 1. Mean (�SD) Drill, Tap and Screw, Torques, Forces, and Times for the 4.5-mm AO and 6.5-mm AP Insertion Equipment

Screw Type Drill, Tap, Screw Mean Torque (Nm) Mean Force (N) Max Torque (Nm) Max Force (N) Max Time (Sec)

AO

3.2-mm drill 0.2 � 0.1 67.3 � 5.9a 0.9 � 0.4 183.5 � 38.6a 36.9 � 6.44.5-mm drill 0.1 � 0.1 56.8 � 12.4a,b 0.5 � 0.2 119.4 � 26.9b 24.3 � 10.94.5-mm tap 0.1 � .04a 7.1 � 2.3a 0.8 � 0.2a 50.3 � 14.8a 54.0 � 14.54.5-mm screw 1.0 � 0.2 23.5 � 14.1 3.6 � 0.5a 62.6 � 34.0 22.7 � 4.8

APAP drill 0.2 � 0.1 58.0 � 8.3b 0.8 � 0.1 164.6 � 17.4a 65.5 � 20.1*AP tap 0.5 � 0.1b 9.6 � 4.2b 2.9 � 0.4b 67.3 � 24.7b 59.5 � 15.8AP screw 1.0 � 0.5 38.5 � 16.8 3.1 � 0.7b 98.0 � 32.0 17.2 � 13.6

For a given variable (eg, maximum torque for AO and AP taps), values with different superscripts are significantly (P � .05) different from one another.*The mean (� SD) combined drilling time for the 3.2-mm and 4.5-mm AO drill bits was not different than the tapered AP drill bit.


tion, and prolonged surgery times. The values forinsertion variables were comparable to those previ-ously reported,2 indicating that use of the AP systemperformed consistently in cadaveric equine bone, andif used with continuous fluid irrigation, was likely tobe comparable to existing techniques for drilling,tapping, and screw insertion in living equine bone.

AO screws had significantly greater mean yieldpushout strength, and similar pushout stiffness andfailure strength compared with AP screws. However,AO screws failed by bone failure at the bone-screwinterface, while the majority of AP screws failedprimarily by screw deformation. Because it was notpossible to tell whether AP screw deformation was atesting artifact or a true failure mode, AP screws mayhave greater pushout strength than AO screws.

AO screws compressed 54% of the total surfacearea compared with 42% by AP screws (Table 2). Thearea under compression, a circular shape extendingoutward from the head (AO) or the base (AP) of thescrews, was associated with a radiating force andpressure gradient. The highest forces and pressureswere located centrally and distally (AO), and centrallyand palmarly (AP) screws, with diminishing valuesextending toward the outer circumference proximally(Fig 2B). With higher pressures located centrally,placing either AO or AP screws more distal in thecondylar fossa should result in higher pressures lo-cally, which may increase fracture stability and there-fore enhance healing at the articular surface. Furtherinvestigations to identify ideal screw placement inrelation to compressive force or pressure generated byAO or AP screws may be warranted.

Peak compression achieved in standardized simu-lated bone material using the Acutrak and Herbertscrews were 65% and 39%, respectively, of thatattained by the AO cancellous screw applied in lagfashion.1 In a previous report using shear stiffness asan indirect assessment of interfragmentary compres-sion,2 6.5-mm AP screws achieved 69% of the meanshear stiffness attained by 4.5-mm AO screws. In ourstudy, mean interfragmentary compression using AP

screws was 44% (compressive force) and 65% (com-pressive pressure) (Table 2) of that of AO screws.Given that the amount of interfragmentary compres-sion required for maintaining reduction and promotinghealing of complete, displaced, longitudinal fracturesof third metacarpal or metatarsal condyles is notspecifically known, further investigations are requiredbefore use of AP screws can be recommended as thesole means of fixation where interfragmentary com-pression is the predominant force responsible formaintaining fracture stability.

