UNIVERSITY OF WASHINGTON Department of Orthopaedics · 2019. 7. 30. · Contents Foreword

University of WashingtonDepartment of Orthopaedicsand Sports Medicine

2003 Research

Report

UNIVERSITY OF WASHINGTON

Department of Orthopaedicsand Sports Medicine

2003 Research Report

Cover Illustration: “La Cathédrale” by Auguste RodinS. 1001, pierre, 69 x 29.5 x 31.8 cmPhotograph: Eric and Petra HesmergMusee Rodin

Department of Orthopaedics and Sports MedicineUniversity of WashingtonSeattle, WA 98195

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

DESIGN & LAYOUT:Fred Westerberg

UNIVERSITYOF WASHINGTON

SCHOOL OFMEDICINE

Contents

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

Neuropathology By Sunrise ...................................................................................................................................... 2

Paul Brand, M.D.

UW Medicine Hand Institute ................................................................................................................................... 5

Visiting Lecturers ....................................................................................................................................................... 6

Back Pain in Former Intercollegiate Rowers: A Long-Term Follow-Up Study ....................................................... 7

Carol C. Teitz, M.D., John W. O’Kane, M.D., and Bonnie K. Lind, M.S.

Corticosteroid Versus Synvisc (Hylan GF-20) Injections for Knee Osteoarthritis: A Prospective, RandomizedTrial, and Results of a Retrospective Study Evaluating Adverse Reactions to Repeated Courses of Synvisc ...... 11

Seth S. Leopold, M.D., Brigham B. Redd, M.D., Winston J. Warme, M.D., Paul A. Wehrle, M.D.,Patrick D. Pettis, L.V.N., and Susan Shott, Ph.D.

The Association Between Smoking and Nonunion of Long Bone Fractures ....................................................... 15

David P. Barei, M.D., F.R.C.S.(C), Sarah Holt, M.S.P.H., Sean E. Nork, M.D., Carlo Bellabarba, M.D.,William J. Mills, M.D., and Bruce J. Sangeorzan, M.D.

The Value of the Ankle-Brachial Index for Diagnosing Arterial Injury Following Knee Dislocation: AProspective Study..................................................................................................................................................... 18

William J. Mills, M.D., David P. Barei, M.D., F.R.C.S.(C), Anthony M. Buoncristiani, M.D.,and Patrick McNair, M.D.

When Do Medical Students Learn Musculoskeletal Medicine? ............................................................................ 21

Gregory A. Schmale, M.D.

RNA Interference Suppresses Expression of EWS/FLI-1 in Ewing’s Sarcoma Cells ............................................ 23

Howard A. Chansky, M.D., Anna Zielinska-Kwiatkowski, M.D., Yoshito Matsui, M.D., Ph.D.,M. Blackburn, B.S., Ernest U. Conrad, M.D., Jim D. Bruckner, M.D., and Liu Yang, Ph.D.

Central Screw Placement in Simulated Scaphoid Waist Fractures: A Biomechanical Study ............................... 26

Wren V. McCallister, M.D., Jeff Knight, B.S., Robert Kaliappan, M.D., and Thomas E. Trumble, M.D.

Surgical Intervention for the Clawed Hallux Deformity ....................................................................................... 29

Fady N. Elias, B.S., William R. Ledoux, Ph.D., Tracy J. Yuen, B.S., Soren L. Olson, B.S.,Randal P. Ching, Ph.D., and Bruce J. Sangeorzan, M.D.

Association Between Funding Source and Study Outcome in Orthopaedic Research ........................................ 33

Seth S. Leopold, M.D., Winston J. Warme, M.D., E. Fritz Braunlich, M.D., and Susan Shott, Ph.D.

Augmenting Skeletal Integrity at Senescence via “Restful” Physical Activity ....................................................... 36

Sundar Srinivasan, Ph.D., Steven C. Agans, B.S., Katy King, B.S., Nicholas Y. Moy, B.S.,Sandra L. Poliachik, Ph.D., and Ted S. Gross, Ph.D.

Regrowth of Amputated Rabbit Digit Tips ............................................................................................................ 39

Christopher H. Allan, M.D., Traci Barthel, M.D., and Brian P. Rubin, M.D., Ph.D.

Diagnosis and Treatment of Craniocervical Dissociation: A Series of 17 Consecutive Survivors Over SevenYears .......................................................................................................................................................................... 42

Carlo Bellabarba, M.D., Sohail K. Mirza, M.D., G. Alexander West, M.D., Ph.D., Frederick A. Mann, M.D.,David W. Newell, M.D., and Jens R. Chapman, M.D.

Acetabular Fractures in Senior Patients .................................................................................................................. 45

Alysandra Schwarz, M.D. and M.L. Chip Routt, Jr., M.D.

Preoperative Factors Associated with Greater Improvement Following Humeral Hemiarthroplasty ................ 49

Carolyn M. Hettrich, B.S., Edward J. Weldon III, M.D., Richard S. Boorman, M.D.,and Frederick A. Matsen III, M.D.

The Effect of Foot Position on Mechanism of Fracture in Automobile Frontal Collisions ................................. 52

M. Assal, M.D., P. Huber, B.Eng., Allan F. Tencer, Ph.D., E. Rohr, C. Mock, M.D., and R. Kaufman, M.Sc.

Effect of Therapeutic Footwear on Foot Reulceration in Patients with Diabetes: A Randomized ControlledTrial .......................................................................................................................................................................... 55

Douglas G. Smith, M.D., Gayle E. Reiber, M.P.H., Ph.D., Carolyn Wallace, Ph.D., Katrina Sullivan, D.P.M.,Shane Hayes, C.Ped., Christy Vath, B.S., Matthew L. Maciejewski, Ph.D., Onchee Yu, M.S.,Patrick J. Heagerty, Ph.D., and Joseph LeMaster, M.D.

COL11A2 Collagen Gene Transcription is Differentially Regulated by EWS/ERG Sarcoma Fusion Protein andWild-Type ERG ........................................................................................................................................................ 58

Yoshito Matsui, M.D., Ph.D., Howard A. Chansky, M.D., Fariba Barahmand-pour, Ph.D.,Anna Zielinska-Kwiatkowska, M.D., Noriyoki Tsumaki, Akira Kyoui, Hideki Yoshikawa, Liu Yang, Ph.D.,and David R. Eyre, Ph.D.

Cross-Linking of Collagens II, IX and XI in Chondrocyte Cultures As a Monitor of Normal MatrixAssembly .................................................................................................................................................................. 60

Russell J. Fernandes, Ph.D. and David R. Eyre, Ph.D.

Department of Orthopaedics and Sports Medicine Faculty ................................................................................. 63

Department Addresses ............................................................................................................................................ 64

Graduating Residents .............................................................................................................................................. 65

Incoming Residents ................................................................................................................................................. 66

New Faculty ............................................................................................................................................................. 67

Research Grants ....................................................................................................................................................... 68

Contributors to Departmental Research and Education ....................................................................................... 73

Alumni ..................................................................................................................................................................... 75

Endowments ............................................................................................................................................................ 77

2003 ORTHOPAEDIC RESEARCH REPORT 1

Foreword

Best wishes,

Frederick A. Matsen III, M.D.Residency Alumnus, 1975

T his year’s edition of theDepartment of Orthopaedicsand Sports Medicine Research

Report is dedicated to our outstandinghand program and to the newlyestablished UW Medicine HandInstitute at Roosevelt. It is fitting toinclude an article from our clinicalfaculty member, Paul Brand.Orthopaedic surgeons get to know Dr.Brand through his classic articles on thehand. Paul Brand was born in 1914 inthe mountains of India, where hisparents were missionaries. He went toLondon, England, for his education andhad his medical and surgical trainingat London University. In 1946 Paul andhis wife, Margaret, who is also a doctor,went to India, where Paul taughtsurgery at the Christian MedicalCollege and Hospital in Vellore. Dr.Brand became the first surgeon in theworld to use reconstructive surgery tocorrect the deformities from leprosy inthe hands and feet. He was electedHuntarian Professor of the RoyalCollege of Surgeons of England in 1952.In 1961 he was honored by QueenElizabeth with appointment as“Commander of the Order of theBritish Empire.” He was also therecipient of the Distinguished ServiceAward from the Department of Healthand Human Services, United StatesPublic Health Service. But, his earlymastery of hand biomechanics is onlypart of the story. Outside orthopaedicshe is known for his spiritualappreciation for the human bodyexpressed in books such as Fearfully andWonderfully Made, and Pain: The GiftNobody Wants. I suspect you will enjoyhis article in this report, and if so, youmay enjoy these books and some of hisother articles, such as “Low Momentsin My Surgical Career” and “ThePursuit of Happiness.” He is truly arenaissance man.

In searching for the cover art tohonor our hand program, I discoveredthis 1908 sculpture at the MuseeAuguste Rodin, 77, Rue de Varenne inParis, called La Cathédrale. Like Brand,Rodin was awed by the human form –you know some of his other works thatcelebrate the body, such as The Thinker,The Kiss, The Age of Bronze, and perhapsMan with a Broken Nose. Rodin was

born in 1840 and died in 1917 whenDr. Brand was three years old. LaCathédrale is composed of two upraisedright hands coming together to evokethe spiritual space of ancient cathedrals.

The U.W. hand program is led bythree orthopaedic hand surgeons, ChrisAllan, Doug Hanel and Tom Trumbleand two plastic surgeons, Nick Vedderand Arshad Muzaffar. This teamincludes the first full-time handsurgeons at the University ofWashington. The members of this teamare prolific in their teaching, theirresearch, their publications and theirclinical care. To learn more about “HowD’ya Sew an Amputated Finger BackOn?,” “Amazing Operation Gives BoyTwo New Thumbs,” “Single-PortalEndoscopic Carpal Tunnel ReleaseCompared with Open Release,” “Man’sHand Reattached after Daring Rescueat Sea,” and to discover the identity ofthe winner of the Bovill Award for thebest paper of the year given at theOrthopedic Trauma Association annualmeeting, please see http://w w w. o r t h o p. w a s h i n g to n . e d u /hand_wrist. The last time we had handson our Research Report Cover was in1994. You can download this PDF athttp://www.orthop.washington.edu/dept. Can you identify the artist beforelooking at the answer on the copyrightpage?

In this year’s report you will meetour three new faculty members, help uswelcome our new residents andcelebrate the graduation of our chiefresidents. You will also get a chance toexperience some of the breadth anddepth of our research program. We haveselected some basic science articles onthe cell and molecular biology ofcartilage and sarcomas. Other articlesshow the importance of rest in buildingbone and the ability of amputated digitsto regenerate. Several articles concernthe foot, ranging from the relation offoot position to injury, and treatmentof big toe deformities, to the value ofspecial footwear for individuals withdiabetes. You can also learn of theimpairment of bone healing bysmoking and the value of special bloodpressure measurements to avoidamputation after knee dislocation. Ourfaculty and residents have presented

answers to other questions, such as“How do folks do when the head isdislocated from the neck?”; “How canyou predict the result of a shoulderreplacement?”; “Does the source offunding affect the outcome ofresearch?”; “Is Synvisc better thancortisone for knee arthritis?”; “How canfractures of the scaphoid bestabilized?”; “How do older people farewith fractures of the hip socket?”; “Doesrowing cause back pain?”; and “Whendo medical students learn about bonesand joints?” While this is only a sample,it presents quite a variety.

We are a bigger and strongerDepartment now than ever before. Ourexcellence is the result of (1) the manysuperior staff, residents, students andfaculty that work in the Department,(2) the patients that trust us to restoretheir comfort and function, and (3)those generous individuals and groupsthat give us the financial support tocarry on our work. With your help wewill continue to pursue new and betterways to help people stay ‘fit for life!’

Dr. Paul Brand

2 2003 ORTHOPAEDIC RESEARCH REPORT

It was about fifty-five years ago. I wasworking at the Christian MedicalCollege at Vellore in south India. I

had completed my own residencytraining in London during the years ofwhat we called ‘the Blitz.’ London wasbeing bombed every night and weyoung surgeons were faced with largenumbers of cases of severe trauma. Ourpatients had been dug up from underfallen houses and suffered from crushedlimbs and from wounds caused bybroken glass. I had become veryexperienced at analyzing the exactcauses of paralysis of the hand and alsothe legs. I had become experienced atusing many types of tendon transferoperations to restore balance to thelimbs of patients who had come in withparalysis of any one or more of themotor nerves in the limbs. I had alsoworked in the Orthopaedic unit of aChildren’s Hospital where I had beenresponsible for treating many cases ofpoliomyelitis, which was common inLondon in those days.

Thus on starting orthopaedicsurgery in India I was on the lookoutfor cases of polio and of severe traumabecause in that area at least I was morecompetent than anybody else.

It wasn’t long before I noticed thatoutside the gates of the hospital therewere usually numbers of beggars whodisplayed deformities of the hands andfeet and of their faces. Several of thesebeggars had deformities of the handthat looked like polio cases that I hadseen in London and many others hadsevere wounds which were beingdisplayed without dressings in order toexcite the pity of passersby who wouldthen put some money in the hat or bowlof these beggars.

I was shocked to see so manydeformities and so many woundsoutside the hospital and I expressed myfeelings to one of the older Indiandoctors as he came through the gate. Isaid, “Why do we allow these wounded,paralyzed patients to stay outside thehospital and collect money by begging?Surely we should take them into thehospital and we could find the cause oftheir problems and operate to treatthem!”

The Indian doctor smiled at me andsaid, “It’s very obvious that you are newto this part of the world, and you don’trealize that all these beggars are lepersand it is leprosy that has caused theparalysis and deformities and woundsthat don’t heal. Another thing you don’tknow is that if we allowed any of theseleprosy patients to come into thehospital, all of the other patients wouldrun away. In fact, many of the staffwould leave the hospital rather than berequired to be in close contact withthese lepers. You won’t find any generalhospital in India that would ever admita leprosy patient if they wanted to stayin business.”

