of 59 /59

UNIVERSITY OF WASHINGTON Department of Orthopaedics 1998 … · 2019. 7. 30. · Benirschke, has established the Jerome H. Debs Endowed Professorship in Orthopaedic Traumatology

  • Author

  • View

  • Download

Embed Size (px)

Text of UNIVERSITY OF WASHINGTON Department of Orthopaedics 1998 … · 2019. 7. 30. · Benirschke, has...


    Department of Orthopaedics1998 Research Report

  • Department of OrthopaedicsUniversity of WashingtonSeattle, WA 98195

    EDITORS:Frederick A. Matsen III, M.D.Peter Simonian, M.D.Fred WesterbergSusan E. DeBartolo

    DESIGN & LAYOUT:Fred Westerberg

    Cover Illustration: “The Champion Single Sculls” by Thomas EakinsThe Metropolitan Museum of Art, Purchase,The Alfred N. Punnett Endowment Fund andGeorge D. Pratt Gift, 1934. (34, 92)Photograph (c) 1994 The Metropolitan Museum of Art



  • Contents

    Foreword .................................................................................................................................................................... 1

    Sports Safety for Women: An Opportunity Whose Time Has Come ..................................................................... 2

    Frederick A. Matsen, III, M.D., Peter T. Simonian, M.D., and Carol C. Teitz, M.D.

    The Effect of Neck Position on Spinal Canal Occlusion in a Cervical Spine Burst Fracture Model ..................... 4

    Randal P. Ching, Ph.D., Nathan A. Watson, B.S., Jarrod W. Carter, B.S., and Allan F. Tencer, Ph.D.

    Differential Localization of Collagen Types I, IIA and III in Human Osteoarthritic Cartilage ............................ 7

    Yong Zhu, M.D., Howard A. Chansky, M.D., Frederick A. Matsen, III, M.D., and Linda J. Sandell, Ph.D.

    Proteolysis of Cartilage Collagens II, IX and XI by Collagenase-3 (MMP-13) ...................................................... 9

    Jiann-Jiu Wu, Ph.D., Vera Knauper, Ph.D., Gillian Murphy, Ph.D., and David R. Eyre, Ph.D.

    Matrix Metalloproteinase-Mediated Release of Immunoreactive Telopeptides from Cartilage Type II

    Collagen ................................................................................................................................................................... 11

    Lynne M. Atley, Ph.D., Ping Shao, B.S., Vince Ochs, B.S., Kathy Shaffer, B.S., and David R. Eyre, Ph.D.

    Peak Growth Age: A New Maturity Indicator for Idiopathic Scoliosis ................................................................. 13

    Kit M. Song, M.D., David G. Little, M.B.B.S., F.R.A.C.S. (Orth), Don Katz, C.O., and John Herring, M.D.

    Biomechanical Study of Foot Function .................................................................................................................. 17

    Bruce J. Sangeorzan, M.D.

    Assessing the Value of a Procedure: A Pilot Study Using Total Shoulder Arthroplasty as an Example .............. 20

    Jay L. Crary, M.D., Kevin L. Smith, M.D., and Frederick A. Matsen, III, M.D.

    The Simple Knee Test: A Quick and Sensitive Self-Assessment of Knee Function .............................................. 23

    Michael H. Metcalf, M.D., Roger V. Larson, M.D., Peter T. Simonian, M.D., and Frederick A. Matsen, III, M.D.

    The Response of Cartilage to an Intra-articular Step-Off: A Sheep Weight Bearing Model ............................... 26

    Thomas E. Trumble, M.D.

    Contact Pressures at Osteochondral Donor Sites in the Knee .............................................................................. 28

    Peter T. Simonian, M.D., Thomas L. Wickiewicz, M.D., and Russell F. Warren, M.D.

    The Enzymatic Preparation of Allograft Bone ....................................................................................................... 30

    David J. Belfie, M.D. and John M. Clark, M.D., Ph.D.

    Angulated Screw Placement in the Lateral Condylar Buttress Plate for Supracondylar Femur Fractures ......... 33

    Peter T. Simonian, M.D., Greg J. Thomson, Will Emley, Richard M. Harrington, Stephen K. Benirschke, M.D.,and Marc F. Swiontkowski, M.D.

    Comminuted and Unstable Iliac Fractures ............................................................................................................ 37

    Julie Switzer, M.D., Sean E. Nork, M.D., and M.L. Chip Routt, Jr., M.D.

    Early vs Delayed Acetabular Fracture Fixation ...................................................................................................... 39

    Oriente DiTano, M.D., William J. Mills, M.D., and M.L. Chip Routt, Jr., M.D.

  • A Biomechanical Comparison of Some Bone-Ligament-Bone Autograft Replacements for the InjuredScapho-Lunate Ligament ........................................................................................................................................ 41

    E.J. Harvey, J.B. Knight, Doug P. Hanel, M.D., and Allan F. Tencer, Ph.D.

    Early “Simple” Release of Post-Traumatic Elbow Contracture Associated With Heterotopic Ossification ........ 43

    Randall W. Viola, M.D. and Doug P. Hanel, M.D.

    A Prospective Multicenter Functional Outcome Study of Arthroplasty in Glenohumeral InflammatoryArthritis .................................................................................................................................................................... 44

    David N. Collins, M.D., Doug T. Harryman, II, M.D. and Michael T. Wirth, M.D,

    Department of Orthopaedics Faculty .................................................................................................................... 46

    Incoming Faculty ..................................................................................................................................................... 47

    Graduating Residents .............................................................................................................................................. 49

    Incoming Residents ................................................................................................................................................. 50

    Contributors to Departmental Research and Education ....................................................................................... 51

    National Research Grants ........................................................................................................................................ 53



    The cover of this year’s researchreport, “The Champion SingleSculls” is a 1871 oil on canvaswhich hangs in the MetropolitanMuseum of Art. Its painter, ThomasEakins (1844-1916), is one of America’smost distinguished artists. Althoughthis painting celebrates a race held onPhiladelphia’s Schuylkill River, thesetting is very reminiscent of theMontlake Cut, the scene of many UWcrew victories. We selected this cover inhonor of the many Husky studentathletes whom we have had the chanceto serve this year and also of our newDean, Paul Ramsey, who himself isfrequently found rowing his single scullon the Cut.

    This has been an important year forthe Department of Orthopaedics. Twoof our senior faculty have gone on toassume prestigious positions at otherschools of medicine. MarcSwiontkowski is now the Chair ofOrthopaedics at the University ofMinnesota. Linda Sandell now holds anendowed professorship at WashingtonUniversity in St. Louis. We wish themboth our very best as they assume theirnew leadership positions.

    We have successfully recruited fiveexciting new regular faculty memberswho will assume their new roles thissummer. Bill Mills, a former UWresident, and Sean Nork, currently atraumatology fellow, will join thefaculty based at Harborview. JohnO’Kane, our first primary care sportsmedicine fellow will join our SportsMedicine team. Mohammad Diab willjoin the Pediatric Orthopaedic groupafter his fellowship in Boston. Finally,Randy Ching will become a member ofthe core faculty at the BiomechanicsLab.

    The excellence of the faculty hasagain been recognized by theDepartment’s inclusion in the top tenOrthopaedic services in the UnitedStates by US News and World Report andby our ranking in the top fiveOrthopaedic Departments in terms ofgrant support from the NationalInstitutes of Health. The AmericanAcademy of Orthopaedic Surgeons andits affiliated specialty organizationsselected over 50 presentations byDepartmental faculty and residents at

    the annual meeting in New Orleans.Fully as important is the ranking we aregiven by medical students applying forOrthopaedic training: this year weagain matched with five sensationalnew residents: Tim DuMontier, ScottHacker, Tim Rapp, Bill Sims, and CarlaSmith, who are profiled later in thisreport.

    This year also saw a most specialgesture of appreciation by one of ourpatients who, in recognition of theoutstanding care rendered him by SteveBenirschke, has established the JeromeH. Debs Endowed Professorship inOrthopaedic Traumatology. The two X-rays on this page show the type ofreconstruction that Dr. Benirschke haspioneered. The upper film shows asevere fracture dislocation involving thesubtalar joint. The second film showsan anatomic reconstruction. TheRegents of the University ofWashington have just approved Steve’sappointment to this Professorship. Thisgenerous philanthropy will assure thatSteve’s academic efforts are wellsupported and that traumatologyresearch will be forever secure in ourDepartment.

    This year’s Research Report displaysonly a sample of the breadth and depthof Departmental research. From themost basic biochemical,immunohistological, biomechanicaland biological research to newdirections of clinical research, includingan attempt to assess the value ofOrthopaedic interventions, our facultyand residents strive to betterunderstand and better treat theconditions which compromise ourpatients’ comfort and function. Wehope you enjoy this brief tour and wewelcome your thoughts, suggestionsand support.

    If you’d like to learn more about theUniversity of Washington Departmentof Orthopaedics, please visit ourinternet address at http://www.orthop.washington.edu/ or dropus an Email message at [email protected] or [email protected]

    We would like to conclude with ahearty thanks to all the friends, staff,faculty, residents and students who

    make the Department of Orthopaedicswhat it has become today. Countlessacts of service and generosity by youenable the team to achieve at the highestlevel as we pursue our missions ofresearch, education and patient care.

    Best wishes,

    Frederick A. Matsen III, M.D. Chairman

    Peter Simonian, M.D.Associate Professor,

    Sports Medicine



    Difficult beginnings. In 1896,when the first modernOlympic games were held inAthens, the only female participant wasan unofficial marathoner namedMelpomene. Despite attempts byofficials to physically remove her fromthe course, Melpomene completed the40 km race in 4.5 hours, faster thanmany men. By 1928, more Olympianswere women, but when severalcollapsed from heat exhaustion at theend of an 800-meter race, theInternational Olympic committeeplaced a ban on women competing inraces longer than 200 meters—a banthat lasted for 32 years! In the UnitedStates the progress women have madein the athletic arena has been slow andanything but steady, both inparticipation and in funding. It was notuntil 1973, in fact, when golfer TerryWilliams signed on at the University ofMiami that a woman athlete received afull-tuition athletic scholarship.