Maximum insertion torque of the AP and AO

Fig 4. A dorsopalmar radiographic projection of a thirdmetacarpal bone (MC3) construct after pushout testing of a4.5-mm AO cortical bone screw (A). Failure occurred in all 6constructs by shearing of bone at the bone-screw interfacealong the entire length of the screw as represented by thecharacteristic load deformation curve (B).

Table 3. Mean (�SD) Mechanical Testing Variables for AO and AP Pushout Tests

FixationMethod

Yield Failure

Load (kN) Displacement (mm) Energy (kNmm) Stiffness (kN/mm) Load (kN) Displacement (mm) Energy (kNmm)

AO 7.2 � 1.2a 1.8 � 0.2a 5.0 � 1.8a 5.6 � 0.6 9.9 � 2.3 2.8 � 0.8a 15.1 � 11.6a

AP 5.4 � 1.5b 1.1 � 0.5b 3.4 � 2.1b 5.9 � 2.3 12.1 � 4.7 7.4 � 3.6b 72.6 � 47.8b

For a given variable, values with different superscripts are significantly (P � .05) different from one another.


Fig 5. Dorsopalmar radiographic projection (left) and corresponding load-deformation curves (right) obtained from pushouttesting of third metacarpal bone (MC3) constructs stabilized with 6.5-mm AP screws. Evaluation of the post-test radiographsdemonstrated that in 3 constructs (A), screw failure occurred at the distal end of the screw (black arrow), followed by deformationin the body of the screw (white arrow). The base of the screw remained stable. In 1 construct (B), screw failure occurred at theapex (black arrow) followed by shearing of bone at the bone-screw interface along the entire length of the screw (white arrow).In 1 construct (C), failure occurred by shearing of bone along the entire length of the screw (white arrows) without permanentdeformation of the screw. In 1 construct (D), screw failure only occurred at the screw apex (black arrow), and the repairedsimulated lateral condylar fracture remained stable. When evaluating the corresponding load-deformation curves, it was likelythat initial construct yielding (black arrows) was associated with deformation of the screw apex (A, B, and D) or shearing of thebone at the bone-screw interface (C). When the shear strength of the bone appeared to be high, yielding and progressive plasticdeformation occurred primarily by plastic deformation of the screw (A, 3 constructs and D, 1 construct). When the shear strengthof the bone was similar to the compressive strength of the screw, the construct yielded by deformation at the apex of the screwand progressive plastic deformation were associated with shearing of bone at the bone-screw interface along the entire length ofthe screw (B, 1 construct). When the shear strength of the bone was lower than the compressive strength of the screw, initialyielding and continued plastic deformation occurred by shearing of the bone at the bone-screw interface (C, 1 construct).

screws represented the perceived maximum torquethat would provide maximum interfragmentary com-pression and prevent torsional failure of the screw orbone-screw interface. The authors elected not to stan-dardize maximum insertion torque to obtain an appre-ciation for the manual force required when tighteningthe screws. The maximum force required to strip therecessed screw driver insert of the AP screw wasbetween 5.8 and 6.0 Nm.2 There was a 47% margin ofsafety for stripping because the mean maximum APscrew torque was 3.1 Nm. However, the margin ofsafety was only 28% for the AO screw because screwhead failure occurred between 5 and 5.5 Nm,2 andmaximum mean insertion torque was 3.6 Nm.