When I continued asking questionsabout leprosy and saying that we surelyought to be able to make a special wardwhere we could study the exact causeand correctability of the paralyses, theolder doctor laughed at me and saidthat if I really wanted to examineleprosy patients, I would have to go toa leper colony. He said that manymissions had set up crude centers wherelepers could be fed and looked after. Noother patients will go there. He went onto say, “But don’t get any ideas aboutoperating on these hands and feet. Youcan see from the wounds and the ulcersthat lepers have non-healing flesh. Ifyou try to operate you will simply makea new wound that also will never heal.Don’t forget they have non-healingflesh.”

I went on making inquiries fromother doctors who were experienced inIndia and it soon became clear thatthere was no future in trying to studyleprosy in our hospital at Vellore. Imade inquiries and found there was aleprosy colony about eighty miles fromus and it was run by a Scottish missionsociety. The doctor in charge, a Dr.Cochrane, had had many years ofexperience caring for lepers. So I madea habit of driving down to theleprosarium and talking to Dr.Cochrane who tried to teach me fromhis own experience all the main featuresof leprosy. Dr. Cochrane’s backgroundwas dermatology and since many of themanifestations of leprosy involveswellings and wounds at the surface of

the body, dermatology was a usefulbackground to have.

There was no cure for leprosy inthose days. There was a drug calledDapsone which was under trial andwhich seemed to be effective inarresting the progress of the disease, butin most cases the germ of leprosy,microbacterium leprae, soon becameimmune to Dapsone and would resultin a recurrence of the disease.

One thing I learned from Dr.Cochrane was that superficial sensorynerves became paralyzed in leprosy andit resulted in the patient being unableto feel either light touch or pain. Dr.Cochrane also agreed that he wouldn’twant to operate on a leprosy patientbecause he was afraid that the operationwound would not heal.

It was in that leprosarium that Ibegan to study the pattern of paralysisand I learned that the most commonmuscles to become paralyzed were theintrinsic muscles of the hand,particularly those that were supplied bythe ulnar nerve. Median nerve paralysisalso occurred in the intrinsic muscles,especially those supplying the thumb. Ifound that the motor branches of theradial nerve were very rarely paralyzedand I found no cases of paralysis of themedian nerve muscles in the forearm.They were very rarely paralyzed,whereas the ulna nerve muscles in theforearm were very frequently paralyzed.

Dr. Cochrane couldn’t tell me whyor in what way various muscles wereparalyzed in leprosy, although he knewthat the big muscles in the thigh and inthe upper arm did not seem to sufferparalysis.

I was getting more and morefascinated by this strange disease anddetermined that somehow we wouldfind a way to operate, using musclesthat were not going to get paralyzed toreplace those which were alreadyparalyzed.

The one thing that I had to discoverbefore starting to operate was to findout what was the cause of paralysis andwhy some nerves seemed so much moreprone to be paralyzed, although surelythe whole body had the same bacteria.Why did they cause paralysis to some

Neuropathology By Sunrise

PAUL BRAND, M.D.


muscles and not to others?It was obviously essential for me to

get an autopsy of somebody who hadsuffered leprosy in a severe form andstudy all the nerves.

Here I had a problem. I couldn’thave an autopsy at the medical collegebecause there were no leprosy patientsthere, and in any case, it was verydifficult to get permission for autopsyin India, where the average familydemanded that they take their familymember home to die, or at least withina few hours to take the body to becremated or buried within 24 hours ofdeath, because in that hot climate adead body would rapidly decay andbecome offensive.

I communicated with severalleprosy colonies and asked them toinform me quickly if they had a patientwho was dying and who did not havean urgent family insistence about thedisposal of the body. After quite a longtime I finally got a message from aleprosy colony saying that a patient hadjust died and that the family wanted totake the body at about sunrise the nextday, but in the meantime they had noobjection to my performing an autopsy,so long as it was finished by aboutdawn.

I could get away from my work onlyin the evening and had problemsfinding the way. I took with me anassistant surgeon and a pathologytechnician with a box of little bottles offormalin so that we could takespecimens from nerves or other tissuesand keep them for histological studylater.

It was after midnight when wefinally found the leprosy colony andmet the guard at the gate who knew wewere coming and who had twohurricane lanterns. He said that hewould take us to the hut where the bodywas lying and would leave onehurricane lantern with us and take theother one back with him. We hadn’texpected to have a formal laboratory towork in, but we were a little shocked tofind that we had to go through theleprosy colony and out the back gateand then walk a distance over sanddunes, almost to the beach. There wefound a mud-walled hut with athatched roof, no electricity and just awooden table on which the corpse ofthis elderly, deformed patient was lying.The guard hung the hurricane lanternfrom the bamboo rafters overhead. It

gave enough light so we could see towalk around the room, but it was of nohelp for a careful dissection.Fortunately, we had brought with ustwo pen-sized flashlights. We foundthat if we held the flashlights in ourmouths we could do a dissection in thesmall circle of light that the flashlightilluminated. The guard left us there inthis weird situation and pulled theheavy wooden door shut behind him,commenting that we should leave itshut, otherwise the jackals might comein, attracted by the smell.

My assistant doctor worked on theright hand side, exposing all the nervesin the right arm and the right leg, whileI worked on the left-hand side, exposingthe nerves in the left arm and the leftleg. Starting at the shoulder I patientlyexposed the median nerve, the radialnerve and the ulnar nerve. I kept thenerves in continuity, whereas myassistant on the other side cut across thenerves, taking specimens at differentmeasured levels above and below theelbow and giving the specimens to ourtechnician for preservation.

I shall never forget that strange,dark, isolated hut, nor shall I forget thedetails of the pathology as it appearedin the nerves during my dissection.Starting at the level of the shoulder andworking my way down the ulner nerveI was able to follow little branches andtwigs from the nerves and all the timethe only light I had was a circle of aboutsix to eight inches wide. The nervelooked quite normal to me as I workedmy way from the axilla towards theelbow, and then gradually as I gotnearer to the elbow, the nerve seemedto be swollen. Then as the nerve wentbehind the medial epicondyle of thehumerus the nerve became very swollenand looked obviously diseased. When Idissected below the elbow, movingtowards the wrist, the ulnar nervesuddenly began to change and by thetime it was two or three centimetersbelow the elbow the nerve again seemedto be quite normal, neither swollen norinflamed. It remained the same wayuntil it was within a few centimeters ofthe wrist. There the nerve began to beswollen again and as it crossed the wristand into the palm, each little nervebranch to the various fingers wereswollen and appeared diseased.

Then I worked on the median nerveand again found that it seemed to benormal at the level of the axilla and then

as it passed down to the elbow itremained normal. It was normal as itcrossed the elbow and it was normal aswe moved down the forearm until wegot quite close to the wrist and now itbegan to be swollen and diseased andswollen again as it broke into thebranches supplying the abductorpollicis brevis.

And so we went on, carefully, slowly,little by little. The room began to beuncomfortably stuffy and the bodyitself began to develop an offensiveodor. We moved down to the legs andbegan to expose a normal lookingsciatic nerve and femoral nerve. We sawsome changes in the nerves around theknee and especially around the neck ofthe fibula and so as we approached theankle some of the nerves becameswollen and diseased and others seemedto be normal in appearance. Butnothing that we saw gave us any clue asto why certain bits of nerves wereswollen and certain bits of the samenerve looked normal. The same nervemight have an area of disease followedby an area that looked normal, followedby another area of disease,but…..WHY? We hoped that perhapsthe histology would explain this to us.

Finally, after we had spent perhapsfour or five hours on the careful,meticulous dissection and when theroom was getting intolerably warm andstuffy, we began to see the light ofsunrise around the cracks of the door.We looked at each other and said, “Nowis the time to get a little bit of rest andthen we can go back and sew up all theincisions.” We felt as if we were largelydefeated because we really had no wayto tell why some parts of the nervesseemed to be diseased and other partsabsolutely normal, but we couldn’tstand to be in that room any longer. Wepushed the room open and went outonto a sand dune and we looked outtowards the sea and now the sunrisebegan to be bright and beautiful. Thepath of the sunrise light came acrossfrom the horizon as a bright, shiningpath across the sea to where we were.

We took great deep breaths of thefresh air of that dawn and we shookourselves to get rid of the sense ofconfinement that we had beenexperiencing for so many hours in thathut. Then we turned back to the hut,having left the door wide open and wesaw that the sunrise, now brilliant withlight, was shining straight through the


door and onto the body and all its openincisions. The way the body was lyingon the table exposed the whole of theleft side to the sun’s rays. To the rightside, where all the specimens had beentaken, was still mostly shade.

As I stood at the entrance to that hutand looked at all the nerves that I hadbeen exposing and had previously onlyseen six inches at a time, I suddenlyexperienced the most amazingsensation of revelation. Now for the firsttime I could see the whole length of thenerves. It became amazingly clear thatwherever a nerve came to lie close underthe skin, it was always swollen, andwhere it came very close it lookeddiseased and totally abnormal. It wasalso clear that wherever any nerve waslying deep in the arm, and especiallywhere it was lying under a muscle, itlooked absolutely normal. Thus theulna nerve, where it crossed the elbowbehind the epicondyle, was obviouslydiseased and swollen, but movingfurther down the forearm the ulnanerve took a dive and lay deep to theflexor carpi radialis muscle and deep tosome of the fibers of profundus - thereit looked absolutely normal. Thentowards the wrist, as the fibers of theflexor carpi ulnaris and the profundusthinned out and gave place to tendons,there the ulna nerve became closer tothe skin and there again it began to beswollen. The twigs of nerve in the palmof the hand and at the side of eachfinger were all swollen and abnormal.

I cried out. I almost shouted, “We’vegot our answer!” Nerves near the skinare paralyzed, nerves deep to musclesare safe.

The median nerve did not comenear the surface of the skin where it wascrossing the elbow joint; it remaineddeep to upper arm muscles and to theorigins of the flexor muscles for thewrist and fingers.

We went back into our little roomand looked back at our dissections,seeing them for the first time as wholenerves, rather than little segments ofnerves. We took photographs and thenwe quickly stitched up the incisions justabout in time for the family to comeand take the body away.

Back at the medical school I spenttime in the library trying to find anyother example of nerve paralysis thatwas dependant on depth under theskin.

I found an account from Australia

where a similar mycobacterium, the m.ulcerans which affects cattle was foundto grow in an incubator only if thetemperature was set at about sevendegrees Celsius below bodytemperature. They had been trying togrow the microbacterium but hadfailed. One day, some agar plates ofmycobacterium ulcerans had been putinto the incubator at ordinary bodytemperature, but by a mistake, theincubator was set at about thirtydegrees Celsius. The next day theyfound the mycobacterium ulcerans hadgrown beautifully in the agar plate atthe cooler temperature.

This completed our study becauseit was obvious that microbacteriumleprae seems to thrive and do all itsdamage in tissues that are cool, and donot rise much above thirty degreesCelsius. Although it has still not provedpossible to grow the microbacteriumleprae, this must be due to some otherfactor, but it is true that in the body onlythose parts of the body that are belownormal body temperature are damagedby leprosy.

We were very happy to identify thisfactor which had never been identifiedbefore, and which has proved to be anexplanation for features thatcharacterize leprosy but do notcharacterize most other mycobacterialdiseases like tuberculosis. For example,tuberculosis can damage deep bonesand joints and causes tuberculomas inthe brain, as well as skin, whereasleprosy never grows deep in the bodyor brain. Leprosy affects the eyes, butonly in the cornea and anteriorchamber, which are cool, and neveraffects the retina which is muchwarmer.

We were able to map the arms andlegs for surgical purposes, knowing thatdeeply placed nerves supply muscleswhich do not ever become paralyzedand which can be used for tendontransfer, whereas nerves which aresuperficial supply muscles which arevery liable to become paralyzed and aretherefore not appropriate to be used fortendon transfers.

I don’t know how long it would havetaken us to work out the pattern ofparalysis in leprosy or the relationshipof m. leprae to temperature, if it hadnot been for that exciting momentwhen the SUNRISE illuminated soeffectively and so much more efficientlythan a whole five hours of flashlight

images that never allowed us to see allthe nerves in continuity.


UW Medicine Hand Institute

Dear Colleagues,

The hand surgery faculty in the Department of Orthopaedic Surgery and Sports Medicineand the Division of Plastic Surgery in the Department of Surgery are pleased to announcethat Paul Ramsey, M.D., Dean of University of Washington School of Medicine, approvedour application for a Hand Surgery Institute on May 7, 2003. The Hand Surgery Instituteformalizes the incredible cooperation between Orthopaedic Surgery and Plastic Surgery.We have four faculty in Orthopaedic Surgery: Thomas Trumble, M.D. Douglas Hanel, M.D. Christopher Allan, M.D. John Sack, M.D.And two faculty in Plastic Surgery: Nicholas Vedder, M.D. Arshad Muzaffar, M.D.They provide 24 hour on call coverage for University of Washington Medical Center,Harborview Medical Center, Children’s Hospital Medical Center and the VeteransAdministration Medical Center. Our hand team is the only group available to managesevere amputation and limb threatening injuries for the five state region of Washington,Montana, Alaska, and Idaho and Wyoming on a full time basis. We perform care fornearly all the major hand and upper extremity injuries in this five state region.

Together we administer an accredited hand fellowship, and we train three fellows inhand surgery every year who have their primary boards in either plastic or orthopaedicsurgery. The fellows are selected through the National Resident Matching Program.During our last accreditation review this spring we won top honors for our academicprogram. We also provide the clinical and didactic education for both the orthopaedicand plastic surgery residents. In addition to the clinical program, our residents andfellows take part in clinical, biomechanical, and microsurgery research programs as youcan tell from this research report.

