1999 Research Report www - UW Orthopaedics and Sports …...1999 ORTHOPAEDIC RESEARCH REPORT 7...

UNIVERSITY OF WASHINGTON

Department of Orthopaedics

1999 Research Report

UNIVERSITY OF WASHINGTON

Department of Orthopaedics1999 Research Report

Department of OrthopaedicsUniversity of WashingtonSeattle, WA 98195

EDITORS:Fred WesterbergFrederick A. Matsen III, M.D.Peter Simonian, M.D.

DESIGN & LAYOUT:Fred Westerberg

Cover Illustration: “Builders No. 1” by Jacob Lawrence1970, Gouache on paperCourtesy of the Artist and Francine Seders GalleryCollection of Henry Art Gallery, SeattlePhotograph: University of Washington Media Services

UNIVERSITYOF WASHINGTON

SCHOOL OFMEDICINE

http://www.orthop.washington.edu/faculty/faculty2.htm#Matsen

http://www.orthop.washington.edu/faculty/faculty2.htm#Simonian

http://www.orthop.washington.edu/

http://www.washington.edu/medical/

Contents

Foreword .................................................................................................................................................................... 7

Chronic Distal Biceps Tendon Ruptures Reconstruction with Autogenous Semitendnosus ............................... 8

Douglas T. Harryman II, M.D, David G. Duckworth, M.D., Michael J. Moskal, M.D., Tim A. DuMontier, M.D.,Kevin L. Smith, M.D., and Frederick A. Matsen III, M.D.

Risk of Early Closed Reduction in Cervical Spine Subluxation Injuries .............................................................. 13

Gerald A. Grant, M.D., Sohail K. Mirza, M.D., Jens R. Chapman, M.D., H. Richard Winn, M.D.,David W. Newell, M.D., Dolors T. Jones, R.N., Mark A. Konodi, M.S., and M. Sean Grady, M.D.

Early Versus Delayed Surgery for Acute Cervical Spinal Cord Injury ................................................................... 16

Jens R. Chapman, M.D., Sohail K. Mirza, M.D., Walter F. Kengel III, M.D., Paul A. Anderson, M.D.,James C. Bailey, M.D., M. Sean Grady, M.D., Mark A. Konodi, M.S., and Hansen A. Yuan, M.D.

Relationship Between Neurological Deficit and Spinal Cord Deformation in Cervical Spine Injuries .............. 18

Gregg K. Motonaga, B.S., Randal P. Ching, Ph.D., Mark A. Konodi, M.S., and Sohail K. Mirza, M.D.

Creatine Supplementation ...................................................................................................................................... 20

John O’Kane, M.D.

Increased Type II Collagen Turnover in Trained Marathon Runners Versus Age-Matched Contols .................. 22

Daniel J. Stechschulte, Jr., M.D., Ph.D., Lynne M. Atley, Ph.D., John M. Wright, M.D., Russell F. Warren, M.D.,Margaret G.E. Peterson, Ph.D., and David R. Eyre, Ph.D.

Subluxation of the Talocalcaneal Joint in Symptomatic Adult Flatfoot ............................................................... 24

Dheera Ananthakrisnan, M.D., Randal P. Ching, Ph.D., Allan F. Tencer, Ph.D., Sigvard T. Hansen, Jr., M.D.,and Bruce J. Sangeorzan, M.D.

Correlation of the Simple Knee Test with the Lysholm Score for Follow-up of ACL Reconstruction................ 27

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

Elbow Medial Collateral Ligament Isometry: Evaluation of the Ulna Insertion Site .......................................... 29

Peter T. Simonian, M.D.

A New Technique for Mapping Articular Cartilage Contour and Thickness ....................................................... 32

Anthony G. Norman, B.S., William M. Dougherty, Ph.D., Howard Chansky, M.D., Peter Simonian, M.D.,John M. Clark, M.D., Ph.D., and John A. Sidles, Ph.D.

Importance of Surface Integrity to Healing of Articular Cartilage Defects .......................................................... 34

Julie Switzer, M.D., Anthony G. Norman, B.S. and John M. Clark, M.D., Ph.D.

Collagen Type II Cross-Linked Telopeptides: A Promising Marker of Cartilage Degradation in Arthritis ........ 36

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

Expression of Cartilage Extracellular Matrix and Potential Regulatory Genes in a New Human

Chondrosarcoma Cell Line ..................................................................................................................................... 38

Howard Chansky, M.D., James R. Robbin, Wendy H. Raskind, Steven Cha, Ernest U. Conrad, M.D., ........................John M. Clark, M.D., James D. Bruckner, M.D., and Linda J. Sandell, M.D.

Does End-to-Side Nerve Repair Work? A Study of Axonal Sprouting Stimulated from Intact Nerves

Across an End-to-Side Coaptation ......................................................................................................................... 40

Wren T. McCallister, B.S., B.A., Peter Tang, M.D., and Thomas E. Trumble, M.D.

Influence of Comorbid Medical Conditions on Pre-Treatment SF-36 Scores ..................................................... 43

Sohail K. Mirza, M.D., Barry Campbell, B.S., Randal P. Ching, Ph.D., Jens R. Chapman, M.D., Mark K. Konodi,M.S., and Frederick A. Matsen III, M.D.

Functional Outcome of In Situ Screw Fixation Versus Flexion Intertrochanteric Osteotomy for

Slipped Capital Femoral Epiphysis ......................................................................................................................... 46

Mohammad Diab, M.D., S. Daluvoy, B.S., J. Kasser, M.D., and B. Snyder, M.D.

Somatosensory Evoked Potential Monitoring During Closed Humeral Nailing ................................................. 47

William J. Mills, M.D., Jems R. Chapman, M.D., Lawrence R. Robinson, M.D., and Jefferson C. Slimp, Ph.D.

Percutaneous Stabilization of U-Shaped Sacral Fractures Using Iliosacral Screws ............................................. 49

Sean E. Nork, M.D., Clifford B. Jones, M.D., Susan P. Harding, M.D., Sohail K. Mirza, M.D.,and M.L. Chip Routt, Jr., M.D.

Screw Fixation of Osteochondral Injuries in Comminuted Acetabular Fractures ............................................... 51

Daniel Segina, M.D., Sean E. Nork, M.D., and M.L. Chip Routt, Jr., M.D.

Practical Approaches to Injury Prevention Using Biomechanics and Epidemiology .......................................... 55

Allan F. Tencer, Ph.D.

Failed Acromioplasty: Investigation of 74 Cases .................................................................................................... 57

David G. Duckworth, M.D., Kevin L. Smith, M.D., and Frederick A. Matsen III, M.D.

The Position of Prosthetic Components in Failed and Successful Shoulder Arthroplasties ............................... 59

Michael J. Moskal, M.D., Barry Campbell, B.S., Kevin L. Smith, M.D., and Frederick A. Matsen III, M.D.

Characteristics of Patients Treated with Proximal Humeral Hemiarthroplasty for Acute Three

and Four Part Fractures and the Correlation with Shoulder Function ................................................................ 62

Richard Rozencwaig, M.D., Shadley C. Schiffern, B.S., John Antoniou, M.D., Hiroshi Takeda, M.D.,and Douglas T. Harryman II, M.D.

Physical Changes in Polyethylene Glenoid Components After In Vivo Use ......................................................... 64

Marius M. Scarlat, M.D., William M. Dougherty, Ph.D., Kevin L. Smith, M.D., John A. Sidles, Ph.D.,James D. Bruckner, M.D., and Frederick A. Matsen III, M.D.

Department of Orthopaedics Faculty .................................................................................................................... 67

Incoming Faculty ..................................................................................................................................................... 68

Graduating Residents .............................................................................................................................................. 69

Incoming Residents ................................................................................................................................................. 70

Contributors to Departmental Research and Education ....................................................................................... 71

National Research Grants ........................................................................................................................................ 73

1999 ORTHOPAEDIC RESEARCH REPORT 7

Foreword

Our cover artist this year is JacobLawrence, an importanttwentieth century artist who

now lives in Seattle. Lawrence’s styleuses flat, strongly outlined shapes inbold geometric patterns to typify thedaily life and history of black people inAmerica. Our cover painting is entitled“The Builders” - a most suitableselection for a Department of faculty,residents and staff dedicated to theconstruction and reconstruction of thehuman frame. This painting evenfeatures some of our traditionalorthopaedic tools: hammers, saws andnails. But in building as inOrthopaedics, the end result isdetermined not by the tools, but whoholds them. The builder, the surgeon,is the method!

This year has been poignant for theDepartment, in that a number of ourfaculty, family and friends sustainedmajor orthopaedic injuries. Ourprayers and best wishes go out to them.These ‘up close and personal’experiences make us recognize that, asproud as we are of what Orthopaedicscan offer, we have a long way to gobefore we are as effective builders as wewould like to be. Our hope lies in theresearch we do, in our commitment tocontribute to the base of knowledge onwhich Orthopaedics is based.

Our research efforts arepermanently enhanced by theendowments that have becomemainstays of our research programs.The Burgess Chair has enabled itsholder, David Eyre, to continue makemajor contributions to ourunderstanding of the development anddecay of the human frame. The HansenChair, held by Ted Hansen, providescore support for our research programin Orthopaedic Biomechanics. TheJerome Debs Chair, held by SteveBenirschke, supports cutting edgeresearch in the management ofcomplex foot and ankle injuries. Theseenduring programs are now joined bythe E.A. Codman/DePuy Professorshipfor Shoulder and Elbow Research. Inrecognition of his consistently excellentresearch, always done with a sharp focuson the End Result, Douglas T.Harryman II has been appointed by theUniversity as the first holder of this

Professorship. The illustrations at theright show Amory Codman above andan illustration of one of Doug’s manyinnovations: the arthroscopic release ofa frozen shoulder.

We are well on the way to aprofessorship in spine surgery inrecognition of the innovative researchof Drs Mirza and Chapman. Finally,we have launched a campaign toestablish a Chair for Women’s SportsMedicine and Fitness – this Chair willhelp us come to grips with thedisproportionately high injury rateamong active women and lack ofgender- specific information on howthese injuries can be best prevented andmanaged.

If you’d like to learn more about theDepartment of Orthopaedics at theUniversity of Washington, visit us atwww.orthop.washington.edu. If you’dlike to learn about how you can supportour research and educational programsor have other questions for us, pleasedrop us a letter or an Email [email protected] [email protected].

As always we must conclude byexpressing our profound appreciationto all the friends, staff, students,residents and faculty who have madethis Department the Best in the West.Thanks for your special contribution toour missions of research, teaching andpatient care.

Our next issue will be in the newmillennium - stay tuned.

Best wishes,

Frederick A. Matsen III, M.D.Chairman

Peter T. Simonian, M.D.Chief, Sports Medicine Clinic

E.A. Codman, 1869-1940

Diagram depicting the typicalsequence of capsular release: (1)postero-superior and posterior, (2)antero-superior, rotator interval andsubcoracoid recess, (3) antero-inferior; middle and inferiorglenohumeral ligaments and (4)postero-inferior; postero-inferiorrecess and inferior sling.

8 1999 ORTHOPAEDIC RESEARCH REPORT

Rupture of the distal end of thebiceps brachii tendon occursinfrequently and usually

presents only in middle-aged males(Figure 1). These tears generally causesubstantial loss of supination and elbowflexion strength leading to significantfunctional impairment.

Most surgeons recommend acuteanatomical repair within a maximumof two or three weeks after injury toavoid difficulty. In chronic distal bicepstears, achieving an anatomic repair isunlikely because of prolongedcontracture of the musculotendinousunit, tendinous attenuation andscarring between the muscular planesof the forearm. For these reasons,clinicians have recommended eithernonoperative management orattachment of the retracted distal bicepstendon stump to the brachialis, amethod which does not restoresupination strength.

In 1991, we began offeringreconstruction of the distal biceps usingan autogenous semitendinosusinterposition graft to patients withdisabling loss of elbow flexion andforearm supination. The purpose ofthis study was (1) to describe oursurgical method for reconstructing achronic rupture of the distal bicepstendon using a free autogenoussemitendinosus graft and (2) todetermine the outcome in terms ofstrength and elbow range, forearmrotation and functional activities ofdaily living after reconstruction.

MATERIALS AND METHODS

The first procedure was performedin 1991. Since then, a total of ninepatients with ten ruptures have hadreconstruction of the distal bicepstendon using a semitendinosus graft(Table 1). All patients were male andtheir average age was 46 years at thetime of repair (range, 29 - 61 years).The mean interval between rupture andtendon reconstruction was 9.0 months(range, 1-19 months). The averagefollow-up for seven of nine patientswith more than two-year follow-up was

39 months (range, 24-74 months). Allinjuries resulted from a suddeneccentric force against a contractedbiceps with a flexed elbow.

All operative procedures wereperformed by one of the senior authors(DTH II and FAM III). Allreconstuctions utilized a doublestranded semitendinosus autograft tobridge the gap between the retractedbiceps tendon and the radial tuberosity.A two incision technique was utilizedin six of ten cases and a one incisiontechnique in four cases.

All patients were available forfollow-up clinical examination andSimple Elbow Test self-assessmentquestionnaire (source SET-FAM III).Clinical examination included flexion,extension, pronation, and supinationmotions in all patients. Flexion andsupination strength measurementsusing a Cybex isokinetic dynamometerwere obtained in seven patients witheight repairs having follow-up ofgreater than two years.

OPERATIVE TECHNIQUE

Under tourniquet control, ananterior curvilinear Henry-typeincision is made over the antecubitalfossa and the radial tuberosity isexposed. An elliptical hole is openedin the radial tuberosity. Surgicaldissection is directed proximally toexpose the distal stump of the retractedbiceps tendon which is usually adherentto the coracobrachialis. The entire distalbiceps muscle is mobilized whileprotecting the musculocutaneous nerveuntil an elastic muscular bounce isrecreated.

The single or two incision techniquecan be used to secure thesemitendinosus to the radial tuberosity.The one-incision technique requiresgreater distal dissection in the forearmand the two-incision technique requiresa separate incision on the extensorsurface. When a single incision ischosen, six drill holes are placed in aring distal to the elliptical tendoninsertion site in the radial tuberosityand three sutures are passed (Figure 2A

and 2B).The semitendinosus graft is

harvested. The looped mid-section ofa folded tendon is implanted into theradial tuberosity. Sutures are passed ina Bunnell fashion through each limb ofthe tendon using a tapered #5 Mayoneedle to avoid cutting fibers. Afterattaching the tendon, full supinationand pronation is performed to insurethat the radial tuberosity andreconstructed autograft tendon has freepassage through the proximal radio-ulnar bursa.

The free proximal ends of the graftare passed through the centralmusculo-tendinous fibers of the distalbiceps and are pulled toward the radialtuberosity while the elbow is flexed toadjust light tension in the reconstructedmusculo-tendinous unit and thensutured in place.

Postoperatively, the elbow isimmobilized in flexion but motionbegins the day after surgery. Each weekafter surgery, passive extension with thesupinated forearm is increased by 10°until full motion is recovered. Passivesupination and pronation is performedwith the elbow flexed at 90° untilmaximal flexion and extension arerecovered. Twelve weeks afterreconstruction, active assisted flexion-extension and supination-pronationare begun. Resistance to flexion isincreased one pound each week to fourpounds by four months after the repair.At six months, slow repetitive curlsunder a comfortable load areperformed until fatigue. The patientshould permanently avoid suddenheavy loads against elbow flexion.

RESULTS

All ten reconstructions remainedintact and were assessed using side toside strength and motionmeasurements (Figure 3). Excellentrange of motion was maintained in 8of 9 patients and elbow strength waswithin a mean of 15% of the oppositeside in the seven patients having a twoyear follow-up. Follow-up examinationresults which include range of motion

Chronic Distal Biceps Tendon Ruptures Reconstruction withAutogenous SemitendinosusDOUGLAS T. HARRYMAN II, M.D., DAVID G. DUCKWORTH, M.D., MICHAEL J. MOSKAL, M.D.,TIM A. DUMONTIER, M.D., KEVIN L. SMITH, M.D., AND FREDERICK A. MATSEN III, M.D.


and strength testing are summarized intable 2.

All patients returned to workfollowing surgery, which in five casesinvolved heavy manual labor. Fivepatients, active in sports pre-operatively, were able to return withminimal functional disability. Threecomplications occurred; one posterior

interosseous nerve neuropraxia whichcompletely resolved, one heterotopicbone formation which was successfullymanaged by excision and irradiationand one case of adhesions between thegraft and brachialis requiring surgicalrelease. Affirmative answers averaged11.6 out of 12 on the Simple Elbow Test(Table 3). The time to reach maximum

Case Age I n j u r e d / Work De lay Time to Max imum Spor t s? Biceps Sx's Leg Sx's RTW? Success?All Male Dominant Comp in Rx Operat ion Recove ry B e f o r e / Persist? ( g r a f t ) Y / N

s ide ( M t h s ) ( M t h s ) ( M t h s ) A f t e r

1 6 0 R/R Yes NA 1 6 2 4 N/N cramps & stiffness parasthesias Retired Yes

2 4 6 L / L Yes 2 3.5 1 2 N/N mild weakness No Yes Yes

3 2 9 L /R No 1 2 4.5 1 2 Y / Y No No Yes Yes

4 - R 6 1 R/R Yes 1 2 1 2 1 2 N/N R-hand weakness occ. cramps Yes Yes

5 - L - L /R - 2 4 1 9 1 8 N/N mild weakness No NA Yes

6 4 3 R/R Yes 2 7 8 Y / Y occ. cramps occ. pain Yes Yes

7 3 4 L /R Yes 3 6 1 8 Y / Y pain and stiffness No Yes Yes

8 4 7 R / R No 2 1 6 Y / Y No No Yes Yes

9 5 4 R / L No 6 1 7 6 N/N No No Yes Yes

1 0 3 6 R / R Yes 2 4 Recovering N/N No Yes Yes Yes

Avg 47yrs Avg ± SD 7 / 9 - Y 7 ± 8 9 ± 6 14 ± 6 4/4 Yes 6/10-Yes 3-Yes 9-Yes 10-Yes

Table 1: Preop/Postop Demographics.

functional recovery was an average of14 months (range: 6 to 24 months). Allpatients regarded their operation assuccessful.

DISCUSSION

Morrey and colleagues, comparedresults of immediate anatomicalreattachment, delayed reattachment,

Case Range of Motion S t reng th Endurance St reng th Endurance SimpleExt° - Flex° Deficit (%) Deficit (%) Deficit (%) Deficit (%) Elbow

Flexion Flex work Supination Supn work T e s t( 6 0 ° / s e c ) ( 1 8 0 ° / s e c ) ( 3 0 ° / s e c ) ( 1 5 0 ° / s e c )

(20 reps) ( 1 5 r e p s )( P r o n / S u p ) Side Nl/Recon Side Nl/Recon Side Nl/Recon Side Nl/Recon # Yes

1 0° -140° (80°-80°) 24% - 4 2 % 15% 0 % 1 0

2 20° - 140° (80°-65°) 40% 50% 0 % 50% 1 1

3 0° / 145° (80°-80°) 17% 31% - 4 4 % 22% 1 2

4 - R 5° / 145° (80°-80°) 18% 15% 0 % 0 % 1 2

5 - L 5° / 145° (80°-80°) Index side Index side Index side Index side 1 2

6 0° / 145° (80°-80°) - 1 1 % - 2 9 % - 1 6 % 17% 1 2

7 0° -140° (80°-80°) 3 % 6 % 28% - 2 3 % 1 2

8 10° - 130° (50°-80°) - 1 % - 3 2 % - 1 0 % - 1 8 % 1 2

9 check * * * * NA

1 0 check * * * * NA

4° - 143° (76°-78°) 15% ± 17% 5 ± 35% -4 ± 23% 7 ± 25% 11.6

* no strength testing until >12 months

Table 2: Functional Assessment.


attachment to the brachialis tendon,and nonoperative treatment of distalbiceps ruptures. Immediatereattachment restored normal strengthby one year in comparison tononoperative treatment, which resultedin a 40% loss of supination strength and30% loss of flexion strength. Attachingthe brachialis muscle made no

significant improvement in supinationstrength.

Recently, a case report using thehamstring autograft in a similar waythat we have described was reported,however, our repairs antedate thisreport and no series has been publishedon using grafts to re-establish the lostlength of biceps tendon. Hamstring

Simple Elbow Test

1 . Is your elbow comfortable at rest?2 . Does your elbow allow you to sleep comfortably?3 . Can you reach the small of your back to tuck in your shirt?4 . Can you place your hand behind your head with your elbow straight out?5 . Will your elbow allow you to pull on socks or stockings?6 . Can you lift one pound to the level of your shoulder?7 . Can you use your elbow to help rise from a chair?8 . Can you carry twenty pounds at your side?9 . Will your elbow allow you to comb your hair?

10 . Will your elbow allow you to throw a ball?11 . Does your elbow allow you to wash the back of your opposite shoulder?12 . Can you work full time at your regular job?

autograft is commonly used for tendongrafting and has minimal donor sitemorbidity.

Chronic ruptures of the distal bicepstendon can be managed successfully byusing a semitendinosus autograft tobridge the defect between the retractedbiceps tendon and radial tuberosity.This technique is relatively simple,

Figure 1: Clinical photograph of a patient with a chronic distal biceps tendon rupture and a retracted biceps (see arm with tattoo).

Table 3: The Simple Elbow Test.


reliable, and reproducible. Patientsrecovered elbow and forearm strengthand returned to normal activitieswithout significant donor sitemorbidity. Overall our patients weresatisfied with the surgical procedureand believed that it improved theirelbow function, relieved discomfort,and restored supination and flexionpower. Patients were able to return towork, sporting activities andcosmetically have a more normalappearing arm. We conclude thatsemitendinosus autograft should beconsidered for patients who have achronic tear of the distal biceps tendon(greater than 3 weeks) which hasretracted so that it can no longer reachthe radial for direct reattachment.

RECOMMENDED READING

Baker, B.E and Bierwagen, D.: Ruptureof the distal tendon of the bicepsbrachii. Operative versus non-operativetreatment. J Bone Joint Surg.1985,(67A), 414-417.

Boyd, H.B. and Anderson, L.D.: Amethod of reinsertion of the distalbiceps brachii tendon. J Bone JointSurg.1961, (43A)7:1041-1043.

Hang, D.W., Bach, Jr., B.R., andBojchuk, J.: Repair of Chronic DistalBiceps Brachii Tendon Rupture usingFree Autogenous SemitendinosusTendon. Clin. Orthop. 1996; 323:188-191.

Morrey, B.F., Askew, L.J., An, K.N., andDobyns, J.H.: Rupture of the distaltendon of the biceps brachii. Abiomechanical study. J Bone Joint Surg.1985, (67A), 418-421.

Figure 2A: Insertion of looped semitendinosus autograft into the radial tuberosity. For the one-incision method, the graft is initially secured into the radialtuberosity. Then the tendon is repaired to the distal biceps tendon stump. Tension is adjusted with the elbow flexed 90°.


Figure 3: Clinical photograph of a patient at two years who recovered excellent strength of elbow flexion and supination.

Figure 2B: Insertion of looped semitendinosus autograft into the radial tuberosity. For the one-incision method, the graft is initially secured into the radialtuberosity. Then the tendon is repaired to the distal biceps tendon stump. Tension is adjusted with the elbow flexed 90°.


The proper treatment of acutetraumatic cervical spinesubluxation injuries remains

controversial. Clinically significant discherniation occurs in association withcervical trauma, however, the exactincidence of acute disc herniationremains unknown. No strict guidelinesexist in the acute management of theseinjuries, with regard to the timing ofclosed reduction and prioritization ofimaging studies. Reduction of adislocation of the cervical spinepotentially increases the diameter of thespinal canal at the level of injury anddecompresses the spinal cord. However,

it has also been theorized that patientswith cervical spine subluxation injuriesmay suffer neurological deteriorationfollowing closed reduction in thepresence of an associated herniated discat the level of the injury. We present ourexperience with 121 patients admittedwith cervical spine subluxation injuries.We address the incidence of cervicaldisc injury and the risk of early closedreduction in these patients.


We reviewed 121 patients admittedto Harborview Medical Center inSeattle, Washington with traumaticlower cervical spine injuries (C3-C7).We excluded patients with gunshotwounds, minor fractures, and injurieswithout subluxation. We retrospectivelyreviewed the medical records andimaging studies (i.e. CT and MRI) ofthe remaining 82 patients (63 men and19 women, average age 42 years).