Maximum insertion torque was greater for AOscrews than AP screws, which may suggest that highermaximum screw torque would likely result in greaterinterfragmentary compression. However, in this study,there was no linear relationship between screw torqueand bone fragment compression for both groups.Although a direct linear relationship between screwtorque and interfragmentary compression achievedwith screws inserted in lag fashion has been reported,3

the mean maximum torque achieved was 1.27 Nm. Itis possible that a direct relationship may exist for alower maximum screw torque; however, once thatmaximum is reached, interfragmentary compressionmay peak as well. For headed screws inserted intopretapped holes, the torque loss because of threadfriction is minimal. However, the torque loss becauseof friction at the countersink interface can be up to40% of the maximum insertion torque.4 It is possiblethat as AO screws were tightened, maximum interfrag-mentary compression was achieved at lower torquevalues, and the final maximum insertion torque repre-sented a large component of the torque to overcomethe frictional forces at the site where the screw headcontacted the bone in the condylar fossa. Furtherinvestigations to determine the optimum torque for APscrew insertion and for AO screws inserted in lagfashion for fixation of third metacarpal or metatarsalcondylar fractures to achieve maximum interfragmen-tary compression may be warranted.

The screw pushout rate of 1 mm/min is much lowerthan previous studies4-13; however, it has been used inanother study assessing pushout strength of multiplescrew types.1 We elected to use this testing rate toevaluate the pushout strength of the AP screw becauseit has been used successfully1 and because the pushouttechnique had not been described in previous equinestudies. Although adjusting loading rates for specific

in vivo loading environments seems warranted, stan-dardizing these rates would enhance comparisonsbetween similar biomechanical studies.

Holding strength, or power, has been demonstratedto be a good measure of screw performance.1,3-18

Considering the configuration of headed screws, me-chanical testing to determine holding strength has beenaccomplished mainly by pullout tests.3-18 However,with the headless design of AP screws, it was neces-sary to perform pushout testing to determine holdingstrength. Although this was not a traditional method,directing the pushout force along the longitudinal axisshould equal the pullout force directed along the sameplane.1

Yield pushout load, displacement, and energy, andstiffness values obtained from pushout testing shouldbe physiologically relevant as they likely relate to theamount of holding power generated by each screwtype before plastic deformation occurs. Although thiswas true for AO screws, which yielded by shear failureof the bone at the bone-screw interface (6/6; Fig 4), itdid not appear to represent the physiological holdingstrength of AP screws. The yield and stiffness valuesfor AP screws appeared to represent the load anddisplacement values that were associated with defor-mation of the screw apex only. Thus, most APstabilized specimens (5/6) remained stable after yield-ing, whereas AO constructs, which yielded at thebone-screw interface along the entire length of thescrew, became unstable (Figs 4 and 5). It is possiblethat the effective holding strength of the AP screwmay be more closely related to the failure loadobtained from mechanical pushout testing, and there-fore they should have much greater holding strengththan AO screws.

Although it was not possible to tell whether APscrew deformation was a testing artifact or a truefailure mode, the failure pattern of the AP constructsmay be explained by the theoretical relationship be-tween the screw type and configuration and shearstrength of the surrounding bone. The effective hold-ing power of the tapered AP screw was likely differentfor each progressive change in diameter and threadpitch. When the shear strength of the bone appeared tobe higher than that of the compressive strength of thescrew, only plastic deformation of the screw occurred(Fig 5A and D). When the shear strength of the boneappeared to be similar to that of the compressivestrength of the screw, after initial yielding at the screwapex, failure occurred by shearing of bone at thebone-screw interface along the entire length of the


screw (Fig 5B). It was likely that progressive defor-mation of the screw occurred until a larger diameterregion was reached, which resulted in a screw com-pressive strength that was greater than the shearstrength of the bone. When the shear strength of thebone appeared to be lower than the compressivestrength of the screw, failure occurred by shearing ofthe bone at the bone-screw interface along the entirelength of the screw (Fig 5C).

The results of this study indicate that the APinsertion equipment should be equivalent to existingtechnology for use in equine bone. No complicationswere recognized during drilling, tapping, and screwinsertion for either fixation method. The 6.5-mm APscrew generated less compressive force and pressurethan AO screws; however, considering the comparablemechanical performance in shear testing,2 further in-vestigations are required to understand the relationshipbetween interfragmentary compression and shearstrength. Although the AP screw system tended togenerate higher pushout strength in this study, andmay have several advantages, which can facilitaterepair of incomplete, nondisplaced longitudinal frac-tures of third metacarpal and metatarsal condyles,2

until further studies are performed to investigate theminimum compression required for stabilization andhealing of complete, displaced condylar fractures,using AP screws as the sole means of fixation of thesefractures in horses must be cautioned.