    Women are becoming better athletes,but getting hurt more often. Each yearmore women of all ages participate inorganized sports and vigorous physicalrecreation. Each year the level ofcompetition among women athletesgets higher: faster running, cycling andskiing times, more three pointers andslam dunks in basketball, and moredifficult dives, vaults, skating and rockclimbing maneuvers. Unfortunately,one fact remains the same: womenathletes have more injuries than theirmale counterparts. For example,women high school cross countryrunners have the highest injury rate ofany group of sports participants, maleor female. Knee problems, shin splintsand stress fractures have not only endedmany promising careers, but have alsotaken away the joy of participation frommany more middle of the pack runners.It is not known if this greaterpredilection to injury is a result ofdifferences in anatomy, running style,diet, strength, or shoe design. It is clearthat the answers will only be found inthe study of the woman runner, ratherthan continuing to apply knowledgegained from research on men to the

    challenges faced by women.The ACL Enigma.Perhaps no gender

    difference is more important and moreperplexing than the repeatedobservation that women basketballplayers have anterior cruciate ligament(ACL) tears at many times the rate oftheir male counterparts (Figures 1, 2,3). In a recent study of almost 12,000varsity basketball players, high schoolwomen were found to be at three timesthe risk of men, while in at the collegiatelevel, the risk for women was six timesthat of men for this season endinginjury! ACL injuries account for 25%of all female basketball-related injuries.In football ACL injuries occur fromheavy physical contact. However inwomen’s basketball, 96% of theseinjuries happen without contact withanother player, usually during a pivot,a sudden stop or in landing a jump.

    The reasons for the disparity ininjury rates between the genders areunknown; however a number oftheories have been suggested.

    Slower neuromuscular response:Some propose that women tend tobecome interested in sports at a laterage than men, and as a result have had

    fewer years to develop their agility.Diminished strength: Others

    propose that women have a lowerstrength-to-mass ratio than men, givingthem less ability to control pivots,sudden stops or awkward landings fromjumps. Decreased quadriceps andhamstring strength can increase the riskof knee injury; average quadricepsstrength in females is known to be 60%that of males.

    Different anatomy: Still otherspropose that the disparity results fromanatomical differences. For example,women have more of a valgus (knock-kneed) alignment of their lowerextremities than men (who are morebow-legged as a rule). Landing on sucha limb is more likely to cause it to buckleto the inside, tearing the medialcollateral ligament and the anteriorcruciate ligament. The valgus limb isalso associated with an increased “Q-angle” which may make the quadricepsless effective in controlling the landing.Radiographic studies have shown thatwomen tend to have narrowerintercondylar notches than men, a factwhich could cause the rim of the notchto stress the ligament in

    Sports Safety for Women:An Opportunity Whose Time Has ComeFREDERICK A. MATSEN, III, M.D., PETER T. SIMONIAN, M.D., AND CAROL C. TEITZ, M.D.

    Figure 1: Intact ACL Being Probed


    hyperextension. The narrower notch inwomen may also reflect a smaller-sizedanterior cruciate ligament less able toresist injuring forces.

    Joint laxity: Loose or hypermobileknees are at higher risk for ACL tears.The increased joint laxity of womenmay enable the knee to be stretchedaway from a stable configuration to onewhere the ligaments are at greater risk.Estrogen influences connective tissue

    metabolism and can cause an elevationof endogenous relaxin which increasessoft tissue laxity. Recent evidencesuggests that during the luteal phase ofthe menstrual cycle the laxity may beat its peak; furthermore birth controlpills prolong the luteal phase increasingthe time of increased laxity. Currentresearch is attempting to correlate thetiming of ACL injuries with phases ofthe menstrual cycle and with the use ofbirth control pills.

    Time for action. While these theoriesare of interest, they have failed tosuggest strategies for making basketball,skiing, soccer and running safer forwomen. The fact that women athletesdiffer from men has been established;it has finally been recognized thatlessons learned from studying mencannot be generalized; it is time to focuson approaches for injury preventionwhich are specific to women.Unfortunately, due to the greatereconomic interest in men’s sports, it isrelatively more difficult to fund aresearch program to benefit the womanathlete. Recognizing these challenges,the UW Sports Medicine and FitnessTeam has dedicated itself to the pursuitof the resources and the knowledgewhich will lead to safer recreation andcompetition for women of all ages.


    Gray, J., Tontine, J., McKenzie, D., andet al.:: A survey of injuries to theanterior cruciate ligament of the kneein female basketball players. Int J SportsMed, 6: 314-316, 1985.

    Houseworth, S., Mauro, V., Mellon, B.,and et al..: The intercondyar notch inacute tears of the anterior cruciateligament: A computer graphics study.Am J Sports Med, 15: 221, 1987.

    Malone, T., Hardaker, W., WE, G., andet al..: Relationship of gender toanterior cruciate ligament injuries inintercollegiate basketball players. J SoOrthop Assoc, In Press: 1993.

    Nordgren, B.: Anthropometricmeasures and muscle strength in youngwomen. J Rehabil Med, 4: 165, 1972.

    Powers, J.: Characteristic features ofinjuries in the knee in women. ClinOrthop, 143: 120-124, 1979.

    Tipton, C., Matthes, R., and Martin, R.:Influence of age and sex on the strengthof bone-ligament junctions in kneejoints of rats. J Bone Joint Surg, 60A:230-234, 1978.

    Arendt E, Dick R: Knee injury patternsamong men and women in collegiatebasketball and soccer: NCAA data andreview of literature. Am J Sports Med;23:694-701, 1995.

    Figure 3: ACL Reconstructed with Tendon Graft

    Figure 2: Probe Behind Torn ACL


    A significant number ofneurological injuries arethought to occur or to beaggravated during emergencyextrication, transport and evaluation ofpatients with spinal injury. In 1980,Cloward reported that, of the patientswith fatal injuries treated at theEdinburgh Royal Infirmary, 25% of thefatal complications were related toevents occurring between the time ofthe accident and that of arrival of thepatient to the emergency room. Asubsequent study by Schriger andassociates examined the ability ofemergency medical personnel toimmobilize the cervical spine of studysubjects in neutral position — thecommonly advocated positionfollowing spinal injury. They reportedthe lack of an accepted standard fordefining neutral position, andconcluded that immobilization on a flatwooden backboard would have placed98% of their subjects into relativecervical extension. Biomechanicalstudies have shown that the spinedisplays a “neutral zone” rather than asingle neutral position, as is true for

    many other joints of the body. Hence,because neutral position can neither beprecisely defined nor maintained,positioning of a patient lacksconsistency and is thereforepredisposed to variability. Perhaps, amore effective approach would be toidentify —and avoid— those spinalpositions that could potentially inducefurther harm to the spinal cord byincreasing canal encroachment orocclusion. Additionally, those positionsthat reduce post-injury canal occlusioncould be advised and pursued.

    In the interest of preventing furtherinjury during the management ofpatients with cervical spine injury, thisbiomechanical study was conducted atthe Orthopaedic BiomechanicsLaboratory to examine the effects ofneck position on canal occlusion(stenosis). Although previous studieshave documented the effects of flexion-extension orientation on canalocclusion in intact (non-injured)specimens, none have explored thisrelationship using an injury model, orin other cervical orientations (e.g.,lateral bending). A burst fracture

    model was chosen because of thesignificant degree of spinal canalinvolvement typically associated withthis fracture pattern. In addition,Fontijne and associates, reported thatincreased spinal canal stenosis wascorrelated with an increased probabilityof neurological deficit in patients withthoracolumbar burst fractures.Therefore, this study enabled theinvestigation of whether neck positionhad a significant effect on canalocclusion —with potential neurologicconsequences.


    Specimen PreparationEight fresh human-cadaver, cervical

    spine specimens (levels C1 to T3) wereobtained for this study through theWilled-Body Program of the Universityof Washington Department ofBiological Structure. The specimenswere prepared by removing the softtissues while leaving theosteoligamentous structures intact.Spinal levels C1-C2 and T1-T3 of eachspecimen were potted in dental plasterproviding rigid attachments to facilitatetesting. Burst fractures were createdusing a drop-weight protocol thatinvolved the release of a 4.5 Kg guidedweight from a drop-height of 1.5meters. Post-injury radiographs andcomputed tomography (CT) scansdocumented the location and severityof the burst fractures (see Figure 1 foran example).

    InstrumentationSpinal canal occlusion was

    measured using a custom spinalocclusion transducer system developedin our lab. This transducer consistedof a fluid-flow system (Figure 2) inwhich water was circulated through asection of highly flexible, peristaltictubing that was passed through thespinal canal —replacing the cord. Anyimpingement on the tubing produceda rise in fluid pressure that wasmonitored by an upstream pressuretransducer. To quantify canal

    The Effect of Neck Position on Spinal Canal Occlusion in aCervical Spine Burst Fracture ModelRANDAL P. CHING, PH.D., NATHAN A.WATSON, B.S., JARROD W. CARTER, B.S.,AND ALLAN F. TENCER, PH.D.

    Figure 1: CT image of a burst-fracture specimen.


    occlusion, the relationship betweentubing diameter and pressure wasestablished through calibration.

    Angular motions across the level ofinjury were recorded using anelectromagnetic three-dimensionaltracking system (Polhemus 3-SpaceFastrak System, Kaiser Aerospace andElectronics Company, Colchester, VT).Motion sensors were attached to pinsinserted into the vertebrae immediatelyabove and below the injury site. Anacrylic loading frame was used toposition the specimens withoutinterfering with the electromagneticmotion sensors.