As a cooperative effort we manage hand surgery clinics at all four medical centers in theUniversity of Washington as well as an office in Bellevue at the Eastside Specialty Center.Our distinguished team includes three full professors with numerous grants and awardsfor research. Because of our clinical productivity, the University of Washington MedicalCenter has sponsored a hand surgery center in the Roosevelt II building complete withdedicated out patient operating rooms, clinic and office space, hand therapy, advanceddiagnostic imaging with magnetic resonance imaging, computerized tomography andarthrography. Our combined clinical productivity has resulted in over 10,000 office visitsper year and nearly 4,000 surgeries. As the first full time hand surgeon at the Universityof Washington, I look on in wonder and awe at the tremendous growth of our handteam that the formation of the Hand Institute celebrates. The timing of this researchreport highlighting the hand surgery service is another brilliant bit of planning by ourChairman in Orthopaedic Surgery, Frederick A. Matsen III, M.D. He is also the chairmanwith the vision to recruit full time hand surgery faculty and crystallize the the role ofhand surgery at the University of Washington.

Sincerely,

Thomas Trumble, M.D.(Top to Bottom) Drs. Trumble,Hanel, Allan, Vedder, andMuzaffar.


Visiting Lecturers

This year at our annual LeCocqLecture on January 23rd and 24th,we were honored to have Dr.

John J. Callaghan as our 2003 LeCocqLecturer. Dr. Callaghan is a Professorof Orthopedic Surgery at the Universityof Iowa and is the Lawrence andMarilyn Dorr Chair in HipReconstruction and Research. He isactive in a number of professionalsocieties and is currently serving as theChairman of the Anatomy and ImagingCommittee for the American Academyof Orthopaedic Surgeons and ProgramChairman for the Association ofArthritic Hip and Knee Surgery. Healso serves as an Associate Editor for theJournal of Bone and Joint Surgery andthe Journal of Arthroplasty as well as aconsultant reviewer for the Journal ofBiomechanics, American Journal ofSports Medicine, and ClinicalOrthopaedics and Related Research. Dr.Callaghan is well known for hisexpertise in knee and hip jointreconstruction. His publications,awards, and grants in this area arenumerous. The faculty, residents, andcommunity physicians were treated to3 innovating lectures from Dr.Callaghan during the 2 days: “Choicesand Compromises in Hip and KneeSurgery,” “Why Did We Leave CharnleyTotal Hip Replacement?,” and “MobileBearing Knee Replacement: Are WeGoing Forward, Backwards, orSideways?”

John J. Callaghan, M.D.

This year at our annual Residents’Research Days on May 29th and30th, we were honored to have

Dr. Joshua J. Jacobs as our OREF HarkLecturer. Dr. Jacobs is the CrownFamily Professor and AssociateChairman for Academic Programs inthe Department of OrthopaedicSurgery at Rush Medical Center,Chicago. He also serves as the Directorof the Section of Biomaterials and theDirector of the Orthopaedic ResidencyProgram. He is active in a number ofprofessional societies and is currentlya member of the Hip Society and isChairman of the Council on Researchand Scientific Affairs of the AmericanAcademy of Orthopaedic Surgeons. Dr.Jacobs is well known for his expertisein the field of Biomaterials. His majorresearch focus is on thebiocompatibility of permanentorthopaedic implants, particularly jointreplacement devices. During the 2 daysof lectures, the faculty, residents, andcommunity physicians were treated to3 lectures from Dr. Jacobs: “The BasicScience of Periprosthetic Osteolysis,”“Systemic Implications of Total JointReplacement,” and “Corrosion ofMetallic Orthopaedic Implants.” Inaddition to Dr. Jacobs’ lectures, the R3sand the R4s presented the progress oftheir research, while the R5s presentedthe completion of their researchprojects.

Joshua J. Jacobs, M.D.


In a previous analysis, we found thatthe prevalence of back pain duringintercollegiate rowing was 32

percent. When college rowers developback pain, often they and their parentsask, “Is this back pain likely to be a life-long problem?” The current outcomeanalysis was undertaken to answer thisquestion by determining whether theback pain associated withintercollegiate rowing resolves orpersists, the variables that might predictresolution or persistence, and theseverity of current back pain. Inaddition, the lifetime prevalence ofback pain in rowers who had no backpain during college was compared tothat in the general population.

Meta-analyses of back pain in thegeneral population report lifetimeprevalence rates ranging widely from11-84% and point prevalence ratesfrom 5.6 to 33 %. The inclusion criteriafor “lifetime” back pain in these studiesvaries from “have you ever had low backpain” to “have you ever had low backpain daily for at least 2 weeks?”Waxman et al completed a two-phasepopulation-based questionnaire studyof low back pain over a 3-year period.Fifty-nine percent of 1455 adults aged25-64 had experienced at least one

episode of low back pain that lastedmore than one day. The likelihood ofhaving pain increased significantly withage. Hellsing and Bryngelsson studied6626 men nearly 40 years of age whohad been examined for the first time atthe age of 18 during examinations formilitary service. The prevalence of lowback pain increased from 38% to 74%during the 20-year period. Asignificantly increased risk for frequentpain problems at follow-up exam wasfound in those subjects whose earlierback pain had caused absence fromwork or reduced activity levels.

METHODS

Surveys were sent to 4680 formerintercollegiate rowing athletes who hadgraduated between 1978 and 1998 from5 schools with strong rowing programs.2165 surveys were returned (46%).Resolution or persistence of back painwas examined in three different ways.Subjects were asked whether they hadever had back pain since college(“any”), if so, whether it had beencontinuous since college, and regardlessof whether or not it had beencontinuous, whether they had backpain at the time of completing thesurvey (“current”). In addition, we

examined the association of continuingto row after college with current backpain. To determine the potential effectof the severity of college back pain onsubsequent back pain, we analyzed theprevalence of any, continuous, andcurrent back pain in relation to timeloss from college rowing and ending thecollegiate rowing career. To assess theseverity of back pain subsequent tocollege, we asked the subjects to ratetheir current pain and the worst painthey had had on a scale of 1-10 with 1being hardly noticeable. We also askedwhether work had been missed becauseof back pain. For comparisons toprevious back-pain literature, weconsidered a positive response to thequestion “Subsequent to college, haveyou ever had back pain lasting morethan one week?” as “lifetimeprevalence.” We consider current backpain as “point prevalence.”

RESULTS

Resolution or persistence790 of 1561 subjects (51.4%)

answered that they had back painlasting at least one week at some pointsince college (lifetime prevalence). Thetime from college graduation tocompletion of the survey ranged from0-20 years with a mean and median of13 years. Subsequent back pain wastwice as likely in subjects who had backpain in college than in those who hadnot had back pain during college(78.9% vs. 37.9%). 479 subjects(32.4%) had back pain at the time ofcompleting the survey (pointprevalence). Those who had back painin college were almost three times aslikely to have current back pain as thosewho had not had back pain duringcollege (57.7% vs. 20.2%). Ninety-eightsubjects reported back pain that hadbeen continuous since college (19.8%).Of 1561 subjects, 662 were female and897 were male. (Two subjects did notreport their sex). Sex was not asignificant predictive variable for thepersistence of back pain by any of thethree measures. Among rowers whohad back pain in college, age was asignificant variable only in the lifetime

Back Pain in Former Intercollegiate Rowers: A Long-TermFollow-Up Study

CAROL C. TEITZ, M.D., JOHN W. O’KANE, M.D., AND BONNIE K. LIND, M.S.

Picture 1: Intercollegiate rowers.


prevalence of back pain (p=0.01).Among rowers who were asymptomaticin college, age was significantlyassociated with prevalence of currentpain as well as with the lifetimeprevalence of back pain (p<0.001)(Figure 1).

Predictive variables502 subjects reported that they were

still rowing at the time of the survey.Based on whether or not they had painin college, there was no significantdifference between the proportions ofsubjects who continued to row. 337 of1046 subjects with no back pain duringcollege rowing were still rowing(32.2%) and 165 of 515 subjects withcollege back pain (32%) were stillrowing. Current pain was noted in32.1% of those still rowing and in32.5% of those who were no longerrowing. There was also no associationbetween the current prevalence of backpain and the intensity of after-collegerowing (defined as competitive,recreational, or sporadic)(p=0.4).

Subjects who missed more than oneweek of practice or competition whilein college were more likely to have

subsequent and current back pain thanthose with less time loss (p<0.01).Rowers who ended their college rowingcareers because of back pain were morelikely to have subsequent episodic,continuous, and current back pain(p<0.002) (Figure 2).

Severity of post-college back painThe mean severity rating for all

rowers whether or not they had pain incollege was 3.5±1.9. Those with collegepain reported a similar average severityof their current pain to those with nocollege pain (3.6 vs. 3.5, p = 0.55). Themean severity rating of worst back painsince college for all rowers whether ornot they had pain in college was6.4± 2.2. Those with college painreported a significantly higher averagerating of their worst pain than thosewith no college pain (6.9 vs. 6.3, p <0.002).

Of 790 subjects reporting back painsince college, 195 (26.2%) reportedmissing work because of back pain. Thepercentage of subjects missing workdue to back pain was not related to thepresence or absence of college back pain(26.1% vs. 26.4%). Among those who

missed work because of back pain,those with physical jobs were no morelikely to miss work than those withsedentary jobs (65.8% vs. 67%).

DISCUSSION

Bending and rotating the spineincreases the risk of low back pain.Although rowing involves similarmotions, the lifetime prevalence ratesof low back pain in former rowers arenot significantly different from those inthe general population. Rowers whodevelop back pain in college have moresubsequent back pain than rowers whodid not have back pain in college(78.9% vs. 37.9%) confirming Biering-Sorensen’s finding that a previousepisode of low back pain is a strongpredictor of future episodes.Nevertheless, even in rowers who hadback pain in college, this lifetimeprevalence rate is similar to rates of 74%reported by Hellsing and Bryngelssonin their subjects who had back pain as18 year olds. Moreover, only 51% of allrowers had ever had back pain sincecollege, similar to the 59% lifetimeprevalence rate noted in subjects of

Figure 1: Lifetime and point prevalence of back pain as a function of age.


similar ages in two other population-based studies and in the range of theoft quoted 60-80% of the generalpopulation that will experience oneepisode of acute back pain at somepoint in life. When looked at by decade,the higher lifetime prevalence of backpain in older rowers is likely due to theolder rowers having more “exposure”time to develop back pain than theyounger rowers. This correlation of ageand lifetime prevalence of back painwas also found in other studies.Furthermore, the increase in back painprevalence in our subjects from 32% atcollege age to 84.3% at age 40 is similarto that in Hellsing and Bryngelsson’sstudy of subjects at age 18 (38%) and40(74%).

Continuous back pain has not beenreported previously. Therefore, wecannot compare findings in otherstudies with the prevalence ofcontinuous back pain in our subjectswho began having back pain in college.

Our point prevalence rate of 32.4%is at the upper end of the prevalencerange previously reported (12-33%).This relatively high prevalence rate mayrepresent a bias on the part of

respondents in that rowers with backpain at the time the survey instrumentarrived might be more likely toparticipate in our study. Nevertheless,if this bias exists in our study, then thelifetime prevalence rates noted abovewould be potentially even lower thanthose in the general population.

An additional factor that mightcontribute to our relatively high pointprevalence rate is continuing to rowsubsequent to college. Based on ourprevious study, we expect at least a 32%prevalence of low back pain in activelycompetitive rowers. Although thecurrent rowers are not all rowingcompetitively, they have the additionalrisk factor of age. Nevertheless, theoverall prevalence of back pain in oursubjects who are still rowing is 32.1%,essentially unchanged from theprevalence of back pain during collegerowing.

Rowers with greater time loss fromcollege rowing and those ending theircollege rowing careers because of backpain had a higher likelihood ofsubsequent back pain than rowers withless time loss who were able to continuerowing in college.

The mean severity of current backpain in our subjects was 3.5±1.9 on ascale of 1-10 with 1 being hardlynoticeable. Hillman et al reported amean severity score of 6.24. They alsonoted that 26.5% of their subjectsreported pain in the mild (grade 1-4)category, 40.8% in the moderate (grade5-7) category, and 32.6% in the severecategory (grade 8-10). Our subjectswith current pain compare favorably.75.6% report mild, 19% moderate, andonly 5.4% severe pain. Twenty sixpercent of our subjects missed workbecause of back pain. This rate is similarto the 22.5% found by Biering-Sorensen in a general population andthe 26% found in the 15 highest-riskmajor industries by Guo et al. In ourstudy, among those who missed workbecause of back pain, those withphysical jobs were no more likely tomiss work than those with sedentaryjobs.

CONCLUSION

Rowers who develop back painduring competitive college rowing areno more likely than the generalpopulation to have episodic back pain

Figure 2: Prevalence of subsequent back pain as a function of time loss in college.


subsequent to college or to miss workbecause of back pain. Also they reportless severe pain than subjects inprevious population-based studies.Moreover, rowers who do not developback pain in college have significantlylower rates of back pain as they age thanthe general population. If one cancomplete a college rowing careerwithout back pain, one’s prognosis forback pain later in life is better than thatin the general population.

RECOMMENDED READING

Biering-Sorenson F: Low back troublein a general population of 30-, 40-, 50-, and 60-year old men and women:study design, representative years, andbasic results. Dan Med Bull 29:289-299,1982.

Guo H, Tanaka S, Halperin W, et al.:Back pain prevalence in US industryand estimates of lost workdays. Am JPublic Health 89(7):1029-35, 1999.

Hellsing A, Bryngelsson I: Predictors ofmusculoskeletal pain in men: A twenty-year follow-up from examination atenlistment. Spine 25(23):3080-3086,2000.

Waxman R, Tennant A, Helliwell P: Aprospective follow-up study of low backpain in the community. Spine25(16):2085-90, 2000.

ACKNOWLEDGEMENTS

The authors wish to thankUniversity of Washington CrewCoaches Jan Harville and Bob Ernst,athletic trainer Chrissy Price, and Dr.Jo Hannafin for their advice, support,and encouragement ; Drs. ArthurBoland, Paul Fadale, Marc Galloway,and Timothy Hosea for assistance inobtaining subjects for participation;and the subjects for completing thedetailed surveys.