ACKNOWLEDGMENT

This study was supported in part by the Marcia Mac-Donald Rivas Research Fund, University of California-Davis. The authors thank the Acumed Corporation, Beaver-ton, OR, for providing insertion equipment, and AcutrakPlus screws.

REFERENCES

1. Wheeler DL, Mcloughlin SW: Biomechanical assessment ofcompression screws. Clin Orthop 350:237-245, 1998

2. Galuppo LD, Stover SM, Jensen DG, et al: A biomechanicalcomparison of headless tapered variable pitch and AO

cortical bone screws for fixation of a simulated lateralcondylar fracture in equine third metacarpal bones. VetSurg, accepted for publication, November, 2000

3. Nunamaker DM, Perren SM: Force measurements in screwfixation. J Biomech 9:669-675, 1976

4. Hughes AN, Jordan BA: The mechanical properties of surgi-cal bone screws and some aspects of insertion practice.Injury 4:25-36, 1972

5. Koranyl E, Bowman CE, Knecht CD, et al: Holding power oforthopedic screws in bone. Clin Orthop 72:283-286, 1970

6. Yovich JV, Turner AS, Smith FW: Holding power of ortho-pedic screws in equine third metacarpal and metatarsalbones part I. Foal bone. Vet Surg 14:221-229, 1985

7. Yovich JV, Turner AS, Smith FW: Holding power of ortho-pedic screws in equine third metacarpal and metatarsalbones part II. Adult horse bone. Vet Surg 14:230-234, 1985

8. Yovich JV, Turner AS, Smith FW, et al: Holding power oforthopedic screws comparison of self-tapped and pre-tappedscrews in foal bone. Vet Surg 15:55-59, 1986

9. DeCoster TA, Heetderks DB, Downey DJ, et al: Optimizingbone screw pullout force. J Orthop Trauma 4:169-174, 1990

10. Leggon R, Lindsey RW, Doherty BJ, et al: The holdingstrength of cannulated screws compared with solid corescrews in cortical and cancellous bone. J Orthop Trauma7:450-457, 1993

11. Blikslager AT, Bowman KF, Abrams FC, et al: Holdingpower of orthopedic screws in the large metacarpal andmetatarsal bone of calves. Am J Vet Res 55:415-418, 1994

12. Serdish SA, Moore RM, Kelly K, et al: In vitro pulloutstrength of screws inserted in adult equine third metacarpalbone after overdrilling a 4.5-mm threaded insertion hole.Vet Surg 27:143-149, 1998

13. Colgan SA, Hecker AT, Kirker-Head CA, et al: A comparisonof the Synthes 4.5-mm cannulated screw and the Synthes4.5-mm standard cortex screw systems in equine bone. VetSurg 27:540-546, 1998

14. Lyon WF, Cochran JR, Smith L: Actual holding power ofvarious screws in bone. Ann Surg 114:376-384, 1941

15. Ansell RH, Scales JT: A study of some factors which affectthe strength of screws and their insertion and holding powerin bone. J Biomech 1:279-302, 1968

16. Uhthoff HK: Mechanical factors influencing the holdingpower of screws in compact bone. J Bone Joint Surg [Br]55:633-639, 1973

17. Uhl RL: The biomechanics of screws. Orthop Rev 18:1302-1307, 1989

18. Daftari TK, Horton WC, Hutton WC: Correlations betweenscrew hole preparation, torque of insertion, and pulloutstrength of spinal screws. J Spinal Disorders 7:139-145,1994


a biomechanical comparison of equine third metacarpal condylar bone fragment compression and screw...

Documents