    Experimental ProcedureEach specimen was tested in eight

    directions of bending including:flexion, extension, right- and left-lateralbending, and four intermediate (45°)positions. To describe the multi-axisbending applied to each specimen, aclock-face orientation was adopted inwhich flexion corresponded to the12:00 o’clock position, while rightlateral bending corresponded with 3:00o’clock. In this study, neutral positionwas defined as the most central positionof the spine aligned within the neutralzone while preserving a normal lordoticcurvature. The first occlusionmeasurement recorded for eachspecimen was with the spine placed inneutral position. This provided abaseline measurement of the residualcanal occlusion resulting from injuryand served as the basis for statisticalcomparison against all other spinalpositions.

    Two occlusion measurements weretaken in each direction of bending, oneat the onset of ligament tensioning (i.e.,the neutral-zone [NZ] boundary), andthe second at full range of motion(ROM). The neutral-zone-boundarypositions were established by loading(bending) the specimen twice to its fullrange of motion before allowing it toreturn to an equilibrium position. Thefull range-of-motion positions wereproduced by fully tensioning the spinalligaments while being careful not toinduce further damage to the specimen.Positioning of the specimens wasperformed manually with a randompositioning order used for eachspecimen. In addition to bending,measurements were also taken whileapplying axial torsion (both clockwiseand counter-clockwise), tension

    (traction) and compression to thespine.

    Data AnalysisThe occlusion data was expressed as

    a percentage reduction from intactcanal midsagittal diameter. Paired t-tests were used to compare the canalocclusion measurements taken at eachspinal position against measurementstaken at the neutral position.Significance was established based on

    a computed p-value of less than 0.05.

    RESULTSA contour plot (Figure 3) was

    prepared showing the relationshipbetween canal occlusion and neckposition in different directions ofbending. This plot uses a clock-faceorientation to describe the bendingdirections and displays an increasingpercentage of canal occlusion as agrayscale gradient. The average canal





    PottingPolhemus Sensors Attached to Spine

    Acrylic Frame


















    7:30 4:30

    ROMNZ Boundary Percent Canal


    Neutral Position

    [13.7° + 6.1][7.7° + 2.9]

    Figure 2: Schematic diagram of the experimental apparatus showing the spinal canal occlusion transducerand sensors for measuring spinal angulation.

    Figure 3: This contour map helps to visualize the relationship between canal occlusion and neckposition at the neutral-zone boundary (NZ) and at full range of motion (ROM). Darker shadingindicates a greater percentage of occlusion.


    occlusion in neutral position was 23%± 20 which corresponded to thebaseline (residual), burst-fracture canalencroachment. Significant increases inocclusion, from neutral position, werefound particularly in extension and inlateral bending combined withextension (i.e., 4:30 and 7:30orientations). For example, at full rangeof motion, the average measuredocclusion in extension increased to 42%± 20 (p = 0.01). Similar increases wereobserved at the 4:30 position (38% ±21, p = 0.04), and at the 7:30 position(39% ± 20, p = 0.02). There was a trendtowards reduced (decreased) canalocclusion for flexion (14% ± 22),however, not at a statistically significantlevel. Figure 4 documents the effect ofaxial torsion, tension (traction) andcompression on canal occlusion.Traction demonstrated no appreciableeffect on occlusion, however occlusionincreased significantly in compression(35% ± 18, p = 0.02). In summary, wefound that compression, extension, andlateral bending combined withextension significantly increased canalocclusion for all eight specimens tested.

    DISCUSSIONThe objectives for this study were

    achieved by documenting the effect ofneck position on canal geometry in aburst fracture model, and subsequentlytesting the hypothesis that differentpositions may affect (by increasing or

    decreasing) canal occlusion. Ourfindings indicate that for burstfractures, positioning the spine intoextension increased canal occlusionconsiderably. At full range of motionin extension, the residual (baseline)canal occlusion increased from 23%(the average occlusion in neutralposition) to 42%, an increase of 19%.Similarly, axial compression resulted ina statistically significant, 12% increasein measured canal occlusion (from 23%to 35%). Although none of the testedneck positions yielded statisticallysignificant decreases in canal occlusion,a trend was observed for specimensplaced into flexion (i.e., the averagecanal occlusion decreased from 23% to14% in flexion). These responses wereobserved in all eight of the spinespecimens tested. Other spinalpositions did not appear to eitherpositively or negatively affect canalocclusion. Therefore, the results of thisstudy suggest that for a patient with acervical spine burst fracture,positioning the neck into eitherextension or compression maysignificantly increase canal occlusionand potentially exacerbate a spinal cordinjury.


    Cloward RB. Acute cervical spineinjuries. Clinical Symposia 1980;32(1):1-32.

    Schriger DL, Larmon B, LeGassick T,Blinman T. Spinal immobilization ona flat backboard: Does it result inneutral position of the cervical spine?Ann Emerg Med 1991; 20(8):878-881.

    Fontijne WP, de-Klerk LW, BraakmanR, Stijnen T, Tanghe HL, Steenbeek R,van-Linge B. CT scan prediction ofneurological deficit in thoracolumbarburst fractures. J Bone Joint Surg 1992;74B(5):683-685.

    Ching RP, Watson NA, Carter JW,Tencer AF. The effect of post-injuryspinal position on canal occlusion in acervical spine burst fracture model.Spine 1997, 22(15): 1710-1715.

    CW CCW Traction Compression0








    al O




    p < 0.05


    Figure 4: Mean and standard deviations for canal occlusion associated with rotation (both clockwise(CW) and counterclockwise (CCW)), tension (traction), and compression.


    P ain and disability caused bydegenerative damage to articularcartilage affect millions ofpeople in the United States. Sincearticular cartilage has little, if anynatural capacity to heal with hyalinecartilage, most current treatments relyon symptomatic care until the patientis considered old enough to receive aprosthetic joint replacement. Bothsymptomatic care and jointreplacement are fraught with sideeffects and are ultimately temporarysolutions. Both can be expensive andneither allow a patient to return tophysically demanding jobs. Centersthroughout the world are developingtechniques to repair or regeneratecartilage in defects. Much publicity andexcitement over the most recent ofthese techniques applied to humans hasbeen tempered as clinical results havenot lived up to expectations. Basicscientific knowledge of the response ofcartilage to injury is incomplete and thedevelopment of successful clinicaltechniques is unlikely without thisknowledge.

    Articular cartilage contains a uniqueassortment of the major and minorcollagens in addition to proteoglycansand other less abundant matrixmolecules. Little is known about theability of chondrocytes in vivo toregenerate a functional extracellularmatrix. A major question is whetherchondrocytes “activated” in the earlystages of osteoarthritis (OA) cansynthesize an extracellular matrix(ECM) that is functionally similar tothe cartilage articular surface producedduring development. Very few studieshave been directed toward themolecular aspects of normaldevelopmental and repair processes ofjoints. Consequently, it is impossibleto determine why the attempts at repairseen in OA are inadequate to regeneratenormal hyaline cartilage.

    It has been known for many yearsthat the metabolic activity of articularchondrocytes increases inosteoarthritis, although the significanceof these changes is poorly understood.The degradation and loss of the

    extracellular matrix components ofosteoarthritic cartilage areaccompanied by increased synthesis oftypes II and III collagen. The newlysynthesized type II collagen mayrepresent an incomplete repairresponse by chondrocytes embedded indegenerating matrix. In addition it hasrecently been shown that certaincollagens are markers of specific stagesof cartilage development. For example,type IIA procollagen is a marker ofdeveloping cartilage whereas type IIBcollagen is synthesized only by maturechondrocytes. As developmentproceeds the predominant collagenbecomes type IIB collagen and the typeIIA procollagen is removed.

    To investigate the possibility thatchondrocytes in osteoarthritic cartilagemay revert to a “younger” phenotypeto facilitate repair of diseased matrix,we used immunohistochemistry toanalyze the synthesis and distributionof type IIA procollagenNH2-propeptide, type II collagen triplehelix, type II collagenCOOH-propeptide and types I and IIIcollagens in osteoarthritic cartilage.

    METHODSImmunohistochemistry is a

    technique that takes advantage of thehumoral immune system’s ability togenerate antibodies against specificmolecules, or portions of molecules.Antibodies can be generated byinjecting animals with specific proteinsof interest. In this study, the proteinsof interest were various collagens andcollagen subunits. To visualize theprimary antibody-antigen (protein)combination, a standard secondaryantibody is attached to the primaryantibody and then a chromogen, orfluorescing molecule, is chemicallyattached to the secondary antibody.The exquisite specificity of thesereactions allow the in vitro detection ofproteins of interest in paraffin-embedded histologic specimens. Whena different chromogen is used for eachantibody, confocal microscopy canreveal the relative distributions of eachprotein in a single histological section.

    We studied 29 samples from 11 patientswith advanced osteoarthiritsundergoing total shoulder (4), knee (3)or hip (4) replacement. Toimmunolocalize the collagens ofinterest, five antibodies were raisedagainst collagen type I, type IIAprocollagen NH-2 propeptide, type IIcollagen triple helix, type II collagenCOOH-propeptide, and type IIIcollagen. Confocal microscopy wasused to assess the intensity anddistribution of staining of the 5molecules of interest.

    RESULTSThere was no significant staining for

    type I collagen in osteoarthriticcartilage. All 11 cases stained stronglyfor type III collagen and stained tovarying degrees for type IIAprocollagen NH-2-propeptide. TypeIIA procollagen was mainly localizedwithin the pericellular matrix ofchondrocytes undergoing cloning inthe middle and deep zones of theosteoarthritic cartilage. Type IIAprocollagen was also found in 2 of 2osteophytes which were studied.Interestingly, the staining for type IIAprocollagen was more intense in thecartilage removed from the shoulderand in non-weightbearing regions ofthe hip and knee than in that removedfrom weightbearing areas of the kneeor the hip. Type III collagen was foundthroughout the fibrillated matrix in all29 specimens. Type III collagen wasfound diffusely within the superficialand middle zones of the cartilage butwithin the deep zone it was localized tothe territorial matrix. The mature typeII collagen triple helix was foundthroughout the cartilage matrix withpeak levels found in the vicinity ofchondrocytes. The type II carboxy-propeptide was only found within thecytoplasm of chondrocytes. Confocalimaging revealed that type IIA NH-2propeptide and type II collagen triplehelix were co-localized in thepericellular matrix within the lacunae.Appropriate positive and negativecontrol tissues and antibodies wereutilized.