We treated all patients immediatelyupon presentation to the emergencyroom according to ATLS (AdvancedTrauma life Support) guidelinesfollowing spinal immobilization. Wegave all patients methylprednisolone.Senior neurosurgical house officersperformed serial clinical examinationsusing the American Spinal InjuryAssociation (ASIA) motor score andFrankel Scale to assess the degree ofneurological impairment. We usedGardner-Wells tongs to apply axialcraniocervical traction withfluoroscopic monitoring. Using

sequential weight applications, weattempted immediate reduction toimprove the spinal alignment andreduce cord compression. Weperformed open reduction whentraction was unsuccessful. Wedetermined the ASIA motor andFrankel scores upon admission to theemergency room, at 6 hours, and 24hours after reduction.

We obtained cervical magneticresonance images (MRIs) within 24hours following closed reduction. Weestablished disc injury from the MRIsas either a herniation or disruption. Ina herniation, the disc extrudeddeforming the thecal sac or nerve roots.In a disruption, the disk space waswidened, but without disk extrusion.

For our statistical analysis wecrosstabulated several dependentfactors with classifications of the discinjury and fracture type to calculate thePearson correlation coefficient. Wecompared pre- and post-reductionASIA Motor scores using a paired-samples t test with p values of p<0.05.

RESULTS

All of our patients sustained cervicalspine subluxation injuries, Table 1. Ourimmediate, rapid closed reduction wassuccessful in 80 of 82 patients (98%).Our average time to achieve closedreduction was 2.1 ± 0.24 hours (mean± SEM) from the time of arrival in theemergency room. The average tractionweight we applied was 37 ± 4% (mean± SEM) of body weight (range, 4% to80%). Despite application of maximalweight, we were unable to adequatelyrealign the spinal column with closedreduction in two patients. Neither ofthese patients suffered neurologicaldeterioration during our closedreduction attempts.

We were able to perform an MRIwithin 24 hours of reduction in 76patients. The incidence of disk injuryat the level of the subluxation was 46%(herniation 22% and disruption 24%).The incidence of disk injury wassignificantly more common in theelderly (greater than 55 years old). The

Risk of Early Closed Reduction in Cervical Spine SubluxationInjuriesGERALD A. GRANT, M.D., SOHAIL K. MIRZA, M.D., JENS R. CHAPMAN, M.D., H. RICHARD WINN, M.D.,DAVID W. NEWELL, M.D., DOLORS T. JONES, R.N., MARK A. KONODI, M.S., AND M. SEAN GRADY, M.D.

Injury Type Number of Patients % of Total

Unilateral facet dislocation 26 32%

Bilateral facet dislocation 15 18%

Compressive burst fracture 21 26%

Extension injury 9 11%

Miscellaneous 11 13%

Table 1: Classification of Skeletal Injury.

Figure 1: Radiographs can clearly showdisplacement in a vertebral dislocation.


presence of a disk injury significantlycorrelated with the presence of a spinalcord contusion on MRI.

The majority of our patientspresented with complete or incompletespinal cord injuries, Table 2.Neurological recovery, assessed by theASIA motor scores, significantlyimproved 24 hours following closedreduction, Table 3. Only one patientdeteriorated following immediateclosed reduction.

DISCUSSION

Our review of the traumatic cervical

spine injuries over the last three yearsindicates that immediate closedreduction is safe in awake andexaminable patients. We successfullyreduced 98% of our patients within 2.1ours of admission to the emergencyroom. Post-reduction MRI indicatedacute disk injury in 46% of the patientsin our series. Although a traumatic diskherniation could increase spinal cordcompression following closedreduction, our experience does notshow this to be a clinically significantconcern. Irrespective of the presence of

a disk injury, the incidence ofneurological deterioration followingclosed reduction was rare in our series.One patient out of 76 undergoing post-reduction MRI deteriorated 4 ASIAmotor points, but did so more than 6hours following closed reduction. Thisminor change in motor score was mostlikely a cervical radicular compromiseand not due to spinal cord compression.

The goal of skeletal traction forclosed reduction is to alleviatecompression of the spinal cord. Severalclinicians have strongly advocatedimaging studies prior to reduction inany patient with a neurological deficitto assess the injured disk. Obtaining anMRI prior to reduction can add asignificant delay and may place thepatient at added risk. We do not delayclosed reduction in a patient with acervical subluxation who presents witha complete motor deficit. We alsoadvocate immediate closed reductionfor neurologically incomplete andintact patients. None of the 22 intactpatients showed any neurologicaldeterioration following immediateclosed reduction.

Injury Type Number of Patients % of Total

Complete 25 30%

Incomplete 21 26%

Radiculopathy 11 13%

Transient 7 9%

Intact 18 22%

Table 2: Classification of Neurological Injury.

Figure 2: MRI of a dislocated cervical spine visualizes the cordcompression inferred from X-rays (Figure 1). This visualization isnot necessary to recognize the need for reduction.

Figure 3: Radiographs after closed reduction show that the spinalcanal is realigned and spinal cord compression reduced (Samepatient as in Figure 1).


CONCLUSION

Rapid closed reduction of cervicalspine subluxation injuries with tractionis safe and effective. In our series, closedreduction was successfully performedwithin a few hours of the injury.Although a disk injury documented byMRI was often associated with thesetraumatic cervical spine injuries, therisk of neurological deteriorationfollowing closed reduction was rare.

Figure 4: Post-reduction MRI is useful for assessing any residual cord compression.

Neurological Deficit Pre-Reduction Post-Reduction

Complete 9.8±2.9 16.9±3.6†

Incomplete 35.8±7.2 47.7±7.8‡

†Significant p=0.001‡Significant p=0.005

Table 3: Neurological recovery, improvement in ASIA motor scores.


Patients with spinal cord injuriescomprise a major portion of thetrauma victims with high

mortality and long term morbidity. Aswith all trauma victims, the majority ofthese patients are between the ages of16 and 32 years, resulting in adisproportionately high loss ofproductive years of life for both theseindividuals and for society. One thirdof these trauma victims have injuriesinvolving the cervical spine.Management of spinal cord injuryremains controversial. Recentlaboratory studies and clinical trialsindicate that early intervention may beassociated with improved results.Previous studies, however, havereported increased risk of neurologicdeterioration with early surgicalintervention. The optimal timing ofsurgical intervention in cervical spinalcord injuries has not been defined.

The goal of this study was to

investigate acute care complicationsassociated with surgery within threedays of cervical spinal cord injuryversus surgery more than three daysfollowing the injury. We specificallywanted to assess the rate of neurologicdeterioration with early surgery. Wealso wanted to determine if the rates ofother events such as pulmonarycomplications were different with earlyand delayed surgery.


We retrospectively reviewed allpatients undergoing surgicalmanagement for an acute cervicalspinal injury with neurologic deficit attwo institutions between March 1989and May 1991. Our study populationconsisted of 30 patients, 15 patientstreated at the State University of NewYork, Syracuse, New York, and 15patients treated at Harborview MedicalCenter, Seattle, Washington.

The goal of surgery was to restorespinal stability, relieve persistentneurologic compression, and allowmobilization of the patient. At the StateUniversity of New York, patients withneurologic spinal injuries had surgicalintervention usually within 72 hours ofinjury. At Harborview Medical Center,patients underwent immediate closedreduction; patients were subsequentlyobserved for neurologic status for 10 to14 days prior to undergoing surgicalstabilization. Postoperative treatmentwas similar at both institutions.

Neurologic status was recorded atinitial presentation, preoperatively, andat time of discharge from the hospitalor transfer to a rehabilitation unit. Inaddition to a neurologic exam,neurologic function was recorded asFrankel Grade and American SpinalInjury Association (ASIA) motor score.Glasgow Coma Score was determinedfor each patient at the time ofpresentation to the emergency room.Associated non-spinal injuries werealso recorded. Injury Severity Score wasretrospectively calculated from thisdata.

Statistical analysis was performedusing a two-sample Students t-testassuming unequal variances and chi-square analysis with the Yatescontinuity correction. Significance wasdefined by a p-value < 0.05.

RESULTS

The Injury Severity Score averaged24.8 for the early surgery group and26.2 for the late surgery group. TheGlasgow Coma Score averaged 14.4 forthe early surgery group and 14.1 for thelate surgery group. Neither differencewas statistically significant. Thedifference in the duration of ICU stayand mechanical ventilation was notsignificant between the two groups.Duration of acute care stay was longerin the late surgery group: 36.8 dayscompared to 21.9 days, consistent withthe longer pre-operative hospitalizationperiod for the patients in the latesurgery group.

Early Versus Delayed Surgery for Acute Cervical Spinal CordInjuryJENS R. CHAPMAN, M.D., SOHAIL K. MIRZA, M.D., WALTER F. KRENGEL III, M.D.,PAUL A. ANDERSON, M.D., JAMES C. BAILEY, M.D., M. SEAN GRADY, M.D., MARK A. KONODI, M.S.,AND HANSEN A. YUAN, M.D.

Figure 1: MRI of an 18 year old male with sudden tetraplegia following a football injury.


The improvement in the ASIAMotor score from pre-operativeassessment to post-operativeassessment was significant in the earlysurgery group but not in the latesurgery. The improvement in theFrankel Grade from the pre-operativelevel to the post-operative level wasagain significant in the early surgerygroup but not in the late surgery group.

The number of majorcomplications and minorcomplications was not statisticallydifferent between the two groups. Nopatient showed neurologicdeterioration during the course oftreatment.

DISCUSSION

The results of this study support theposition that patients who sustain acutetraumatic injuries of the cervical spinewith associated neurologic deficit maybenefit from surgical decompressionand stabilization within 72 hours ofinjury. Surgery within this early periodon patients with acute traumaticinjuries of the cervical spine was not

associated with increasedcomplications. Patients who underwentsurgery within this relatively early timeperiod showed a significantimprovement in neurologic statusduring their acute hospitalizationwhereas patients who underwentdelayed surgery did not show anysignificant change in neurologic status.

Our study raises the question thatfactors other than the timing ofmechanical decompression may alsoinfluence neurologic outcome.Specifically, an unstable spine protectedby traction and bed rest may sustainsome degree of continued abnormalmotion. This skeletal instability at theinjured segment may have a deleteriouseffect on the healing potential ofinjured neural tissue. Skeletalstabilization in spinal cord injury maysimilarly protect the neural tissue fromfurther injury and facilitate neurologicrecovery. We noticed a difference inneurologic recovery in our two smallpatient populations even though bothof these groups had early mechanicaldecompression by closed reduction.

The group of patients kept in tractionand bed rest for two weeks showed lessneural recovery.

Timing of intervention has beenshown to be important in laboratorymodels of spinal cord injury.Laboratory studies show evidence of awindow of opportunity in the initialfew hours following a spinal cord injurywhere treatment interventions may befruitful in minimizing the neurologicinsult

CONCLUSIONS

Surgery within 72 hours of injuryin patients sustaining acute cervicalspinal injuries with neurologicinvolvement is not associated with ahigher complication rate and earlysurgery does not result in neurologicdeterioration. Early surgery mayimprove neurologic recovery. Thisassessment requires more research.

Figure 2: Anterior discectomy and stabilization was performed within 3 hours of injury. The patienthad full neurologic recovery.


Several recent studies of traumaticspinal injury have explored therelationship between the

morphology of the spinal canal and theresulting neurological deficit.Researchers have investigated thisrelationship in two ways. The firstapproach focused on the variations ofnormal spinal anatomy that predisposeindividuals to increased risk of injuryor decreased potential for neurologicalrecovery. In these studies theinvestigators evaluated the risk ofneurologic injury measurements of theresidual space around the spinal cord.The second approach focused on theneurological consequences associatedwith spinal canal geometry changes

resulting from traumatic injury. Inthese studies the investigators indirectlymeasured spinal canal morphologyfrom the contour of displaced bonyfragments impinging on the cord.

Regardless of approach, previousstudies have represented the amount ofpre-existing stenosis or post-injurycanal compromise using measurementsof the bony canal rather than the spinalcord. The purpose of our study was todocument the geometry of both the softtissue components and bony elementsof the cervical spine following injury.This enabled us to test the hypothesisthat the severity of neurological deficitin acute cervical spine injuries is relatedto the degree of residual, post-

reduction deformation of the spinalcord.


Our study was a retrospectiveanalysis of 24 patients admitted toHarborview Medical Center, Seattle,Washington for acute cervical spinetrauma between January 1995 andFebruary 1997. We reviewed both themagnetic resonance imaging (MRI)scans of the cervical spine and medicalrecords of these patients. All patientshad T2-weighted MRIs andneurological evaluations based on theAmerican Spinal Injury Association(ASIA) motor scale. Ten patients hadneurologically complete spinal cordinjuries and 14 patients had incompleteinjuries.

We used a personal computer andflatbed scanner to capture images of thecervical spine from the MRI scans. Wescanned one mid-sagittal image from alateral MRI and seven axial images fromthe axial MRIs (one axial image at thelevel of injury and three each above andbelow this level). Once we captured acomplete set of images, we measuredseveral morphological variables fromthe digitized images using the imageanalysis program NIH Image. Table 1describes our initial measurements andFigure 1 shows those measurements inan axial MRI scan. From our initialmeasurements, we next determined aset of parameters describing spinal corddeformation. We normalized theseparameters to account for body-sizeand gender effects (Table 2).

For statistical analysis we used theStatView software package. Weperformed linear regression analysis onthe relationship between each of ournormalized spinal cord deformationparameters and the patient’s ASIAmotor scores. We also used non-parametric analysis to examine anydifferences in deformation parametersbetween the patients with complete andincomplete neurological deficits. Wedefined significance as a p-value < 0.05.

Relationship Between Neurological Deficit and Spinal CordDeformation in Cervical Spine InjuriesGREGG K. MOTONAGA, B.S., RANDAL P. CHING, PH.D., MARK A. KONODI, M.S.,AND SOHAIL K. MIRZA, M.D.

Measurement MRI Source Description

Length of Injury Sagittal Reconstruction Total longitudinal length ofdeformation to the thecal sac

C4 Body Height Sagittal Reconstruction Height of C4 vertebra measured at thecenter of the body

Thecal Sac Area Axial Scan Cross-sectional area of the spacewithin the boundaries of the thecalsac

Thecal Sac SagittalDiameter

Axial Scan Sagittal diameter of the thecal sac

Thecal Sac TransverseDiameter

Axial Scan Transverse diameter of the thecal sac

Spinal Cord Area Axial Scan Cross-sectional area of the spinal cord

Spinal Cord SagittalDiameter

Axial Scan Sagittal diameter of the spinal cord

Spinal Cord TransverseDiameter

Axial Scan Transverse diameter of the spinal cord

Parameter Description

Length of Deformation Length of injury divided by the body height of C4

Cord Area Ratio Spinal cord area at maximum deformation divided by thearea at uninjured levels

Cord Sagittal Diameter Ratio Spinal cord sagittal diameter divided by the sagittaldiameter at uninjured levels

Cord Transverse Diameter Ratio Spinal cord transverse diameter divided by the transversediameter at uninjured levels

Max Area Ratio Maximum value of spinal cord area divided by the thecalsac area along the length of the cervical spine

Max Sagittal Diameter Ratio Maximum value of spinal cord sagittal diameter dividedby the thecal sac sagittal diameter along the lengthof the cervical spine

Max Transverse Diameter Ratio Maximum value of spinal cord transverse diameterdivided by the thecal sac transverse diameter alongthe length of the cervical spine

Table 1: Initial Measurements From Cervical MRI Scans.

Table 2: Summary of Normalized Spinal Cord Deformation Parameters.


RESULTS

There was an inverse relationshipbetween each of our seven parameterswhich measured spinal corddeformation and the neurologicaldeficit. Larger spinal cord deformationswere associated with greaterneurological deficits, although none ofthe linear regressions had an R2-valuegreater than 0.50 (Table 3). Patientswith complete neurological deficitsdiffered from those with incompletedeficits with respect to length ofdeformation, cord area ratio, cordtransverse diameter ratio, maximumarea ratio, maximum sagittal diameterratio, and maximum transversediameter ratio. The cord sagittaldiameter ratio was the only normalizeddescriptor of deformation that did notdiffer between the two groups.

DISCUSSION

The regression analyses failed toshow a linear relationship betweenASIA motor scores and our measuresof spinal cord deformation. Ourpatients had spinal injuries at differentcervical levels. The motor scores couldpossibly have been affected by thedifferences in injury level. For example,upper cervical (C2-C3) injuries couldpresent with different degrees of motorimpairment than lower cervical (C6-C7) injuries. We were unable to accountfor this potentially confoundinginfluence on neurological deficit in ouranalysis.

Our non-parametric analysisindicated that six out of seven of ourspinal cord deformation measures weredifferent between the patients withcomplete neurological deficits andthose with incomplete deficits. A

Parameter R2 Value

Length of Deformation 0.46

Cord Area Ratio 0.28

Cord Sagittal Diameter Ratio 0.09

Cord Transverse Diameter Ratio 0.23

Max Area Ratio 0.31

Max Sagittal Diameter Ratio 0.42

Max Transverse Diameter Ratio 0.16

Table 3: Relationship Between Neurological Deficit and Deformation Parameters. Results of LinearRegression.

relationship existed betweenneurological deficit, measured by theASIA motor score and the spinal corddeformation, measured by ourdeformation parameters. In futureresearch we plan to study the predictivepower of our deformation measures indetermining potential post-reductionneurological deficit.

CONCLUSION

We found a relationship betweenthe degree of spinal cord deformation(post-reduction) and neurologicaldeficit in acute cervical spine injuries.We were able to identify six parametersfrom MRI scans that described thedeformation of the spinal cord and maybe useful in predicting whether patientshave complete or incomplete injuries.Further research will allow us to betterpredict the neurological outcome fromspinal cord deformationmeasurements.

RECOMMENDED READING

Eismont FJ, Clifford S, Goldberg M,Green B. Cervical sagittal spinal canalsize in spine injury. Spine 1984; 9:663-666.

Fontijne WPJ, De Klerk LWL,Braakman R, Stijnen T, Tanghe HLJ,Steenbeek R, van Linge B. CT scanprediction of neurological deficit inthoracolumbar burst fractures. J BoneJoint Surg 1992; 74B(5):683-685.

Hashimoto T, Kaneda K, Abumi K.Relationship between traumatic spinalcanal stenosis and neurologic deficits inthoracolumbar burst fractures. Spine1988; 13:1268-1272.

Matsuura P, Waters RL, Adkins RH,Rothman S, Gurbani N, Sie I.Comparison of computerizedtomography parameters of the cervicalspine in normal control subjects andspinal cord-injured patients. J BoneJoint Surg 1989; 71A:183-188.

Pavlov H, Torg JS, Robie B, Jahre C.Cervical spinal stenosis: Determinationwith vertebral body ratio method.Radiology 1987; 164:771-775.

Rasmussen PA, Rabin MH, Mann DC,Perl II JR, Lorenz MA Vrbos LA.Reduced transverse spinal areasecondary to burst fractures: Is there arelationship to neurologic injury? JNeurotrama 1994; 11:711-720.


Oral creatine supplementationhas become extremely popularamong adult and adolescent

athletes in an attempt to “get an edge”over the competition. Its popularitycomes despite limited informationregarding the safety of long termsupplementation and often conflictingdata regarding its effectiveness. Thisarticle reviews creatine’s biochemistry,potential benefits, side-effects, safety,and recommended dosage. Results ofa survey of common dosage and side-effects among division 1 college athleteswill be discussed.

OVERVIEW

Creatine is an organic compoundnormally found in skeletal muscle. Itcombines with phosphorus to formphosphocreatine (PC). PC serves as adonor of phosphate to make andregenerate ATP which is the energysource for contracting muscle. ATP isnormally generated for short intensityactivity by glycolosis or anaerobic(without oxygen) metabolism. Longerduration or aerobic activity is fueled byATP from aerobic metabolism. Beforeeither form of metabolism startsproducing ATP, in the first few secondsof exercise, energy is supplied byexisting ATP in the muscle.Theoretically, by increasing theconcentration of PC in the muscle, theamount of ATP to fuel initial exerciseis increased. A second theoreticalbenefit of increased PC in muscle is theability to regenerate ATP stores fasterbetween episodes of exercise.

Creatine is normally synthesized inthe body by the liver, kidneys, andpancreas at a rate of 1-2 g per day. About2 g per day is lost by the body throughthe kidneys. Dietary intake frommostly fish and meat is about 1 g perday but varies significantly. Mostindividuals store approximately 125mmole/kg (range 90-160 mmoles/kg)of creatine in skeletal muscles. Thosewith diets low in meat and fish tend tobe in the lower range. The maximumconcentration obtainable in skeletalmuscle is 160 mmoles/kg. Creatineconsumed after muscles are maximallyloaded is filtered by the kidneys and lost

in the urine.Creatine is produced synthetically

for supplementation. It is not currentlybanned by the NCAA or IOC. It isviewed by the FDA as a supplement andis therefore not regulated.

BENEFITS

In certain studies, subjects takingcreatine have shown improvement inperformance. The improvement isgenerally in the form of increased timeto exhaustion or increased poweroutput performing short burst, highintensity, repetitive activities. Most ofthe studies demonstrating positiveresults use a cycle ergometer and theimprovement is noted only in therepeated bouts of exercise.

It is hypothesized that creatine mayincrease the ability to gain strengththrough weight training, largely bydecreasing recovery time between sets.This is not proven.

Studies examining swimming andrunning performance are mixed at best.It is hypothesized that weigh gain (seeside effects) from creatinesupplementation may offset anypotential gains from small incrementalincreases in strength.

Studies examining performance inendurance exercise do not show anybenefit from creatine.

There are no studies assessing “onthe field” performance.

One issue that has caused confusionis that certain individuals, and certainstudy groups, show a benefit while

others do not despite similar testprotocols. It is hypothesized thatsubjects with little creatine in theirregular diet (i.e. vegetarians) maybenefit more than those with highdietary intake. This could account forsome of the mixed results.

SIDE EFFECTS

There is very little research on sideeffects. Weight gain is widelyrecognized which could be a good orbad effect depending on the sport. Theweight gain is likely secondary toincreases in both intra and extracellularwater. Studies have shown that the earlyweight gain is not from increased sizeor number of muscle cells.

Muscle cramps and strains arereported by athletes taking creatine.This evidence is anecdotal only andthere is no proof that creatine isresponsible. It is hypothesized thatincreases in intracellular water mayhave adverse physiologic or mechanicaleffects on the muscle cells resulting inincreased strains.

Side effects reported by 52 collegebaseball and football players over awinter conditioning season arereported in Table 1.

SAFETY

Studies on the short term safety ofcreatine have not demonstrated anysignificant irreversible changes in bloodchemistry, liver function, or kidneyfunction. The kidneys have to “workharder” filtering high concentrations of

Creatine Supplementation

JOHN O’KANE, M.D.

Total percentagemuscle cramps/strains 33%GI-diarrhea 31%GI-other 27%undesirable weight gain 13%dehydration/thirst 13%other 23%none 27%

Table 1: Side effects reported by 52 college baseball and football players over a winter conditioningseason are reported.


creatine particularly during loading,but there is no evidence thus far thatdamage is done. There is no clearpattern of adverse health effects withthe extensive short term experience.

There are no data to support or callinto question the long term safety ofcreatine. This is arguably the mostconcerning feature of creatinesupplementation.

Creatine is unregulated by the FDAin its production, distribution, or sale.Thus the quality control of theingredients is not specified.

DOSAGE

Currently a loading dose of 5 g, 4times a day for 5 days followed by 2 gper day has been recommended. At thisdosage the muscles ability to absorbcreatine is saturated. Taking a higherdose increases the cost and potential forside effects.

In the baseball and football survey,it was common for the recommendedmaintenance dosage to be exceeded.Overall, 75% of the athletes exceeded 5grams/day. Of those, 35% took morethen 9 grams/day and 6% took 17-20grams/day. The unnecessarily highdosages used by these athletes mayexplain the high incidence of side-effects reported in the survey.

CONCLUSION

Creatine may result in smallergogenic gains in some athletes doingspecific types of short burst, repetitiveactivities especially if used inconjunction with a training program.There are many questions that remainregarding long term safety, side-effectsand dosage. Because of the widespreaduse of creatine supplementation, it isimportant for health professionals tocontinue to seek more informationregarding creatine so that they mayprudently and accurately advise theirpatients.