This paper was presented at the2002 meeting of the Magellan Society,Gorgonza, Italy.

This paper has been accepted forpublication by the American Journal ofSports Medicine.


B oth corticosteroid andhyaluronic acid injections arewidely used to palliate the

symptoms of knee osteoarthritis;however, little research has been donethat compares the two interventions.The present report tests the hypothesisthat there are no significant differencesbetween hylan GF-20 (Synvisc) and thecorticosteroid betamethasone sodiumphosphate / betamethasone acetate(Celestone-Soluspan, admixed withlidocaine and marcaine) in terms ofpain relief or improvement in function,as determined by validated scoringinstruments.

This report also summarizes theresults of a separate retrospective studythat used a contemporaneous control

group to evaluate the likelihood ofpainful adverse reactions to repeatedcourses of Synvisc compared to patientstreated with a single course of thatproduct.

METHODS

The prospective trial enrolled 100patients with knee osteoarthritis (Table1), who were randomized to receiveeither intra-articular Synvisc orCelestone-Soluspan (CS), and whowere followed for six months. Patientswith end-stage (“bone-on-bone”)arthrosis visible on weight-bearingradiographs (Figure 1) were excludedfrom the study, as prior work has shownSynvisc to be much less effective in thispatient population. Synvisc patients

received one course of 3 weeklyinjections. CS patients received oneinjection at study enrollment, andcould request one more injection anytime during the study. An independent,blinded evaluator assessed patients withthe WOMAC index, the Modified KneeSociety Score (KSS), and the visual-analog pain scale (VAS).

The retrospective study on repeatedtreatments with Synvisc involvedauditing the records of all patients atthe study site who received more thanone course of treatment with hylan GF-20 (N=19 patients), and comparing thenumber of painful local reactions theyexperienced to that of a group ofpatients who received only one courseof treatment during the same 15-

Corticosteroid Versus Synvisc (Hylan GF-20) Injections for KneeOsteoarthritis: A Prospective, Randomized Trial, and Results of aRetrospective Study Evaluating Adverse Reactions to RepeatedCourses of Synvisc

SETH S. LEOPOLD, M.D., BRIGHAM B. REDD, M.D., WINSTON J. WARME, M.D., PAUL A. WEHRLE, M.D.,PATRICK D. PETTIS, L.V.N., AND SUSAN SHOTT, PH.D.

Treatment Age(years)

Weight(lbs.)

BMI* Gender(% female)

NSAID Use(% using)

CorticosteroidN=50 patients

64 (40-83) 86 (40-143) 29.3 (17-58) 56 56

SynviscN=50 patients

66 (39-79) 86 (48-141) 28.8 (21-51) 52 64

p-value 0.97 0.96 0.82 0.69 0.41

Table 1: Prospective study: Baseline demographic and clinical parameters. Values presented as median (range) except where noted. *BMI=Body Mass Index.

Single-Course GroupN=42

Multiple-Course GroupN=19

p-value

Mean Age, years (range) 64.4 (range, 39-79) 61.0 (range, 38-77) 0.13Gender 21 female (50%) 12 female (63.2%) 0.34Severe DJD 15 (35.7%) 6 (31.6%) 0.75Mean BMI (range) 30.9 (range, 21.2-51.0) 30.9 (range, 23.0-41.1)0.73Bilateral Injections 15 (35.7%) 6 (31.6%) 0.75

Table 2: Retrospective study: Baseline demographics and clinical parameters.


month period at the same center (N=42patients, Table 2). An acute localreaction to hylan GF-20 was defined asacute onset of pain and swelling in theknee, that occurred in temporalproximity (within 72 hours) of a hylanGF-20 injection, in the absence ofanother cause such as acute trauma.Painless effusions that were noted onroutine follow-up were not consideredto be adverse reactions. In all cases, theacute local reactions were rather severeand not difficult to distinguish frompatients’ typical arthritic effusions andbaseline arthritic pain levels. Allpatients noted severe pain andlimitation of activity, and all underwentaspiration and corticosteroid injectionwith prompt amelioration ofsymptoms. The single-course groupwas prospectively enrolled andfollowed, as part of the prospectiverandomized trial mentioned above. Thetwo groups were compared with respect

to several demographic and clinicalparameters, in order to make certainthat no factors other than the additionalcourse(s) of Synvisc could account forany observed differences in the rates ofpainful local reactions.

All injections in both studies wereperformed using the identicalsuperolateral injection technique by thesame two fellowship-trainedorthopaedic knee surgeons. All patientswho received intra-articularcorticosteroid injections weredocumented to have had immediaterelief of symptoms, which tended tovalidate the efficacy of thoseinvestigators’ needle placement. Allpatient evaluations in both studies wereperformed by the same trained studynurse, without the involvement of thetreating surgeons.

RESULTS

In the prospective, randomized trial,

both CS and Synvisc patientsdemonstrated improvements frombaseline WOMAC scores (55 to 40points, p<0.01; and 54 to 44 points,p<0.01, respectively). Scores on the KSSdid not show statistically significantimprovement for CS patients or Synviscpatients (58 to 70 points, p=0.06; and58 to 68, p=0.15, respectively). VASscores improved for Synvisc patients,but not for CS patients (70 to 52 mm,p<0.01; vs. 64 to 52 mm, p=0.28).However, there were no significantdifferences in WOMAC, KSS, or VASresults between the two treatmentgroups, despite 80 percent power todetect clinically relevant differences(Table 3). Women demonstrated astatistically significant improvement inonly one of the six possible outcomes/treatment combinations (the WOMACscale for female Synvisc patients), whilemen demonstrated significantimprovements in five of six (allmeasures except the KSS for maleSynvisc patients). The gender-relateddifferences could not be explained bydifferences in age or disease severity.

Painful local reactions to hylan GF-20 occurred significantly more often inpatients who had received more thanone course of treatment (4 of 19patients, 21.1 percent), than in patientswho received only a single course oftreatment (one of 42 patients, 2.4percent; p=0.029). All of the reactionswere severe enough to cause the patientto seek care on an unscheduled basis.Following corticosteroid injection, thereaction abated without apparentsequelae. With the numbers available,no significant differences were detectedbetween the multiple-course andsingle-course groups in terms of age,gender, body-mass index, diseaseseverity or bilaterality of disease (Table4).

CONCLUSIONS

The prospective randomized trialdetected no differences in pain orfunction between intra-articularinjections with Synvisc or CS at sixmonths’ follow-up. The difference inpharmacy cost between the twotreatments is over 100-fold, with acorticosteroid injection costing about$5, and a course of Synvisc costingnearly $600. To our knowledge, this isthe first trial to compare CS and anyviscosupplement that was not fundedby the manufacturer of the hyaluronic

Figure 1: Anteroposterior (AP) radiograph of a knee, obtained with weight bearing, which demonstratesend-stage (“bone-on-bone”) arthrosis of the medial compartment. Such patients were excluded fromthe present report, as prior work has found Synvisc to be much less effective in the presence of diseaseof this severity.


acid product in question. This study isalso the largest prospective randomizedtrial comparing any hyaluronic acidproduct to any corticosteroid, the firststudy of this type to use validateddisease-specific outcomes instrumentssuch as the WOMAC index or the KSS,and the first comparative trial toevaluate Synvisc, the most widely usedintra-articular hyaluronic acid productin the United States, against acorticosteroid. Women demonstratedsignificantly less response to treatmentthan did men for both CS and Synviscon all three outcome scales. Suchsignificant gender-related differenceswarrant further investigation.

Results from the retrospective studysuggest it may be reasonable to counselpatients who have been treated with acourse of hylan GF-20, and who desirean additional course of treatment, thatthe likelihood of a painful acute localreaction to the medication appears tobe increased.

Given the additional pain andpotential risk associated with the three-injection course of Synvisc, and giventhe approximately 100-fold differencein pharmacy cost at our institution, theauthors do not consider Synvisc a first-line treatment for OA patientsconsidering intra-articular kneeinjections for palliation of symptoms.

Further study of the frequency of acutelocal reactions following repeatedcourses of hylan GF-20, and inquiryinto the mechanisms of those reactions,are warranted.

RECOMMENDED READING

Adams, M. E.; Lussier, A. J. and Peyron,J. G.: A risk-benefit assessment ofinjections of hyaluronan and itsderivatives in the treatment ofosteoarthritis of the knee. Drug Saf, 23:115-30, 2000.

Bernardeau, C.; Bucki, B. and Liote, F.:Acute arthritis after intra-articularhyaluronate injection: onset of

Treatment Time Point WOMAC KSS VASCorticosteroid Before treatment 55 58 64

3 months 42 72 526 months 40 70 52

p-value* <0.01* 0.06 0.28Synvisc Before treatment 54 58 70

3 months 41 69 456 months 44 68 52

p-value* <0.01* 0.15 <0.01*

Table 3: Prospective study: Changes in median outcomes scores over time. *p-values refer to changes over time within each treatment group, based on theFriedman Test; p-values <0.05 were considered statistically significant. N.B. No significant differences were detected between treatment groups by any of thethree outcomes instruments at either follow-up interval, using the Mann-Whitney test (p=0.29-0.61 at 3 months and p=0.69-0.98 at 6 months).

Severe DJD

Gender

BilateralInjections

Age BMI Time to reaction

Severe-Course Group(1 of 42 patients*)Patient 1 No M Yes (reaction

unilateral)49 26.2 <6 hours of 1st injection

Multiple-Course Group(4 of 19 patients*)Patient 1 Yes M Yes (reaction

unilateral)62 24.4 <6 hours of 1st injection,

3rd coursePatient 2 Yes F No 58 37.2 <24 hours of 1st injection,

2nd coursePatient 3 No F No** 52 23.0 36 hours after 2nd injection,

2nd coursePatient 4 No F No** 60 25.7 <24 hours of 3rd injection,

2nd course

Table 4: Retrospective study: Characteristics of patients with acute local reactions. *Frequency of reaction: 2.4% (single-course group) versus 21.1% (multiple-course group); p=0.029. **Patient had bilateral injections done at the first course of treatment, and unilateral injections done at the second course oftreatment.


effusions without crystal. Ann RheumDis, 60: 518-20, 2001.

Jones, AC; Pattrick, M; Doherty, S andDoherty, M: Intra-articular hyaluronicacid compared to intra-articulartriamcinolone hexacetonide ininflammatory knee osteoarthritis.Osteoarthritis Cartilage, 3: 269-73,1995.

Leardini, G; Mattara, L; Francheschini,M and Perbellini, A: Intra-articulartreatment of knee osteoarthritis — Acomparative study between hyaluronicacid and 6-methyl prednisolone acetate.Clin Exp Rheumatol, 9: 375-381, 1991.

Leopold, S.S.; Warme, W.J.; Braunlich,E.F.; Shott, S.: Association betweenfunding source and study outcome inorthopaedic research. Clin Orthop (inpress), 2003.

Puttick, M. P.; Wade, J. P.; Chalmers, A.;Connell, D. G. and Rangno, K. K.: Acutelocal reactions after intraarticular hylanfor osteoarthritis of the knee. JRheumatol, 22: 1311-4, 1995.

ACKNOWLEDGEMENTS

The authors thank the WilliamBeaumont Army Medical CenterDepartment of Clinical Investigation(Grant WBAMC #00/22), and theWilliam Beaumont Army MedicalCenter Pharmacy Department forsupport of the material presented inthis report.


Failure of fracture union is anuncommon occurrence. Thediagnosis is made primarily by

clinical and radiographic assessments.Delayed union suggests a slowed paceof fracture repair outside that which isexpected for the injury. Nonunionrepresents complete absence ofprogression of healing that is predictedto persist indefinitely. The designationof delayed or nonunion is not based onabsolute time criteria, but represents atime continuum individualized for eachpatient’s injury. Increased local fracturecomminution and instability, bone loss,soft tissue disruption, and/or openwounds characterize high-energyfractures. These features, among othervariables, are particularly associatedwith altered fracture healing.

The detrimental effect of cigarettesmoking on fracture repair iscontroversial, and the relative risk ofsubsequent fracture delayed ornonunion is unknown. Clinically,increased rates of pseudarthrosis afterspinal fusions, and an increased timeto union after fracture of the tibia havebeen reported. Recently, however,authors have been unable todemonstrate an association betweensmoking and fracture union. Thepurpose of this report is to define therisk of altered long bone fracturehealing requiring operativeintervention in tobacco smokers.

MATERIALS AND METHODS

A retrospective case-control studydesign was employed using a traumadatabase at an urban level-1 Universitytrauma center. Patients were excludedif they were skeletally immature, hadinsufficient follow-up, or had theirinitial fracture stabilization performedat an outside institution. Between 1995and 2000 inclusive, 197 operativelymanaged long bone delayed andnonunions were identified. Over thesame time period, 197 control patientswith healed long bone fractures wereidentified. The cases and controls werematched on age (within 5 years), and

fracture pattern using the AO/OTAFracture Classification System. Patientrecords were reviewed to obtaindemographic data, fracture specificdata, and associated injuries. Cigarettesmoking behavior was obtained fromthe hospital and trauma registries, andfrom a review of emergency room,social work, clinic, and hospitaldischarge reports. Patients werecategorized either as smokers ornonsmokers and no attempt was madeto quantify the amount or numbers ofcigarettes smoked over any timeinterval. The Injury Severity Score(ISS) is a validated measure of thedegree of systemic injury. Eachpatient’s ISS was recorded and used asa surrogate for the magnitude ofassociated polytrauma. Open fractureswere defined as those fractures thatcommunicated with an open wound tothe outside environment.