    Differential Localization of Collagen Types I, IIA and III inHuman Osteoarthritic CartilageYONG ZHU, M.D., HOWARD A. CHANSKY, M.D., FREDERICK A. MATSEN, III, M.D.,AND LINDA J. SANDELL, PH.D.


    DISCUSSIONThese results demonstrate that

    chondrocytes from osteoarthriticcartilage synthesize significant amountsof type IIA procollagen, a type ofcollagen which is normally found onlyin developing cartilage. This isconsistent with the hypothesis thatchondrocytes from diseased cartilagemay “dedifferentiate” in an attempt torecapitulate the developmentalpathways that result in the formationof articular cartilage. The absence oftype I collagen indicates that thesephenotypic changes are not thosewhich are commonly seen during the“dedifferentiation” of chondrocytesgrown in vitro. It is plausible that thephenotypic changes documented inthis study reflect an aborted attempt bychondrocytes within damaged matrixto generate a healing response. Thetype IIA procollagen generated by thesechondrocytes remains confined to thepericellular matrix as opposed to thesituation that exists duringdevelopment when the type IIAprocollagen is deposited throughoutthe matrix. The local pericellularenvironment of damaged matrix maycontain inhibitory factors, or may bemissing stimulatory factors, which areneeded to sustain the repair response.The local mechanical environment mayexplain why chondrocytes from theshoulder are able to synthesize greaterquantities of type IIA collagen thanthose in weight-bearing joints such asthe hip and knee.

    Unlike many, if not most tissues,articular cartilage lacks the intrinsiccapacity to heal with like-tissue. Basedupon many studies it appears that thenatural response of a chondrocyte todamage of surrounding matrix is tosynthesize new matrix with a structureor ultrastructure different from that ofhealthy adult cartilage. The findings ofthis study suggest that the repair tissueassumes a phenotype resembling thatof developing cartilage. This abnormalmatrix is not able to withstand thenormal stresses placed upon thearticular surface and is unable to beincorporated into the surroundingmatrix, and thus eventually fails. Beforescientists and surgeons can developeffective techinques to restore damagedhyaline cartilage we must understandthe normal response of joints todamage. This study is one step towardunderstanding these processes. In the

    future it may become possible tomanipulate these native responses toour advantage with various cytokines,growth factors or genes.


    Aigner T, Gluckert K, von-der-MarkK: Activation of fibrillar collagensynthesis and phenotypic modulationof chondrocytes in early humanosteoarthritic cartilage lesions.Osteoarthritis-Cartilage. 5(3):183-9,1993

    Sandell, L.J., Morris, N., Robbins, J.R.and Goldring, M.R.: Alternativelyspliced type II procollagen mRNAsdefine distinct populations of cellsduring vertebral development:differential expression of the amino-propeptide. J. Cell Biol. 114: 1307-1319,1991.

    Aydelotte, M.B. and Kuettner, K.E.:Heterogeneity of ArticularChondrocytes and Cartilage Matrix. In:Joint Cartilage Degradation: Basic andClinical Aspects, J. Frederick Woessner,Jr., and David S. Howell, eds., MarcelDekker, Inc., New York., 37-66, 1993.

    Gerich TG, Fu FH, Robbins PD, EvansCH: Prospects for gene therapy insports medicine. Knee Surg SportsTraumatol Arthrosc 4(3):80-7, 1996.


    Collagenase-3 (MMP-13) is oneof the more recently identifiedhuman matrixmetalloproteinases. First found inbreast carcinomas (1), collagenase-3 isalso produced by chondrocytes inhuman osteoarthritic cartilage.Collagenase-3 produced bychondrocytes is believed to be a keyenzyme in the degradation of cartilagein osteoarthritis.

    The collagenous framework ofcartilage consists of cross-linkedcopolymers of types II, IX and XIcollagens. In this study we haveinvestigated the ability of recombinantcollagenase-3 to degrade anddepolymerize these three collagentypes.

    MATERIALS AND METHODSHuman recombinant

    procollagenase-3 was expressed intransfected NSO mouse myeloma cellsand purified from conditioned cellculture medium (2). The proenzymewas activated by p-aminophenylmercuric acetate.

    Slices of fetal bovine epiphysealcartilage were extracted at 4°C in 1MNaCl, 0.05M Tris-HCl, pH 7.5 for 24hr then in 4M GuHCl for 24 h toremove non-cross-linked proteins andproteoglycans. The washed residueswere minced and digested with pepsinand salt fractionated into types II, IXand XI collagens. From the 1M NaClextract, newly made intact types II, IXand XI collagen molecules werepurified (4).

    Each collagen substrate wasdissolved at 2mg/ml in 0.5M acetic acidand dialyzed against 0.2 M NaCl, 50mMTris-HCl, 10mM CaCl

    2, 0.05% Brij 35,

    pH 7.5. Aliquots of each were incubatedwith activated collagenase-3 (substrate/enzyme=200:1, w/w) at 30°C for 24 h.The digests were analyzed by SDS-7.5%PAGE. Protein bands were transblottedto PVDF membrane and identified byamino-terminal microsequencing.

    RESULTS AND DISCUSSIONSimilar to collagenase-1,

    collagenase-3 degrades type II collageninto 3/4 and 1/4 fragments, and has noactivity against triple-helical domainsof type XI collagen. However,collagenase-3 was more active thancollagenase-1 in degrading native typeII collagen. Amino-terminal sequenceanalysis identified cleavage sitesbetween bonds at Gly 778-Gln 779 andGly 781-Ile 782 which are three and sixresidues C-terminal to the collagenase-1 cleavage site, respectively. Cleavage at

    Proteolysis of Cartilage Collagens II, IX and XI by Collagenase-3(MMP-13)JIANN-JIU WU, PH.D., VERA KNÄUPER, PH.D.*, GILLIAN MURPHY, PH.D.*, AND DAVID R. EYRE, PH.D.

    3/4 1/4N C




    Collagenase -3








    Figure 1: Molecular model summarizing sites of collagenase-3 cleavage in type II collagen


    the Gly 775-Leu 776 bond (collagenase-1 site) could only be identified insamples incubated with enzyme undermilder conditions (i.e. 22°C not 30°Cfor 6 hr). Sequence analysis of the 3/4fragment produced from salt-soluble,intact type II collagen showed a limitedcleavage of the N-telopeptide domain,but only exterior (i.e. N-terminal) tothe cross-linking lysine residue.Similarly, digestion of syntheticpeptides matching the type II C-telopeptide showed no cleavage interiorto the cross-linking lysine (Figure 1).

    When incubated with intact typesIX and XI collagens, collagenase-3trimmed the N-propeptide extensionsfrom all three chains of intact type XIcollagen. It also shortened the a1(IX)chain, consistent with the removal ofthe NC4 domain. However, unlikestromelysin (3) collagenase did not cutthe type IX collagen molecule in itsNC2 domain. This may be importantin understanding the role ofstromelysin in cartilage matrixdegradation under conditions whencollagenases are not expressed.

    SUMMARYCollagenase-3 rapidly cleaves type

    II collagen at the Gly-Leu bond ascleaved by collagenase-1. The resulting1/4 fragments are further cleaved bycollagenase-3 removing three or sixmore amino acids from the aminoterminus. Although some cleavage ofthe type II collagen telopeptide domaincan occur, this action exterior to thecross-links has no depolymerizingpotential. Therefore, after the initialattack of collagenase 3 on the triple-helix, other metalloproteases (e.g.gelatinase) are probably important incompleting the process ofdepolymerization. The action ofstromelysin in cleaving in the interiorof type IX collagen may bephysiologically important in mediatingcollagen network swelling in theabsence of collagenase(s).


    Freije, J.P.M. et al. Molecular cloningand expression of collagenase-3, anovelhuman matrix metalloproteinaseproduced by breast carcinomas. J. Biol.Chem. 269:16766-16773, 1994.

    Knauper, V. et al. Biochemicalcharacterization of human collagenase-3.J Biol Chem 271:1544-1550, 1996.

    Wu, J.J. et al. Sites of stromelysincleavage in collagen types II, IX, X andXI of cartilage. J. Biol. Chem. 266:5625-5628, 1991.

    This work was supported by NIHgrants AR 36794 and AR 37318.

    *Dept. of Cell and MolecularBiology, Strangeways ResearchLaboratory, Cambridge, UK


    Proteolytic degradation of thecollagenous matrix of cartilagecan lead to loss of tissue andimpaired joint function. The full rangeof proteinases involved in thedegradative process and their preciseroles are poorly understood, thoughmatrix metalloproteinases (MMPs) arestrongly implicated (1). MMPs areover-expressed in the joint tissues andsynovial fluid of patients with arthritis(2, 3). In a recent study, MMP activitywas correlated with type II collagendegradation in articular cartilagecultures that had been stimulated withinterleukin-1a (4).

    The aim of this study was toevaluate whether an immunoassaybased on a monoclonal antibody(mAb) 2B4, which recognizes a domainof the a1(II) C-telopeptide EKGPDP(5), measures MMP cleavage productsof cartilage. This 6-amino-acid corepeptide was identified in cross-linkedform in urine where it appears to be aprotease-resistant end product of typeII collagen fibril degradation.Immunoassays that quantitate markersof type II collagen degradation couldbe useful for understanding normalcartilage turnover and the pathogenesis

    of arthritic disease.

    MATERIALS & METHODSI m m u n o a f f i n i t y - p u r i f i e d

    stromelysin (MMP-3) was from IL-1-stimulated human dermal fibroblasts.Collagenase 2 (MMP-8), matrilysin(MMP 7), membrane-type (MT) - 1-MMP (MMP 14) were all recombinanthuman preparations (generous giftsfrom Dr. R Martin, Roche Bioscience).Recombinant human collagenase 3(MMP-13) was kindly provided by Dr.G Murphy, Strangeways Research Lab.,Cambridge.