RECOMMENDED READING

Juhn MS, Tarnopolski M. Oral creatinesupplementation and athleticperfomance: a critical review. Clin JSport Med 1998; 8(4): 286-297.

Juhn MS, Tarnopolski M. Potential sideeffects of oral creatinesupplementation: a critical review. ClinJ Sport Med 1998; 8(4): 298-304.

Stricker PR. Other ergogenic agents.Clinics in Sports Medicine, SportsPharmacology 1998; 17(2): 284-286.


N ormal articular cartilage,composed of chondrocytes,collagens, proteoglycans,

water, and non-collagenous matrixproteins, has unique biological andmaterial properties that enable it towithstand significant loads withoutapparent detrimental effect. Althoughthe maximal tolerable repetitive impactload is unknown, it is thought thatarticular cartilage can withstand asingle impact load of up to 25megapascals (MPa) withoutdemonstrable injury. Normal physicalactivities, including running andjumping, are not thought to exceed thisthreshold in normal joints. However,if this capacity is exceeded by repetitivesubmaximal loading over time or by asingle overloading event, damage to thejoint may result.

The relationship between articularcartilage and weight-bearing exercise iscomplex. While motion and moderateactivity is probably beneficial, certainathletic pursuits in general and manyactivities at an elite level may lead tooverload of articular cartilage andpredispose to osteoarthritis. Severalstudies have investigated therelationship between long distancerunning and osteoarthritis in bothhumans and animals. While some ofthe epidemiological results arecontradictory, analysis of the publisheddata as a whole suggests that elite levelrunning (high miles per week or highlifetime miles) leads to an increased riskof osteoarthritis. As such, the presentstudy was undertaken to assess theeffect of pre-marathon training(repetitive loading) and completion ofa competitive marathon (concentratedrepetitive load) on the breakdown/turnover of type II collagen, theprincipal collagen of articular cartilage,and type I collagen, the principalcollagen of bone.


Healthy, age-matched volunteerswithout known orthopaedic orrheumatologic disease were recruitedfrom the orthopaedic department of a

teaching hospital and divided into twogroups based upon their intention toparticipate (runners, N=7) or notparticipate (controls, N=6) in amarathon. All subjects were males withthe exception of one runner. Noparticipant had any history of articularinjury or known familial predispositionto osteoarthritis, although one runnerand one control had undergonearthroscopic partial medialmeniscectomy in the remote past. Allsubjects were residents, fellows, orgraduate students not takingmedication. An exercise and runninghistory was obtained from each. Bodymass indices were calculated for allparticipants and reported as kg/m2.

Second morning void urine sampleswere obtained at four weeks pre-raceand at one, four, eight, and fifteen dayspost race. Urinary levels of a cross-linked type II collagen C-telopeptidefragment were assayed with amonoclonal antibody by an enzyme-linked immunosorbent assay (ELISA)and reported as Col 2 CTx / mgcreatinine (2CTx). This experimentalassay utilizes an antibody thatrecognizes a proteolytic neoepitope oftype II collagen generated in vitro bymatrix metalloproteinases and isthought to have the potential to followcartilage degradation in degenerativejoint disorders. Type I collagenbreakdown was measured by acommercially-available urinary assayand reported as nM bone collagenequivalents / mM creatinine (NTx).The data were analyzed using repeatedmeasures analysis of variance(ANOVA) techniques.

RESULTS AND DISCUSSION

Runners ran an average estimated21 miles per week (range 10 - 40) in theyear preceding the marathon, whilecontrols ran an average estimated twomiles per week (range 0 - 12). Runnersalso had completed a significantlygreater number of previous marathonsthan the controls. All participants werenon-obese, height and weightproportionate with body mass indices

in the normal range. No differencebetween the average BMI of the runnersand the controls was noted. As a group(all time points combined), the runnershad increased 2CTx levels as comparedto the controls, and the difference wasstatistically significant (mean increase= 42%, p = .035). No difference in NTxlevels was noted (controls = 43.9 + 21.1;runners = 39.2 + 17.9 nM bone collagenequivalents / mM creatinine). Thedifference between 2CTx and NTxlevels as a function of time was notstatistically significant (p = 0.6).However, the difference in the 2CTxlevels between the runners and thecontrols at the eight day post race timepoint was large (p = 0.6, uncorrectedfor multiple comparisons).

The present pilot study is the firstto demonstrate increased turnover orbreakdown of cartilage type II collagenin trained athletes, as assessed by a novelmolecular method. Although thisdifference may be a result of an adaptiveor selective mechanism peculiar toathletes who choose to participate indistance running, this finding alsocould be a consequence of repetitivetrauma to the articular cartilage of thelower extremities and spine. If thislatter explanation of the current dataproves correct, it could have significantimplications for long distance runners.Interestingly, the urinary 2CTx (type IIcollagen breakdown product) values forrunners and controls in this study arevery similar to those previouslyreported for normal adults and adultswith arthritis, respectively. Given thenotoriously poor sensitivity ofradiographic and clinical parametersfor all but the most advanced stages ofosteoarthritis, a non-invasive markerfor early degradation of cartilage typeII collagen as an indicator of pre-clinical articular cartilage pathology ispotentially beneficial. Such a tool onlyfacilitate monitoring of the naturalhistory of active joints as well as theeffectiveness of medical management.In athletes, appropriate activityrestriction or modification of exercise/rehabilation protocols might have the

Increased Type II Collagen Turnover in Trained MarathonRunners Versus Age-Matched ControlsDANIEL J. STECHSCHULTE, JR., M.D., PH.D., LYNNE M. ATLEY, PH.D., JOHN M.WRIGHT, M.D.,RUSSELL F. WARREN, M.D., MARGARET G. E. PETERSON, PH.D., AND DAVID R. EYRE, PH.D.


potential to prevent or delay the onsetof osteoarthritis in at risk individuals,possibly prolonging their careers andenhancing retirement years.

RECOMMENDED READING

Buckwalter, J.A. and H.J. Mankin,Articular cartilage. Part II.Degeneration and osteoarthrosis,repair, regeneration, andtransplantation. J Bone Joint Surg,1997. 79A: p. 612-632.

Buckwalter, J.A. and N.E. Lane,Athletics and osteoarthritis. Am JSports Med, 1997. 25(6): p. 873-881.

Atley, L.M., Shao, P., Ochs, V., Shaffer,K., Eyre, D.R. Matrixmetalloproteinase-mediated release ofimmunoreactive telopeptides fromcartilage type II collagen. Trans OrthopRes Soc, 1998. 23: p. 850.


Though a significant percentageof the population has a foot thatcould be described as flat, only

a small number of these flatfeet havefunctional limitations or manifestclinical problems.(23,29,31) Whatdistinguishes painful flat foot from asymptom free flatfoot is unknown. It isknown that the shape of the bone isaltered in some hindfootdeformities.(2) Soft tissue injury orabnormality may also contribute to anacquired deformity. (4,11,17,21,24) Inaddition, patients with a symptomaticflatfoot are likely to have a diminishedarch on the contralateral side.(8) Thissuggests that there may be an anatomicshape that predisposes a foot to clinicaldeformity. However, there is nounifying objective method ofunderstanding different types of flatfeet. The use of two dimensionalradiographs has been sufficient neitherto distinguish the subtle differencesbetween flatfoot and normal foot, nor

between normal and abnormal flatfoot.Because of the suggestion on plane

radiographs of some of flatfeet that thesubtalar and talonavicular joints aresubluxed, the term ‘peritalarsubluxation’ has been coined todescribe this category of severesymptomatic flatfoot. Actualsubluxation has never beendocumented, however. If there is actualsubluxation, it may provide someinsight into the pathology and itstreatment.

This project was designed toquantitatively define the relationshipbetween the articular surfaces of thetalus and calcaneus in a simulatedweight bearing position in patients withsevere symptomatic flatfoot.

METHODS

Computed Tomography scans ofthe feet of eight patients withsymptomatic flatfoot were used toreconstruct a model of the talo

calcaneal articulation. The scans hadbeen performed on a custom loadingframe developed to simulate weightbearing in a neutral position with a 75Naxial compressive load. The digitalComputed Tomography image datawere used to make three dimensionalcomputer models of the articularsurfaces of the talus and calcaneus ofeach foot. These were used to calculatethe percentage of the articular surfacethat was in contact, and conversely, thepercentage subluxed. Two surfaces werebuilt for each bone. The posterior facetwas one surface and the anterior andmiddle facets were combined to formthe second surface. The data werecompared to scans from feet of fourcontrol patients without hindfootdeformity, using Mann-Whitney non-parametric analysis (Figure 1 and 2).

RESULTS

The contact was measured andreported individually from theperspective of both the calcaneus andtalus because the joint surfaces wereslightly different sizes. The posteriorfacet was analyzed alone while themiddle and anterior facets wereanalyzed as a single surface. In each casethe data are reported as the averageratio of the area of the joint in contactto the entire surface area. The area notin contact is considered subluxed(Figure 3 and 4).

Posterior FacetThe ratio of the part of the posterior

facet of the calcaneus in contact withthe talus to the entire area of the jointsurface was .68 +/- .092 in flatfeet and.92 +/- .022 in the control surfaces(p=0.0066). Alternatively, it could bestated that 32% of the surface ofposterior facet of the calcaneus issubluxed in the flatfeet and 8% issubluxed in the controls. Viewed formthe perspective of the posterior facet ofthe talar articular surface, the ratio ofcontact to entire surface area in flatfeetwas .69 +/- .116 versus 0.89 +/- .046 inthe controls (p=0.0415). Conversely, itcould be said that 31% of the talarposterior facet is subluxed in the study

Subluxation of the Talocalcaneal Joint in Symptomatic AdultFlatfootDHEERA ANANTHAKRISNAN, M.D., RANDAL P. CHING, PH.D., ALLAN F. TENCER, PH.D.,SIGVARD T. HANSEN, JR., M.D., AND BRUCE J. SANGEORZAN, MD.

Figure 1: The surface objects were modeled using computer-aided design (CAD) software (FORM Z,Autodessys Inc., Columbus Ohio) which was used to render the surface of each facet. This is amodel of a right control foot viewed in a CAD environment from posteriorly to display the curve ofthe joint surface. The dark gray surface on top is the talar joint surface. The lighter gray surface onthe bottom is the calcaneal surface. The middle surface (white) is an artificial surface that representsthe area of overlap on CT. Note that the saddle shaped posterior facet and the more planar middlefacet surfaces are nearly coincident.


population versus 11% in the controls.These differences were significant.

Anterior/middle Facet The ratio of contact of the anterior/

middle facet of the calcaneus to totaljoint surface was .51 (+/- .23) in theflatfeet versus .95 (+/- .06) in thecontrol group (p=0.0066). Alternatively

this could be stated that 49% of theposterior facet of the calcaneus issubluxed in the flatfeet while only 5%is subluxed in the normal group. Thetalar articular surface of the flatfeet hada contact ratio of .44 (+/- .21) versus.87 (+/-.08) in the control group(p=0.0066). This could alternatively be

stated the 56% of the talar middle facetsurface is subluxed in the flatfeet whileonly 13% is subluxed in the normalgroup. These differences weresignificant.

The stimulus for this investigationwas the sense that treatment ofsymptomatic flatfoot in adults is notalways based on clear objective data.Many treatments have been proposed,each of which may effect change at adifferent anatomic location. A possiblestarting point to evidence basedtreatment is to better define thepathological condition. In this case, weattempted to define the relationship ofthe major tarsal bones to one another.

CONCLUSIONS

There is greater subluxation of thetalo calcaneal joint during simulatedweight bearing in patients withsymptomatic flatfoot than there is inthe talo calcaneal joint in people with anormal arch. The subluxation is greaterin the anterior and middle facet that itis in the posterior facet. Though it doesnot preclude other causes, least onecomponent of the altered shape of asymptomatic flat foot is a the loss of theintergrity of the talocalcanealarticulation.

RECOMMENDED READING

Anderson, J.G., Harrington, R.E.,Ching, R.P., Tencer, A.F., Hansen, S.T.Jr., Sangeorzan, B.J.: Alterations in TalarMorphology Associated with Adultflatfoot. Foot and Ankle, 18:705-709,1997.

Cowan, D. N.; Robinson, J. R.; JonesPolly, D. W., Jr. and Berry, B.H.:Consistency of visual assessments ofarch height among clinicians. Foot andAnk. Int., 15:213-217, 1994.

Dyal, CM, Feder, J, Deland, JTThompson, FM Pes Planus in patientswith posterior tibial tendoninsufficiency: Asymptomatic VersusSymptomatic Foot. Foot and Ankle,18:85-88, 1997.

Evans, D. E.: Calcaneo-valgusdeformity. J. Bone and Joint Surgery,57-B:270-278, 1975.

Gadzag, A.R.; Cracchiolo, A. 3rd.:Rupture of the posterior tibial tendon.Evaluation of injury of the springligament and clinical assessment of

Figure 2: The rendered surface of a right control foot viewed as if looking down at the subtalar joint.The dark gray shade represents the surface of the talus, light gray, the calcaneus. Note that thesurfaces are aligned.

Figure 3: The rendered surface of a right flatfoot viewed as if looking down at the subtalar joint.The light shade represents the surface of the talus, the darker represents the articular surface ofthe calcaneus. Note that the calcaneal surface images are posterior and lateral to the talar surfaces.The middle/anterior facet surface is nearly dislocated.


Calc:Tal OL:Calc0.0

0.5

1.0

1.5Flatfoot

Control

Are

a R

atio

Middle Facet Posterior Face

**

* *

OL:Tal Calc:Tal OL:Calc OL

Figure 4: Graphic display of data. All data were expressed as a ratio to normalize for the foot size.A ratio of 1.0 would mean that one articular surface is perfectly coincident with the other. Calc/Tal= ratio of the surface area of the calcaneus to that of the talus; OL/Calc= ratio of the overlaparea (area in contact) over the total joint surface the calcaneus OL/Tal = overlap area of the surfaceover the total joint surface the talus. Subluxed area is the converse of the overlap area. The datafor the anterior and middle facet are displayed on the left , the posterior facet on the right. Therewere no significant differences in size between the surfaces of the flatfeet versus those of thenormal feet or between the surface of the talus versus that of the calcaneus. There were significantdifferences in subluxation in the both middle/anterior and posterior facets. Note also the very smallerror bars in the control feet versus the large bars in the flat feet.

tendon transfer and ligament repair. J.Bone and Joint Surg., 79A:675-81, 1997.

Jones, B. S.: Flatfoot: A preliminaryreport of an operation for severe cases.

Pedowitz, W. J. and Kovatis, P.: Flatfootin the adult. J. Am. Acad. Orthop. Surg.,3(5):293-302, 1995.


There is need for orthopaedicsurgeons to document thefunctional status of their

patients before and after treatment todetermine the efficacy of theirmanagement. Because the surgeon isan important variable in the result, it isimportant that surgeons know theirown results rather than relying onoutcomes reported in the literature –“The surgeon is the method”.

Some approaches to outcomedocumentation consume substantialtime and resources; others do notcapture the patients’ perception of theircomfort and function ; still others arenot sufficiently 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.

The Simple Knee Test (SKT) wasdeveloped as a tool to facilitate thedocumentation of treatmenteffectiveness based upon changes inself-assessed patient function. Its simplefunctional questions facilitatecommunication between patients andphysicians regarding pre and posttreatment status. The SKT is sensitive

to common knee pathologies and becompleted within a few minuteswithout assistance.

This study uses the SKT todemonstrate functional outcome of acommonly performed procedure: ACLreconstruction. Our goals were 1) todemonstrate the correlation of the SKTscore with that of the more traditionalLysholm score and 2) to use the SKT tocompare results of ACL reconstructedpatients with age-matched controls.


148 consecutive patients who hadan ACL reconstruction by a singlesurgeon (RVL) from September, 1989to September, 1991 were retrospectivelyreviewed at an average of 6 years aftersurgery. Each reconstruction utilized adouble-looped semitendinosus andgracilis autograft. The average age atsurgery was 29 years old (range 13-51).There were 95 males (64%) and 53females (36%). Each patient wascontacted via telephone and completeda Lysholm score and the Simple KneeTest.

The total number of SKT questionsthat each patient answered affirmativelyat follow-up was divided by 15 (the totalnumber of SKT questions) and thisratio was compared to the ratio of eachpatient’s follow-up Lysholm score

divided by 100 (the highest Lysholmscore possible). A correlation coefficientwas computed for these ratios.

Responses to each SKT questionwere then compared to responses by360 age and activity level-matchedcontrols. In addition, patient responseswere compared to those of an age andactivity level-matched group of knownACL-deficient patients prior toreconstruction. An unpaired t-test wasused to compute a p value for each SKTquestion. Significance was defined as p–value < 0.05.

RESULTS

109 of the original 148 could belocated for follow-up (74%). Fourpatients were eliminated from the studybecause they had required a subsequentACL reconstruction on the same knee.An additional patient was eliminatedbecause he had sustained a severe lateralplateau fracture.

The average patient with an ACLreconstruction could perform 13 of the15 Simple Knee Test questions atfollow-up. This contrasts to the typicalpatient prior to reconstruction whocould perform only 7 of the 15 SKTquestions. Normal control individualstypically can perform 15 of the 15 SKTquestions. (Table I)

The SKT results were highlycorrelated with the Lysholm Score(correlation coefficient 0.80 (p<.0001))(Figure 1).. The average Lysholm scoreof the reconstructed group was 92(range 63-100). However, despite goodLysholm scores, there were a significantnumber of patients who were not freefrom swelling, were not able to run,could not participate fully in their sportand could not kneel comfortably

DISCUSSION

Reported Lysholm scores of patientswho are more than two years after anACL reconstruction range between 84and 95. Our patients’ average Lysholmscore of 92 is comparable to other seriesof ACL reconstruction.

Although the results of the SKTcorrelated with the Lysholm Score, the

Correlation of the Simple Knee Test With the Lysholm Score forFollow-up of ACL ReconstructionMICHAEL H. METCALF, M.D., ROGER V. LARSON, M.D., PETER T. SIMONIAN, M.D.,AND FREDERICK␣ A. MATSEN III, M.D.

1 Does your knee allow you to sleep?2 Will your knee straighten fully?3 Will your knee bend all the way?4 Is your knee free from swelling?5 Does your knee move smoothly without catching or locking?6 Can you sit comfortably with your knee bent for 1 hour?7 Can you enter and leave a car without difficulty?8 Can you stand and walk without your knee giving away?9 Does your knee allow you to walk two blocks without difficulty?

1 0 Does your knee allow you to do your usual work?1 1 Will your knee allow you to go down stairs easily?1 2 Will your knee allow you to run?1 3 Will your knee allow you to participate fully in your sport?1 4 Will your knee allow you to kneel comfortably?1 5 Can you hop on one leg?

Table 1: The Simple Knee Test.


SKT provided a more specificfunctional analysis. Each of the 13 SKTquestion has previously been shown tobe statistically independent andtherefore focuses on an independentaspect of patient function. For example,in this study only 57% of patients couldcomfortably kneel on their knee. Theinability of 43% of the patients beingunable to kneel comfortably may berelated to the technique of graft fixationto the tibia. This potential problemcould not have been identified from theLysholm score.

Patient education andcommunication is facilitated by theSKT. Despite good Lysholm scores,there were a significant percentage ofpatients that were not free fromswelling, were not able to run, could notparticipate fully in their sport and couldnot kneel comfortably. Patients canunderstand functional data present inthis way, however few patients wouldunderstand what it means to have anaverage Lysholm score of 92.

In summary, the SKT has beenshown to be highly correlated to themore established Lysholm score. TheSKT has the additional advantage ofisolating specific areas of concern thatmay be overlooked by a singlenumerical score. Finally, the SKTfacilitates communication with patientsregarding the likely results of treatment.

RECOMMENDED READING

Metcalf MH, Larson RV, Simonian PT,Matsen FA, III. The Simple Knee Test:A quick and sensitive self-assessment ofknee function. University ofWashington Orthopaedic ResearchReport. 1998.

Arciero RA, Scoville CR, Snyder RJ,Uhorchak JM, Taylor DC, Huggard DJ.Single versus two-incision arthroscopicanterior cruciate ligamentreconstruction. Arthroscopy 1996;12:462-9.

Novak PJ, Bach BR, Jr., Hager CA.Clinical and functional outcome ofanterior cruciate ligamentreconstruction in the recreationalathlete over the age of 35. Am J KneeSurg 1996; 9:111-6.

Simonian PT, Harrison SD, Cooley VJ,Escabedo EM, Deneka DA, Larson RV.Assessment of morbidity ofsemitendinosus and gracilis tendonharvest for ACL reconstruction. Am JKnee Surg 1997; 10:54-9.

5

6

7

8

9

10

11

12

13

14

15

40 50 60 70 80 90 100 110

L y s h o l m

SK

T

Patient*

Figure 1: Graph of correlation between SKT results and the Lysholm score. (*Each point may representmore than one patient.)


The ulnar collateral ligament ofthe elbow (UCL) consists of ananterior bundle, a posterior

bundle and transverse fibers. Theanterior bundle has been shown to bethe primary restraint to valgus stress atthe elbow. The anterior bundle of theUCL is the most discrete of the threeportions of the ligament complex. It’smargins are readily distinguished fromthe surrounding joint capsule and it’sfibers are often intimately associatedwith the deep surface of the flexor mass.The posterior bundle consists of a lessdistinct fan shaped thickening of theposterior capsule. The transverse fibersoriginate and insert on the ulnacovering a bony depression on themedial portion of the trochlear notchand contribute little or nothing to thestability of the elbow.

The UCL originates from theanterior inferior surface of the humeralmedial epicondyle. The width of the

origin varies but in most cases occupiesthe middle two thirds to three quartersof the epicondyle in the coronal plane.This UCL origin is a relativelyreproducible area to find duringligament reconstruction whencompared to the osseous UCLinsertion. The anterior bundle insertson the medial border of the coronoidat the sublime tubercle.

The origin of the anterior bundle ofthe UCL lies slightly posterior to therotational center of the elbow. Theanterior bundle is further divided intoan anterior band and a posterior band.The eccentric origin of thesecomponents of the anterior bundle inrelation to the rotational center throughthe trochlea creates a cam effect duringflexion and extension. The anteriorband tightens during extension andposterior band tightens during flexion.This reciprocal tightening of the twofunctional components of the anterior

bundle allows the ligament to remaintaught throughout the full range offlexion despite it’s eccentric origin inrelation to the rotational center of theelbow.

This complex anatomy andinteraction of the bands of the anteriorbundle of the UCL may limit thesuccess of achieving a successfulisometric reconstruction with a singlegraft. Current UCL reconstructiontechniques describe a single humeraltunnel to emulate the origin and twoseparate ulnar tunnels to emulate theinsertion. The purpose of this studywas to attempt to find the optimalposition or positions for reconstructionof the ulnar insertion of the UCL.


Five fresh, cadaveric elbowspecimens were used for this study.Displacements during elbow range ofmotion were measured with a liquid

Elbow Medial Collateral Ligament Isometry: Evaluation of theUlna Insertion SitePETER T. SIMONIAN, M.D.