Statistical analysis consisted ofperforming an overall odds ratio toestimate the relative risk of fracturenonunion in smokers versusnonsmokers. The presence ofpolytrauma, and open fractures areindependently known to be potentialconfounding factors. Comparison ofISS between cases and controls wasperformed using the t-test. The rate ofoccurrence of open fractures betweencases and controls was assessed usingthe Chi-square test. Stratum-specificodds ratios were obtained for cases andcontrols with open fractures, and anadjusted odds ratio using the Mantel-Haenszel method was then performed.

RESULTS

One hundred and ninety-seven longbone delayed/nonunion fracture caseswere identified and consisted of 2clavicles, 19 humerii, 17 radii/ulnae, 61femurs, and 98 tibias. Cases were thenmatched by age and the AO/OTAFracture Classification to 197 healedcontrol fractures. The total studypopulation, therefore, consisted of 394total fractures. The mean age of theentire group was 38.3 years.

The mean ISS for the cases andcontrols was 14.2 and 13.8, respectively(p=0.6327). One hundred and threeopen fractures occurred in the cases(52.3%) as compared with 69 openfractures (35.0%) in the control group.This difference was statisticallysignificant using the Chi-square test(p=0.001).

Ninety-nine of 197 cases versus 74of 197 controls were identified assmokers. Calculation of the overallodds ratio demonstrated that smokerswere 70% more likely to develop adelayed/nonunion than nonsmokers(OR, 1.7; 95% CI=1.12-2.51). Becausea statistically significant number ofcases had sustained open fractures ascompared to the control group,stratum-specific odds ratios wereperformed for cases and controlsaccording to the presence or absence ofan open fracture. The evaluation ofonly closed fractures established thatsmokers were 96% more likely todevelop a delayed/nonunion thannonsmokers (OR, 1.96; 95% CI=1.14-3.37). However, no statisticallysignificant increased risk of delayed/nonunion could be shown in smokersversus nonsmokers with open fractures(OR, 1.41; 95% CI=0.76-2.62),suggesting that this variable was not asignificant confounder. The Mantl-Haenszel adjusted odds ratioaccounting for the presence of an openfracture confirmed a 70% increasedlikelihood of delayed/nonunion insmokers as compared to nonsmokers(OR, 1.7; 95% CI=1.13-2.55).

DISCUSSION

The impact of cigarette smoking onsurgical procedures on the spine havebeen well documented and include anegative clinical impact in terms ofoutcomes, union rates, and clinicalrecovery after both cervical and lumbararthrodesis procedures.

The effect of smoking on fractureunion has been controversial, andconflicting reports have been published.Recently, Adams et al., compared 140

The Association Between Smoking and Nonunion of Long BoneFracturesDAVID P. BAREI, M.D., F.R.C.S.(C), SARAH HOLT, M.S.P.H., SEAN E. NORK, M.D.,CARLO BELLABARBA, M.D., WILLIAM J. MILLS, M.D., AND BRUCE J. SANGEORZAN, M.D.


smokers with 133 nonsmokers withopen tibia fractures. Despite theincreased rates of flap failure andsurgical procedures for nonunion insmokers, they were unable to establisha relationship between smoking andnonunion with statistical significance.Giannoudis was also unable todemonstrate an association betweenunion rates and smoking in acomparison of 32 femoral diaphysealnonunions and 67 healed controlpatients.

The goal of the present study wasto determine the relative risk of longbone fracture delayed and nonunion insmokers as compared to nonsmokers.A retrospective case-controlmethodology was employed, with casesand controls matched by age andfracture pattern using the AO/OTAFracture Classification System. Thisalpha-numeric radiographic fractureclassification system is the standardsystem supported in the orthopaedicliterature and codes each fractureaccording to a consistent segmentwithin each bone. Each patient’sfracture can then be subtyped andgrouped according to the degree ofcomminution and orientation of majorfracture planes. Thus, this classificationsystem not only allows for theconsistent direct grouping of fracturesbased on their radiographicappearance, but also indirectly poolsfractures according to the injuriousmechanism, severity, and prognosis.Matching of fracture types based onthis classification was performed toeliminate discordant fracture patternswith dissimilar union rates. Thedefinition of delayed union ornonunion was at the attendingsurgeon’s discretion. We chose,however, to include only those patientswith delayed/nonunion that weremanaged operatively as the endpoint forselecting cases. We felt that this mostconsistently defined the presence of thecondition being examined, andsimilarly, identified those cases thatwere felt to have clinical relevance.Because our institution is a referralcenter for acute injuries and complexpost-traumatic reconstructiveproblems, cases and controls wererequired to have had their initial anddefinitive acute fracture stabilizationsperformed at our institution tominimize bias related to their initialcare.

Figure 1: A and B: Five months post medullary nailing, these anteroposterior and lateral imagesdemonstrate an atrophic nonunion of a mid-diaphyseal humeral shaft fracture. C and D: Treatmentconsisted of removal of the previously placed medullary implant, plate fixation, autograft bone grafting,and smoking cessation measures.

(A) (B)

(C) (D)


Our data suggests that cigarettesmokers are 70% more likely to havean acute long bone fracture result indelayed/nonunion as compared tononsmokers. A statistically significantrate of open fractures was found in thecases as compared to the controls.Because the presence of an openfracture is an independent risk factorfor delayed/nonunion, stratum-specificodds ratios were performed. Thisdemonstrated that smokers with closedfractures were 96% more likely toproceed to delayed/nonunion thannonsmokers with closed fractures. Anincreased risk of delayed/nonunion insmokers with open fractures, however,was not demonstrated (OR, 1.41; 95%CI=0.76-2.62), suggesting that smokingand open injuries may not have acumulative effect. Given that thepresence of an open fracture was not aconfounding variable, the Mantel-Haenszel adjusted odds ratio (OR, 1.7;95% CI=1.13-2.55), is similar to theoverall odds ratio.

This study has demonstrated a neartwo-fold risk of delayed/nonunion ofacute closed long bone fractures inpatients who smoke. Smoking shouldbe considered a substantial biologic riskfactor for altered fracture healing andpatients should be counseled as such.The dose-response relationship, and theeffect of smoking cessation in theacutely injured scenario are areas forfurther research.

RECOMMENDED READING

Adams CI, Keating JF, Court-BrownCM. Cigarette smoking and open tibialfractures. Injury 2001;32(1):61-5.

Brown CW, Orme TJ, Richardson HD.The rate of pseudarthrosis (surgicalnonunion) in patients who are smokersand patients who are nonsmokers: acomparison study. Spine1986;11(9):942-3.

Giannoudis PV, MacDonald DA,Matthews SJ, Smith RM, Furlong AJ, DeBoer P. Nonunion of the femoraldiaphysis. The influence of reaming andnon- steroidal anti-inflammatorydrugs. J Bone Joint Surg Br2000;82(5):655-8.

Kyro A, Usenius JP, Aarnio M,Kunnamo I, Avikainen V. Are smokersa risk group for delayed healing of tibialshaft fractures? Ann Chir Gynaecol1993;82(4):254-62.

Porter SE, Hanley EN, Jr. Themusculoskeletal effects of smoking. JAm Acad Orthop Surg 2001;9(1):9-17.


Vascular injury associated withknee dislocation is a commonand potentially limb

threatening complication. A greaterthan eight-hour delay inrevascularization can lead to aboveknee amputation. Controversy remainsas to the optimal method of diagnosingpossible vascular injury. While routinearteriography has long been advocated,recently some authors have questionedthe need to perform arteriography onall patients. These authors suggest that‘hard’ findings on physical examination(active hemorrhage, expandinghematoma, absent pulse, distalischemia or bruit) are effective evidencefor identifying arterial injury. Theycriticize routine arteriography for thedelay it causes in revascularization, itspotential complications and for its highcost.

Critics of physical examinationalone cite the potential for misseddiagnosis leading to devastatingvascular compromise. A recent meta-analysis of the accuracy of the pulseexamination in patients with kneedislocation concluded that anabnormal pedal pulse is not sensitiveenough to detect a surgical vascularinjury, and these authors recommendedthe “liberal use of angiography”.

At Harborview Medical Center, oururban level-1 trauma center, the Ankle-Brachial Index (ABI) has long beenused to diagnose or rule out arterialinjury in both blunt and penetratingextremity trauma. However, nopublished series has described theefficacy of the ABI in a pure populationof patients with knee dislocation.

We present a consecutive series ofpatients with knee dislocationevaluated by both ABI and physicalexamination to assess the reliability ofthese methods for identifying arterialinjury.


Fifty-one patients ranging in agefrom 15 to 74 years with 52 kneedislocations admitted to Harborview

Medical Center over a 40-month period(October 1998 - February 2002) wereenrolled in a standardized treatmentprotocol. Systolic blood pressures forall four extremities were obtained witha Doppler probe and standardizedblood pressure cuff. In the lowerextremity the systolic pressure wasmeasured at the posterior tibial anddorsalis pedis arteries. To calculate theankle-brachial index, the highestmeasured arterial pressure in the ankleor foot was divided by the higher of thebrachial arterial pressures from bothupper extremities.

By palpation, pulses weredetermined to be either normal,diminished (compared to thecontralateral limb), or absent.Hypotensive patients were resuscitatedin the Emergency Department (ED)prior to assessing ABI’s, pulses, orclinical signs of perfusion. In all cases,an assessment of vascular status wasdocumented within one hour of arrivalto the ED. For those patients whopresented frankly dislocated (n=18),vascular evaluation was made afterclosed reduction. Patients wereexcluded from this study if theypresented to our institution more than24 hours from injury (n=7), if they hada vascular injury treated at an outsideinstitution prior to transfer to ours(n=5), or if they had bilateral upperextremity injuries precluding adequatebrachial pressure measurements (n=1).This left a total of 38 patients with 38knee dislocations available for studyevaluation.

Nineteen patients were involved inmotor vehicle accidents, eleven werepedestrians struck by automobiles, twowere injured in industrial accidents,three fell from a significant height, twowere injured during athletics, and onemorbidly obese patient sustained adislocation stepping from bed.

Those patients with an ABI ≥ 0.90were immobilized and admitted forserial examinations, delayed arterialduplex examination and eventualligament reconstruction. Patients withan ABI < 0.90 underwent either

emergent arteriography (Figure 1) in aradiology suite adjacent to ED, orunderwent immediate surgicalexploration. Three patients with non-dopplerable pedal pulses in the injuredlimb were considered to have an ABIof 0.0. All patients were seen in follow-up at two weeks and six weeks, thenthree, six and twelve months frominjury or last surgery, and then annuallywhen possible.

RESULTS

Eleven patients had ABIs < 0.90(range 0.0 - 0.74) in the limb ipsilateralto the knee dislocation (Figure 2). ABIsfor the contralateral limb ranged from0.92 - 1.2 (average 1.05). Two patientswith expansile knee hematomasunderwent emergent exploration andrevascularization with reversesaphenous vein grafting (RSVG) for atransected popliteal artery. Thesepatients did not undergo preoperativearteriography. The nine remainingunderwent emergent arteriography andwere found to have arterial injuryrequiring surgical intervention. Theirarterial injuries included 6 poplitealartery occlusions, one popliteal arterytransection, one common femoralartery thrombosis with peroneal arterythrombosis, and one superficial femoralartery (SFA) high-grade chronicstenosis with an intimal flap that alteredthe popliteal artery flow. All underwentsurgical revascularization and RSVG.One patient also required angioplastyof his SFA stenosis. Follow-up for thesepatients averaged twelve months (range8-24 months) and at their most recentfollow-up, ten of eleven patients had nocomplication of revascularization. Onepatient airlifted from a remote locationarrived in the ED sixteen hours afterinjury. This patient had an ABI of 0.25and a popliteal artery occlusionapparent on arteriogram. Despiteemergent revascularization andmultiple fasciotomies, this patientdeveloped progressive necrosis of allfour leg compartments and ultimatelyrequired above knee amputation.

The Value of the Ankle-Brachial Index for Diagnosing ArterialInjury Following Knee Dislocation: A Prospective Study

WILLIAM J. MILLS, M.D., DAVID P. BAREI, M.D., F.R.C.S.(C), ANTHONY M. BUONCRISTIANI, M.D.,AND PATRICK MCNAIR, M.D.


For this series of patients with kneedislocations, an ABI of < 0.90 indicatedsurgical arterial injury in 100%.Similarly, the sensitivity, specificity andPPV of an ABI < 0.90 was 100%.

Twenty-seven patients had ABI’s ≥0.90 in the injured limb (range 0.90 -1.46 - see Figure 2). ABI’s for thecontralateral limb ranged from 0.92 -1.46 (average 1.08). None had evidenceof vascular injury detectable by dailyserial clinical examination and/orarterial duplex ultrasonography.Clinical follow-up in this groupaveraged 19 months (range 4-36months). No patient in this group hadsigns of delayed vascular compromise.In this study population, the negativepredictive value of an ABI > 0.90 was100%.

Of eleven patients with vascularinjury and ABI < 0.90, one had normalpalpable pulses (matching thecontralateral limb), and an ABI of 0.74.This patient had arteriographic findingof a chronic SFA 90% stenotic lesionand a popliteal artery flow limitingintimal flap. Ten patients in this grouphad diminished (n=1) or absent (n=9)pulses. Of 27 patients with ABI > 0.90,three had decreased (n=2) or absent(n=1) pulses, and 24 had normal pulses.

The sensitivity, specificity and PPV foran abnormal pulse examination indetermining arterial injury weretherefore 91%, 89% and 77%respectively. The NPV of a normal(symmetric) pulse examination was96%.

DISCUSSION

The utility of the Ankle-BrachialIndex in assessing chronic arterialinsufficiency is well established. Its rolein assessing acute arterial injury is alsosupported for both penetrating andblunt upper and lower extremitytrauma.

Our data suggest that the ABI isextremely useful as a non-invasivescreening tool for assessing acutevascular injury in the patient with kneedislocation. In our series, the ABI hada greater sensitivity, specificity andpositive predictive value (all 100 %)than our clinical pulse evaluation. Hadwe relied only on clinical presentation,three patients would have undergonean unnecessary arteriogram. Moreimportantly, one patient with normalpulses but acute arterial injury wouldhave been excluded altogether fromarteriographic study.