    Enzyme digests: 100 mg CNBr-digested human cartilage type IIcollagen, or 12.5 mg synthetic C-telopeptide domain, were incubatedwith 0.75 mg enzyme in digestionbuffer (0.2M NaCl, 50 mM Tris-HClpH 7.5, 10 mM CaCl

    2, 0.05% Brij).

    Reactions were stopped with 1.10phenanthroline (or soybean trypsininhibitor for trypsin) immediately (0hr), or after incubation for 24 hr at37∞C (24 hr).

    2B4 immunoassay: 2B4immunoreactivity was quantitated byinhibition ELISA. 96-well microtitre

    plates were coated overnight at 4∞Cwith EKGPDP linked to BSA, thenblocked for 2 hr at room temperaturewith buffer containing 0.25% BSA.Digestion buffer or calibrator(glutaldehyde cross-linked EKGPDP)were added, followed by 2B4-horseradish peroxidase conjugate.After incubation for 1.5 hr at 4∞C, theplates were washed and the peroxidasesubstrate TMB was added. After 15 mincolor development was stopped withH


    4 and absorbance monitored at 450

    nm.Reverse-phase (RP)-HPLC:

    Peptides were resolved by RP-HPLCusing a macroporous C8 column and agradient of acetonitrile and 1-propanolcontaining 0.1% TFA. The UVabsorbance of the eluant wasmonitored at 2 wavelengths; 280 nm(primarily for the tyrosine functionalgroup) and 220 nm (primarily forpeptide bonds). Peptides wereidentified by amino-terminalmicrosequencing.

    RESULTSCompared with the buffer control

    and trypsin digests, all of the MMPstested released 2B4 epitope fromdenatured cartilage type II collagen(Table 1). We then determined theMMP-specific cleavage sites in the C-telopeptide using the synthetic peptideAFAGLGPREKGPDPLQYMRA. Initialscreening established that mAb 2B4recognized MMP-digests of thesynthetic peptide, but gave no signalwith the intact molecule. Whenmatrilysin digestion products of thepeptide were resolved by RP-HPLCseveral fragments were identified, themost abundant (by 220 nmabsorbance) eluting at 40 ml (Fig. 1).Sequence analysis identified thispeptide as the product of P-L cleavage.Inhibition ELISA showed that it was thesource of all the generated 2B4immunoreactivity. Digestion withstromelysin or collagenase 3 producedessentially the same range of peptidefragments, including the 2B4-immunoreactive AFAGLGPREKGPDP,

    Matrix Metalloproteinase-Mediated Release Of ImmunoreactiveTelopeptides From Cartilage Type II Collagen


    Protease 2B4 epitope(ng/ml)

    0 h 24 hBuffer control 2 12Stromelysin 4 500Collagenase 3 2 430Collagenase 2 2 113Matrilysin 4 3100MT-1-MMP 3 145Bacterial collagenase 2 170Trypsin 2 10

    Table 1: Release of 2B4 epitope from denatured cartilage type II collagen following digestion witha panel of proteases.


    but the yield of the latter was muchlower than with matrilysin.

    DISCUSSIONAs part of a study to identify useful

    biochemical markers of cartilagecollagen degradation, we have shownthat a proteolytic neoepitopeoriginating in the collagen type II C-telopeptide is generated in vitro byseveral MMPs. Of the MMPs so fartested, matrilysin appeared to be themost potent. In preliminary studies, wehave detected 2B4 immunoreactivity insynovial fluid, serum and urine (5).This 2B4 epitope, therefore, has thepotential to be a useful marker of typeII collagen degradation in vivo. We arecurrently testing this in clinical studiesand in animal models of osteoarthritis.




    0 20 40 60






    Elution Volume (ml)

    Synthetic PeptideSynthetic PeptideSynthetic PeptideSynthetic PeptideSynthetic Peptide



    A 220nm

    A 280nm

    Matrilysin DigestMatrilysin DigestMatrilysin DigestMatrilysin DigestMatrilysin Digest

    2B4 Immunoassay2B4 Immunoassay2B4 Immunoassay2B4 Immunoassay2B4 Immunoassay

    A 220nm

    A 280nm



    Figure 1: RP-HPLC resolution and 2B4 immunoreactivity of the synthetic peptide following proteolyticcleavage with matrilysin.


    Murphy, 1995 Acta Orthop Scand 66:55-60.

    Okada et al., 1992 Lab Invest. 66: 1069-1073.

    Lohmander et al., 1993 ArthritisRheum. 36:181-189.

    Kozaci et al., 1997 Arthritis Rheum 40:164-174.

    Eyre et al., 1996 Proceedings of theASMBR, Seattle, WA.

    Supported by grants from NIH (AR37318 & 36794) and OstexInternational, Inc.

    ** Ostex International, Inc., Seattle,WA.


    I t is widely accepted that curveprogression in idiopathic scoliosisis related to growth.Recommendations for bracing andoperative intervention rely upon theability to assess the growth potential ofthese patients. Chronologic age, Rissersign and menarcheal status arecommonly used to estimate growthpotential and the potential for futurecurve progression. These maturityscales aim to group children in such away that they behave similarly withregards to growth and development.Buckler performed a longitudinal studyof normal adolescents and reported amethod of grouping children by theyears around the age at which peakheight velocity occurred. He found thatthese methods produced a growthvelocity curve most representative ofindividual growth patterns, and thatmales, females, early and late developersbehaved very similarly with respect tothis scale. We have termed this maturityscale as Peak Growth Age (PGA). The

    purpose of this investigation was toapply this maturity scale to girls withidiopathic scoliosis followed throughadolescence. We aimed to see if itpredicted growth of the patient andlikelihood of curve progression moreaccurately than existing maturity scales.

    MATERIALS AND METHODSWe reviewed the charts and

    radiographs of 130 girls from a databaseof 371 patients braced for adolescentidiopathic scoliosis from 1977 to 1995.All patients had a curve magnitude ofat least 25 degrees, idiopathic scoliosis,at least 6 visits to our institution over aminimum of 2 years with heightmeasurements recorded through theadolescent growth spurt; progressionfrom Risser 0 to at least Risser 4, andtreatment with either a Boston TLSOor Charleston bending brace.

    Height measurements for eachpatient were used to calculate heightvelocities in centimeters per year. Anexample of the calculation for an

    individual patient is shown in Table I.A minimum 6 month interval was usedfor calculations. The time of maximalheight velocity was designated PeakGrowth Age zero (PGA 0). For thepatient in Table I, this was in February1990 when the height velocity peakedat 10.1 centimeters per year. Weassigned a positive or negative PGA tothe years preceding and following peakheight velocity (PGA -1, PGA +1, etc.).If the first height velocity available fora patient was greater than nine cm peryear, this was designated as PGA 0.Cessation of growth was a heightvelocity less than 2 cm/year. We plottedheight velocity against PGA, Risser sign,chronologic age, and menarcheal age.We looked at the subset of patients withcurve progression greater than 10degrees and examined which maturityscale most closely correlated with curveprogression and the likelihood of curveprogression to 45 degrees.

    Peak Growth Age: A New Maturity Indicator for IdiopathicScoliosisKIT M. SONG, M.D., DAVID G. LITTLE, M.B.B.S., F.R.A.C.S. (ORTH), DON KATZ, C.O.,AND JOHN A. HERRING, M.D.

    Figure 1: Plot of height velocity versus Risser Sign. Height velocity is greatest at Risser 0 which is not a defined point in time.


    RESULTSWe could assign PGA in all cases.

    The plots for height velocity versuschronologic age, PGA, Risser sign, andmenarcheal age are shown in figures 1-4. The median plot for PGA is the onlyscale which retains the characteristicsof an individual plot, with a high peakand sharp decline. Menarche occurredin the year around PGA 0 in 41% ofgirls in our study. Sixteen patients(12%) had menarche two years or moreafter PGA 0 and 8% had menarche twoyears or more before PGA 0. Thegrowth pattern of these patients clearlyfollowed PGA rather than menarchealage. PGA 0 was an earlier identifiablelandmark than the onset of menses orRisser 1 which occurred at a median of7 months and 9 months after PGA 0.

    Ninety-seven curves progressed atleast 10 degrees and 61 progressed to45 degrees or more. The point ofmaximal curve progression wassignificantly more closely groupedaround PGA 0 than a specificchronological age (F test p=0.0001),Risser sign or menarche (p=0.02).Fifty-nine of 86 curves (69%) greaterthan 30 degrees at PGA 0 progressed to45 degrees or more, whereas only 2 of44 curves (5%) 30 degrees or lessprogressed to 45 degrees or more. Thisdifference was significant by Chi square(p


    range in this population of bracedpatients while very few patients withcurves less than 30 degrees at PGA 0progressed to curves of 45 degrees ormore.

    It may not be as easy to assign PGAprospectively for patients as it was toassign it in this retrospective review.Height information is often availablefrom families and pediatricians. If only2 height measurements are available,and the velocity is greater than 9 cm/year, the patient is very likely to be inthe growth peak. For smaller velocities,previously utilized methods such as theRisser sign, menarcheal status, andchronologic age can be used to place thepatient along the growth curve. Thetiming of these events relative to PGA0 is shown in figure 5.

    In summary, we believe that PGA isan easily calculated, accurate methodfor estimating growth remaining forchildren with idiopathic scoliosis. It ismore accurate than the use of indirectmethods such as chronologic age, Rissersign, and menarchal status and can helpphysicians prognosticate cessation ofgrowth and likelihood of curveprogression.


    Anderson, M.; Hwang, S. ; and Green,W. T.: Growth of the normal trunk inboys and girls during the second decade

    of life. J. Bone and Joint Surg.,47A:1554-64, 1965.

    Buckler, J. M. H.: A longitudinal studyof adolescent growth. London,

    Figure 4: Plot of height velocity versus Peak Growth Age. The plot for the population studied is similarto that of an individual with a well defined peak and rapid decline.