0 deg 20 deg 40 deg 60 deg 80 deg 100 deg 120 deg 140 deg Mean

Position A

(stnd dev)

0 mm 0.45 mm

(0.48 mm)

1.47 mm

(0.94 mm)

2.36 mm

(1.02 mm)

2.77 mm

(1.19 mm)

2.89 mm

(1.45 mm)

2.85 mm

(1.48 mm)

2.34 mm

(1.61 mm)

1.89 mm

Position B

(stnd dev)

0 mm 1.58 mm

(2.17 mm)

2.56 mm

(2.77 mm)

3.58 mm

(3.39 mm)

4.76 mm

(4.26 mm)

5.41 mm

(5.00 mm)

5.44 mm

(5.34 mm)

5.06 mm

(5.17 mm)

3.55 mm

Position C

(stnd dev)

0 mm 1.05 mm

(1.07 mm)

2.21 mm

(1.27 mm)

3.55 mm

(1.50 mm)

4.73 mm

(2.13 mm)

5.46 mm

(2.64 mm)

5.57 mm

(3.00 mm)

5.40 mm

(2.70 mm)

3.49 mm

Position D

(stnd dev)

0 mm 1.08 mm

(0.54 mm)

2.10 mm

(0.57 mm)

2.84 mm

(0.87 mm)

3.53 mm

(0.79 mm)

3.65 mm

(1.14 mm)

3.37 mm

(1.29 mm)

3.01 mm

(1.44 mm)

2.45 mm

Position E

(stnd dev)

0 mm 1.92 mm

(1.41 mm)

3.34 mm

(1.77 mm)

4.58 mm

(1.76 mm)

5.38 mm

(1.81 mm)

5.60 mm

(2.24 mm)

5.31 mm

(2.43 mm)

4.79 mm

(2.49 mm)

3.86 mm

Position F

(stnd dev)

0 mm 1.08 mm

(0.76 mm)

2.74 mm

(0.80 mm)

4.65 mm

(1.30 mm)

6.16 mm

(1.27 mm)

7.35 mm

(2.04 mm)

7.23 mm

(2.31 mm)

7.39 mm

(2.91 mm)

4.58 mm

Position G

(stnd dev)

0 mm 1.43 mm

(0.84 mm)

2.67 mm

(1.20 mm)

3.87 mm

(1.43 mm)

4.58 mm

(1.53 mm)

4.31 mm

(1.27 mm)

3.52 mm

(0.96 mm)

2.79 mm

(1.09 mm)

2.90 mm

Position H

(stnd dev)

0 mm 1.08 mm

(0.26 mm)

2.96 mm

(1.05 mm)

5.14 mm

(1.66 mm)

6.58 mm

(1.76 mm)

6.82 mm

(1.87 mm)

6.28 mm

(1.68 mm)

5.34 mm

(1.85 mm)

4.28 mm

Position I

(stnd dev)

0 mm 1.50mm

(0.99 mm)

3.86 mm

(1.94 mm)

6.25 mm

(2.34 mm)

9.08 mm

(3.78 mm)

9.74 mm

(3.81 mm)

9.97 mm

(4.06 mm)

9.41 mm

(3.62 mm)

6.23 mm

Table 1: Average displacements and standard deviations measured at the nine various points on the ulna at various degrees of elbow flexion.


mercury strain gauge. The gauge wasfixed to the humeral origin of themedial collateral ligament with astraight needle drilled into the originat the epicondyle. The gauge wassutured to the eyelet of the needle. Theother end of the gauge was sutured tothe eyelet of a second straight needle.Nine drill holes spaced 5 mm apart werethen made in the ulna with a drill ofthe same diameter as the straightneedle. The central hole (position E)was drilled in the center of the sublimetubercle of the proximal ulna. The eightother positions (positions A-D and F-I) were drilled around the centralposition E. Positions A-C wereproximal, position D was directlyanterior and position F was directlyposterior to the central point E.Positions G-I were most distal.Positions A, D, and G were anterior,positions B, E, and H were central inthe anteroposterior plane, andpositions C, F, and I were posterior. Theneedle with the mounted strain gaugewas then placed into each of the drilledulna insertion positions (Figure 1). Foreach of these nine insertion points,strain gauge measurements wererecorded at 20 degree intervals as the

elbow was moved from 0 to 140 degreesof flexion. A total of 72 measurementswas recorded for each specimen. Thestrain gauge was zeroed at 0 degrees offlexion each time the strain gauge wasmoved to a new position on the ulna.

RESULTS

The degree of excursion betweenthe fixed origin and multiple insertionsites studied increased up to 100 or 120degrees before decreasing with furtherflexion. The degree of excursion wasthe least for the most proximal andanterior points studied.

The following data averaged theexcursion for all degrees of elbowflexion tested for each position on theulna (table 1, graph 1). The mostanterior positions (A,D,G) had the leastaverage excursion through the range ofmotion. Point A, the most proximal ofthese anterior locations, had the leastmean excursion at 1.89 mm. PositionI, the most distal and posterior position,had the greatest averaged excursion at6.23 mm, more than three times thatof position A. Interestingly, position E,the central point, had a mid-rangeamount of average excursion at 3.86mm.

DISCUSSION

The spectrum of UCL injuries canbe managed conservatively in non-throwing athletes, allowing even elitelevel participation in most sports. Theoverhead athlete represents anexception. The extreme valgus loadsgenerated by the throwing motion placetremendous stress across the medialjoint. Subtle injury to the ulnarcollateral ligament can lead to disablinginstability in the throwing athlete in thesetting of repetitive valgus stress.Reconstruction is indicated for those inhigh demand upper extremity sportssuch as baseball pitching, javelinthrowing and serving sports. Overheadathletes with medial elbow pain or asense of movement in the elbow withvalgus loads during sports participationwho also have UCL laxity on physicalexam may be candidates for ligamentreconstruction.

Sixty eight percent of overheadathletes who underwent reconstructionin one report were able to return totheir previous level of participation.Those with previous operations on theelbow had a lower chance of returningto their previous level of participation.A possible factor for the limited successof this surgery may be related to thedifficulty in finding a single isometriculnar insertion point for the complexanterior band of the UCL.

The anterior bundle of the ulnarcollateral ligament has been shown tobe the primary restraint to valgus stress.The anterior bundle is further dividedinto an anterior band and a posteriorband. The eccentric origin of thesecomponents of the anterior bundle inrelation to the rotational center throughthe trochlea creates a cam effect duringflexion and extension with the anteriorband tightening during extension andposterior band tightening duringflexion. This action of reciprocaltightening of the two functionalcomponents of the anterior bundle hasbeen thought to keep the ligamenttaught throughout the full range offlexion despite it’s eccentric origin inrelation to the rotational center of theelbow. However, the results of thepresent study would suggest thatcurrent reconstruction techniquesusing a common tunnel for UCL originand any combination of ulnar insertionsites tested do not achieve the reciprocalstabilizing action of the native ligament.

Graft fixation for UCL

Figure 1: Isometry displacement during elbow range of motion was measured with a liquid mercurystrain gauge. The gauge was fixed to the humeral origin of the medial collateral ligament with astraight needle drilled into the origin (arrow). The gauge was sutured to the eyelet of the needle.The other end of the gauge was sutured to the eyelet of second straight needle. Nine drill holesspaced 5 mm apart were then made in the ulna with a drill the same diameter as the straight needle.The central hole (position E) was drilled in the center of the sublime tubercle of the proximal ulna(where the gauge is shown in the picture). The eight other positions (positions A-D and F-I) weredrilled around the central position E. Positions A-C were proximal, position D was directly anteriorand position F was directly posterior to the central point E. Positions G-I were most distal. PositionsA, D, and G were anterior, positions B, E, and H were central in the anteroposterior plane, andpositions C, F, and I were posterior.


reconstruction is achieved throughbone tunnels in the ulna and medialepicondyle. The technique originallydescribed by Jobe places convergentdrill holes at the anterior and posteriormargins of the sublime tubercle of theulna separated by a 1 cm bone bridge.The hard cortical surface of the tubercleprovides an excellent bone bridge forgraft fixation. For graft fixation in themedial epicondyle, Jobe originallydescribed fashioning divergent drillholes from a single entry hole on theanterior distal epicondyle. Based on theresults of this study, a more optimalposition for these two drill holes shouldbe as anterior and proximal relative tothe sublime tubercle as possible.Another option that might improve theemulation of the reconstructed UCLwhen compared to the native UCLmight be a two tunnel origin as well asinsertion; however, this technique wasnot addressed in the present study.

No insertion point tested on theulna demonstrated isometry. Thedegree of excursion between the fixedorigin and multiple insertion sitesstudied increased up to 100 or 120degrees before decreasing with furtherflexion. The degree of excursion wasthe least for the most proximal andanterior points studied.

0

1

2

3

4

5

6

7

A B C D E F G H I

Location

Me

an

E

xcu

rsio

n

(mm

)

Graph 1: The mean strain gauge excursion at eight different degrees of elbow flexion for each ulna location tested. The most anterior positions (A,D,G)had the least averaged motion through an elbow range of motion. Point A, the most proximal of these anterior locations, had the least mean excursion at1.89 mm. Point I, the most distal and posterior position, had the greatest averaged excursion at 6.23 mm, more than three times that of point A. Point E,the central point, had a mid-range amount of averaged excursion at 3.86 mm.

The most anterior positions(A,D,G) had the least average excursionthrough an elbow range of motion.Point A, the most proximal of theseanterior locations, had the least meanexcursion at 1.89 mm. Point I, the mostdistal and posterior position, had thegreatest averaged excursion at 6.23 mm,more than three times that of point A.Point E, the central point, had a mid-range amount of averaged excursion at3.86 mm.

The complex anatomy of theanterior band of the UCL may limit thesuccess of achieving a successfulisometric reconstruction. The optimalposition for reconstruction of theinsertion of the UCL with a singleorigin from the medial epicondyle ofthe humerus was 5 mm anterior and 5mm proximal to the sublime tubercleof the uln

RECOMMENDED READING

Morrey, B. F., and An, K. N.: Articularand ligamentous contributions to thestability of the elbow joint. Am J SportsMed, 11: 315-19, 1983.

Morrey, B. F., Tanaka, S., and An, K. N.:Valgus stability of the elbow: Adefinition of primary and secondary

constraints. Clin Orthop, 265: 1991.

O’Driscoll, S. W., Jaloszynski, R.,Morrey, B. F., and An, K. N.: Origin ofthe medial ulnar collateral ligament. JHand Surg, 17A: 1992.

Schwab, G. H., Bennett, J. B., Woods,G. W., and Tullos, H. S.: The role of themedial collateral ligament. Clin Orthop,146: 42-52, 1980.

Andrews, J. A., and Whiteside, J. A.:Common elbow problems in theathlete. J Orthop Sports Phys Ther, 17:289-95, 1993.

Conway, J. E., Jobe, F. W., Glousman, R.E., and Pink, M.: Medial instability ofthe elbow in throwing athletes.Treatment by repair or reconstructionof the ulnar collateral ligament. J BoneJoint Surg, 74A: 67-83, 1992.

Jobe, F. W., Stark, H., and Lombardo, S.J.: Reconstruction of the ulnar collateralligament in athletes. J Bone Joint Surg,68A: 1158-63, 1986.

Wilson, F. D., Andrews, J. R., Blackburn,T. A., and McClusky, G.: Valgusextension overload in the pitchingelbow. Am J Sports Med, 11: 83-88,1983.


A ccurate measurements ofarticular cartilage contour andthickness are a valuable adjunct

to cartilage research. For example,indentation testing to determinecartilage compliance requiresknowledge of cartilage thickness;measurements of cartilage swelling canserve as a parameter of degradation.Techniques used to measure cartilagethickness include histology, ultrasound,needle probe, optical examination andmagnetic resonance imaging. Each ofthese techniques has specific advantagesand limitations. Here we describe a newmethod which has potentially betteraccuracy and versatility compared tothose techniques in current use.

METHOD

Five pairs of tibial plateaus wereremoved from 5 adult New Zealand

white rabbits after sacrifice followingour IACUC guidelines. Three 0-80machine screws were placed throughthe plateau into bone using tapped drillholes for rigid fixation. A cast of eachplateau was immediately made usinghydrophilic silicone dental castingcompound (Aquasil, Dentsply, Milford,DE). The material sets within 6 min.with maximal linear dimensionalchange of 0.03% and yields 20 microndetail reproduction. Each cast wasmounted on a precision, two axismotorized table (Parker) and scannedusing a laser displacement sensor(Keyence, model LK-081, Lake Oswego,OR) at a frequency of 2000 acquisitionsper mm. The sensor measures distanceswith a nominal resolution of 3 microns.The table motors and data acquisitionsequence were run through a 100kHz16 bit board (National InstrumentsPCI-MIO-16XE-10) using Labviewsoftware and a Macintosh PowerMac8500. Measurements were taken in agrid covering the entire plateauincluding both condyles. Each scanrequired 15 min. to complete, and thefinal data sets were decimated to 0.1mm spacing between sampled points,yielding a grid of 300 x 300 measuredheights for each tibial surface. Then,all the plateaus were immersed for 48hin a 400 IU/100 ml papain solution atpH 7 and 38°C, completely removingall the articular cartilage down to thetidemark. A second grid of distancemeasurements was made from casts of

the treated plateaus.Using the screw impressions as

fiducial landmarks that remainedunaltered in the pre-digestion and post-digestion impressions, the pre- andpost-digestion data sets weresuperimposed. This step requiredalignment of the two data sets in threetranslational and three rotationaldegrees of freedom. The requiredrotations and translations weredetermined by numerically minimizingthe root-mean-square differencebetween the two data sets within 1.5mm diameter discs centered on eachscrew impression. Repeated scans ofthe same impression, with theimpression mounted in varyingalignment on the scan table, showedthat the software adjusted thealignment to within an root-mean-square surface error of 40 microns orless. All numerical analysis wasconducted in Mathematica 3.0, usingMathematica’s built-in two-dimensional interpolating functions,and the resulting subtracted surfaceswere exported as digital files to thepublic domain program NIH Image.

Soft tissue thickness over the plateauwas calculated as the difference betweenthe pre- and post-digested data sets.The five pairs of knees were evaluatedby making right-left comparisons ofjoint contour and cartilage thicknessalong a standardized line across theplateau (fig. 1). Maximal cartilagethickness in the medial and lateral

A New Technique for Mapping Articular Cartilage Contour andThicknessANTHONY G. NORMAN, B.S., WILLIAM M. DOUGHTERY, PH.D., HOWARD CHANSKY, M.D.,PETER SIMONIAN, M.D., JOHN M. CLARK, M.D., PH.D., AND JOHN A. SIDLES, PH.D.

Figure One. Maps of the plateau casts in whichdensity corresponds to surface elevation (A)and cartilage thickness (B). Screws (S) are usedto align pre- and post digestion scans. ACL(A) is visible.

Figure Two. Plot of plateau surface contour along line P in fig. 1A. Eminences (E) are visible ashighest points.


plateau was also calculated.

RESULTS

This method produced maps ofcartilage surface contour and thicknessover the entire plateau. The data couldbe viewed en face as a map in whichelevation or thickness was representedby density (fig. 1). The joint contouror cartilage thickness in any specifiedarea could then be plotted (figs. 2, 3).Maximum thickness was 1.50 ± .19 mmon the medial plateau and 1.1±.27 mmon the lateral. Right to left variationwas small ( 1%).

DISCUSSION

The method presented here isunique in its ability to produce detailedmaps of articular cartilage surfaceanatomy and thickness. Such maps areespecially helpful when comparing onejoint to another, or when looking forfocal areas of cartilage alteration. Themaximum thickness values herecompare closely to those reportedelsewhere for the rabbit; no previoushigh-accuracy maps of global cartilagethickness have been reported. Pendingproof of their accuracy, thesetechniques will have many applicationsin orthopaedic research relating tocartilage anatomy and response toinjury.

The distance measurementsreported here have three main errorsources: the ± 3 micron accuracy of thesensor; the 0.03% deformation of thecasting material; and numerical errorresulting from superimposition of thepre- and post-digestion scans. By usingcasts, we eliminated optical error dueto translucency of the cartilage, at theprice of introduced error due to

variation in casting. We anticipate that,with careful casting technique, 10micron absolute accuracy may beobtainable over an entire tibial plateau.Direct scanning of the cartilage ispossible if the surface of the cartilage iscovered with an opaque powder, andthis alternative technique wouldeliminate the need for casting.

RECOMMENDED READING

Jukka, et al, J. Biomech. 28:32

Clark et al, JBJS 72B:866

Clark, et al, J Anat 191:215.

Figure Three. Plot of cartilage thickness along line P in fig. 1B. Thickest cartilage is medial adjacentto eminence.


The standard experimental modelfor articular cartilage injury is adrill hole made through the

articular surface. Such superficialdefects of cartilage do not healspontaneously1, leading to speculationthat properties of the synovial fluidinterfere with healing. In this study, weexamine the repair response in cartilagedefects with and without an intactoverlying surface layer. Our hypothesiswas that surface integrity influences thehealing response.

METHODS

Using a drill guide and fluoroscopy,1.5 mm diameter drill holes were madein the articular cartilage of the centralpatella of one knee, (randomized rightor left), in 24 adult New Zealand whiterabbits.

Six animals were in group A, and 9in groups B and C. In 2 animals of eachgroup, the hole in the cortex wasoccluded with wax (bw in figure 1b) toblock periosteal invasion. Animals weresacrificed at 3, 4 or 6 weeks post surgeryand the patellae were fixed in 2%glutaraldehyde, decalcified in EDTA,embedded in epoxy resin and sectionedfor light and electron microscopy.Animal care was approved by ourIACUC.

RESULTS

New bone formation and osteoidwere observed at the unobstructedexternal opening of all drill holes, wherean extensive, periosteum-based bloodsupply entered. When the hole wasoccluded by bone wax, new boneformation was impeded, but cartilage

Importance of Surface Integrity to Healing of Articular CartilageDefectsJULIE SWITZER, M.D., ANTHONY G. NORMAN, B.S., AND JOHN M. CLARK, M.D., PH.D.

repair was not. However, by 6 weeks,all defects in cartilage which did nottransgress the surface (figure 2a) werefilled in by a combination of swellingof surrounding cartilage and formationof new, chondroid tissue containingboth round cells in lacunae andabundant metachromatic groundfibrous substance. This tissue washistologically distinct from the morefibrous cartilage associated with newbone formation, and never formed ifthe hole did not penetrate into articularcartilage. If a hole passed through thearticular surface, the surface did notheal in the period examined.(figure 2b)

DISCUSSION

This experiment demonstrates thatdrill hole defects in articular cartilagepromptly fill with hyaline cartilagewhen the overlying surface layer isintact. These reverse drill holescomplement the typical defects createdby drilling through the surface whichproduce small volumes of fibrous tissuecompared to the abundant hyalinetissue seen in our closed lesions in thesame time frame. Surface integrityappears to influence the healingresponse. Longer term follow-up anddetermination of collagen types arenecessary to demonstrate thecompetency of the repair tissue.

ACKNOWLEDGMENT

Drills were supplied by Synthes(USA)

RECOMMENDED READING

Caplan et al CORR, 342:255, 1997

Figure One: The holes were made through the anterior cortex of the patella, passed through thebody and ended in one of three locations illustrated in figure 1: A) In the subchondral bone (b); B)In the articular cartilage (ac); C) In the synovial space, having passed completely though the cartilage.


Figure Two: Lesions at 6 weeks. a) Drill hole (dh) into cartilage. Defect has filled with hyalinecartilage; only sign of injury is absence of calcified layer. New bone (nb) is forming in hole. b) Drillhole through surface (s). Surface remains disrupted. Undermined articular cartilage (ac) hasherniated into hole.


N C

C

C

N

N

N

α2(I) QYDGKGVGN

α1(I) (Y)DEKSTGGN

OH

OH(H)

K

+

NTx

N

α1(I) EKAHDGGRC

α1(I) EKAHDGGRC

OH

OH(H)

K

+

CTx

A specific cartilage degradationmarker has potential value inchondroprotective drug

development and clinical managementof arthritis patients.

INTRODUCTION

There is a need for minimallyinvasive biochemical assays that canassess the rate of cartilage degradationin patients with degenerative jointdiseases. The collagen framework ofcartilage turns over extremely slowly inthe normal adult and its grossdegradation in articular cartilage is

believed to be a critical, irreversibleevent in osteoarthritis. A degradationassay for cartilage collagen isparticularly desirable. We describe anadvance in developing an immunoassaydesigned to measure pyridinoline-cross-linked telopeptides from type IIcollagen in human urine.


Monoclonal antibodies (mAbs)were raised in mice to a peptide,EKGPDP, the specific sequence in theC-telopeptide domain of human typeII collagen, that contains a cross-linking

Collagen Type II Cross-Linked Telopeptides: A Promising Markerof Cartilage Degradation in ArthritisLYNNE M. ATLEY, PH.D., PING SHAO, B.S., KATHY SHAFFER, B.S.*, VINCE OCHS, B.S.*, J.D. CLEMENS*,AND DAVID R. EYRE, PH.D.

lysine. An antibody, mAb 2B4, showedhigh binding affinity to the syntheticpeptide, to protease digests of cartilagecollagen and to pyridinoline-containing cross-linked C-telopeptidefragments isolated from human urinethat embodied this core sequence. Amicrotiter-plate ELISA was developedthat could quantify the natural antigen(col2CTx) in cartilage explant culturemedia (1), urine (2), serum andsynovial fluid. Urines were comparedfrom growing children, normal adultsand arthritic patients, expressing resultsas ng of cross-linked EKGPDP/mg

Figure 1: This figure depicts the two molecular sites of cross-linking in type II collagen and shows the C-telopeptide core sequence recognized by mAb 2B4 asa cross-linked degradation product.


creatinine. Recombinant preparationsof metalloproteinases (MMPs) andother proteinases were tested for theirability to generate the colCTx antigenfrom human type II collagen. MMPswere generously supplied by Dr. G.Murphy, Strangeways Research Lab.,Cambridge, UK and Dr. R. Martin,Roche BioScience, Palo Alto, CA.

RESULTS AND DISCUSSION

Fig. 1, depicting the two molecularsites of cross-linking in type II collagen,shows the C-telopeptide core sequencerecognized by mAb 2B4 as a cross-linked degradation product. Thispyridinoline-based fragment was themajor form of 2B4-immunoreactiveantigen isolated from human urine.The profile of such immunoreactivepeptides of low molecular weight (<2kDa) was similar in urine of childrenand adults. The excretion rate wasabout 100-fold higher in children thanin normal adults consistent with theactive degradation of growth platecollagen in children. Initialcomparisons with arthritis patientsindicate a wide range of individualvalues and a significantly higherexcretion than in control subjects.

Collaborative studies are ongoing tocorrelate levels with clinicalmeasurements of disease activity. The2B4 epitope proved to be an MMP-generated proteolytic neoepitope oftype II collagen degradation. All theMMPs tested (MMP1, 3, 7, 8, 13 and14) were active but matrilysin was themost potent. Using synthetic peptidesas substrates, it was shown that cleavageof the P/L bond was essential for 2B4epitope recognition (illustrated in Fig.2).

REFERENCES

Atley et al. Arth Rheum 40:584, 1997

Eyre et al. J Bone Miner Res 11:S413,1996.

3/4 1/4

Triple-helicaldomain

Collagenase cleavage-site

N C

C

C

N

N

Cross-linkedC-Telopeptide

neoepitope(col2CTx)

-EKGPDP-EKGPDP

X

LQYMRALQYMRA

MMPs

Figure 2: Using synthetic peptides as substrates, cleavage of the P/L bond was shown to be essential for 2B4 epitope recognition.

*Ostex International , Inc., Seattle,WA


Chondrosarcomas are malignanttumors of cartilage that mayarise de novo or within a

preexisting benign cartilage tumor,from periosteal or perichondral tissues,or within other soft tissues. They arethe second most common primarymalignant neoplasm of bone in adults,with a mean age at diagnosis of fortyyears. High grade chondrosarcomashave a relatively poor prognosis and allgrades of tumors have significant ratesof local recurrence. Attempts to devisehistological or biochemical criteria forthe classification of chondrosarcomashave met with limited success. The lackof a reliable grading system hinders thedevelopment of improved treatmentprotocols and the ability to providepatients with adequate prognoses.

The cellular origin ofchondrosarcomas and the mechanismof tumorigenesis are unknown. The

development and progression ofchondrosarcoma is likely a multi-stepprocess in which multiple geneticalterations accumulate in a cell. p53 isa tumor suppressor gene located onchromosome 17 that is inactivated intumors of many types; loss of p53activity is thought to be a primary eventin the development of manymalignancies including sarcomas.

Study of the development ofchondrosarcomas has been hamperedby the lack of suitable stable models thatcan be evaluated on a molecular level.In order to create such a model wemaintained cells derived from humantumors in long-term culture andestablished several chondrosarcoma celllines. This report summaries some ofstudies conducted in our Departmentseeking to characterize the extracellularmatrix phenotype of one of these celllines, Ch-1, as well as the expression of

molecules potentially involved in theregulation of cell differentiation andproliferation.

RESULTS

The chondrosarcoma cell line, Ch-1, has a unique phenotype that sharesimportant characteristics with thephenotype of normal chondrocytessuch as abundant production of thelarge proteoglycan aggrecan and thesmall proteoglycan biglycan and theexpression of types IX and XI collagens.Cytogenetic analysis of cell culturesfrom different passages found aremarkably consistent karyotype whichis consistent with the presence of asingle cell type.