Our data also supports the role of

preoperative arteriography in patientswith ABI scores < 0.90. Thoughevidence reveals that the poplitealartery is the most likely site of vascularinjury, in two of our study patients,other than popliteal artery injuryoccurred. One patient with multipleinjuries including a bowel injury, scalplaceration and ipsilateral thighdegloving injury had a commonfemoral artery occlusion with ipsilateralperoneal artery thrombosis. His ABIscore was 0.0. A second patient with asubarachnoid hemorrhage and kneedislocation following MCA had both achronic high-grade SFA stenosis and anacute popliteal artery injury. His ABIscore was 0.74. Given our currentdiagnostic tools, only arteriography,whether done in the ED or at the timeof surgery would enable rapididentification of these anatomicalvariations.

CONCLUSION

Our data support the conclusionthat routine arteriography is notnecessary to accurately diagnose avascular injury. Though the proponentsof physical examination alone providepromising clinical results, our studyindicates that the ABI should be

Figure 1: Arteriograms in a patient with an ABI of 0.25 demonstrating complete popliteal artery occlusion.


included as an integral part of theexamination process. The ABI wasinitially developed as a means tominimize the rate of negativearteriography, to avoid arteriographythat failed to alter overall management,to minimize potential complications, toavoid treatment delays and to lowercosts. Our results suggest that the ABIeffectively accomplishes all these goalsin the patient with knee dislocation.The ABI is neither time consuming,invasive, costly, nor is it associated withcomplications. We feel strongly that theABI should be considered not a separateprocedure, but an extension of thephysical examination in patients witha knee dislocation.

ACKNOWLEDGEMENT

The authors wish to thank SarahHolt, M.S., for her assistance in thestatistical analysis of the data presented.

RECOMMENDED READING

Green NE, Allen BL. Vascular injuriesassociated with dislocation of the knee.J Bone Joint Surg Am 1977;59(2):236-9.

Miranda FE, Dennis JW, Veldenz HC,et al. Confirmation of the safety andaccuracy of physical examination in theevaluation of knee dislocation forinjury of the popliteal artery: aprospective study. J Trauma2002;52(2):247-51; discussion 251-2.

Vascular Injury Vs. ABI

0

0.25

0.5

0.75

1

1.25

1.5

No Vascular Injury Vascular Injury

0.90

Figure 2: Graphic representation of the ABI in those patients with vascular injury all below 0.90 (range0.0 - 0.74), and those without vascular injury, all above 0.90 (range 0.92- 1.46).

Barnes CJ, Pietrobon R, Higgins LD.Does the pulse examination in patientswith traumatic knee dislocation predicta surgical arterial injury ? A Meta-analysis. J Trauma 2002; 53(6): 1109-1114.

Johansen K, Lynch K, Paun M, CopassM. Non-invasive vascular tests reliablyexclude occult arterial trauma ininjured extremities. J Trauma1991;31(4):515-9; discussion 519-22.

Lynch K, Johansen K. Can Dopplerpressure measurement replacearteriography in the diagnosis of occultextremity arterial trauma? Ann Surg1991;214(6):737-41.

Cole PA, Campbell R, SwiontkowskiMF, Johansen KH. Doppler arterialmeasurements reliably exclude occultarterial injury in blunt extremitytrauma. (Abs) Orthopaedic TraumaAssociation, October 1998.


A recent study revealed that 80%of incoming residents across avariety of medical fields failed

to demonstrate basic competency inmusculoskeletal medicine on a 25-question exam whose content wassupported by Orthopaedic ProgramChairs and Medicine DepartmentChairs across the country. It is unclearwhether the failure of these newresidents was a result of curricularinadequacies at the pre-clinical level ordeficiencies in clinical experiences.

BACKGROUND

The University of Washington trainsapproximately 180 students per year.The major pre-clinical experience inmusculoskeletal medicine takes placeduring winter quarter of the secondyear. This course, entitled theMusculoskeletal System, lasts 8 weeksand includes approximately 40 totalhours of contact time. Of this, 16 hoursare spent in the gross lab doingdissection, 4 hours are spent in smallgroup physical exam, and 20 hours arespent in anatomy and clinical

correlation lectures. During this samequarter of second year, 25 hours arespent in a problem-based curriculumwith an emphasis on orthopaedicissues, and additional 2 hours are spentlearning physical exam skills in anintroductory clinical medicine course.Clinical rotations in Orthopaedics areelective, routinely lasting four weeks,though all students are required tocomplete a four-week rotation inRehabilitation Medicine.

The caliber of University ofWashington students is high. The meanPart I and Part II National Board scoresfor the medical school are above thenational average, and the pass rate forthe exam is also above average. Overthe past five years, University ofWashington students have averaged0.2-0.3 standard deviations above thenational mean on the musculoskeletalsub-test.

METHODS

Surveys of second-, third-, andfourth-year medical students at theUniversity of Washington were

administered to evaluate theeffectiveness of the curriculum inteaching key concepts inmusculoskeletal medicine. Second-year students were surveyed twice,before and after their primaryintroductory course in themusculoskeletal system. The surveyinstrument was a two-form, web-based,anonymous questionnaire, eachconsisting of 12 of 24 short-answerquestions on musculoskeletal medicine,previously validated at the Universityof Pennsylvania. This report containsdata from one year’s survey of a secondyear class, pre- and post-course, a thirdyear and a fourth year class.

RESULTS

Survey results revealed markedimprovement in understanding of basicmusculoskeletal medicine in second-year students as a result of the pre-clinical course, with the mean scoreimproving from 19% to 60% correct(pre-test standard deviation of 12, posttest standard deviation of 20) (Figure1). Yet, only 20 of 74 (27%) second year

When Do Medical Students Learn Musculoskeletal Medicine?

GREGORY A. SCHMALE, M.D.

Picture 1: Second year medical students practice physical examination of the knee during their musculoskeletal system course.


Figure 1: Histogram of student scores by year in medical school.

NumberofStudents

Score Number Passing Percent Passing

(mean) (std dev) (range) Score > 70%Second YearPre-Course

79 19% 12 0-54 0 0

Second YearPost-Course

74 60% 17 23-91 20/74 27%

Third Year 34 53% 13 29-90 5/34 15%Fourth Year 59 62% 17 27-100 20/59 34%

Table 1: Study groups.

students post-course demonstratedcompetency, deemed as a score greaterthan or equal to 70%. Likewise, onlyfive of 34 third-year students (15%) and20 of 59 fourth-year students (34%)demonstrated competency (Table 1).Though less than 50% participationwas achieved for each survey, theseresults support prior reports of the highincidence of inadequate preparation ofmedical students in musculoskeletalmedicine.

No differences were found whencomparing types of questions -anatomic vs. clinical - by year. An itemanalysis revealed the lowest frequenciesof correct answers were on questionsrelated to disc disease and low backpain, across all years of students.

DISCUSSION AND CONCLUSION

These results call into question the

appropriateness of our pre-clinicalcurriculum in musculoskeletalmedicine. Comparison of questions tocurriculum for the second yearintroductory course has revealed thateach topic represented by the 24 surveyquestions is represented by directinstruction in the curriculum. It maybe the case that without routine clinicalreinforcement of the issues first raisedin the musculoskeletal introductorycourse, learning will not reliably occur.The routine use of standardizedmultiple choice tests as a measure ofcompetency may also disguise a higherlevel of misunderstanding thansuggested by scores on thesestandardized examinations.

RECOMMENDED READING

Freedman KB. Bernstein J. Educationaldeficiencies in musculoskeletal

medicine. Journal of Bone & JoinSurgery - American Volume. 84-A(4):604-8, 2002 Apr.

Freedman KB. Bernstein J. Theadequacy of medical school educationin musculoskeletalmedicine.[comment]. Journal of Bone& Joint Surgery - American Volume.80(10):1421-7, 1998 Oct.

Pinney SJ, Regan WD. Educatingmedical students aboutmusculoskeletal problems. Arecommunity needs reflected in thecurricula of Canadian medical schools?J Bone Joint Surg Am 2001;83-A(9):1317-20.

Craton N, Matheson GO. Training andclinical competency in musculoskeletalmedicine. Identifying the problem.Sports Med 1993;15(5):328-37.


Despite progress in treatment,Ewing’s sarcoma remains anoften fatal skeletal cancer of

adolescents. In over 90% cases ofEwing’s sarcoma and the related

peripheral primitive neuroectodermaltumors (PNET), a highly specific andrecurrent t(11;22) balancedchromosomal translocation results inthe fusion of a portion of the gene EWS

on chromosome 22 to a portion of thechromosome 11-derived gene FLI-1.This translocation is believed to be thefundamental genetic defect that causesEwing’s sarcoma. The resultant EWS/FLI-1 fusion protein causes abnormalgene expression (transcription andsplicing). Despite the discovery of thefusion protein over 10 years ago thetreatment of Ewing’s sarcoma remainsa combination of surgery,chemotherapy and radiotherapy. Mostimportantly, even with currenttreatment regimens a significantnumber of children and young adultswill develop untreatable pulmonarymetastases.

Recent advances in tumor biologyhave led to an effective molecularlytargeted treatment, the Bcr-Abl kinaseinhibitor imatinib mesylate (Glivec),for chronic myelogenous leukemia andgastrointestinal stromal tumors. Similartranslocations are frequently present inmusculoskeletal sarcomas and as such,offer an attractive rationale for targetedmolecular therapies. Several studieshave reported on the use of antisenseribonucleic acid (RNA) technology tomodulate the growth of Ewing’s cells inculture and in mice by targeting themessenger RNA (mRNA) that encodesthe chimeric protein EWS/FLI-1. Inessence, antisense RNA binds to andinactivates the complementary strandof mRNA, leading to a specificreduction in EWS/FLI-1 mRNA andprotein. RNA interference (RNAi) is arecently described endogenousbiological process that like antisenseRNA destroys specifically targetedmessenger RNAs. RNAi using doublestranded “small interfering” RNA(siRNA) is a more robust process thanantisense RNA in that lowerconcentrations of RNA are required,suppression of the targeted message ismore complete, and the requirementsto identify a specific susceptible regionof the targeted mRNA are less stringent.The double stranded RNA effectssequence specific gene suppression oftargeted mRNA of identical sequence.EWS/FLI-1 messenger RNA possesses

RNA Interference Suppresses Expression of EWS/FLI-1 inEwing’s Sarcoma Cells

HOWARD A. CHANSKY, M.D., ANNA ZIELINSKA-KWIATKOWSKA, M.D., YOSHITO MATSUI, M.D., PH.D.,M. BLACKBURN, B.S., ERNEST U. CONRAD, M.D., JIM D. BRUCKNER, M.D., AND LIU YANG, PH.D.

Figure 1: Western blot of SK-ES cells treated with double-stranded small interfering RNAs targetingEWS/Fli-1 mRNA. EWS/Fli-1 protein was detected with anti-Fli-1 antibody. siRNA-2 is small interferingRNA EFT11 (negative control), siRNA-2 is small interfering RNA targeted against the EWS/Fli-1breakpoint, and siRNA-3 is small interfering RNA Fli1-1282 targeted against the 3' end of EWSS/Fli-1.As shown in lane 3, siRNA Fli1-1282 results in a nearly complete absence of expression of EWS/Fli-1protein. There was no detecable normal (“wild type”) Fli-1 protein detected in the control orexperimental cells. The row labeled “rabbit IgG” refers to the antibody used to concentrate the EWS/Fli-1 fusion protein.

Picture 1: Gross example of Ewing’s sarcoma tumor, seen at end of tweezers.


a unique genetic sequence in the regionof the fusion (breakpoint). This specifictarget makes EWS/FLI-1 an attractivetarget for RNA interference. Previousstudies with antisense technology havedemonstrated that targeting EWS/FLI-1 can revert malignant Ewing’s cells toa more normal phenotype, evenshrinking tumors in a mouse model.Here we describe our initial attemptsto “knock-down” expression of EWS/FLI-1 in the widely studied SK-ES(Sloan Kettering - Ewing’s Sarcoma)Ewing’s sarcoma cell line.

RESULTS

The antisense and sense strands ofthe siRNA targeted against the junctionof EWS and FLI-1 and the the 3-primeend of FLI-1 mRNA were chemicallysynthesized. SK-ES cells were grown toconfluence in a 15 cm plate. The cellswere trypsinized and resuspended insterile saline with various double-stranded siRNAs at a concentration of200 nM. A standard process referred toas electroporation (literally, shockingthe cell membrane) was used tofacilitate the passage of the siRNAthrough the cell membrane. A randomsequence of double-stranded siRNAwas used as a negative control (EFT11).The cells were incubated for anadditional 36 hours in McCoy’s tissueculture medium and then wereharvested for Western blotting. Westernblotting permits the separation andquantitation of proteins of differentsizes by the speed through which theytravel through a polyacrylamide gel.

The proteins in the gel were transferredto a nylon membrane and EWS/FLI-1was detected with a commerciallyavailable monoclonal anti-FLI-1antibody.

As demonstrated in Figure 1, Lane3, there was a dramatic reduction inexpression of the EWS/FLI-1 fusionprotein associated with the doublestranded siRNA (FLI-1 1282) targetingthe 3' end of the fusion protein.Surprisingly, the siRNA (EFT21, Lane2) that specifically targeted the fusionbreakpoint did not appreciably affectexpression of EWS/FLI-1. The negativecontrol (EFT11, lane 1) did not affectexpression, nor did combinations ofsiRNAs result in additional repressionof expression (Lanes 4-6).

One subset of SK-ES cells waselectroporated under identicalconditions as for the siRNA with DNAencoding green fluorescent protein(GFP). Green fluorescent protein issynthesized only by cells that have takenup the DNA and is readily detectablewith a fluorescent microscope. Thisyields what is referred to as transfectionefficiency, or the percentage of cells thathave successfully taken up the GFPDNA and are synthesizing GFP. As seenin Figure 2, the transfection efficiencywas approximately 50%.