    Figure 5: Relationship of PGA 0 to median menarche, menarche +2, Risser 1, Risser 4, and Risser 5. The onset of menses or being Risser 1 implies that the childis past their major growth spurt.


    Springer-Verlag, 1990.

    Bunnell, W. P.: The natural history ofidiopathic scoliosis before skeletalmaturity. Spine, 11:773-6, 1986.

    Carman, D. L.; Browne, R. H. ; andBirch, J. G.: Measurement of scoliosisand kyphosis on radiographs.Intraobserver and interobservervariation. J. Bone and Joint Surg.,72A:328-33, 1990.

    Howell, F. R.; Mahood, J. K. ; andDickson, R. A.: Growth beyond skeletalmaturity. Spine, 17:437-440, 1992.

    Little, D. G.; Sussman, M. D.: The Rissersign: A critical analysis. J. Pediatr.Orthop., 14:569-75, 1994.

    Loncar-Dusek, M.; Pecina, M. ; andPrebeg, Z.: A longitudinal study ofgrowth velocity and development ofsecondary gender characteristics versusonset of idiopathic scoliosis. Clin. Orth.and Rel. Research., 270:278-282, 1989.

    Peterson, L.; Nachemson, A. L.:Prediction of progression of the curvein girls who have adolescent idiopathicscoliosis of moderate severity. J. Boneand Joint Surg., 77A:823-27, 1995.

    Shuren, N.; Kasser, J. R.; Emans, J. B. ;and Rand, F.: Reevaluation of the useof the Risser sign in idiopathic scoliosis.Spine, 17:359-61, 1992.

    Veldhuizen, A. G.; Baas, P. ; and Webb,P. J.: Observations on the Growth of theAdolescent Spine. J. Bone and JointSurg., 68B:724-728, 1986.

    Table I: Sample calculations for height velocity. One should use a minimum of 6 month interval between height for calculation. As an example between 4/12/90 and 8/13/90 there is an interval of only 4 months. The interval from 2/12/90 to 8/13/90 is instead selected for an interval of 6 months. The change in heightwas 3 cm. which is a height velocity of 6 cm. The peak height velocity is highlighted and is 10.1 cm.


    The Orthopaedics BiomechanicsLaboratory has an ongoingseries of research projectsdesigned to better understand footfunction, and the surgical repairs offoot trauma and foot deformities. Thisis a summary of some of our recentactivity.

    The effect of foot position on tarsalload transfer During gait, load istransmitted down the tibial shaft, andinto the talus, where it is transferred toboth the hindfoot and forefoot. Wehave determined the relative loaddistribution between the posterior andanterior facets of the talocalcaneal joint,and studied the role of the talonavicularjoint as part of this mechanism (Figure1). Using a cadaveric model of thenormal foot, we tested the hypothesisthat load transmission between thecalcaneus and navicular, measured bycontact pressure and area, wouldchange with foot position.

    Mechanisms for suport of the talusThe mechanisms of inferior and medialsupport of the talus were determinedto provide insight into the mechanism

    of loss of peritalar stability with flatfoot.In the intact foot the tibialis posterior(TP) tendon forms an articulation withthe medial plantar aspect of the talarhead. When the TP tendon ruptures,there is loss of dynamic support in themidfoot and a gradual change in theposition of the tarsal bones. The talusplantar flexes, the calcaneus everts, thenavicular abducts, and the foot loses itsarch (Figure 2). This change ispresumed to occur as a result of loss ofstatic support as the calcaneonavicularligaments elongate. The goal of thisstudy was to determine (i) thecharacteristics of interaction betweenthe talar head and thecalcaneonavicular ligament, (ii) itsrelative contribution to support of thetalus compared to those of the inferiorarticulations, and (iii) the effect of theposterior tibialis tendon, by measuringcontact pressures using a staticallyloaded cadaveric foot model.

    Reconstruction of the mediallongitudinal arch in flatfoot. As asurgical treatment for flatfoot,lengthening the lateral column of the

    foot restores the anatomic positions ofthe calcaneus and talus, tensions theplantar tissues and restores the mediallongitudinal arch. Two methods havebeen advocated, the Dillwyn-Evanswhich lengthens the calcaneus byinsertion of a bone graft between theanterior and medial facets, violating thetalocalcaneal joint; and the Seattletechnique, which places the bone graftat the calcaneocuboid joint, fusing itbut maintaining the talocalcaneal jointintact. These studies were designed tocompare the joint contact propertiesafter restoration by either method, andto study the effect on kinematics of thefoot. For surgical treatment ofsymptomatic flatfoot, calcaneallengthening restores the position of thesubluxed talus and tensions the plantarsoft tissues of the hindfoot. Fusing thecalcaneo-cuboid joint maintains thelengthened lateral column and thecorrected position of the calcaneus.However, the effect of these procedureson hindfoot kinematics is unknown.Therefore, the kinematics of thehindfoot were measured first in normal

    Biomechanical Study of Foot Function


    Anterior-middle facet

    Talo-navicular Posterior facet


    Figure 1: Approximate contact areas on the talus and estimated directions of resultant force on each articular area.


    cadaveric feet, then after simulatedlateral column lengthening and fusionin different positions.

    Treatment of acquired flatfootAcquired flatfoot deformity has beenattributed to the loss of the tibialisposterior tendon. We studied the effectof releasing the tibialis posteriortendon (simulating rupture) on thekinematic positions of the hind footbones. Another study examined theeffect of restoring posterior tibialtendon function in an acquired flatfootmodel. Treatments include variousfusions, lateral column lengthening,ligament reconstruction and tendontransfer. This study used an acquiredflatfoot model to test the hypothesis:restoring the tibialis posterior tendonfunction (simulating repair) willsignificantly affect the kinematicorientation of the hindfoot complex inheel strike, stance, and heel off.

    METHODSWe have developed several

    techniques which we use in various

    combinations. Joint contactmeasurements are made usingtransducers made of pressure sensitivefilm (Prescale Film, Superlow and Low,Fuji, Tokyo, Japan) with an assembledthickness of 0.3 mm. After insertion ofthe film between the joint surfaces,increasing load, maintaining it for 30secs, and unloading, the pressure filmis imaged along with a set of calibrationprints using a flat bed scanner. Theimaged films are analyzed for totalcontact area (>0.5 MPa) and meanpressure. Overall errors are 3.8% forcontact area and 9.3% for pressure.

    The loading apparatus consists of ashaft applying axial load to the tibia andfibula through a plate which allowsflexion/extension of the tibia, andclamps to attach 8 tendons of the footto cables connected to individualpneumatic cylinders (Figure 3). Eachfoot is tested in three positions: heelstrike (5% of the gait cycle), stance(30%), and near toe off (45%). Thecontributions of various muscles inthese three phases of gait were

    determined from information on theirphysiological crossectional area andpercent of maximum voluntarycontraction in each phase

    The motion tracking systemconsists of a transmitter which definesthe global axis system, located on thetibial shaft, and a set of receivers. Anacrylic shaft is placed into theintramedullary canal of the tibia tosupport the transmitter of anelectromagnetic motion transducer(Fastrak 3-D, Polhemus Systems,Colchester, VT) and transmit load tothe specimen. Methacrylate cement isplaced into holes drilled in thecalcaneus, talus, navicular, and cuboidof each foot. A carbon fiber pin isplaced into each hole, anchored in thecement, and an acrylic mount fixed tothe pin on which is located a receiverof the motion tracking system. Outputconsists of the direction cosinesrequired to define the orientation ofeach receiver axis. In this method, theinitial and final positions of eachreceiver axis are known and areprojected into each of three planes(sagittal, frontal, horizontal) defined bythe global axis system, similar to thosewhich might be obtained by viewingplanar radiographs. The transmitter Zaxis is aligned with the long axis of thetibial shaft, the X axis is directedtowards the medullary canal of thesecond metatarsal, and its Y axis isapproximately medial-lateral.Verification of sensor measurementaccuracy (ratio of measured output toknown input in the same direction)ranges from 0.944 to 1.006 in rotation,and 0.996 to 1.045 in translation.

    RESULTS AND CONCLUSIONS(i) Contact area and force on the

    talus are greatest in the posterior facet,followed by the talonaviculararticulation, the anterior and middlefacets, and the calcaneonavicularligament.

    (ii) The contact force in thecalcaneonavicular ligament articulationis best aligned to oppose medialdisplacement of the talar head.

    (iii) Loss of posterior tibialis tendontension has no effect on contact forcein the calcaneonavicular ligamentarticulation with the talus.

    (iv) Flatfoot decreases talonavicularjoint area, which is not restored byeither reconstruction. The Dillwyn -Evans procedure decreases

    Figure 2: Lateral radiographs of normal foot (top), and the same foot after creation of a flatfootdeformity by cyclic loading.


    Regulated, PneumaticCylinders

    Tibial Load(Compressive)

    Tendon Loads (Tensile)






    anteromedial facet contact area andincreases posterior facet pressure. TheSeattle method has less effect uponhindfoot joint contact characteristics.

    (v) Calcaneo-cuboid fusion withthe calcaneus lengthened and the footin neutral position has no effect onhindfoot kinematics. However, fusingthe foot in other orientations limitstalocalcaneal and talonavicular jointmotion.

    (vi) The increases in joint loadbetween heel strike and toeoff are morelikely due to differences in total loadapplied than due to differences inorientation of the talus since differencesin joint angle are small (10 degdorsiflexion for toeoff compared with6 deg of plantarflexion for heelstrike).High forces in toeoff are due primarilyto a combination of higher axial loadand larger Achilles force required tocreate thrust.

    Figure 3: Lateral schematic of experimental loading frame with foot in heel off condition.


    B ecause of the large sums ofmoney spent on health care, itis important to assess the valueof various therapeutic interventions.Recent attempts to define the value ofan intervention have been based on theresources necessary to provide it, suchas the Resource Based Relative ValueScale (RBRVS). Although thismethodology may be useful forallocating reimbursement, it does notdetermine the value of the interventionto the patient. For example, the“relative value” can be high if the serviceis expensive to provide, even if it isineffective. The authors suggest that‘value’ is more typically defined as thebenefit divided by the charge, that iswith the resource expense in thedenominator, rather than in thenumerator as it is in the RBRVS.Specifically, we propose that the value

    of a procedure be determined bymeasuring the differences theprocedure makes in the patient’scomfort, function, and quality of lifeand then dividing these by the chargeof the procedure.