Messenger RNAs encoding tworecently discovered cartilage-derivedmolecules, Cart-1 and CD-RAP, are alsosynthesized by Ch-1. Cart-1 is ahomeobox gene that may be involvedin gene regulation during cartilagedifferentiation; it is expressed in pre-chondrogenic mesenchyme and inchondroblasts. CD-RAP is a smallsecreted protein expressed in cartilageat all stages of maturation and in theSwarm rat chondrosarcoma. CD-RAPis also synthesized by melanoma celllines and is thought to be involved incell-matrix interactions and regulationof DNA synthesis.

In contrast to what occurs inchondroblasts and maturechondrocytes, Ch-1 cells express nodetectable type II procollagen mRNAand consequently no type II collagen issecreted into the extracellular matrix.The synthesis of type II procollagen isdifficult to maintain in long-term cellculture of chondrocytes andtransformed chondrocytes. In normalchondrocytes there is evidence thatexpression of type II collagen is stronglyregulated as opposed to expression ofaggrecan which appears to beconstitutive This pattern of collagenregulation is consistent with thephenotype of Ch-1, but as with any lineof transformed cells such changes maybe the result of genomic instability andmultiple random mutations. The

Expression of Cartilage Extracellular Matrix and PotentialRegulatory Genes in a New Human Chondrosarcoma Cell LineHOWARD CHANSKY, M.D., JAMES R. ROBBIN, WENDY H. RASKIND, STEVEN CHA, ERNEST U. CONRAD, M.D,JOHN M. CLARK, M.D., JAMES D. BRUCKNER, M.D., AND LINDA J. SANDELL, M.D.

Figure 1: Analysis of p53 in Ch-1. A: Loss of heterozygosity. Autoradiograph of DNA amplified bypolymerase chain reaction at the p53 pentanucleotide alu sequence repeat in the first intron. B:Autoradiograph of the sequencing gel showing mutation in cDna amplified from the mRNA of Ch-1p53. Mutation is indicated by the arrow. C: DNA and protein sequence of normal and mutant p53.


proteoglycan decorin was not detectedin Ch-1. Decorin is generallyconsidered to interact with the fibrillarcollagens, types I and II. As little or notype I or type II collagens are made,there is no production of a fibrillarextracellular matrix; consequently it ispossible that the lack of decorinsynthesis is related to the lack of fibrillarcollagen synthesis. The lack of type Icollagen mRNA synthesis also indicatesthat these cells are not“dedifferentiated” chondrocytes whichswitch their fibrillar collagen synthesisfrom type II to type I.

It has been reported that unculturedhigh grade chondrosarcomassynthesize lower levels of type IIcollagen than do tumors of lower grade.Kawashima et al showed an inversecorrelation between expression of typeII collagen and the ability of welldifferentiated chondrosarcomas togrow faster and more extensively inathymic mice. Taken together, theseobservations indicate that high gradechondrosarcoma tumors arecharacterized by reduced expression oftype II collagen. This again is found tobe the case in Ch-1.

p53 acts as a tumor suppressor genebecause of its inhibitory control of thecell cycle and its role in programmedcell death or apoptosis. Loss of p53 wasidentified in 3 of 6 informative grade 2chondrosarcomas we studied, and innone of 7 grade 1 tumors. Reductionto homozygosity, associated withmutation of the remaining allele, is ahallmark of tumor suppressor genes,and is found for p53 in up to 60% ofhuman cancers. Ch-1 conforms to thispattern with a missense mutation of thesingle identifiable p53 allele at amutational “hotspot” - codon 175.Although there is yet no evidence thatp53 mutations are of prognostic valuein chondrosarcomas, there are reportsof increased expression of mutant p53in a subset of high gradechondrosarcomas. The p53 mutationin our source chondrosarcoma tissuemay have contributed to the biologicalaggressiveness of the tumor as well asrendered the tumor amenable tomaintenance of the chondrosarcomaphenotype in long term cell culture.

In summary, a molecularphenotype of an aggressivechondrosarcoma cell line derived froma high grade human chondrosarcomais beginning to emerge. The cell line

reported here synthesizes abundantaggrecan and no detectable levels oftypes I and II collagen mRNAs. TheCh-1 cell line and the Swarm ratchondrosarcoma exhibit similarexpression of fibrillar collagen types IXand XI, aggrecan, biglycan and CD-RAP. Preliminary studies show that Ch-1 will be a useful cell line in which tostudy expression of type XI collagenand CD-RAP. Further evaluation of thep53 tumor suppressor gene, CD-RAPand other regulatory genes may yieldinsight into both chondrosarcomapathogenesis and potential adjuvanttherapies.

RECOMMENDED READING

Chansky H, Robbins JR, Cha S, et al:Expression of cartilage extracellularmatrix and potential regulatory genesin a new human chondrosarcoma cellline [In Process Citation]. J Orthop Res,16:521-30, 1998.

Dietz UH, Sandell LJ: Cloning of aretinoic acid-sensitive mRNA expressedin cartilage and during chondrogenesis.J Biol Chem, 271:3311-6, 1996.

Dobashi Y, Sugimura H, Sato A, et al:Possible association of p53overexpression and mutation withhigh-grade chondrosarcoma. DiagnMol Pathol, 2:257-63, 1993.

Evans HL, Ayala AG, Romsdahl MM:Prognostic factors in chondrosarcomaof bone: a clinicopathologic analysiswith emphasis on histologic grading.Cancer, 40:818-31, 1977.

Gitelis S, Bertoni F, Picci P, et al:Chondrosarcoma of bone. Theexperience at the Istituto OrtopedicoRizzoli. J Bone Joint Surg [Am],63:1248-57, 1981.

Kawashima A, Ueda Y, Tsuchiya H, etal: Immunohistochemical localizationof collagenous proteins in cartilaginoustumors: characteristic distribution oftype IX collagen. J Cancer Res ClinOncol, 120:35-40, 1993.


Despite increased interest in theregeneration of peripheralnerves due to continued

improvement in microsurgicaltechniques, the ability to manipulatenerve regeneration remains limited.The 1992 report by Viterbo et al. thatproposed successful end-to-side axonalsprouting represented a significantadvance in the repair of peripheralnerve injuries because an intact nervecould “donate” axons to the distal endof an injured nerve when the injurycreated a large gap or a very proximallesion that would severely limitsuccessful nerve regeneration. Theinitial conclusions appeared todemonstrate an intrinsic ability ofaxons to sprout across the end-to-sidecoaptation. More recently, Lundborget al. reported successful collateralsprouting across an end-to-sidecoaptation using predegeneratednerves.

An end-to-side attachment of thedistal end of a proximally injured nerveto a nearby intact nerve could provideregenerating axons to preserve themotor and sensory targets, enablingotherwise unobtainable functionalrecovery. The end-to-side axonalsprouting theory is dependent uponaxonal sprouting occurring at the siteof nerve coaptation.

In addition to end-to-side nerverepair, investigative efforts in the area

of peripheral nerve regeneration havefocused on neurotrophic factors,peptides shown to be influential in thedifferentiation, function and survival ofneurons. Nerve growth factor (NGF),derived from end-organ targets, andciliary neurotrophic factor (CNTF),produced by neuron-associatedSchwann cells, are two neurotrophicfactors which have received significantattention in recent years. The exactmechanisms by which these factorsinteract are not well understood,however, there is considerable evidencethat each has a prominent role in theregeneration of peripheral nerves.Evidence exists for a synergistic effectbetween CNTF and both basicfibroblast growth factor (bFGF) andbrain derived neurotrophic factor(BDNF). In the central nervous system,recent studies have suggested thepossibility of a positive interactionbetween CNTF and NGF. To ourknowledge, there are no publishedreports investigating potentialinteractions in the peripheral nervoussystem.

Our goal was to identify evidence ofnerve regeneration after end-to-sidenerve coaptation in the rat model usingelectrical stimulation, axonal countingand muscle weights. This study differsfrom previous studies in two importantrespects. First, we attempted to enhancethe regeneration while previous studies

only addressed the presence or absenceof regeneration. The ability to stimulateregeneration would improve theclinical effectiveness of end-to-siderepair. Second, we sought to identifywhether the source of the regeneratedaxons was from true sprouting of theintact tibial nerve or from the injuredproximal peroneal stump.

METHODS

In male Harlan Sprague Dawleyrats, an end-to-side repair was made inthe right leg (with the left unoperatedleg serving as a control) as outlined inFigure 1.

The experimental groups: in GroupB, after coaptation, the growth factors(NGF 25 nanograms and CNTF 10nanograms in a 100 microliter aliquot)were injected into the epineural sheath1.0 centimeter proximal to thecoaptation. Group A received only thecoaptation while in Group C theperoneal nerve was transected distally,coiled upon itself and buried in muscleproximally to prevent reinnervation.

Twenty weeks after surgery, theperoneal nerves and tibialis anteriormuscles were harvested from theoperative and contralateral sides andprocessed for qualitative andquantitative analysis including axonalcounting, light microscopy and wetmuscle weights. Before harvesting,nerve stimulation with 0.5 milliampsoccurred 1.0 centimeter proximal tocoaptation and 1.0 centimeter distal tothe coaptation on both the tibial nerveand the peroneal nerve. The footresponse (flexion or extension) wasrecorded.

Statistical AnalysisStandard two-sided tests with

significance level a and power 1-b wereused for power analyses using thevalidation data of previous studies. Thepopulation sizes (6-7 animals per studygroup) were computed for 80% powerwith 95% confidence. To provide anextra margin of safety we elected to useeight animals per group. The data wasanalyzed using ANOVA with a two-

Does End-to-Side Nerve Repair Work? A Study of AxonalSprouting Stimulated from Intact Nerves Across an End-to-SideCoaptationWREN V. MCCALLISTER, B.S., B.A., PETER TANG, M.D., AND THOMAS E. TRUMBLE, M.D.

Sciatic n. Sciatic n.

proximal peroneal stump

Tibial n. Peroneal n. Tibial n.

Peroneal n.

Figure 1: Diagram of end-to-side coaptation technique in the rat sciatic nerve model. The peronealnerve was coapted to the intact tibial nerve after epineurotomy 1.5 cm distal to its division from thesciatic nerve and secured by two 10-0 nylon sutures. Note the proximal peroneal stump, elongatedfor illustrative purposes.


tailed student’s t test to determinesignificant differences between thegroups.

RESULTS

Axonal CountingFigure 2 reports the results of axonal

counting performed on sections takenfrom the segment of the peroneal nervethat was transferred end-to-side to thetibial nerve. The results are reportedas the ratio of the density (axons permicrometer squared) of the fibers of theexperimental side to the control side.These results are statistically significantat p < 0.002. Qualitatively, theregenerated fibers were well myelinated,of smaller diameter and higher densitywhen compared to the unoperatedcontrol fibers (Figure 3). Thehistological appearance of sectionstaken proximal to the site of coaptationbut distal to the proximal peronealstump revealed a significant number ofaxons traveling on the surface of theouter epineurium (Figure 4). Suchaxons were not observed in controlsections.

Electrical Stimulation Nerve stimulation in the normal

unoperated animal results in apredictable foot response. Withstimulation of the sciatic nerve above,and the tibial nerve below, the divisionof the peroneal nerve the normalanimal displays dominant foot flexionwith contraction of the gastrocnemiusmuscle. Stimulation of the peroneal

nerve results in foot extension withcontraction of the tibialis anteriormuscle. After nerve transfer, thenormal response was observed in bothexperimental groups with someexception in the growth factor group.There was an anomalous mixture offoot flexion and extension withstimulation above the coaptation site infour out of eight animals. Two out ofeight animals displayed foot flexionwith stimulation of the peroneal nerve.

Muscle WeightsReported as the ratio of the operated

to the contralateral unoperated side, thewet muscle weights recorded at the timeof harvesting were: Group A, 0.539 ±0.024; Group B, 0.596 ± 0.022; andGroup C, 0.220 ± 0.003 (p<0.002).

DISCUSSION

In this animal model we havedemonstrated axonal sprouting in thecontext of end-to-side coaptationstimulated by the injection of NGF andCNTF in combination. Electricalstimulation, axonal counting andtibialis anterior muscle weights indicatethe presence of axonal sprouting andsuccessful reinnervation of targetmuscle. Our results support the workof previous studies in the centralnervous system proposing a potentialinteraction between NGF and CNTFthat facilitates nerve regeneration.Most importantly, this study providesevidence that the intact tibial nervefunctions as a conduit for the

regenerating axons that travel in theouter epineurium of the nerve. If theseaxons are derived from the proximalperoneal stump, site “Y” Figure 1, thischallenges the proposed mechanism ofend-to-side axonal sprouting at the siteof coaptation. Further, such analternative explanation to successfulend-to-side axonal sprouting questionsthe potential application of end-to-sidenerve repair in the clinical setting.

RECOMMENDED READING

He, C.; and Chen Z.: Enhancement ofmotor nerve regeneration by nervegrowth factor. Microsurgery, 13(3):151-154, 1992.

Lundborg, G.; Zhao, Q.; Kanje, M.;Danielsen, N.; and Kerns, J.M.: Cansensory and motor collateral sproutingbe induced from intact peripheral nerveby end-to-side anastomosis? J. of HandSurgery (Br) , 19B: 277-82, 1994.

Newman, J.P.; Verity, A.N.; Hawatmeh,S.; Fee, W.E.; and Terris, D.J.: Ciliaryneurtotrophic factor enhancesperipheral nerve regeneration. Arch.Otolaryngol. Head Neck Surg., 122:399-403, Apr. 1996.

Pollock, M.: Nerve regeneration. Curr.Opin. Neurol., 8: 354-358, 1995.

Viterbo, F.; Trindade, J.C.; Hoshino, K.;and Neto, A.M.: Latero-terminalneurorrhaphy without removal of theepineural sheath: Experimental studyin rats. Sao Paulo Medical Journal,110(6): 267-275, 1992.

0.96

2.80

0

1

2

3

4

Group A Group B

Experimental Group

Rat

io o

f ax

on

den

sity

Figure 2: Axonal counting data from peroneal nerve sections twenty weeks after nerve transfer. ForGroups A and B, the standard error was 0.11 and 0.15, respectively (p<0.002).


(C)

(A) (B)

Figure 3: Histological appearance of sectioned peroneal nerves twenty weeks after nerve transfer (toluidine blue, 40X). (A) normal nerve (B) controlcoaptation only, Group A (C) growth factor injection, Group B. Note the increased density of smaller, well-myelinated nerves, when compared to normalnerve, consistent with axonal sprouting and regeneration.

Figure 4. Histologic section taken proximal to the site of coaptation but distal to the proximal peronealstump (toluidine blue, 40X). Note the significant number of axons traveling on the surface of theouter epineurium distinct from the tibial nerve proper.


Medical care is undergoing arevolution in the name of“Quality of Care”. Quality is

often assumed to be a measurable andvalid means of comparison inallocating increasingly limited healthcare resources. This focus on quality isso intense that quality of care is beingconsidered as an accreditation criterionto ensure compliance to standards ofcare in a coercive fashion. In this settingof competing physicians, treatments,organizations on the one side and finiteresources on the other, it is imperativethat we recognize the characteristics ofthe tools we use to define and measurequality.

A concurrent development is tofocus the measurement of health carequality on the patients’ perspective. Inthis outcomes driven approach, thebenefits of health care are assessed usingstandardized patient-oriented outcomeassessment instruments. The impact ofmedical intervention is assessed interms of the extent to which it meetsthe patients’ expectations andfunctional needs.

This study was designed to measurethe effect of comorbid conditions on

the functional status and well being ofpatients presenting for orthopaedicconsultation at a regional academicmedical center. The goals of the studywere to identify the functional deficitsexperienced by these patients using astandardized functional assessmentinstrument, the Short Form (SF)-36,and assess the relative health-statuslimitation experienced by these patientsdue to coexisting diseases.


This study is a prospectiveassessment of all new outpatientspresenting for evaluation to ourinstitution over a three-year intervalfrom January 1, 1995 to December 31,1997. We asked all new patientsevaluated at the Bone and Joint Center,University of Washington, to completea questionnaire containingidentification data, medical historyinformation, and an SF-36 form priorto their evaluation. The treatingphysician reviewed these forms forcompleteness during the initialinterview and examination. Wesubsequently scanned the SF-36 formsinto an electronic database and placed

the paper copy in the patient’s file. Ourstudy population consisted of 8941patients.

We used the Total CharlsonComorbidity Score to quantifycomorbidity in our patients. Weselected comorbid diagnoses identifiedas being important in determining thefuture health status of patientsaccording to criteria reported by. Weused the technique described by Deyoto review patients’ electronic medicalrecord for these comorbid diagnoses.

We constructed a Paradox 5.0database containing the SF-36information. This database containedan event record, demographic data,diagnosis data, and procedure data. Weused a customized scoring routine tostandardize SF-36 scoring. This routineused the procedures established in theSF-36 Scoring Manual to producenormalized scores for the eight SF-36functional domains. Our routineexpressed the final domain score foreach patient as a percentage of age andsex matched control values.

We stratified the patients into threegroups according to their comorbidityscore (Table 1). We evaluated the effectof comorbidity on each SF-36 domainscore using the software package SPSS7.5. We calculated correlationcoefficients between the Charlson Scoreand each of the SF-36 functionaldomain scores. We defined significanceas a p-value <0.05. We performed a oneway analysis of variance (alpha = 0.05)between the comorbidity groups andeach functional domains measured bythe SF-36 health assessmentinstrument. We also conducted posthoc comparisons using the Bonferroni/Dunn test.

RESULTS

The mean age of the studypopulation was 47 years (SD 16).Results of the one-way ANOVA showeddifferences in age among the threegroups. The patients with higherCharlson scores were older. Correlationwas significant at the 0.01 level forphysical function, general health,

Influence of Comorbid Medical Conditions on Pre-TreatmentSF-36 ScoresSOHAIL K. MIRZA, M.D., BARRY CAMPBELL, B.S., RANDAL P. CHING, PH.D., JENS R. CHAPMAN, M.D.,MARK A. KONODI, M.S., AND FREDERICK A. MATSEN III, M.D.

Group 1 Group 2 Group 3

Charlson Score "0" "1-3" ">3"

No. of patients (%) 6897 (77.1%) 1857 (20.8%) 187 (2.1%)

Age 44.4 ± 16.1 53.6 ± 16.4 57.3 ± 16.6

Physical function 0.708 ± 0.343 0.635 ± 0.398 0.544 ± 0.382

Role-physical 0.407 ± 0.508 0.424 ± 0.569 0.346 ± 0.589

Bodily pain 0.518 ± 0.333 0.536 ± 0.362 0.537 ± 0.374

General health 0.907 ± 0.304 0.819 ± 0.338 0.652 ± 0.354

Vitality 0.830 ± 0.375 0.765 ± 0.395 0.623 ± 0.340

Social function 0.732 ± 0.342 0.707 ± 0.351 0.642 ± 0.356

Role-emotional 0.813 ± 0.468 0.798 ± 0.500 0.815 ± 0.505

Mental 0.931 ± 0.262 0.917 ± 0.262 0.879 ± 0.240

Table 1: Patient stratification by total Charlson comorbidity score.


vitality, social function and mentalhealth domains of SF-36. Each of thesedomains showed a decreased scoreassociated with higher Charlsoncomorbidity scores. Role-physical,bodily pain, and role-emotionaldomains did not show statisticallysignificant correlation withcomorbidity (Table 2).

Results of the one-way ANOVAsupported the hypothesis that the levelof comorbidity has a differential effecton the measures of physical function,general health, vitality, social function,and mental health scores of the SF-36health assessment instrument. Each ofthese domains showed a progressivedecrease with an increasing presence ofcoexisting medical conditions. As anexample, Figure 1 shows this effect forthe physical function domain. Thepatient groups with greatercomorbidity had lower physicalfunction scores. The level ofcomorbidity did not have a differentialeffect on the measure of role-physical,bodily pain, and role-emotional scoresof the SF-36 health assessmentinstrument (Table 3).

DISCUSSION

The outcomes movement hasgenerated optimistic prospects forimproving medical care of our patients.Systematically measuring quality ofcare will help patients, physicians, andthird-party payers make better health-care decisions. Development ofstandardized and validated instrumentsof health-status assessment has allowedlarge scale deployment of outcomeassessment techniques. Instrumentssoliciting subjective data from patientscan provide useful information thatmay be more important in determiningthe patients’ perception of well-beingthan traditional measures of clinical orphysiological indexes. Development ofshorter questionnaires has made largescale application more practical. Theseshorter forms have been demonstratedto be equally responsive in surgicalpatients when compared with longerinstruments.

Interpretation of treatmentinterventions requires identification ofspecific patient populations evaluated.Excluding patients with significantcomorbidity limits the interpretation ofresults. Randomization of subjectscontrols for both known and unknownconfounding factors. Controlling for

Domain Correlation Coefficient* Significance

Physical function** -0.095 < 0.0001**

Role-physical -0.003 0.761

Bodily pain 0.020 0.066

General health** -0.134 < 0.0001**

Vitality** -0.097 < 0.0001**

Social function** -0.037 < 0.0001**

Role-emotional 0.016 0.138

Mental** -0.035 < 0.0001**

*Spearman’s rho**Significant at the 0.01 level (two-tailed)

Domain ANOVA

P Value

Group 1

versus 2*

Group 1

versus 3*

Group 2

versus 3*

Physical function <.0001** <.0001** <.0001** .0010**

Role-physical 0.1290 .2290 .1250 .0589

Bodily pain 0.1120 .0450 .4559 .9645

General health <.0001** <.0001** <.0001** <.0001**

Vitality <.0001** <.0001** <.0001** <.0001**

Social function <.0001** .0047** .0006** .0170

Role-emotional 0.5014 0.244 .9457 .6486

Mental 0.0066** .0456 .0093** .0653

*Bonferroni/Dunn, significance level 0.05**Significant difference

18318316783N =

Comorbidity Groups (n)

Group 3Group 2Group 1

Phy

sica

l Fun

ctio

n (m

ean

+_9

5%C

I)

.8

.7

.6

.5

.4

Figure 1: Physical function scores for the three Charlson Score groups expressed as a percentage ofage and sex matched control values.

Table 2: Correlations between Charlson score and SF-36 domains.

Table 3: Levels of significance between comorbidity groups.


potential confounding factors in non-randomized trials, however, requiresconsideration of coexisting diseases andseverity of illness. Our workdemonstrates that coexisting medicalconditions need to be considered in thecomparisons of functional assessmentin large populations.

CONCLUSION

Comorbid diseases significantlyinfluence the functional assessment asmeasured by SF-36 in patientspresenting with musculoskeletalcomplaints. This effect is independentof age, sex, and the primarymusculoskeletal diagnosis.


In slipped capital femoral epiphysis(SCFE) the proximal femoralmetaphysis displaces posteriorward

to produce an apex anterior deformity.In situ screw fixation (ISSF) accepts thisdeformity. Flexion intertrochantericosteotomy (FIO) has the potential torestore anatomy, and is theoreticallysafe in terms of the proximal femoralblood supply. In a retrospective,nonrandomized study of unilateral,chronic, severe, stable SCFE, wecompared the outcome of ISSF versusFIO at a minimum of 2 yr. follow-up.


39 patients met the inclusioncriteria, of whom 20 agreed toparticipate. Ten patients (5 female, 5male; age 10.6-16.5 yr., average 13.0 yr.;follow-up 3.3-13.0 yr., average 6.8 yr.)underwent ISSF and 10 patients (4female, 6 male; age 11.6-25.5 yr.,average 15.5 yr.; follow-up 2-13.8 yr.,average 8.0 yr.) underwent FIO (seefigures). Three components enteredinto the measurement of functionaloutcome: patient perception, physicalexamination (range or motion,strength, body mass index [BMI], gait,performance) and radiographicassessment (slip angle, deformity, jointspace, and sphericity by AP, frog andtrue lateral röntgenograms). A HipOutcome Instrument was derived fromestablished validated instruments:Harris Hip Score (HHS),Musculoskeletal Function Assessment(MFA), Short Form – 36 (SF-36) andNursing Health Evaluation (NHE).

RESULTS

FIO patients demonstrated asignificantly greater BMI (p<0.05) andslip angle (p<0.05) than ISSF patients.There were no significant differences inMFA, SF-36 or NHE. The HHS washigher for the ISSF group comparedwith the FIO group (one tailed t-test p= 0.04). The FIO group demonstratedsignificantly improved hip internalrotation compared with the ISSF group(p<0.05). There was no significantdifference in hip flexion, abduction,

adduction, external rotation, walkingspeed, stair climbing speed or footprogression angle. There was noevidence of osteonecrosis in eithergroup. Chondrolysis followed FIO in1 patient (who sustained andintraoperative fracture of the femoralneck), but did not occur after ISSF.