DISCUSSION

These results suggest that it ispossible to specifically suppressexpression of EWS/FLI-1 chimericmRNA and protein with doublestranded siRNA. The newly discovered

process of RNA interference is a potentmethod to specifically target individualproteins. The sequence of DNA at thebreakpoint of EWS/FLI-1 is foundnowhere else in the human genome andthus represents the ideal target forsiRNA. It is estimated that there is abouta 50% likelihood that any given 21-basesequence is susceptible to RNAinterference thus we are nowdeveloping addition siRNAs targetedagainst the breakpoint. As seen inFigure 1, Ewing’s cells do not synthesizedetectable quantities of FLI-1 protein.Thus the 3' end of EWS/FLI-1, whileshared with the normal FLI-1 protein,still may be a legitimate target for RNAinterference. We are now evaluating theeffects of our siRNA consructs on thegrowth and phenotype of Ewing’s cellsin vitro. At least in the earlier stages ofthe disease, Ewing’s sarcoma cells likelyrequire continued expression of EWS/FLI-1 RNA and protein to maintain acancerous phenotype. As techniques forviral delivery of siRNA improve topermit long-term stable expression ofspecific siRNAs, RNA interferencemight one day be used in amultimodality treatment strategy thatincludes gene therapy.

RECOMMENDED READING

Delattre O, Zucman J, Plougastel B, etal. Gene fusion with an ETS DNA-binding domain caused bychromosome translocation in humantumours. Nature 1992 359(6391):162-5.

Figure 2: Transfection efficiency of SK-ES cells with electroporation. Light microscopy (magnification 100X) of SK-ES cells on left half of image with the samefield illuminated with fluorescence microscopy on right side. Ther SK-ES cells have taken up green fluorescent protein (GFP) DNA after electroporation.Expressed GFP appears green when exposed to fluorescent light. Transfection efficiency, or the percentage of cells that express GFP is approximately 50percent.


Yang L, Chansky HA and Hickstein D:EWS/Fli-1 fusion protein interacts withhyperphosphorylated RNA polymeraseII and interferes with serine-arginineprotein-mediated RNA splicing. (2000)J Biol Chem. 275(48):37612-8.

Ouchida M, Ohno T, Fujimura Y, et al.Loss of tumorigenicity of Ewing’ssarcoma cells expressing antisense RNAto EWS-fusion transcripts. (1995)Oncogene 11(6): 1049-54.

Fire A, Xu S, Montgomery MK, KostasSA, et al. Potent and specific geneticinterference by double-stranded RNAin Caenorhabditis elegans. (1998)Nature 391: 806-811.

Elbashir SM, Harborth J, Lendeckel Wet al. Duplexes of 21-nucleotide RNAsmediate RNA interference in culturedmammalian cells. (2001) Nature 411:494-498.


The internal fixation of scaphoidfractures remains a challengingtask. Use of the Herbert screw

for the internal fixation of acutefractures of the scaphoid waist gainedpopularity after its introduction in1984.

It has been shown that the technicalaspects of screw insertion are importantin the clinical outcome of acutefractures of the scaphoid waist. Inaddition, higher rates of centralplacement in the proximal fragment ofthe scaphoid have been reported whenusing cannulated screws. In this study,we investigate a model of internalfixation for scaphoid waist fracturesusing the Herbert-Whipple cannulatedscrew. Biomechanical properties of thisfixation are evaluated, under single loadto failure conditions, in centrallyaligned versus eccentrically alignedscrews. Thus, we test our hypothesisthat there exists an optimal positionalplacement for screws used in theinternal fixation of acute fractures ofthe scaphoid waist, centrally locatedwithin the proximal fragment of thescaphoid.


Each of eleven bilateral pairs ofscaphoids from fresh cadaveric wristswas used in this study. The density ofeach scaphoid was determinedaccording to Archimedes’ principle.

One scaphoid in each pair hadcentral screw placement and the othereccentric. A smooth transverseosteotomy was made at the narrowestaspect of the scaphoid waist using acustom manufactured jig and theosteotomy fixed using standardtechnique for Herbert-Whipplecannulated screws. This resulted in allspecimens having the same length ofthread contact area and a reproducibleeccentric position with respect to thecentral axis of the scaphoid. Screwplacement was confirmed withorthogonal radiographs and centralpositional placement was defined as ascrew position within the central one-third of the proximal fragment of thescaphoid on both the antero-posteriorand lateral radiographs (see Figures 1and 2).

For mechanical testing, theproximal fragment of each fixed

specimen was potted inpolymethylmethacrylate and orientedat a 45-degree angle to the horizontalplane to mimic its normal attitude in awrist held in the neutral position. Thisenabled delivery of a dorsal-to-volarcantilever load via pneumaticallydriven plunger, which represents theprimary physiologic load encounteredby the scaphoid (see Figure 3). Aconstant load was applied twice, firstthe load was increased until twomillimeters of displacement wasrecorded, then the load was returned tozero and increased again to record loadat failure and mechanism of failure.Stiffness was calculated using the loadand displacement measurements.

STATISTICAL METHODS

Statistical analysis was performedusing software to calculate theWilcoxon signed rank non-parametricanalysis to compare the specimen’sstiffness and strength. Power analysisresulted in a computed sample size of7.84. To provide an extra margin ofsafety, we have elected to use elevenmatched pairs of scaphoids.

RESULTS

When measured according toArchimedes’ principle, the meandensity (standard deviation) for leftscaphoids was 1.67 grams per cubiccentimeter (.84), while right scaphoidshad a mean density of 1.98 grams percubic centimeter (1.03). There was nosignificant difference between thedensity of the right and left scaphoidswithin each matched pair. Tables 1 and2 show the data for biomechanicaltesting.

For all specimens, the mode offailure was screw migration andfracture at the screw-bone interfacewith dorsal wedge opening. In no casewas there either fracture at less than twomillimeters of displacement or fractureof the polymethylmethacrylate. Uponretrieval of all screws after failure, therewas no evidence that any of the screwssustained a plastic bending deformity.

Central Screw Placement in Simulated Scaphoid Waist Fractures:A Biomechanical StudyWREN V. MCCALLISTER, M.D., JEFF KNIGHT, B.S., ROBERT KALIAPPAN, M.D.,AND THOMAS E. TRUMBLE, M.D.

A B

Figure 1: Central screw position. Antero-posterior (A) and lateral (B) radiographs showing centralplacement within the central one-third of the proximal pole of the scaphoid.


DISCUSSION

Our hypothesis was that centralscrew position in the proximalfragment of the scaphoid would offerbiomechanical advantage whencompared to eccentric positioning.There are no directly comparablestudies available for comparison. Weselected a transverse osteotomy modelfor this study because it resembles theoriginally reported fracture by Herbert

and Fisher and is an accepted model inpreviously published investigations. Itis for this fracture type that internalfixation is advocated in the acute settingwhen displaced or unstable and forthose with non-displaced fractures whodesire a quicker return to function.

Technical aspects of screw insertionare critical to the success of internalfixation when treating acute fracturesof the scaphoid waist. Accurate

placement of non-cannulated screws inthe proximal fragment of the scaphoidis difficult and the use of a cannulatedsystem improves accuracy ofplacement.

Biomechanical testing hasdemonstrated that cannulated screwsare superior to the non-cannulatedHerbert screw in their ability togenerate compressive forces across thefracture and resist cyclical bendingloads.

Central placement of the screw inthe proximal fragment of the scaphoid,using either cannulated or Herbertscrews, is associated with significantlydecreased times to union. In addition,there is evidence that post-operativerange of motion is correlated with thedegree of scaphoid alignment achievedby internal fixation.

The results of our investigationsuggest that when screws are positionedin the central portion of the proximalfragment of the scaphoid, significantlygreater force is required to generatefracture displacement when comparedto screws that are eccentricallypositioned (see Tables 1 and 2).

Limitations of this study include theuse of cadaveric specimens, use of alinear osteotomy, and lack ofmeasurement under cyclical loadingconditions. In addition, this study issubject to the usual limitations thatapply to biomechanical investigationswhen attempting to extrapolate to theclinical setting including the impact ofthe local tissue environment and forcesrealized in vivo.

The advent of the Herbert screwgreatly improved the internal fixationof scaphoid fractures. The introductionof cannulated screws demands that thesurgeon make an informed decisionregarding their use. Currentexperimental evidence supportsimproved compressive forces and betterresistance to cyclical bending loadswhen using cannulated screws. Inaddition, improved post-operativealignment and range of motion areassociated with central placement of thescrew in the proximal fragment of thescaphoid and using cannulated screwsresults in a higher rate of centralplacement. These data, in combinationwith the results of this study, suggestthat central screw placement in theproximal fragment of the scaphoidoffers both clinical and biomechanicaladvantages. Clinical efforts and

A B

Figure 2: Eccentric screw position. Antero-posterior (A) and lateral (B) radiographs showing eccentricplacement with respect to the central one-third of the proximal pole of the scaphoid.

Figure 3: Schematic depicting the potted scaphoid specimen and testing apparatus for delivery ofdorsal-to-volar cantilever bending load.


techniques that facilitate screwplacement in the central portion of theproximal fragment of the scaphoidshould be encouraged whenundertaking the open reduction andinternal fixation of acute fractures ofthe scaphoid waist.

RECOMMENDED READING

Adams, B. D.; Blair, W. F.; Reagan, D.S. and Grundberg, A. B.: Technicalfactors related to Herbert screwfixation. J Hand Surg [Am],13(6):893-899, 1988.

Carter, F. M., 2nd; Zimmerman, M. C.;DiPaola, D. M.; Mackessy, R. P. andParsons, J. R.: Biomechanicalcomparison of fixation devices inexperimental scaphoid osteotomies. JHand Surg [Am],16(5):907-912, 1991.

Herbert T. J. and Fisher, W. E.:Management of the fractured scaphoidusing a new bone screw. J Bone JointSurg Br,66(1):114-123, 1984.

McCallister WV, Knight J, Kaliappan R,Trumble TE: Central Screw Placementof the Screw in Simulated Fractures ofthe Scaphoid Waist. A BiomechanicalStudy. Journal of Bone and JointSurgery, Vol. 85-A, No. 1, p.72-77,January 2003.

Trumble, T. E.; Clarke, T. and Kreder,H. J.: Non-union of the scaphoid.Treatment with cannulated screwscompared with treatment with Herbertscrews. J Bone Joint SurgAm,78(12):1829-1837, 1996.

Trumble, T. E.; Gilbert, M.; Murray, L.W.; Smith, J.; Rafijah, G. andMcCallister, W. V.: Displaced scaphoidfractures treated with open reductionand internal fixation with a cannulatedscrew. J Bone Joint Surg Am,82(5):633-641, 2000.

Screw PositionParameter % Central > Eccentric Significance

Stiffness 43%* p < .01

Load at 2 mm displacement 113% p < .01

Load at failure 39% p > .05

Table 1: Results of biomechanical testing (n=11).

Table 2: Comparing biomechanical properties of central versus eccentric screw placement in the proximal pole of the scaphoid. *Central screw positionresulted in 43% greater stiffness when compared to the eccentric screw position.

Screw PositionParameter Central Eccentric

Stiffness (N/mm) 12.7 ± 4.47 8.88 ± 3.44

Load at 2mm displacement (N) 126 ± 74.9 59.1 ± 47.8

Load at failure (N) 712 ± 412 513 ± 354


The term clawed hallux refers toa specific deformity of the greattoe. It is clinically defined as the

extension of the firstmetatarsophalangeal joint (MTPJ)combined with flexion of theinterphalangeal joint (IPJ) (Figure 1).It causes 2 clinical problems; anincreased pressure beneath the firstmetatarsal head and deformity thatcauses the toe to rub on the shoe, eitherof which may lead to the developmentof an ulcer (Figure 2). This imbalanceis presumed to be a result of afunctional “overpull” (adisproportionate load) of one or moreof the three extrinsic muscles of the firstray: the peroneus longus (PL), extensorhallucis longus (EHL), and flexorhallucis longus (FHL). This deformityhas been reported in several patientpopulations including those withcerebral palsy, SCI, head injury, anddiabetes mellitus (Figure 1). Only onesurgical procedure is described fortreatment of the clawed hallux and it isblind to the deforming force. Thepurpose of our research was to assessthe effectiveness of different surgicalinterventions (the modified Jonesprocedure and the flexor hallucis longustransfer) in correcting both thedeformity and its mechanicalconsequences. We examined theangular changes at the MTPJ and IPJas well as the plantar pressures beneath

the first metatarsal and the distal hallux.

METHODS

Six fresh-frozen cadaver feet (84years of age, SD 3.5 years) wereobtained for this IRB-approved study.The tibia and fibula of each foot weresectioned 12 cm proximal to the ankleand the extrinsic tendons weredissected free to the superior extensorretinaculum. The feet were randomlyselected for either the modified Jonesprocedure (EHL transfer into the dorsalfirst metatarsal) or transfer of the FHLinto the proximal phalanx. These twoprocedures address different etiologies.

In the modified Jones procedure,the IPJ of the great toe was exposedthrough an L-shaped incision on themedial side of the foot. The tendon ofthe EHL was dissected and cuttransversely 1cm proximal to the IPjoint. The synovial sheath around theEHL was excised. A transverse hole 3mm was drilled on the inferomedialaspect of the first metatarsal neck andcontinued along the long axis of thebone until the dorsolateral aspect of theneck. The tendon was passed throughthe hole and sutured to itself and theperiosteum for reinforcement.