    Standardized self-assessment toolsare now available for documentingpatients’ health and function before andafter treatment. The SF- 36 is one ofthe most widely used general healthstatus assessments. The SimpleShoulder Test is a commonly usedinventory of shoulder functions. Bothof these tools have been shown to reflectthe health and functional deficitsassociated with shoulder problems aswell as the change in these parametersafter treatment. Because they can beused without the patients having toreturn for follow-up visits, these self-assessment tools are useful for repeated

    assessments over time. This allowstime-dependent changes in the benefitexperienced by patients to beincorporated into the analysis. Theactual lifetime benefit experienced bythe patient will be the summation ofbenefit measured throughout thepatient’s life.

    In this pilot study, the authors usedthe SF-36 and the Simple Shoulder Testfor patient self-assessment before andsequentially after shoulder arthroplasty.These results were used to predict the“natural history” of a patient’s responseto the procedure. The value of theprocedure relative to the self-assessment parameters is determinedby dividing the lifetime benefitexperienced by the patient by the totalcharges for the procedure.

    METHODSPatient population. 125 patients

    who underwent 137 primary totalshoulder arthroplasties completed theShort-Form 36 and Simple ShoulderTest self-assessment questionnaires ontheir initial visit and at 6 monthintervals after their procedure. Allpatients were from the practice of anindividual surgeon and the procedurescarried out with standardized selectioncriteria, technique, prostheses andrehabilitation.

    Benefit. The benefit of theprocedure is defined in this study as thepercent change in a parameter, i.e. thedifference between the post-operativescore and the pre-operative scoredivided by the maximal possible scoreon the SST or the SF-36. Thus if apatient could perform 4 of the 12 SSTfunctions pre-operatively and 8 of the12 post-operatively, the benefit for theSST would be (8-4)/12 or 33%.Similarly if the SF-36 comfort scoreimproved from 32/100 to 76/100, thebenefit for SF-36 comfort would be44%.

    If the benefit determined at eachfollow-up is plotted against number ofyears after the procedure, the areabeneath this curve is the realized

    Assessing the Value of a Procedure: A Pilot Study Using TotalShoulder Arthroplasty as an Example


    Figure 1: The average change in scores from the pre- to post-operative period was statistically significantfor the Simple Shoulder Test and some of the subsections of the SF-36 (p


    Figure 2: The average percentage change in SST, Comfort and Physical Role for each post-operative year.

    benefit, expressed in units of %-years.If this curve is extrapolated out for thelife of the patient, the area beneath thecurve is the lifetime benefit of theprocedure, again in units of %-years.The lifetime value of the procedure isthe lifetime benefit divided by thecharge of the procedure, expressed inpercent-years/$1000.

    The sums of hospital and physiciancharges were normalized to 1992dollars using the Consumer Price Indexfor medical care for urban areas.

    RESULTSThe average age of our study

    population was 61 years (range 29-84).38 shoulder replacements were done infemales and 98 in males. Each totalshoulder was followed for an averageof 2.0 years. 125 patients whounderwent 137 total shoulderscompleted 481 follow-up SF-36 andSST questionnaires during the post-operative period, for an average of 3.5follow-ups per shoulder.

    Benefit of TSAThe average change in scores from

    the pre- to post-operative period wasstatistically significant for the Simple

    Shoulder Test and some of thesubsections of the SF-36 (p


    night. This effect diminished at a rateof 1.9% per year during the five yearsof follow-up. Given this slope, theestimated years of benefit with regardsto comfortable sleep was 43.2. Theaverage life expectancy of these patientswas 20.3 years. Therefore benefit wastruncated at that point. Using themethodology described above, theaverage patient who could not sleepcomfortable pre-operatively because ofshoulder discomfort could expect 13.9years (4919 days) of comfortable sleepattributable to this procedure. Thecharge per night of comfortable sleep($22,843/4919 days) would be $4.64.

    CONCLUSIONWe have illustrated a new

    methodology for measuring the valueof a medical intervention. In thisexample, we used self-assessmentquestionnaires—the Short-Form 36and the Simple Shoulder Test—toquantitate the benefit and value apatient receives from total shoulderarthroplasty. This type of analysis is,to our knowledge, unique in that itattempts to measure the lifetime benefitand value of the procedure.

    Our data demonstrate that SF-36scores and the SST are significantlyimproved after shoulder arthroplasty.As assessed by the SST and SF-36, thebenefit tends to lessen with time afterthis surgical procedure. The SST andSF-36 percent benefit scores showed anear linear decrease over time. We usedthese slopes to predict scores in thefuture. This allowed us to predict thelifetime benefit that a patient willreceive. Dividing the lifetime benefitby the expense of the procedure yieldsan estimate of the value of theprocedure, which is comparable toother measures of value in commonuse, such as miles per gallon.

    The purpose of this investigationwas to demonstrate an approach to

    measuring the benefit and value ofinterventions which could be appliedto other procedures and other practices.Because successful orthopedicprocedures yield benefits that areapparent to the patient, orthopedicsurgeons are fortunate that patient self-assessments can facilitate such ananalysis.


    Bayley KB, London MR, GrunkemeierGL, and et al: Measuring the success oftreatment in patient terms. Med Care33(4 Suppl):AS226-235, 1995.

    Beaton DE, and Richards RR:Measuring function of the shoulder. JBone Joint Surg 78A:882-890, 1996.

    Bergner M, Bobbitt RA, Carter WB, andet al: The Sickness Impact Profile:Development and final revision of ahealth status measure. Med Care19:787-805, 1981.

    Cotton P: Orthopedics research nowasks, “does it work?” rather than just,“how is the procedure performed?”.JAMA 265:2164-2165, 1991.

    Emery DD, and Schneiderman LJ:Cost-effectiveness analysis in healthcare. Hastings Cent Rep 19(4):8-13,1989.

    Gillespie WJ, and Daellenbach HG:Assessing cost-effectiveness inorthopaedic outcome studies.Orthopaedics 15:1275-1277, 1992.

    Levinsohn DG, and Matsen FA, III:Evaluation of nine commonmechanical shoulder disorders with theSF-36 health status instrument.Presented at the 64th Annual Meetingof the American Academy ofOrthopaedic Surgeons, San Francisco,CA, 1997.

    Lippitt SB, Harryman DT, II, andMatsen FA, III: A Practical Tool forEvaluating Function: The SimpleShoulder Test. In The Shoulder: ABalance of Mobility and Stability.Matsen FA, III, Fu FH, and Hawkins RJ,(eds.). American Academy ofOrthopaedic Surgeons, Rosemont, Il,pp. 501-518, 1993.

    Matsen FA, III: Early effectiveness ofshoulder arthroplasty for patients whohave primary glenohumeraldegenerative joint disease. J Bone JointSurg 78A:260-264, 1996.

    Matsen FA, III, Ziegler DW, andDeBartolo SE: Patient self-assessmentof health status in glenohumeraldegenerative joint disease. J ShoulderElbow Surg 4(5):345-351, 1995.

    Nerenz DR, Repasky DP, WhitehouseFW, and et al: Ongoing assessment ofhealth status in patients with diabetesmellitus. Med Care 30(5):MS112-124,1992.

    Parker MJ, Myles JW, Anand JK, and etal: Cost-benefit analysis of hip fracturetreatment. J Bone Joint Surg 74B:261-264, 1992.

    Radosevich DM, Wetzler H, and WilsonSM: Health status questionnaire (HSQ)2.0: Scoring comparisons and referencedata. Health Outcomes Institute,Bloomington, MN, 1994.

    Segel MJ, and Heald RB(editors):Medicare RBRVS: ThePhysician’s Guide.Chicago, AmericanMedical Association, 1996.

    Vital Statistics of the United States. LifeTables. U.S. Department of Health andHuman Services. 1992.

    Table 1: A summary of first year benefit, rate of decrement in benefit, years of benefit and total lifetime benefit versus SST, Comfort and Physical Role.

    SST Comfort Physical Role

    First year benefit 44% 25% 26%

    Rate of decrement in benefit 4.8%/yr 3.3%/yr 7.9%/yr

    Years of benefit 9.2 yrs 7.6 yrs 3.3 yrs

    Total lifetime benefit 246% - yrs 120% - yrs 69% - yrs


    The Simple Knee Test: A Quick and Sensitive Self-Assessment ofKnee Function


    There is an increasing demand forall orthopaedic surgeons todocument the functional statusof their patients before and aftertreatment as well as the efficacy of theirtreatment approaches to differentdiagnoses. Existing approaches to thisdocumentation consume substantialtime and resources and often do notcapture the patients’ perception of theircomfort and function. Others are notsufficiently sensitive to clinicalproblems with the anatomic area ofinterest. The ideal method would be arapid self-assessment that patients caneasily complete within a few minuteswithout assistance, whether in the officeor at home. The tool must discriminatedependably between normal andpathological knee function.

    Our goals were (1) to develop a self-assessment questionnaire sensitive to

    common knee pathologies whichwould be practical in the context of abusy office practice, (2) to documentits results in asymptomatic andsymptomatic knees and (3) todemonstrate its sensitivity to a range ofknee pathologies.

    METHODSInitial questionsA series of questions was developed

    from established knee scores. Initialquestions were taken from the Lysholm,Revised Hospital for Special SurgeryKnee Score, Cincinnati Knee score, andthe 1993 International KneeDocumentation Committee KneeLigament Standard Evaluation.Subsequently the questions werereformatted in a “yes/no” format witha “normal” response in the positivesense. All questions were reviewed with

    six experienced knee surgeons andedited to develop 23 questions (Table1).