DISCUSSION

The rationale for treatment of SCFEwith proximal femoral osteotomy isbased upon the inferred relationship ofdeformity to early dysfunction and latecoxarthritis. Proof of the latter requireslonger follow-up. We found that onlyhip internal rotation was significantlyincreased by FIO. Moreover, our resultssuggest that this may be associated witha diminution in HHS. Within thelimited power of this study, we can statethat improved outcome after FIO wasnot realized despite the theoreticaladvantages. While internal rotation wasimproved with physical exam, the FIOpatients tended to have more pain anda significantly worse HHS. From these

Functional Outcome of In Situ Screw Fixation Versus FlexionIntertrochanteric Osteotomy for Slipped Capital FemoralEpiphysisMOHAMMAD DIAB, M.D., S. DALUVOY, B.S., J. KASSER, M.D., AND B. SNYDER, M.D.

“pilot” data, we plan a multicenter studyto further test the hypothesis that thereis no significant advantage to treatmentwith FIO compared with ISSF.

Figure 1: Lateral Radiograph. FlexionIntertrochanteric Osteotomy.

Figure 2: AP Radiograph. FlexionIntertrochanteric Osteotomy.


Most closed, isolated humeralshaft fractures can besuccessfully treated non-

operatively, and the indications forsurgery are reasonably well established.Surgical stabilization generally involveseither plating or intramedullary nailing.In contrast to open reduction andinternal fixation with compressionplating, locked intramedullary nailingoffers the advantages of increasedbiomechanical strength in load tofailure and three point bendingstiffness, while closed nailing leaves thefracture hematoma relativelyundisturbed, theoretically enhancingfracture healing. However, closedinsertion of a locked, reamedintramedullary nail may putneurovascular structures, especially the

radial nerve, at risk. Recently, intra-operative somatosensory evokedpotential (SSEP) monitoring hasproven to be a reliable method ofidentifying abnormal neural responses,both in instrumentation of the spinalcolumn and acetabular fracture surgery.The purpose of this study was to assessthe role of intra-operative SSEPmonitoring during closed, lockedintramedullary nailing of humeral shaftfractures.

MATERIAL AND METHODS

Between April 1991 and February1994, 81 patients met indications forinternal fixation of humeral shaftfractures at Harborview MedicalCenter. Of these, 45 were treated withinterlocked humeral nailing. Thirteenpatients treated in this fashion were feltto be at risk for iatrogenic nerve injurysecondary to placement of a lockedintramedullary nail. This increased riskwas perceived because of a)unobtainable preoperative neurologicalexamination due closed head injury (6patients), or b) anticipation of difficultreduction maneuver due to fracturecomplexity or displacement likelyendangering the radial nerve duringmanipulation (7 patients). Indicationsfor surgical stabilization included aninability to maintain satisfactoryreduction with closed treatment (6patients), polytrauma (5 patients),floating elbow (1 patient), andmalaligned delayed union (1 patient).

SSEPs were used for intra-operativemonitoring. Baseline SSEP signal waselicited in all patients prior to beginningthe surgical procedure. The radial andmedian nerves were monitored withrecording from cortical and subcorticalsites. After sterile preparation of theentire upper limb, pairs of sterile 1/2"needle electrodes were placed forstimulation using 2-3 cm inter-electrode separation. The superficialradial nerve was stimulated at the wristjust proximal to the radial styloid, themedian nerve between the flexor carpiradialis and palmaris longus tendons,and the posterior interosseous nerve 2

cm distal to the radial head. Responseswere recorded at Erb’s point, the C7spinous process and the mastoidprocess (sub-cortical sites) and thecontralateral scalp (cortical site).Several hundred responses wereaveraged to obtain a response andrecordings were made from each nerveat least once every 20 minutes, but morefrequently during fracturemanipulation and guide wire or nailinsertion and interlocking. Criteria forintra-operative abnormality wereamplitude decrease of more than 30%for subcortical sites or more than 50%at the scalp.

The surgical technique consisted ofclosed manipulation and fracturereduction followed by placement of aguide rod across the fracture site intothe distal humerus under fluoroscopicvisualization. For the eleven nailsinserted in an antegrade fashionproximal interlocking with a singlescrew was always utilized, while theneed for distal interlocking wasdetermined by the fracture pattern anddistal canal filling. Distal interlockingwas performed with a free handtechnique in an anterior-posteriorfashion. Both retrograde nails werelocked proximally and distally.

RESULTS

Baseline signals were obtainable in12 of 13 patients for both radial andmedian nerves. Signal from the medianbut not radial nerve was obtained inone patient with a severe closed headinjury and a displaced fracture of thehumeral shaft. A preoperativeexamination of upper limb functionwas unobtainable due to his closedhead injury. The planned closednailing was abandoned, a lateralapproach to the humerus performedand the fracture site exposed. Theradial nerve was found in the fracturesite, released and protected, and opennailing continued without furtherincident.

Intra-operative SSEP changesoccurred in 2 of the remaining 12patients. In one, the superficial radial

Somatosensory Evoked Potential Monitoring During ClosedHumeral NailingWILLIAM J. MILLS, M.D., JENS R. CHAPMAN, M.D., LAWRENCE R. ROBINSON, M.D.,AND JEFFERSON C. SLIMP, PH.D.

Figure 1: A pre-operative x-ray of asegmentally unstable fracture of thehumeral shaft.


nerve response amplitude diminishedbelow established thresholds duringfracture manipulation prior toattempted guide rod insertion. Fiveminutes after altering the manipulationtechnique and discontinuinglongitudinal traction, the amplitudereturned to baseline. Two centimetersshortening of the segmentallycomminuted humeral shaft wereaccepted, nailing proceededuneventfully, and the postoperativeneurological examination was normal.Union occurred in a timely fashion.The second patient with intra-operativeSSEP changes had a significant drop inthe median nerve amplitude duringattempted anterior to posterior distalinterlocking screw placement.Interlocking was discontinued, andafter 5 minutes 75% recovery of themedian nerve signal was noted. Distalinterlocking was then performeduneventfully from posterior to anterior.No postoperative neurological deficitswere identified. SSEP monitoringtherefore impacted our surgicaltechnique in 3 of 13 patients (23 %) .No patient monitored during humeralnailing had new neurological deficitspostoperatively.

DISCUSSION

For those fractures in which surgicaltreatment is indicated, closedintramedullary nailing of the humeralshaft may offer distinct biomechanicaland soft tissue advantages over openreduction and plate fixation. Onepossible disadvantage of closed nailingis that the radial nerve is not routinelyidentified and protected. Many authorshave reported postoperativeneurapraxia with locked humeralnailing and appropriately recommendcaution during fracture reduction anddistal interlocking. In patients with apreoperative radial nerve palsy, or otherupper extremity neurological deficit,closed nailing risks iatrogenic injury ifthe nerve is in the fracture site. Clearly,closed nailing in that setting couldproduce devastating radial nerve injury.The role of SSEP monitoring in closed,interlocked humeral nailingdemonstrated here is for that specificgroup of patients in whom a closedhead injury or other factors make upperextremity neurologic status unknown,or in those patients with an intact radialnerve in whom a difficult closedreduction is anticipated.

Somatosensory evoked potentialsallow for a sterile, virtually real-time

assessment of neural status. At ourinstitutions as well as others, the use ofSSEP monitoring has become routinefor surgery involving instrumentationof the spine. This study demonstratesthe technical feasibility and efficacy ofintra-operative upper extremity SSEPmonitoring in a select group of patientsundergoing closed, interlockedhumeral nailing. In three of thirteenpatients (23 %) signal changesprompted a change in surgical plan.

Our experience suggest that intra-operative SSEP monitoring duringclosed humeral nailing in select patientsis beneficial. When baseline signals areunobtainable, a limited exposure of thefracture site and identification andprotection of the radial nerve mayminimize the risk of iatrogenic injury.Similarly significant signaldeterioration with manipulation,traction or interlocking attemptssuggest the need to release traction,alter reduction maneuvers and possiblymodify interlocking technique.

RECOMMENDED READING

Blasier RB and Graham TJ: Lockingintramedullary nailing of fractures ofthe humeral shaft. Am J Orthop 1996;25(5): Suppl. 31-37

Foster RJ, Dixon GL, Bach AW,Appleyard R: Internal fixation ofhumeral shaft lesions. Indications andresults. Orthop Trans 1983; 7:69

Henley MB, Monroe M, Tencer AF:Biomechanical comparison of methodsof fixation of a midshaft osteotomy ofthe humerus. J Orthop Trauma 1991;5(1): 14-20

Mast JW, Spiegel PG, Harvey JP,Harrison C: Fractures of the humeralshaft. A retrospective study of 240 adultfractures. Clin Orthop 1975; 112: 254-262

Robinson LR, Slimp JC, Anderson PA,Stolov WC. The efficacy of femoralnerveintra-operative somatosensoryevoked potentials during surgicaltreatment of thoracolumbarfractures.Spine 1993; 18(13): 1793-97

Russell TA, Simard J, Taylor JC, LavelleDG, Nichols L, Walker B: Interlockingintramedullary nailing of humeral shaftfractures. Orthop Trans 1992; 16(2):334

Figure 2: A post-operative x-ray of thefracture.

Figure 3: Another post-operative x-ray of thesame fracture of the humeral shaft.


S acral fractures are commoninjuries in trauma victims andare associated with a high rate of

neurologic deficits. Despite numerousreports of zone III sacral fractures,transverse sacral fractures, andspondylolisthesis of the sacralsegments, U-shaped sacral fractureshave received little mention. The actualpathoanatomy of these fractures ispoorly understood but is characterizedby bilateral transforaminal sacralfractures with a transverse componentin the upper sacrum. Depending on thepattern of the fracture, the caudal

segment may be flexed (morecommon) or extended relative to thecephalad component. This results inthe typical kyphotic deformity of thesacrum seen in these fractures (Figures1 and 2). The goal of this study is toevaluate a percutaneous method ofstabilization of these unusual fracturesand their early clinical results.


Over a 38 months period, 442patients with pelvic ring disruptionswere treated at Harborview MedicalCenter, a level one trauma center. Onlythirteen (2.9%) of these patients haddisplaced U-shaped sacral fracturestreated with percutaneous iliosacralscrew stabilization. There were 11 maleand 2 female patients, ranging in agefrom 21 to 60 years (mean 39 years).Mechanism of injury includedaccidental falls from heights in fourpatients, suicide attempts by jumpingin three, motor vehicle accidents in two,industrial accidents in two, motorcycleaccident in one, and pedestrian versusmotor vehicle in one. The averageInjury Severity Score (ISS) was 18.1(range, 9-34). The sacral injuries werefurther classified according to Roy-Camille based on plain pelvicorthogonal radiographs and routinetwo-dimensional computerizedtomography scans. There were onetype 1, twelve type 2, and zero type 3fracture patterns. All of the fractureswere through the upper two sacralsegments. Lateral sacral plainradiographs were obtained and in allpatients and best identified thesefractures.

Clinical signs of hemodynamicinstability at admission were present inseven patients. Associated injuriesincluded other spinal fractures in fivepatients, traumatic brain injuries infour patients, and calcaneal fractures inthree patients. A detailed neurologicexamination was documented in allpatients upon arrival and includedmotor strength testing, sensoryexamination, evaluation of rectal tone,

and the presence of lower extremityreflexes. Neurologic abnormalities ofthe lumbar or sacral roots were initiallynoted in eight patients. Another patientdeveloped progressive paresthesias andmotor weakness in the S1 nerve rootdistribution while in the hospital. Twopatients were intubated andpharmacologically paralyzed uponarrival and complete preoperativeneurologic examinations were notpossible.

TECHNIQUE

These thirteen patients were treatedoperatively according to a managementprotocol for their U-shaped sacralfractures. Preoperative CT scans wereevaluated to determine the safe zone forscrew insertion. Neurodiagnosticmonitoring was not used during screw

Percutaneous Stabilization of U-Shaped Sacral Fractures UsingIliosacral Screws

SEAN E. NORK, M.D., CLIFFORD B. JONES, M.D., SUSAN P. HARDING, M.D., SOHAIL K. MIRZA, M.D.,AND M.L. CHIP ROUTT, JR., M.D.

Figure 1: Sacral deformity.

Figure 2: Sacral deformity (outlined).Figure 4: Lateral after fixation.

Figure 3: AP after fixation.


insertions. Fracture stabilization wasaccomplished using fluoroscopicallyguided iliosacral screws insertedpercutaneously. Orthogonal pelvicinlet and outlet images, as well as lateralsacral views were obtainedintraoperativley and correlated with thepreoperative CT scan to assess sacralmorphology. The guide pins andcannulated drills were oriented to allowinsertion of the screws across bothcomponents of the transforaminalsacral fractures (Figures 3 and 4).

RESULTS

Twenty-five fully threadedcancellous 7.0 mm cannulated screwswere used. Eleven patients had bilateralscrew fixations, one patient hadunilateral double screw fixation, andone patient had unilateral single screwfixation. Operative time for screwinsertion averaged 48 minutes, with 2.1minutes of fluoroscopy. Subsequentsacral neurological decompressionswere not performed in any patients. Allof the patients were immobilized aftersurgery with an HTLSO orthosis for 12weeks. Patients were mobilized safelyto a cardiac chair immediately aftersurgery and could be transported in awheelchair postoperatively.

Accurate and safe screw insertionswithout foraminal or canal violationswere confirmed in all patients withpostoperative pelvic plain radiographsand computerized tomograms. Aparadoxical inlet view of the uppersacral segments on the injury AP pelviswas seen in 12 of 13 patients (92.3%)and the diagnosis was confirmed withthe lateral sacral view in 100% ofpatients. Computed tomography wasuseful in confirming the diagnosis in allcases.

There were no deaths.Preoperatively, sacral kyphosis averaged29 degrees (range, 6 – 75 degrees), whilepostoperative sacral kyphosis averaged28 degrees (range, 6 – 63 degrees). Allof the fractures healed clinically andradiographically. There were no woundinfections. One partial screwdisengagement occurred withoutfixation failure in the only patienttreated with a single unilateral screw.This screw was removed twelve weeksafter injury.

Of the nine patients withpreoperative neurologic abnormalities,only two patients had residualneurologic deficits. Of these two

patients, both had multiple lumbarburst fractures which requireddecompression and instrumentedfusions.

DISCUSSION

The diagnosis of U-shaped sacralfractures is difficult. A paradoxical inletof the upper sacrum on the APradiograph should alert both theradiologist and the orthopaedicsurgeon to this injury and suggests theneed for lateral sacral imaging. In thisseries, the lateral sacral radiograph andthe CT scan demonstrated the injuryin all patients.

Surgical stabilization allows safe andearly mobilization of the patient fromrecumbency. Percutaneous fixationshould diminish potential blood lossand operative times compared to opentechniques. In addition, a percutaneousapproach allows for rapid stabilizationin the supine position and does notinterfere with future sacraldecompression. Iatrogenic neurologicinjuries and screw errors were avoidedin this series despite the use of extralong iliosacral screws (up to 150 mm).This series was limited to kyphoticdeformities which allow in situ fixation.Progressive kyphosis was not observeddespite this limited internal fixation.

Neurologic abnormalities arecommonly associated with thesefractures and occur in approximately85% of cases. In this series, completeneurologic recovery was observed inseven of nine patients without sacraldecompression. Because of concernover progression neurologic deficitswhich was observed in one patient inthis series prior to surgery, we feel thatstabilization of these injuries isnecessary to protect the localneuroanatomy and to allow neurologicrecovery. This technique is especiallyuseful in polytraumatized patientswhich may require multiple proceduresand may need to be moved from totalrecumbency.

CONCLUSIONS

Delayed diagnosis is avoided by ahigh clinical suspicion, early lateralsacral radiographs, and pelvic CT scans.Surgical stabilization allows earlymobilization of the patient fromrecumbency. Percutaneous fixationshould diminish potential blood lossand operative times compared to opentechniques, and still allows sacral

decompression of the local neuralelements. Early percutaneous iliosacralscrew fixation is an effective for theseinjuries.

RECOMMENDED READING

Ebraheim NA, Biyani A, Salpietro B.Zone III fractures of the sacrum. A casereport. Spine 1996;21:2390-2396.

Gibbons KJ, Soloniuk DS, Razack N.Neurological injury and patterns ofsacral fractures. J Neurosurg1990;72:889-893.

Roy Camille R, Saillant G, Gagna G,Mazel C. Transverse fracture of theupper sacrum. Suicidal jumper’sfracture. Spine 1985;10:838-845.

Sabiston CP, Wing PC. Sacral fractures:classification and neurologicimplications. J Trauma 1986;26:1113-1115.

Schmidek HH, Smith DA, KristiansenTK. Sacral fractures. Neurosurgery1984;15:735-746.

Strange Vognsen HH, Lebech A. Anunusual type of fracture in the uppersacrum. J Orthop Trauma 1991;5:200-203.


T reatment of periarticularfractures is difficult. Accuratearticular reduction and stable

fixation allows early joint motion andshould improve articular cartilagehealing. This is especially true foracetabular fractures, where clinicalseries have demonstrated the directrelationship between patient outcomeand the quality of articular reduction.Certain acetabular fractures, especiallyposterior wall patterns present achallenging problem due to thepresence of marginal impaction and/orosteoarticular loose bodies. Varioustechniques have been described forrestoring the acetabular articularsurface when loose or marginallyimpacted osteochondral fragments areencountered. We evaluated smallfragment intraosseous screws to secureimpacted or loose osteochondralacetabular fragments in acetabularfractures.

METHODS

Between January 1996 andDecember 1998, 85 patients withacetabular fractures involving the

posterior wall, were operatively treatedat Harborview Medical Center inSeattle, Washington. Eighteen of the 85patients (21%) had osteochondralimpacted or loose fragments amenableto intraosseous screw fixation. Therewere 15 men and 3 women with anaverage age of 41.7 years (range 20 to68). The cause of injury was a motorvehicle accident in all 18 patients.Seventeen of the eighteen patients hadposterior dislocations associated withtheir fractures.

All patients were admitted throughthe emergency room following astandard protocol, which included ananteroposterior pelvic plainradiograph. After initial resuscitation,the posteriorly dislocated femoral headswere urgently relocated usingintravenous sedation and controlled(gentle) manual manipulation. Repeatradiographs confirmed reductions.Additional radiological evaluationsconsisted of iliac and obturator obliqueradiographs (Figure 1). A routine twodimensional pelvic computedtomography (CT) scan was obtained(Figure 2). All 18 fractures were

classified according to Letournel. Therewere 6 elementary posterior wallfractures, 2 associated fractures of theposterior column and posterior wall,and 10 associated transverse andposterior wall fractures. Marginalimpaction, as described by Brumback,and /or osteoarticular loose bodies werepresent in all patients. The patients wereplaced in distal femoral skeletal tractionand admitted to the hospital. Deepvenous thrombosis prophylaxisconsisted of bilateral thigh highsequential compression devices ( KCIInc.) on the lower extremities andsubcutaneous heparin.

A Kocher Langenbeck surgicalapproach was used for each patient.The posterior wall fragment wasretracted superolaterally to reveal theunderlying fracture sites. The hip jointwas distracted and the osteochondralfragments were removed and themarginal impaction zones wereassessed. For those patients withassociated fracture patterns, thetransverse or posterior columncomponent was reduced, clamped, andsecured with cortical lag screws. Theosteoarticular fragments were thenreduced using the intact acetabulumand femoral head as templates.Depending on their size, theosteochondral fragments were thenstabilized using either a 2.7mm, 2.0mm,or 1.5mm extraarticular intraosseousscrew directed away from the joint. Thescrews were inserted through thesubchondral cancellous bone of thefragment into the intact acetabulum.Impacted fragments were treated withelevation to the femoral head. Theresultant cancellous bone defect wasfilled with autograft bone obtainedfrom the ipsilateral greater trochanter.If the fragment demonstratedinstability after bone grafting of thedefect, it was secured with aintraosseous screw. The femoral headwas distracted and the articularreduction inspected. The posteriorwall was then reduced and securedusing a contoured 3.5mm pelvicreconstruction plate. Orthogonalfluoroscopic images demonstrated

Screw Fixation of Osteochondral Injuries in ComminutedAcetabular FracturesDANIEL SEGINA M.D., SEAN E. NORK, M.D., AND M.L. CHIP ROUTT, JR., M.D.

Figure 1: Oblique view showing fracture fragment.


with crutch assist protected the repairfor six weeks after surgery. Progressivepartial weight bearing was permittedthereafter, followed by full weightbearing three months after operation.Serial oblique pelvic plain radiographswere taken at six , twelve and fifty twoweeks post-operatively to asses fracturehealing, implant changes, ectopic boneformation, and joint abnormalities.

RESULTS

Fifteen of the seventeen patientswith posteriorly dislocated hips hadsuccessful closed manipulativereductions. Two patients were noted tohave nonconcentric reductions due tointrarticular large osteochondral loosebodies. Both were taken urgently to theoperating room for open debridementof the hip joint and repair of theirfracturse.

Postoperative acetabular CT scanscharacterized articular incongruity asgapping, step off, or both. Gaps greaterthan 2 millimeters were noted in 6patients, while articular step-off wasseen in 4 patients. The intraosseousscrew position was also noted. Nopatient demonstrated an intraarticularposition of the intraosseous screw orother implants.

Two of the 18 patients, both withloose osteochondral fragments,required intraosseous screw removalduring the procedure. Both patientshad malreduced osteochondralfragments. This was realized once theposterior wall reduction was attemptedand noted to be inaccurate. After screwand fragment removal, an accuratereduction of the posterior wall wasachieved.

One patient developed a deepwound infection 3 months after injurythat required operative debridement ofthe wound and implant removal. Hisinitial postoperative course wascomplicated by a gangrenousgallbladder with subsequent abdominalwound infection requiring openpacking. At reoperation, the intraosseusscrew was unchanged in itssubchondral extrarticular location. Hehad progressive hip pain associatedwith joint space loss, and underwentuncomplicated total hip arthroplasty.

DISCUSSION

Judet and Letournel first describedsupporting marginally impactedosteochondral fragments with

implant safety and fracture reductionquality. The wound was closed in layersover suction drains. Post-operativeoblique radiographs and a CT scan ofthe pelvis and acetabulum were

obtained (Figure 3, 4, and 5).Continuous passive hip motion deviceswere used post-operatively untildischarge with hip flexion limited to 60degrees. Weight of leg weight bearing

Figure 3: Post fixation x-ray.

Figure 2: CT scan showing fracture fragments.


autograft cancellous bone. Theyadvocated that the reduction beachieved using the femoral head as a

template. While this technique mayinitially provide for an accuratereduction, assurance of this accuracy

can be lost once the posterior wallfragment is reduced. This is a directconsequence of the inability to visualizethe joint after reduction and fixation ofthe wall fragment. Brumback et. al.addressed this potential pitfall,recommending intraosseous Kirschnerwire fixation of large or detachedosteochondral fragments. McCardelreported on the potential seriouscomplication of Kirschner wiremigration to the heart using thistechnique. They advocated Herbertscrew fixation of large osteochondralacetabular fragments. Biomechanicalevaluations acetabular fractures havedemonstrated the adverse effects ofarticular incongruity . Olson et al.showed the detrimental effects ofarticular malreduction in transverseand posterior wall fractures. Both Keithet. al. and Olson et. al. examined theeffect of relatively small posterior wallfracture fragments. They demonstratedpotential instability and contactpressure changes with a little as 20%involvement of the posterior wall.These studies validate the belief thatimproved quality of articular reductiontranslates into improved potential forbetter patient outcome. Thisbiomechanical data is complementedby clinical studies that demonstrateimproved patient outcome withimproved quality of articular reduction.Supplemental rigid fixation ofosteochondral fragments, in addition tothe removal of known intraarticulardebris, also offers the best prophylaxisagainst third body wear.