    ControlsThis initial self-assessment was

    administered to 360 persons whoserved as controls. This included 200medical students, residents ororthopaedic staff, as well as 160 patientsseen in the orthopaedic clinics for non-knee related symptoms. In addition tothe 23 questions, each person was askedtheir age, gender, activity level, priorhistory of knee surgery, prior history ofseeing a doctor regarding kneesymptoms, and to comment about eachof the 23 questions. Any individual whohad either a previous knee surgery orhad visited a doctor regarding a kneeproblem was excluded from the controlgroup.

    Patients150 consecutive new patients

    presenting to either the sports medicineor the adult reconstruction clinics forevaluation of a knee problem wereasked to respond to the series of 23questions as well as indicate theimportance of each question to themin their daily activities as eitherimportant or not important. The age,gender and activity level for eachpatient was recorded.

    Diagnostic groupsEach patient was given a clinical

    diagnosis based upon their history,physical examination, radiographs,MRI and/or arthroscopy. Patients withsimilar diagnoses were grouped and the“group response” to each question wascalculated as a percentage.

    Statistical analysisA correlation matrix for all patient

    responses was created (Statview 4.5).Significant intercorrelation waspredetermined to be any set ofquestions with r > 0.50. The meanresponse of each diagnostic relatedgroup to each question was comparedto the mean response of the controlgroup for the same question using a t-test. Significance was defined as p–value< 0.05.

    Question Result

    1 Does your knee allow you to sleep? kept

    2 Will your knee straighten fully? kept

    3 Will your knee bend all the way? kept

    4 Is your knee free from swelling? kept

    5 Does your knee move smoothly without catching or locking? kept

    6 Can you sit comfortably with your knee bent for 1 hour? kept

    7 Does your knee allow you to get up from a chair with out difficulty? deleted

    8 Can you enter and leave a car without difficulty? kept

    9 Can you stand and walk without your knee giving away? kept

    10 Does your knee allow you to walk without a cane or crutch? deleted

    11 Does your knee allow you to walk down the hall without difficulty? deleted

    12 Does your knee allow you to walk two block without difficulty? kept

    13 Will your knee allow you to walk one mile without difficulty? deleted

    14 Does your knee allow you to do your usual work? kept

    15 Will your knee allow you to go down stairs easily? kept

    16 Will your knee allow you to go up stairs easily? deleted

    17 Will your knee allow you to do a deep squat (beyond 90 degrees)? deleted

    18 Will your knee allow you to run? kept

    19 Will your knee allow you to pivot to the outside while running? deleted

    20 Will your knee allow you to pivot to the inside while running? deleted

    21 Will your knee allow you to participate fully in your sport? kept

    22 Will your knee allow you to kneel comfortably? kept

    23 Can you hop on one leg? kept

    Table 1: Initial 23 Questions of the Simple Knee Test and Reduction to the Final 15


    RESULTSControlsFrom the original set of 360

    individuals, 7 persons were excludedbased upon having a prior knee surgeryand 40 others were excluded becausethey had seen a doctor regarding a kneeproblem. This left a total of 313 personsin the control group. Of these, 145(46%) were female and 168 (52%) weremale. The average age was 30. Theaverage Tegner activity level was 5.

    PatientsOf the 150 patients, 79 (53%) were

    female and 71 (47%) were male. Theaverage age was 35 and the averageTegner activity level was 5.

    Diagnostic groupsPatients were placed in one of eight

    diagnostic groups: 1) anterior cruciateligament insufficiency, 2) degenerativejoint diseases, 3) isolated meniscal tears,4) patellar instability, 5) patellarirritation, 6) tendinitis, 7) otherdiagnosis, and 8) no diagnosis.

    Reduction of questionsOf the original 23 questions, 8 were

    eliminated because of: 1) significantintercorrelation with other questions (r≥ 0.50), 2) lack of differential sensitivity

    (≥ 95% of patients within a diagnosticgroup gave the same response to thequestion), or 3) lack of clarity, based onpatient comments. This left a total of15 questions which comprised theSimple Knee Test.

    Question # 11 (Does your kneeallow you to walk down the hall withoutdifficulty?) and Question #13 (Willyour knee allow you to walk one milewithout difficulty?) were both highlycorrelated with Question #12 and weredeleted.

    Question #7 (Does your knee allowyou to get up from a chair withoutdifficulty?) was highly correlated withfour questions (#8, 11, 13 and 16) andwas deleted.

    Question #16 (Will your knee allowyou to go up stairs easily?) was highlycorrelated with questions #7 and #8 andwas deleted.

    Question #10 (Does your knee allowyou to walk without a cane or a crutch?)was not highly correlated with anyother questions; however, it lackeddifferential sensitivity in that greaterthan 95% of patients with ACLdeficiency, meniscal tears, degenerativejoint disease, or patellar irritation

    responded that they could ambulatewithout assistive devices. This questionwas deleted.

    Question #17 (Will your knee allowyou to do a deep squat?), Question #19(Will your knee allow you to pivot tothe outside while running?) andQuestion #20 (Will your knee allow youto pivot to the inside while running?)were both highly correlated with otherquestions and were confusing to asignificant portion of the patients andwere deleted.

    Variance within diagnostic groupsFor each question, the mean

    patients’ response was significantlydifferent from the mean controls’response (p


    Test is unique. All questions that werefound to be highly correlated, lackingdifferential sensitivity, or unclear havebeen deleted. The remaining 15questions measure different aspects ofa patient’s knee function.

    The Simple Knee Test is sensitive todiffering knee pathologies. Forexample, patients with DJD and ACLinsufficiency gave significantly differentanswers to “Does your knee allow youto sleep?,” “Can you sit comfortablywith your knee bent for one hour?,”“Can you enter and leave a car withoutdifficulty?,” “Can you stand and walkwithout your knee giving away?,” and“Can you hop on one leg?” (see Figure1).

    The Simple Knee Test provides aconvenient method for regular follow-up without having the patients returnto the office. The ability to assesspatients via the mail allows a substantialsavings of both patient and officeresources. With a systematic follow-up,surgeons can conduct personal qualitycontrol, determining which techniquesare most effective in their hands.

    The simplicity of the SKT questionsfacilitates communication of results topatients. The SKT reflects the status ofthe knee in functional terms rather thanin degrees of motion, appearance ofradiographs, millimeters ofdisplacement, or scores. Patients caneasily compare their responses to thoseof other patients and, using thiscomparison, help determine theircandidacy for different treatment.

    Continuing clinical research willestablish the characteristic SKT profilesof patients who are most likely tobenefit from treatment as well as theeffectiveness of different therapeuticapproaches in improving patients’assessment of their knee function. Thesimplicity and standardization of theSimple Knee Test facilitates comparisonof these results among the practices ofdifferent surgeons.


    Kelley, R. B., Rudicel, S. A., and Liang,M. H.: Outcomes research inorthopaedics. J Bone Joint Surg75A(10):1562-1574, 1993.

    Matsen, F. A., III, Lippitt, S. B., Sidles, J.A., and Harryman, D. T., II: Practicalevaluation and management of theshoulder. W. B. Saunders Company,

    Philadelphia, 1994.

    Noyes, F. R., Matthews, D. S., Mooar, P.A., and Grood, E. S.: The symptomaticanterior cruciate-deficient knee. Part II:the results of rehabilitation, activitymodification, and counseling onfunctional disability. J Bone Joint SurgAm 65(2):163-74, 1983.

    Tegner, Y., and Lysholm, J.: Ratingsystems in the evaluation of kneeligament injuries. Clin Orthop 198:43-9, 1985.

    Windsor, R. E., Insall, J. N., Warren, R.F., and Wickiewicz, T. L.: The hospitalfor special surgery knee ligament ratingform. Am J Knee Surg 1:140-145, 1988.


    To evaluate the in vivo responseof cartilage to an intra-articularinjury with residual step-off,examining both the biomechanicaleffects as well as the histomorphologicchanges is the objective of this study.

    METHODSAn intra-acticular step-off was

    created in medial tibial plateaus of theknees of 12 adult, domestic sheep. Asagittal osteotomy was made throughthe middle of the weight-bearingsurface, with displacement of themedial fragment distally by onemillimeter. After rigid fixation withscrews, the animals were allowedactivity ad libitum. The contralateralknee was used as a control. Fouranimals were terminated at six weeks;the others at twelve weeks. All animalswere labelled with oxytetracyline (50milligrams/kilogram) three weeks priorand three days prior to termination andwell as calcein (12 milligrams/kilogram) two weeks prior totermination.

    After termination, the knees wereloaded in a materials testing machineusing pressure sensitive film to record

    joint contact pressures. Subsequently,the articular surfaces were sectionedand prepared for both calcified andnon-decalcified histology as well asimaging by scanning electronmicroscopy.

    RESULTSQualitative analysis of the pressure

    sensitive film demonstrated a single,broad area of contact in the controlknees. The knees with intra-articularstep-off had two major contact areaswith an intervening zone of reducedload corresponding to the edge of thedepressed fragment. Quantitativeanalysis revealed reduced overallcontact area in the knees with step-off.

    Histologically, coronal sectionsthrough the articular surfacedemonstrated the presence of thinningand fibrillation on the high side of thestep-off. In addition, at the edge of thehigh side compensatory bending of thenormally vertical fibrils towards the lowside was observed. Some compensatoryhypertrophy of the cartilage at the edgeof the low side was seen, but notadequate to level the articular surface.No significant remodelling was seen in

    The Response of Cartilage to Intra-articular Step-off: A SheepWeight Bearing ModelTHOMAS E. TRUMBLE, M.D.

    the subchondral bone, with the planeof the osteotomy still being readilyidentified.

    Scanning electron microscopydemonstrated partial cartilageremodeling by deformation of the highside cartilage to partially overlap thelow side cartilage and by behind thevertical collagen fibrils on the high sidecartilage even in the unloaded state(Figure 1).

    CONCLUSIONSIn this animal model, a one

    millimeter intra-articular step-offchanged the contact distribution bothqualitatively and quantitatively duringin vitro loading, with a net loss ofcontact area. Examination by both lightand electr