Small screw fixation ofosteochondral fracture fragments incertain acetabular fractures provides anoperative adjuvant when reducing andstabilizing these important fragments.The subchondral small intraosseousscrew usually allows for stable fixationof loose or marginally impactedosteochondral fragments. Thetechnique avoids discarding potentiallyimportant articular components andeliminates the potential problems ofwire embolism. In light of the clinicaland biomechanical data supporting theimportance of accurate articularreduction, the use of a intraosseousscrew provides rigid fixation andassurance of reduction prior to thedefinitive repair of the posterior wall.This hopefully translates into improvedlong term patient outcome byincreasing hip joint congruity.

Figure 4: Post fixation x-ray.

Figure 5: Post fixation CT scan.


RECOMMENDED READING

Letournel, E., Acetabulum fractures:classification and management. ClinOrthop, 1980(151): p. 81-106.

Matta, J.M., Fractures of theacetabulum: accuracy of reduction andclinical results in patients managedoperatively within three weeks after theinjury. J Bone Joint Surg Am, 1996.78(11): p. 1632-45.

Brumback, R.J., et al., Acetabulardepression fracture accompanyingposterior fracture dislocation of the hip.J Orthop Trauma, 1990. 4(1): p. 42-8.

Goulet, J.A., et al., Comminutedfractures of the posterior wall of theacetabulum. A biomechanicalevaluation of fixation methods. J BoneJoint Surg Am, 1994. 76(10): p. 1457-63.

McCardel, B.R., L.E. Dahners, and J.B.Renner, Kirschner wire migration fromthe pelvis to the heart and a newmethod of fixation of articular fracturefragments, acetabular reconstruction. JOrthop Trauma, 1989. 3(3): p. 257-9.

Olson, S.A., et al., Biomechanicalconsequences of fracture and repair ofthe posterior wall of the acetabulum. JBone Joint Surg Am, 1995. 77(8): p.1184-92.


Biomechanics can provide insightinto the mechanisms oftraumatic injury. Epidemiology

addresses injury risk factors. Thecombination of these two disciplinescan make significant contributions toinjury prevention. The three projectsdescribed below demonstrate howcollaboration between the OrthopaedicBiomechanics Lab and the HarborviewInjury Prevention Center can addressthese types of problems.

PREVENTION OF CERVICAL SPINE

WHIPLASH IN REAR-END AUTOMOBILE

ACCIDENTS

Cervical spine hyperextension-

flexion (whiplash) injuries in rear-endautomobile collisions representbetween 65% and 85% of all medicalinsurance claims. In this project, acushion is being developed to be placedbetween the occupant and the seat backand head restraint, that (1) stiffens theseatback-head restraint connection, (2)provides a closer fit to the driver’s spine,neck, and head, and most importantly,(3) absorbs energy during impactwhich would otherwise be transferredto the driver’s head and neck. Thesemodifications are proposed to reducethe complex relative motions betweenthe driver’s head and torso during arear-end impact which results in thewhiplash (actually, shearing, extension,

and flexion) motions of the cervicalvertebrae. The anti-whiplash cushiondesign was based using bothepidemiological and biomechanicalstudies. The seated positions of 722drivers relative to their head restraintswere studied, which showed than about25% had head restraints close (within3” of their heads) while 60% had headrestraints which were either too low ortoo far back, and 10% had no headrestraints in their vehicles. A completeanalysis of 238 low speed rear-endaccidents showed that risk factors fordeveloping whiplash symptomsincluded, prior degeneration of thecervical spine, a rotated position of thehead prior to impact ,and a struckvehicle speed change of at least 7 mph.A biomechanical study of humanvolunteer response to rear-end impactdemonstrated that the car seat iseffective at transferring considerableenergy of impact to the occupant.There is a time lag between impact ofthe torso with the seat back, which isthrust forward, and the head, which isstill falling backwards into the headrestraint. Based on these observations,prototype anti-whiplash cushions arecurrently being tested. (Funded by theHarborview Injury Prevention Center,the Centers for Disease Control andPrevention, and the Physical MedicineResearch Foundation).

PREVENTION OF FEMUR FRACTURES IN

FRONTAL AUTOMOBILE COLLISIONS

Direct impact of the knee with thedashboard can result in supracondylaror midshaft femur fractures or patellarinjuries. Federal Motor Vehicle SafetyStandard (FMVSS 208) defines apermissible maximum axial force ofcontact between the knee and thedashboard of 1700 lbs. This is wellbelow the observed mean fracture loadof the femur (about 2500 lbs in an axialdistal to proximal direction), however,isolated fractures of the femur, even inshoulder belt, seat belt, and air bagrestrained occupants continue to occur.To study the epidemiology of actualaccidents, the National Highway

Practical Approaches to Injury Prevention Using Biomechanicsand Epidemiology

ALLAN F. TENCER, PH.D.

Figure 1: AP x-ray of femur fracture with interlocking nail.


Transportation Safety Administration(NHTSA) has developed 8 regionalcenters where complete engineeringanalyses to determine forces of impactand occupant contact with the interior,detailed medical descriptions ofinjuries received, and biomechanicalanalyses of mechanisms of injury arebeing performed. To this point, 15 casesof isolated severe femur injuries havebeen identified. The approximate peakaxial force acting at impact on thefemur was estimated , in these cases, tobe at least 2500 lbs. This is well beyondthe peak permissible dashboardloading. The problem may be thatdashboard stiffness varies and is greaterin corners such as junctions with thedoor, sterring wheel or center consolewhere many of these contacts takeplace. In contrast, testing of dashboardimpact force with crash dummies isusually done in the center of thedashboard. Using a biomechanicsapproach , we are focusing on thelocations of contact of these femurfracture cases and testing the loadcapacity of the dashboard in these areas.

Ultimately, we will propose redesign ofthese stiff areas of the dashboard toreduce knee contact loads. (Funded bythe National Highway TransportationSafety Administration , DavidGrossman, M.D., P.I.)

PREVENTION OF INJURIES IN FALLS BY THE

ELDERLY

Falls are a significant source ofinjury in the elderly, accounting for themajority of fractures of the distal radiusand the hip in this population. Theseinjuries have been reported to accountfor up to 50% of all hospital beds usedfor fracture care. We hypothesized thatcharacteristics of the faller’s shoes maybe partly responsible for the fall. Inparticular, there is a range of shoe-floorinterface friction below which theinterface is too slippery and may causea slip, and above which, is too sticky,and may cause a stumble. Using anepidemiological approach, a largegroup of elderly, healthy people is beingfollowed. These people have hadextensive screening for cognitive and

medical function as part of a largerstudy. When a fall is reported, abiomechanical survey is performed ofthe site, the shoe wear, and the shoe/floor interface at the site of the fall.These measurements include; (1) shoe/floor dynamic frictional coefficient, (2)shoe sole stiffness, (3) shoe sole contactarea, (4) shoe lateral stability, and (5)floor impact characteristics. A total of400 falls are being surveyed along witha matched control group. We may beable to identify characteristics of shoesthat are more likely to result in falls. Inan adjunct study, a new type of energyabsorbing flooring is being studied asan alternative fall injury preventionapproach. This floor system consists oftwo rigid tiles separated by verticalrubber columns (made of recycled shoesole and tire rubber) which retainnormal stiffness under normal loadingbut collapse in buckling under injurycausing loads. This flooring is currentlybeing tested for its ability to attenuateimpact loads from jumping inbasketball and may find use in nursinghomes where falls are prevalent.

(Funded by the National Instituteon Aging, NIH, Marsha Wolf PhD , PI,the Washington Technology Center,and by SATECH Inc, Kirkland, WA).

Figure 2: Lateral x-ray of femur with interlocking nail.


Anterior acromioplasty is nowone of the commonest surgicalprocedures performed on the

shoulder. While the literature describesgeneral satisfaction with this procedure,approximately one in ten of the patientsin the published series sustain anunsatisfactory result.

The goal of this report is tocharacterize the common features of agroup of patients who were unsatisfiedwith their result followingacromioplasty.


This study concerned 74consecutive patients who presented toour consultation service between Juneof 1993 and April 1998 because ofdissatisfaction with their shoulder atleast one year after an acromioplasty.Essentially all of the acromioplasties inthis series had been performed by opensurgery. Patients with acromioplastycarried out as part of other surgicalprocedures (such as cuff repair) werenot included in this series.

There were 40 males and 34 femalesin the study with an average age of 45± 11 years. Seventy six percent of thecases were covered by workers’compensation insurance. Fifty-threepercent of the patients’ occupationsincluded heavy manual work. Theaverage patient smoked two packs ofcigarettes per week. Nineteen percenthad a distal clavicle excision at the time

of their acromioplasty.Each patient was evaluated to

determine factors associated with thefailure, such as postoperative stiffness,surgical complications and diagnosesnot treated by acromioplasty.Standardized anteroposterior, scapulaoutlet and axillary radiographs wereperformed to judge the adequacy of theacromioplasty and to seek additionalpathology.

Each patient completed the SimpleShoulder Test inventory of shoulderfunctions and the SF-36 health statusassessment.

RESULTS

These patients with failedacromioplasty had poor self-assessedhealth status; the average SF 36 scorefor physical role function was only 22%of age and gender matched controlswhile the comfort score was 36% ofcontrols. They also had poor shoulderfunction as documented by the SimpleShoulder Test (Table I). Only 4% couldsleep comfortably. On average thesepatients could perform only 3.5 ± 2.7of the 12 standardized functions.

Our clinical assessment of these 74patients revealed a number of factorsfrequently associated with patientdissatisfaction following acromioplasty(see Table II). Aside from workers’compensation coverage, the mostcommon was significant limitation ofmotion - especially internal rotation

and cross body adduction. Failure of adeltoid reattachment was present in13%. A substantial number of patientshad diagnoses not treated by theacromioplasty. In some patients,multiple factors were present. Nopatient had clinical or radiographicevidence of an inadequate acromialresection. Eleven percent of the patientsdemonstrated a diffuse chronic painsyndrome around the shoulder whichwe could not explain.

DISCUSSION

In 1972 Neer originally describedthe resection of the anteriorundersurface of the acromial processand the coracoacromial ligament totreat what he referred to as‘impingement syndrome,’ whichincluded a variety of pathologies of thesubacromial bursa and rotator cuff.Open and more recently arthroscopicacromioplasty has become the surgicalprocedure most frequently performedon the shoulder.

The most common feature of thisgroup of patients with failedacromioplasty was coverage of theshoulder condition by workers’compensation insurance. Thisassociation has been noted by Ogilvie-Harris, Hawkins and others. It serves toremind us that individuals who dophysical shoulder work and who arehurt on the job have complexoccupational and economic issueswhich cannot be easily separated fromtheir shoulder condition and itsmanagement.

The second most common featurewas persistent shoulder stiffness. In his1934 book, The Shoulder, AmoryCodman spoke of the subacromial areaas being ‘in constant use. It is in fact ajoint without articular cartilages butquite as indispensable. When itssurfaces are thoroughly adherent, halfof the extent of elevation is lost....’. (seefigure 1). If surgery is performed in thisarea, it seems essential to assure thesmoothness of these articulatingsurfaces and to initiate immediatepostoperative motion before restrictiveadhesions begin to form.

Failed Acromioplasty: Investigation of 74 Cases

DAVID G. DUCKWORTH, M.D., KEVIN L. SMITH, M.D., AND FREDERICK A. MATSEN III, M.D.

Table ISimple Shoulder Test Function Percent ableIs your shoulder comfortable while sleeping? 4%Can you lift 8 lbs to shoulder level? 8%Can you toss a softball overhand 20 yards? 9%Can you carry 20 lbs at your side? 19%Does your shoulder allow you to work? 19%Can you toss a softball underhand 20 yards? 27%Can you wash your opposite shoulder? 31%Can you lift 1 lb to shoulder level? 41%Is your shoulder comfortable at rest? 42%Can you reach your back to tuck in your shirt? 49%Can you place your hand behind your head? 50%Can you place a coin on a shelf? 50%


The third most commonassociation of failed acromioplasty wasfailure of deltoid reattachment. Thiscomplication can be functionallydevastating and is very difficult tomanage because of difficulty inachieving secure purchase on theretracted muscle. This complication isavoided by using the ‘deltoid-on’approach and by minimizing theamount of the deltoid origin that isresected. In this regard it is importantto note that in none of these case offailed acromioplasty was the amount ofacromial resection found to beinadequate.

Finally, failures were associated withdiagnoses not treated by acromioplasty.This observation speaks to theimportance of a thorough preoperativeevaluation of all patients with shoulderpain before a surgical procedure iscarried out.

Table II

Associated condition PrevalenceWorkers' compensation claim 76%Stif fness 69%Deltoid Failure 13%Rotator cuff tears 10%Persistent Subacromial Roughness 10%Cervical spine pathology 7%Glenohumeral Arthritis 6%A-C Joint Instability 4%Biceps/Labral Pathology 3%Inadequate acromial excision 0%

CONCLUSION

This investigation provides insightinto conditions which are commonlyassociated with patient dissatisfactionfollowing acromioplasty. It suggeststhat attention to patient selection,complete preoperative evaluation,careful surgical technique, andimmediate postoperative mobilizationof the shoulder may help minimizesimilar failures in the future.

RECOMMENDED READING

Neer C.S., Anterior Acromioplasty forthe Chronic Impingement Syndrome inthe Shoulder. J Bone JointSurg.1972;54A:41-50.

Neer C.S., Impingement Lesions. ClinOrthop. 1983;173:70-77.

Ogilvie-Harris D.J., Wiley A.M.,Sattarian J., Failed Acromioplasty forImpingement Syndrome. J Bone Joint

Surg. 1990;72B:1070-1072.

Seltzer D.G., Wirth M.A., RockwoodC.A., Complications and Failures ofOpen and ArthroscopicAcromioplasties. Operative Tech SportsMed. 1994;2(2):136-150.

Hawkins R.J., Chris T., Bokor D., KieferG., Failed Anterior Acromioplasty: AReview of 51 Cases. Clin Orthop.1989;243:106-111.

Matsen, FA III, Lippitt, SB, Sidles, JA,Harryman DT II, Practical Evaluationand Management of the Shoulder. WBSaunders, Philadelphia 1994.

Codman, E .A The Shoulder, KriegerMalabar, Florida 1934

Figure 1: Codman drawing from his book The Shoulder. Permission granted by Krieger PublishingCompany.


B iomechanical studies havesuggested that non-anatomicpositioning of the humeral

articular surface can adversely affectglenohumeral kinematics.Identification of simple surgical targetsand landmarks for prosthesis positioncould potentially decrease the incidenceof non-anatomic prosthesis placement.As part of our ongoing effort to identifyand minimize factors which contribute

to the failure of shoulder arthroplasty,we implemented a clinical andradiographic analysis of the positionand orientation of the glenoid andhumeral prosthetic components insuccessful and failed shoulderarthroplasties. Further, we attemptedto indentify simple surgical targets forprosthesis placement. We postulate thatthe tuberosity offers an accessible guideto correctly judging the height of thehumeral component, and the glenoidcenterline offers an accessible guide tocorrectly judging the version of theglenoid component.


The inclusion criteria for the“success” group were (1) a totalshoulder arthroplasty performed by thesenior author AND (2) a follow-upevaluation that included excellent self-assessed shoulder function—specifically, the patient’s opinion thathe/she could perform at least ten of thetwelve functions of the Simple ShoulderTest (SST). The inclusion criteria forthe “failed” group were 1) a previousshoulder arthroplasty (by any surgeon)AND 2) patient dissatisfaction with thefunctional outcome from the previousshoulder arthroplasty promptingreferral to our shoulder service.

The average patient age in thesuccess group was 60 (range, 48 to 72)years and in the failure group was 62(range, 49 to 73) years. In the successgroup, there were 67 men and 13women, with 46 right and 34 leftoperated shoulders. In the failuregroup, there were 55 men and 45women, with 54 right and 46 leftoperated shoulders. All of the cases inthe success group were total shoulderarthroplasties. In the failure group, 53were total shoulders and 47 were hemi-arthroplasties.

Both groups of patients hadstandardized radiographic views oftheir shoulder arthroplasty, includingan anteroposterior (AP) of theglenohumeral joint in the plane of thescapula, an AP of the humerus in 35∞of external rotation, and a true axillary

lateral view. Each patient’s radiographswere evaluated by a single observerusing standardized criteria.

Humeral component positions werecompared for all patients in each of thetwo groups. Each humeral prostheticcomponent was characterized as havingthe superior aspect of its articularsurface either five millimeters below,within five millimeters, or fivemillimeters above the greatertuberosity. Each prosthetic componentwas evaluated for gross anteversion orretroversion.

Glenoid component positions in the53 failed total shoulder arthroplastieswere compared to the glenoidcomponent positions in the 80 totalshoulders in the success group. Theglenoid components were evaluated inthe axillary projection in terms of theirversion in relation to the plane of thescapula. Glenoids were classified aseither more than zero degrees ofanteversion, between zero and 20∞ ofretroversion, or more than 20∞ ofretroversion with respect to the planeof the scapula.

Unstable articulations were those inwhich the radiographs showed either afrank dislocation or translation of thehumeral head center more than fivemillimeters away from the glenoidcenter. The direction (anterior,posterior, superior, inferior oranterosuperior) of glenohumeralinstability was noted.

The frequency of the differentobservations between the success andfailure groups was statistically analyzedusing the nonparametric Mann-Whitney U test for unmatched groups.

RESULTS

Our results indicated that thepatients in the failure group couldperform an average of only 2 (range, 0to 4) of the twelve functions of the SST,whereas the patients in the successgroup could perform an average of 11(range, 10 to 12) functions (p<0.0001).Each of the individual patient’sresponses on the SST was significantly

The Position of Prosthetic Components in Failed and SuccessfulShoulder ArthroplastiesMICHAEL J. MOSKAL, M.D., BARRY CAMPBELL, B.S., KEVIN L. SMITH, M.D.,AND FREDERICK A. MATSEN III, M.D.

Figure 1B: Radiograph showing humeralprosthesis within 5 mm of the greatertuberosity.

Figure 1A: Radiograph showing ahumeral prosthesis more than 5 mmabove the greater tuberosity.


different between the two groups(p<0.0001). The greatest differenceswere in strength (lifting a coin toshoulder level, lifting a pound toshoulder level, lifting eight pounds tohead level), range of motion (placinghand behind head with the elbow at theside), and comfort (sleepingcomfortably).

The failed arthroplasties were threetimes more likely (59 %) to have thehumeral head more than fivemillimeters above the tuberosity thanin the successful arthroplasties (20 %)(p<0.0001) (Fig. 1). Gross humeralmalversion (ante- or retroversion) was

present in 18 % of the failures and in 4% of the successes (p<0.05) (Fig. 2).Lucent lines were noted in 60 % of thefailures and in 39 % of the successes(p<0.005).

The glenoid components of failedarthroplasties (n=53) were malverted(angled more than zero degreesanterior or more than 20 degreesposterior to the plane of the scapula)in 46 % of cases, whereas none of theglenoid components was malverted inthe success group (p<0.0001) (Fig. 3).Of the failed arthroplasties, glenoid

components had lucent lines in 76 %,were loose in 66 %, and had moved in26 % in contrast to 29 %, 0 % and 0 %,respectively, in the successfularthroplasties (p<0.0001).Glenohumeral instability was present in43 of the 100 failed cases. Of these, 35% were posterior, 28 % anterior, 33 %anterosuperior, 5 % superior and 0 %inferior. The correlation coefficients foranterior glenohumeral instability withanterior glenoid and humeralcomponent malversion were 0.75 and0.58, respectively. The correlation

Figure 2B: In this radiograph, the patient’sarm is further rotated internally, relative tothe radiographic beam, to reveal the fins,thus demonstrating the gross humeralanteversion..

Figure. 2A: Radiograph of the humerus istaken in 35∞ of external rotation relative tothe radiographic beam. Lateral fins of thehumeral prosthesis are not seen.

Figure 3B: Radiograph of proper glenoid component orientation.

Figure 3A:. Radiograph of an anteverted glenoid component.


coefficient for posterior glenohumeralinstability with posterior glenoidcomponent malversion was 0.5.

DISCUSSION

Suboptimal humeral and glenoidcomponent position was morecommon in the 100 failed arthroplastiesthan in the 80 successful ones. Theseobservations are consistent withbiomechanical studies which suggestthat non-anatomic positioning of thehumeral articular surface can adverselyaffect glenohumeral kinematics. Anexcessively high humeral prosthesisshifts the center of humeral rotationupwards and lead to excessive soft tissuetension which limits humeroscapularmotion. It is also recognized that non-anatomic positioning of the glenoidcomponent predisposes toglenohumeral instability. Excessivetipping of the component increases therisk of instability in the direction oftipping.

This study suggests an importantrelationship between the functionaloutcome of shoulder arthroplasty andthe position of the prostheticcomponents. The tuberosity offers anaccessible guide to correctly judging theheight of the humeral component. Theglenoid centerline offers an accessibleguide to correctly judging the versionof the glenoid component. Thiscenterline is determined at surgery bypalpating the glenoid centering point(Fig.4A). Unaffected by arthritis, theglenoid centering point is a point at thejunction of the anterior glenoid neckat the lateral aspect of the subscapularisfossa, midway between the upper and

lower crura. A line connecting theglenoid centering point to the center ofthe glenoid fossa is the normalizedglenoid centerline which isapproximately 10 degrees retrovertedfrom the plane of the scapula (Fig. 4B).

We hope that the use of simplesurgical targets such as placing the topof the humeral component within fivemillimeters of the top of the tuberosityand the version of the glenoid withinzero degrees and 20 degrees ofretroversion may prove to be usefulguidelines for minimizing problemsrelated to suboptimal componentplacement.

RECOMMENDED READING

Connor, P.M., Levine, W.N., Arroyo,J.S., Ejnisman, B., Pollock, R.G., Flatow,E.L., Bigliani, L.U.: The SurgicalManagement of Failed ShoulderArthroplasty. (Abstract) J. ShoulderElbow Surg. 7 (2):178, 1998.

Lippitt, S.B.; Harryman, D.T.; Matsen,F.A. III.: A Practical Tool for EvaluatingFunction: The Simple Shoulder Test. In:The Shoulder: A Balance of Mobility

and Stability, pp. Edited by F.A. MatsenIII, F.H. Fu, and R.J. Hawkins.Rosemont, IL, American Academy ofOrthopaedic Surgeons, 1992.

Matsen, F.A. III; Lippitt, S.B.; Sidles,J.A.; Harryman, D.T. II.: PracticalEvaluation and Management of theShoulder. Philadelphia, W.B. Saunders,1994.

Matsen, F.A. III; Rockwood, C.A. Jr;Wirth, M.A.; Lippit, S.B.:Glenohumeral Arthritis and ItsManagement. In: The Shoulder, pp.840-964. Edited by C.A. Rockwood Jr,and F.A. Matsen III. Philadelphia, W.B.Saunders, 1998.

Neer, C.S. II, and Kirby, R.M.: Revisionof humeral head and total shoulderarthroplasty. Clin. Orthop., 170: 189-195, 1982.

Wirth, M.A., and Rockwood, C.A. Jr.:Complications of shoulderarthroplasty. Clin. Orthop., 307: 47-69,1994.

Figure 4B: A line drawing of intraoperative determination of glenoid center line approximately 10degrees retroverted in reference to the plane of the scapula.

Figure 4A:. Line drawing of scapula cross-section, perpendicular to the glenoid fossa.This cross-section serves as a point ofreference for 4B.


Characteristics of Patients Treated with Proximal HumeralHemiarthroplasty for Acute Three and Four Part Fractures andthe Correlation with Shoulder Function

RICHARD ROZENCWAIG, M.D., SHADLEY C. SCHIFFERN, B.S., JOHN ANTONIOU, M.D.,HIROSHI TAKEDA, M.D., AND DOUGLAS T. HARRYMAN II, M.D.

Fractures of the proximal humerusare common, occurring in 3 ofevery 10,000 persons. While

approximately 90% of proximalhumerus fractures are minimallydisplaced and do well with conservativemanagement, severely displacedproximal humerus fractures can have adevastating effect on upper extremityfunction. While some authors haverecommended open reduction andinternal fixation for certain proximalhumerus fractures, early techniqueswere generally unsatisfactory andresulted in poor outcomes. Even withthe improved understanding of boththe vascular anatomy andbiomechanics of the shoulder, theresults of open reduction and internalfixation of severely displaced fractureswith the more modern techniques andfixation devices are often unpredictable.Prosthetic hemiarthroplasty hasbecome an accepted method oftreatment for several well-definedsubsets of these fractures, includingfour-part fractures, fracture-dislocations, head-splitting fractures,and fractures with an impression defectencompassing more than 50% of the

humeral head. Also, some patients withthree-part fractures may be consideredfor a hemiarthroplasty as a result oftheir age, poor bone quality, orextensive comminution.

Several groups have reported theresults of humeral head replacement forthese injuries. Neer’s experienceremains the largest in the orthopaedicliterature, attesting to the relativeinfrequency of these injuries andperhaps a general lack of experience intheir treatment. The Neer prosthesishas been used most widely forreconstruction of the severely fracturedshoulder. However, the results andfrequencies of complications followinghemiarthroplasty for the treatment offractures have varied widely. Despitethese varied functional outcomes,primary proximal humeralhemiarthroplasty for acute three- andfour-part fractures usually prevents thedevelopment of a painful shoulder andavoids the potential complications ofclosed treatment and open reductionwith internal fixation.

The purpose of this retrospectivestudy was to characterize the long-termresults with Neer prosthetic

hemiarthroplasty for three- and four-part proximal humerus fractures. Wealso assessed the correlation betweenthe various ingo parameters andsurgical outcome.


The records from the University ofWashington Medical Center werereviewed revealing 37 patients who hada hemiarthroplasty for an acute three-or four-part fracture of the proximalhumerus treated with the Neerprosthesis. Twenty-eight of those 37patients responded to surveys thatassessed personal demographic data,general function, and the presence ofstandard comorbidities. Furthermore,general health status was evaluatedusing the SF-36 Health StatusInstrument; the function of the injuredand contralateral shoulders wasdetermined using the Simple ShoulderTest (SST).

RESULTS

The mean surgical delay was 20days. The average hospitalization was5.6 ± 2.6 days. The most commonmechanism of injury was a fall (70%)followed by a motor vehicle accident(14%). Ten percent of the injuriesinvolved a legal or industrial claim.Greater tuberosity comminution wasnoted in 50% of the fractures, whilelesser tuberosity comminution wasidentified in 31% of the fractures.Nineteen percent of the patientssustained a rotator cuff tear at the timeof fracture. Extension of the fractureinto the calcar region of the proximalhumerus was present in 19% of cases.

Two superficial infections werenoted postoperatively; these did notnegatively influence the outcome oftreatment. Heterotopic ossification wasidentified in 13% of the postoperativeradiographs evaluated; greatertuberosity union was noted in only 50%of patients. More than half (55%) ofpatients felt that they were stillimproving in their function. Twenty-

Mean Correlation p value

Coefficient

Analog pain scale 3.4 ± 3.1 -0.729 0.0005

# of medications 3.2 ± 2.7 -0.600 0.0095

# of major comorbidities 2.6 ± 1.8 -0.499 0.04

SF-36 Bodily Comfort Score 111.2 ± 34.7 0.654 0.003

SF-36 Mental Health Score 80.4 ± 36.0 0.605 0.0087

Current satisfaction scale

(1=least, 5=most) 3.4 ± 1.6 0.540 0.023

Table 1: Correlation of patient characteristics with shoulder function.


eight percent of patients stated that theyhad difficulty with personal hygiene.One half of patients in this studyresumed their usual pre-injury work oractivities. Eighty-six percent (86%) ofthe patients in this study were satisfiedwith the results of their surgery.

Of the twelve shoulder functionsassessed with the SST, respondents wereable to perform an average of 10.3 ±2.7 functions with their uninjuredshoulder. On the other hand, patientswere only able to perform an averageof 6.6 ± 3.7 functions with theirrepaired shoulder (p < 0.0001) (Figure1).

The following six items were foundto have statistically significantassociation with the alteration inshoulder function following fracturetreatment: analog pain scale, numberof medications, number of majorcomorbidities, SF-36 Bodily ComfortScore, SF-36 Mental Health Score, andpatient satisfaction (Table I).

DISCUSSION

Proximal humeral hemiarthroplastyfor the treatment of three- and four-part fractures is a technicallydemanding procedure. The successfollowing this procedure is partydependent on sound operativetechnique based on the principles of

restoration of humeral length,anatomical fixation of the tuberositieswith ultimate healing to the shaft, andplacement of the prosthetic componentin the appropriate degree ofretroversion. Increased attention andtuberosity fixation, bone grafting, andearly restoration of motionwillhopefully improve our results in thefuture.

RECOMMENDED READING

Cuomo F, Flatow EL, Maday MG, MillerSR, McIlveen SJ, Bigliani LU. Openreduction and internal fixation of two-and three-part displaced surgical neckfractures of the proximal humerus. JShoulder Elbow Surg 1992;1(6):287-295.

Green A, Barnard L, Limbird RS.Humeral head replacement for acute,four-part proximal humerus fractures.J Shoulder Elbow Surg 1993;2(5):249-254.

Moeckel BH, Dines DM, Warren RF,Altchek DW. Modularhemiarthroplasty for fractures of theproximal part of the humerus. J BoneJoint Surg 1992;74A(6):884-889.

Tanner MW, Cofield RH. Prostheticarthroplasty for fractures and fracture-

dislocations of the proximal humerus.Clin Orthop 1983;179:116-128.

Zyto K, Wallace A, Frostick SP, PrestonBJ. Outcome after hemiarthroplasty forthree- and four-part fractures of theproximal humerus. J Shoulder ElbowSurg 1998;7(2):85-89.

0

2

4

6

8

10

12

Injured Shoulder Contralateral Shoulder

SS

T 6.6 ± 3.7

10.3 ± 2.7

Figure 1: Shoulder function as assessed by the SST.


Glenohumeral arthroplasty is aneffective procedure formanaging conditions in which

the normal glenohumeral joint surfaceis lost. Proper functioning of thearthroplasty depends on the glenoidjoint surface retaining the shape inwhich it was designed. The purpose ofthis investigation was to characterizethe changes in glenoid joint surfacewhich occur after in vivo use. Thisreport summarizes our findings on 37glenoids retrieved from patients at thetime of revision surgery for failedshoulder arthroplasty.


This study included 19 glenoids

retrieved by the authors in their practiceof revision arthroplasty as well as 18glenoids retrieved by other surgeons.Each retrieved glenoid was visuallyinspected and a standard set ofobservations were recorded. Eachglenoid was then mounted in a holdingdevice, coated to optimize reflectivityand then scanned using a laser scanneraccurate to 0.1 mm using a system. Thesupport for the mounted glenoids wasmotorized by a 2 axis table with ahorizontal resolution of ± 10micrometers. The scanner system (fig.1) measured distances with a nominalresolution of 3 microns and a frequencyof 2000 acquisitions per mm.

Physical Changes in Polyethylene Glenoid Components After InVivo Use

MARIUS M. SCARLAT, M.D., WILLIAM M. DOUGHERTY, PH.D., KEVIN L. SMITH, M.D.,JOHN A. SIDLES, PH.D., JAMES D. BRUCKNER, M.D., AND FREDERICK A. MATSEN III, M.D.

Scanning data were displayed ingraphic form using the LabViewsoftware. Contour maps wereconstructed from the surface scans. Thedifference between the scanned contourand the spherical contour of a newglenoid was displayed using NIH image.Each glenoid was compared to theshape of an unimplanted componentof the same manufacturer and size. Thetheoretical balance-stability angle(BSA) is defined as the maximal anglethat the net humeral joint reaction forcecan make with the glenoid center linebefore glenohumeral dislocation occurs(fig. 2). The BSA angle was calculatedfor each retrieved polyethylene glenoidfrom the slope of the glenoid surfaceand plotted as contour lines. Thedifferences between the BSA plots forthe new and used glenoids weredisplayed for each specimen.

RESULTS

The findings in the retrievedglenoids included: (1) surface changesof the polyethylene glenoid (polishing,scratches, pits, delamination, anddiscoloration) and (2) changes in theshape of the polyethylene glenoid (lossof the glenoid lip, flattening, andfracture). The frequency of the surfaceand shape changes are graphicallydisplayed in the Figure 3.

The most common surface changeswere polishing (33 out of 37) andscratching (28 out of 37). These surfacelesions are similar to those described forhip and knee polyethylene components.Delamination was found in 6 casesalways being associated with fracture ofthe components. Discoloration wasfound in 11 cases.

The changes in shape revealed bythe laser scanner were not alwaysobvious to the naked eye. The radiusof curvature of used glenoids wasalways greater than that of unusedglenoids of the same size. In vivo useoften transformed the spherical shapeof the glenoid into an ellipsoidal shape.In many cases the changes in the jointsurface were of sufficient magnitude toFigure 1: The scanner system used in this experiment.


Figure 3: The frequency of the surface and shape changes are graphically displayed.

34

29

20

17

6

1110

56

0

5

10

15

20

25

30

35

Polishing FineScratches

CoarseScratches

Pits Delamination Discoloration Rim Erosion Flattening Fractures

Fig 3 : Physical changes in the glenoid surface and shape

affect the calculated balance stabilityangle.

Figure 4 (a, b) shows the findingsfor an unused glenoid. Figure 5 (a,b,c)shows the changes in a used glenoid asrepresented by a) inspection, b) laser-scanner, c) the BSA. The deviation fromsphericity and the modified BSA areapparent.

DISCUSSION

Substantial literature describes thechanges in of polyethylene componentsin hips and knees, but changes thepolyethylene of shoulder arthroplastyhave not been thoroughly described.This study presents the surface andshape changes for a group of glenoidcomponents retrieved after in vivo use.While the observed changes in theglenoid surface were similar to thoseseen in hip and knee components(polishing, scratches, pits,delamination, and discoloration) thesechanges would appear unlikely to alterthe function of the polyethylenecomponent unless there was anaccompanying change in glenoid shape.The shape changes were sometimes

Figure 2: The BSA angle was calculated for each retrieved polyethylene glenoid from the slope of theglenoid surface and plotted as contour lines.


obvious to the eye and at other timeswere sufficiently subtle that they wereonly detectable by the laser scanner. Invirtually all cases there was a loss ofsphericity which was frequently of asufficient magnitude to change thetheoretical balance-stability anglecalculated from the surface contour.The importance of the balance-stabilityangle data lies in the demonstration ofthe potential mechanical consequencesof the changes in shape.

This study was limited by ourinability to collect clinical data on manyof the specimens. As a result we wereunable to correlate the changes insurface and shape with clinical use,

patient symptoms, duration ofimplantation, and other importantattributes.

The strength of this investigationlies in the fact that it is the firstquantitative observation of changes inshape occurring in polyethyleneglenoids after in vivo use. This study hasindicated that 1) these glenoidcomponents which had sphericalsurfaces when they are implantedusually did not have spherical surfaceswhen they were retrieved after in vivouse, 2) the changes in glenoid surfaceshape were often not evident on visualinspection, 3) the magnitude of thechanges in shape may be sufficient to

affect the stability offered by the glenoidcomponents.

Figure 5 (a,b,c): The changes in a used glenoid as represented by a) inspection, b) laser-scanner, c) the BSA. The deviation from sphericity and the modifiedBSA are apparent.

Figure 4 (a, b): The findings for an unused glenoid.


Department of Orthopaedics Faculty

Frederick A. Matsen III, M.D.Professor and Chair

Christopher H. Allan, M.D.Assistant Professor

Carlo Bellabarba, M.D.Assistant Professor

Stephen K. Benirschke, M.D.Associate Professor

Stanley J. Bigos, M.D.Professor

James T. Bruckner, M.D.Assistant Professor

Howard A. Chansky, M.D.Assistant Professor

Jens R. Chapman, M.D.Associate Professor

Randal P. Ching, Ph.D.Assistant Professor

John M. Clark, Jr., M.D., Ph.D.Associate Professor

Ernest U. Conrad III, M.D.Professor

Mohammad Diab, M.D.Assistant Professor

David R. Eyre, Ph.D.Professor

Theodore K. Greenlee, Jr., M.D.Associate Professor

Douglas P. Hanel, M.D.Associate Professor

Sigvard T. Hansen, Jr., M.D.Professor

Children’s Hospital andMedical Center4800 Sand Point Way NESeattle, WA 98105(206) 526-2000

Harborview Medical Center325 Ninth AvenueSeattle, WA 98104(206) 731-3462

University of WashingtonMedical CenterBone and Joint Center4245 Roosevelt Way NESeattle, WA 98105(206) 598-4288

VA Puget Sound Health Care System1660 South Columbian WaySeattle, WA 98108(206) 764-2215

Department of OrthopaedicsUniversity of Washington1959 N.E. Pacific StreetBox 356500Seattle, Washington 98195-6500

Phone: (206) 543-3690Fax: (206) 685-3139

Affiliated Institutions

Douglas T. Harryman II, M.D.Associate Professor

M. Bradford Henley, M.D.Associate Professor

Roger V. Larson, M.D.Associate Professor

William J. Mills, M.D.Assistant Professor

Sohail K. Mirza, M.D.Assistant Professor

Vincent S. Mosca, M.D.Associate Professor

Sean E. Nork, M.D.Assistant Professor

John O’Kane, M.D.Assistant Professor

Milton L. Routt, Jr., M.D.Associate Professor

Bruce J. Sangeorzan, M.D.Professor

John A. Sidles, Ph.D.Professor

Peter T. Simonian, M.D.Associate Professor

Douglas G. Smith, M.D.Associate Professor

Kevin L. Smith, M.D.Assistant Professor

Kit M. Song, M.D.Assistant Professor

Lynn T. Staheli, M.D.Professor Emeritus

UNIVERSITYOF WASHINGTON

SCHOOL OFMEDICINE

Michael D. Strong, Ph.D.Research Professor

Carol C. Teitz, M.D.Associate Professor

Allan F. Tencer, Ph.D.Professor

Thomas E. Trumble, M.D.Professor

Jiann-Jiu Wu, Ph.D.Research Associate Professor

Adjunct FacultyBasia R. Belza, R.N., Ph.D.

Associate Professor, Physiological Nursing

Anne-Marie Bollen, Ph.D.Assistant Professor, Orthodontics

Charles H. Chesnut, M.D.Professor, Nuclear Medicine

Gregory C. Gardner, M.D.Associate Professor, Rheumatology

Thurman Gillespy, M.D.Associate Professor, Radiology

Daniel O. Graney, Ph.D.Associate Professor, Biological Structure

John C. Hunter, M.D.Assistant Professor, Radiology

Frederick A. Mann, M.D.Professor, Radiology

Susan M. Ott-Ralph, M.D.Associate Professor, Division ofMetabolism

Wendy Raskind, M.D., Ph.D.Associate Professor, General InternalMedicine

Michael L. Richardson, M.D.Professor, Radiology

Robert Schoene, M.D.Professor, Medicine

Peter A. Simkin, M.D.Professor, Medicine

Tony J. Wilson, M.D.Professor, Radiology

Joint FacultyJohn E. Olerud, M.D.

Professor, Division of Dermatology

Nicholas B. Vedder, M.D.Associate Professor, Plastic Surgery

Affiliate FacultyDavid A. Boone, C.P.

Director and Co-Principal Investigator,Prosthetics Research Study

Sarah E. Jackins, R.P.T.Rehabilitation Medicine


1999 Department of OrthopaedicsIncoming Faculty

Carlo Bellabarba, M.D.

Upon completion of his one yearSpine Fellowship atNorthwestern University in

Chicago, Carlo Bellabarba will bejoining the University of Washington asan Assistant Professor specializing inspine surgery.

Dr. Bellabarba received his B.Sc. inBiochemistry from McGill University inMontreal, Quebec, Canada, where healso attended medical school. Hisinternship and residency were spent atRush-Presbyterian-St. Luke’s MedicalCenter in Chicago.

As well, he has completed afellowship with the OrthopaedicTrauma Service of Tampa GeneralHospital in Tampa, Florida.

Dr. Bellabarba sees patients at theUniversity of Washington MedicalCenter, at UWMC-Roosevelt,Harborview Medical Center, andVeteran’s Hospital.

Christopher Allan, M.D.

Dr. Allan received his medicaldegree from NorthwesternUniversity in 1992 and

completed his training in orthopaedicsurgery at the University of Chicago in1997. He pursued subspecialty trainingin hand surgery and microsurgery atBaylor College of Medicine with Dr.David Lichtman, an internationallyknown leader in surgery of the wrist.Dr. Allan has an interest in wristproblems, as well as in general hand andmicrosurgery. He is also active inresearch aimed at improving woundhealing after injury.

Dr. Allan sees patients at Universityof Washington Medical Center,UWMC-Roosevelt, HarborviewMedical Center, and Children’sHospital and Regional Medical Center.


Graduating ResidentsClass of 1999

John Michelotti, M.D., will be travellingto Vail, Colorado where he willparticipate in a one year SportsMedicine fellowship and thereafterbegin private practice in Montana.

Craig Boatright, M.D., has accepted aposition as a fellow in spine surgery atEmory University in Atlanta, Georgia.

Jeff Garr, M.D., will be travelling toDaly City, California where he willparticipate in a one year Spinefellowship.

Julie Switzer, M.D., will be spending ayear in Vail, Colorado with VailOrthopaedics on a fellowshipspecializing in sports.

Tom Chi, M.D. and his family will begoing to Kenya to work at a non-profithospital, the Kikuyu OrthopedicRehabilitation Center, for six months.Afterward, he will start a six month footand ankle fellowship in Birmingham,Alabama.


Incoming Residents

Ben DuBois: Ben attended OregonState University where he received hisB. S. in biology. He received his medicaldegree from the University of SouthernCalifornia. He is currently finishing hisinternship under the direction of theDepartment of Surgery at theUniversity of Washington. His outsideinterests include weightlifting, running,basketball and photography.

Andrew Howlett: Andy attended theUniversity of Wisconsin at Madisonwhere he obtained a B.S. inbiochemistry. He attended theUniversity of Washington MedicalSchool where he received his medicaldegree. He is currently a surgical internat the University of Washington.Outside interests include fly fishing,running and basketball.

Guy Schmidt: Guy attended BrownUniversity where he received a B. S. inbiochemistry. He received his medicaldegree from Johns Hopkins University.He is currently completing his surgicalinternship at the University ofWashington. His interests include rockclimbing and alpine skiing.

Brian Shafer: Brian received his B. S.in chemistry from the University ofCalifornia at San Diego. He attendedJohns Hopkins University where heearned his medical degree. He iscurrently a surgical intern at theUniversity of Washington. Brian enjoyssports in his spare time.

Emma Woodhouse: Emma attendedStanford University where she receiveda B. S. in biology. Following herundergraduate education she enteredthe University of Southern Californiawhere she received her medical degree.She is currently a surgical intern at theUniversity of Washington. Personalinterests include tennis and hiking.


MAY 1998 THROUGH APRIL 1999

Friends of Orthopaedics

AFZ Atrrosef-zentrumAnthony AgtarapAircastChristopher AllanFrank AlvineDavid and Jean AndersonCarl AndrewsSamuel BakerWilliam BarrettTimothy BealsSteve BenirschkeGary BergmanBill BoettcherDavid BrokawJohn BrownJames BrucknerErnest BurgessSusan CeroHoward ChanskyJens ChapmanChristensen & JensenGary ClanceyJohn M. ClarkJohn ColettiErnest ConradRichard ConradJay CraryJim CrutcherMark DalesJerome DebsDavid DenekaDePuy, Inc.Richard DimondAnne ElliottLeonard ElyDon EricksenEd Farrar

We express our appreciation to all who have contributed to the work of the Department of Orthopaedics over the past year.Your assistance makes possible special research activities, educational programs, and other projects that we could not offerwithout this extra support from our alumni, faculty, and friends in the community. We owe a special thanks to the Universityof Washington Resident Alumni who have made significant contributions to help further the education of our currentresidents.

Contributors to Departmental Research and Education

Maribeth FischerHarold ForneyJonathan FranklinMike GannonThomas GreenSandy and Ted GreenleeDoug P. HanelJeff HansenSigvard T. HansenDouglas T. HarrymanHealthwiseM. Brad HenleyScott HormelHowmedicaRobin IrishSarah JackinsDavid KargesCarleton KeckCathy KeatingRichard KirbyHoward KingJo KnightReggie KnightPhil KregorWally KrengelRon KristensenDennis KvideraDavid LevinsohnNeil MakiPeter MandtRick and Anne MatsenKeith MayoDaniel McClureSohail MirzaPeter MitchellVince MoscaMark Murphy


Friends of Orthopaedics

Northwest BiometBill OmbremskyJohn O’KaneMark OlsonWilliam OppenheimOrthologicLarry PageEli PowellSteven RatcliffeSteve ReedKenneth ReesmanRohne-poulec Rorer PharmaceuticalDonald RoonChip RouttJohn SackMichael SailerRonald SandlerBruce SangeorzanRichard SemonStewart SchamPeter SimonianKevin L. SmithKit SongLyle SorensonLynn T. StaheliJohn StriplinSurgical DynamicsJames SwansonMarc SwiontkowskiSynthesWilliam ThiemeMichael ThorpeThomas TrumbleCarol TeitzHugh ToomeyMartin TullusUniversity of CalgaryEric VanderhooftTheodore WagnerKarin WahlstromDavid WaxmanJohn WestRichard WesterbrookThomas WilderJay WinzenreidRobert Wood

Brodie WoodZimmerJoseph Zuckerman


NATIONAL RESEARCH GRANTS

DEPARTMENT OF ORTHOPAEDICS

National Institutes of Health (NIH)

Collagens of Cartilage and the Intervertebral DiscDavid R. Eyre, Ph.D.

Imaging of Molecules by Oscillator-Coupled ResonanceJohn A. Sidles, Ph.D.

Pathology of Inborn Skeletal DiseasesDavid R. Eyre, Ph.D.

Skeletal DysplasiasDavid R. Eyre, Ph.D.

Salvage vs. Amputation Following Limb Threatening InjuryDouglas G. Smith, M.D.

Veterans Affairs Medical Center Review Grants

Rehabilitation Research and Development Center of Excellence forLimb Loss Prevention and Prosthetic Engineering

Bruce J. Sangeorzan, M.D.

Biomechanics of Foot Deformities and Alternatives for Surgical CorrectionBruce J. Sangeorzan, M.D.

Centers for Disease Control

Low Speed Cervical Whiplash InjuryAllan F. Tencer, Ph.D.

Orthopaedic Research and Education Foundation (OREF)

Bristol-Myers Squibb/Zimmer Institutional GrantFrederick A. Matsen III, M.D.

Randomized Multicentered Clinical Trial of Distal Radius Fracture TreatmentDoug P. Hanel, M.D.




Orthopaedic Research and Education Foundation (OREF)

Predictable Spine and Occupational Health DataStanley J. Bigos, M.D.

The Shoulder Function and Health Status of Individuals with DocumentedRotator Cuff Tears Before and After Treatment: A MulticenteredProspective Study

Douglas T. Harryman II, M.D.

American Society for Surgery of the Hand

Prospective Randomized Clinical Trial of Hand Therapy Following CarpalTunnel Surgery

Thomas E. Trumble, M.D.

Axonal Sprouting from Intact Peripheral NervesThomas E. Trumble, M.D

Cervical Spine Research Society

Cervical Spine Instability as Measured by Neural Space OcclusionSohail Mirza, M.D.

Genetics Institute

A Quality of Life Assessment in Patients with Open Tibial FracturesSohail Mirza, M.D.

National Highway Traffic Safety Administration

Age-Dependent Properties of the SpineRandal P. Ching, Ph.D.

Ostex International, Inc.

Molecular Markers of Connective Tissue DegradationDavid R. Eyre, Ph.D.




Pfizer, Inc.

Pfizer StudyDavid R. Eyre, Ph.D.

Spinal Dynamics Corp.

Spinal Dynamics Baboon TestingRandal P. Ching, Ph.D.

BRD Functional Disc Space Prosthesis Development ProjectRandal P. Ching, Ph.D.

Surgical Dynamics, Inc.

Biomechanical Testing of a Spinal Cage for Lumbar Spine Interbody FusionAllan F. Tencer, Ph.D.

AO/ASIF Research Commission

Outcome Following Fracture of Tibial PlafondSohail K. Mirza, M.D.

Washington Technology Center

Impact Force of Flooring SurfacesAllan F. Tencer, Ph.D.

1999 Research Report www - UW Orthopaedics and Sports …...1999 ORTHOPAEDIC RESEARCH REPORT 7...

Documents

Transcript of 1999 Research Report www - UW Orthopaedics and Sports …...1999 ORTHOPAEDIC RESEARCH REPORT 7...