For the FHL transfer, a medialincision was made over the MTPJ andcarried distally to the IPJ. The FHLtendon was exposed under theproximal phalanx. The incision was

carried distally to expose the FHLattachment at the distal phalanx. Thetendon was cut at its most distalattachment to the distal phalanx. Thesynovial sheath of the FHL was excised.A 3 mm transverse hole was madethrough the base of the proximalphalanx. The tendon was passedthrough the hole and sutured to itselfand the periosteum.

The specimens were tested pre- andpost-surgical intervention in a loadingframe capable of statically simulatingdifferent phases of the gait cycle. Theposition of the bones of interest wasmeasured using the Fastrakelectromagnetic motion analysissystem; sensors were fastened to carbonfiber rods that were rigidly attached totalus, 1st metatarsal, proximal phalanxand distal phalanx. The Pedar insolepressure measurement system was usedto measure the plantar pressure. Plastictendon clamps were attached to the freetendons of the extrinsic musculatureand the loading protocol developed byOlson et al was used to simulateoverpull of the PL, FHL, and EHL.Paired t-tests were conducted todetermine statistical significancebetween pre-operative and post-operative conditions. Due to smallsample, differences between surgicalprocedures were not statisticallyanalyzed.

Surgical Intervention for the Clawed Hallux Deformity

FADY N. ELIAS, B.S., WILLIAM R. LEDOUX, PH.D., TRACY J. YUEN, B.S., SOREN L. OLSON, B.S.,RANDAL P. CHING, PH.D., AND BRUCE J. SANGEORZAN, M.D.

Figure 1: A medial/lateral photograph of a clawed hallux patient (a) pre- and (b) post-op (6 weeks).

(A) (B)


RESULTS

Both surgical procedures weresuccessful at straightening the toe, bothat the IPJ and MPJ. A significantreduction was found in the averageMTPJ angles when comparing pre-surgical vs. post-surgical data (Figure3a) with the overpull of PL (0.59˚ ±0.25˚ vs. 0.17˚ ± 0.34˚, p = 0.0118), FHL(3.11˚ ± 1.47˚ vs. 0.22˚ ± 0.47˚,p=.0114), and EHL (3.73˚ ± 1.87˚ vs.0.48˚ ± 1.30˚, p = 0.0103). A significantreduction was also found post-surgically with the average IPJ angles(Figure 3b) when the PL (0.22˚ ± 0.23˚vs. 0.13˚ ± 0.21˚, p = 0.0144) and FHL(10.33˚ ± 6.05˚ vs. 1.70˚ ± 2.35˚, p =0.0363) were overpulled. Although theaverage IPJ angle was reduced with theoverpull of the EHL post-surgically(8.26˚ ± 6.13˚ vs. 1.39˚ ± 2.47˚, p =0.0819), this reduction was notstatistically significant. Surgery alsoresulted in an observed reduction ofplantar pressure beneath the 1stmetatarsal (Figure 4a), but only the PLoverpull was significant (8.83 N/cm2 ±5.21 N/cm2 vs. 4.25 N/cm2 ± 5.84 N/cm2, p = 0.0438). The only trend withthe hallux pressure (Figure 4b) was adecrease in post-surgical pressure withthe FHL overpull, but the difference wasnot significant. Surgical procedures aresuccessful at correcting the deformity

and at reducing the pressure beneaththe metatarsal. Not surprisingly thedegree to which the changes occurreddepend upon which of the muscles wascausing the problem.

DISCUSSION

Our laboratory has demonstratedthat a disproportionate load applied tothe first ray extrinsic muscles creates aclawed hallux. To measure theeffectiveness of surgical correction, themodified Jones procedure and the FHLtransfer were performed on 6 randomlyassigned feet. A reduction in theangular changes at the MTP and IPjoints were observed with bothinterventions as well as a reduction inthe plantar pressure beneath the firstmetatarsal. Both procedures correctedthe deformity.

Aberrant pressure beneath the firstmetatarsal has been associated withdiabetic foot ulcers. Thus optimalplantar pressure reduction is animportant surgical outcome. The trendof the means suggests that the FHLtransfer reduces first metatarsalpressure more than the modified Jones,but the data suggest that the FHLtransfer does not decrease halluxplantar pressure as well as the modifiedJones (data not shown).

Limitations of this study include the

(A) (B)

inability of simulating post-surgicalhealing; we did, however, examine thefeet after testing and found no surgicalfailures with regard to tendon slippage.Another limitation associated with thisstudy is the inability to model anormally dynamic and chronic disorderwith the use of static cadaver feetobtained from advanced age specimens.

The peroneus longus has beenshown to both stabilize and plantar flexthe first metatarsal; it is also the primarymuscle involved in increasing theplantar pressure under the firstmetatarsal. Therefore, future surgicaltreatment should take this into account.Our future investigations will allow usto quantitatively assess which surgicalintervention is best at alleviating theclawed hallux as measured by thereduction of angular changes at theMTP and IP joints as well assignificantly reducing plantar pressureunder the first metatarsal. In addition,this study provides some guidance forselecting a surgical procedure that bestfits the clinical problem.

ACKNOWLEDGEMENTS

Department of Veterans AffairsA0806C and The Medical StudentResearch and Training Program at theUniversity of Washington

Figure 2: A plantar ulcer under the first metatarsal head of the same clawed hallux patient (a) pre- and (b) post-op (6 weeks).


RECOMMENDED READING

Coughlin, M.J. and R. A. Mann, Surgeryof the Foot and Ankle, Mosby, 1999.

Hansen Jr., S T., Lippincolt Williams &Wilkins. 2000.

Olson et al., Foot & Ankle International,in press.

(A)

(B)

Figure 3: (a) Average MTPJ angle (+/- SD) for pre- and post-surgical data. (b) Average IPJ angle (+/- SD) for pre- and post-surgical data. An asterisk (*) denotesa statistically significant difference.


Figure 4: (a) Average first metatarsal pressure (+/- SD) for pre- and post-surgical data. (b) Average hallux pressure (+/- SD) for pre- and post-surgical data. Anasterisk (*) denotes a statistically significant difference.

(A)

(B)


Numerous external factors havebeen associated with studyoutcome in non-surgical

specialty journals. Studies havesuggested that funding from thepharmaceutical industry, publicationfrom certain countries, attitudes of theeditors of the journal in question, andrelationship with the tobacco or alcoholindustries all may be associated withpositive outcomes in the publishedmedical literature. To our knowledge,the potential associations between suchexternal factors and study outcomes inorthopaedics has not been reported.

Orthopaedic journals are placing agreater emphasis on meta-analysis andevidence-based medicine. Both of theseclinically-relevant avenues of inquirydepend upon the absence of systematicbiases and external factors in thesubmission and selection of articles forpublication. For example, if there is abias towards the publication of positiveresults over negative ones in clinicaltrials, meta-analyses will systematicallyoverestimate the success rates oftreatments investigated.

The present report tests thehypotheses that funding source,country of origin, and presence of a co-author with training in epidemiologyor statistics are related to the likelihoodof a published orthopaedic studyarriving at a positive conclusion.

METHODS

Material for the study consisted ofarticles published over a 1-year periodduring 1999-2000 in the AmericanEdition of The Journal of Bone andJoint Surgery (JBJS), The AmericanJournal of Sports Medicine (AJSM),and The Journal of Arthroplasty (JA).

All articles reviewed werereproduced and redacted such that theauthors’ names, departmental andinstitutional affiliations, academicdegrees, and funding disclosures wereobscured from view during the reviewprocess.

A pool of three orthopaedicreviewers analyzed the articles

according to predetermined criteria.Two reviewers provided the primaryevaluation of each article, and a thirdreviewer provided the deciding vote incases when the two initial reviewers didnot agree.

The reviewers also collecteddescriptive data on all articles analyzed,including presence or absence of acontrol group, use of randomization,study design (prospective,retrospective, other), and statement ofa testable hypothesis.

The reviewers then classified eachstudy into one of two groups, withrespect to study outcome: (1) Positive/Favorable/Significant Difference Noted;(2) Negative/Unfavorable/NoSignificant Difference/Other. Thereviewers applied strict, standardizedcriteria to the articles to make thedetermination of outcome, which havebeen published and validatedelsewhere. Briefly, a positive study is onethat emphasizes a beneficial resultobtained by the studied variable. If nobeneficial result was emphasized (or ifthe study was not designed to detect abeneficial result), a positive result isdefined as a study that detects astatistically significant difference of aprincipal study endpoint, regardless ofdirection. To minimize the likelihoodthat any bias toward positive outcomeswould be overstated, studies that didnot clearly meet criteria as positivestudies were counted in the negativegroup.

Three potential predictors of thelikelihood of a positive outcome wereinvestigated: funding source, country oforigin of the study, and presence of acoauthor with particular expertise instatistics or epidemiology. Reviewerswere blinded to these variablesthroughout the review process.

RESULTS

A total of 315 studies were reviewedfrom the three journals. Only 3.5% ofpublished studies were randomized,10.5% clearly stated an experimentalhypothesis, and 21.0% were

prospective. About half of the studiesreviewed (50.5%) included anexperimental control of some type; inprinciple, most or all of those studiescould have articulated a hypothesis.

A majority of the studies that couldbe evaluated for outcome reached apositive conclusion (66.4%). Clinicalstudies (studies involving humansubjects) were not significantlydifferent from basic science studies withrespect to the likelihood of arriving ata positive outcome (p > 0.4).

The receipt of commercial fundingwas the only variable analyzed that wasfound to be significantly associatedwith a positive outcome. Of studies thatreceived any funding from industrysources, 78.9% concluded with apositive outcome, compared with only63.3% of studies that received no suchfunding (p = 0.025; Table 1).

Country of origin and presence ofa statistician as a co-investigator werenot significantly associated withpositive outcome in this analysis(p=0.25, and p=0.94, respectively).

Agreement generally was goodbetween the two primary reviewers

Association Between Funding Source and Study Outcome inOrthopaedic Research

SETH S. LEOPOLD, M.D., WINSTON J. WARME, M.D., E. FRITZ BRAUNLICH, M.D.,AND SUSAN SHOTT, PH.D.

“The human intellect…is more moved andexcited by affirmatives than by negatives.”

Sir Francis Bacon (1561-1626)Renaissance author, philosopher, and father ofinductive reasoning.


during the blinded evaluation phase ofthe study. In only 0.6% of studiesreviewed (two of 315 studies) did thetwo primary reviewers disagree on oneor more study end points; in thosecases, the third reviewer served as theadjudicator.

DISCUSSION

The present report found thatpublished studies receiving fundingfrom commercial parties weresignificantly more likely to present apositive conclusion than studies thatwere either self-funded or fundedthrough non-profit sources. Othervariables, such as the involvement of astatistician with the research, or thecountry of origin of the study, were notfound to be significantly related tooutcome.

The fact that industry-fundedresearch published in orthopaedicjournals was more likely to concludewith a positive message about theimplants or pharmaceuticals evaluateddoes not necessarily imply that thosestudies are flawed, or that theinvestigators have been corrupted.Pharmaceutical companies performconsiderable in-house research,presumably guiding them towardsmore promising medications, whichthey then investigate in collaborationwith university- and community-basedinvestigators, perhaps increasing thelikelihood that their publiccollaborations will result in positivestudies. Sizeable grants fromcommercial interests also may allow theinclusion of more patients, perhapsdecreasing the likelihood of Type-IIstatistical error from insufficientsample size, again tending to increasethe likelihood of studies concluding

positively.On the other hand, there is evidence

that other factors may have an impact.Grants from commercial interests oftenare accompanied by contracts thatinclude restrictive covenants, effectivelygiving the corporate sponsor the rightto prevent publication of the results, ifthose results are deemed detrimental tothe goals of that sponsor. These “gagclauses” have delayed or preventedpublication of numerous studies, andhave resulted in threats of expensivelegal action against investigators andinstitutions attempting to publishdespite those contracts. There arenumerous examples of commercially-funded trials that have been halted bythe sponsor before completion, whenit became evident that the results beinggenerated would be financiallyunfavorable to the sponsoringcorporate interest. Although it is widelyagreed that this practice is unethical, itappears to be common. Psychosocial oreconomic influences affecting theinvestigator have also been suggested ascontributing to the disproportionatelypositive outcomes observed incommercially-funded trials.

The present report also found thata large majority of studies (66.4%)published in three widely-readorthopaedic journals over a one-yearperiod concluded with positiveoutcomes. However, in order to knowwith certainty that the preponderanceof positive studies observed in thepresent report indeed represents a bias,there would need to be an assessmentof both published and unpublishedmanuscripts. This would allow thedetermination of whether there is anincreased conditional probability ofpublication for manuscripts that report

positive outcomes over those thatreport negative ones. Evaluation of this“denominator” was beyond the scopeof the present report. There is, however,strong evidence from other work thatthe high proportion of positive studiesobserved in analyses like the presentreport may represent a bias, and not amere coincidence or a natural by-product of the research process.

Positive-outcome bias is harmful forseveral reasons. First, even the mostcareful reader cannot detect its presenceby reading individual articles. Positiveoutcome bias can only be suspectedwhen a large number of studies areevaluated, as was done in the presentreport. Second, and more importantly,the effect of publication bias in journalsis magnified when clinical trials areselected for inclusion in meta-analysesand clinical reviews. If the publishedliterature consists disproportionately ofpositive studies, the resultant meta-analyses will demonstrate a substantialupward bias in the overall estimates oftreatment effects.

If other investigators confirm thefindings of the present study, and ifsome of the influences and less-than-ethical behaviors of commercial partiesdocumented in non-surgical specialtyjournals are prevalent in orthopaedics,then provision of financial disclosuresalongside published articles may not besufficient. In that circumstance, it mightbe worth considering adjusting thepeer-review process to providedisclosure of potential conflicts ofinterest to reviewers when manuscriptsare reviewed, rather than merelydisclosing such relationships to readersat the time of publication.

Table 1: Comparison of Positive and Negative Studies, by Study Variables. *p < 0.025 compared with studies that received no commercial funding; no othervariables showed statistical significance at the p < 0.05 level.

Outcome Whole Cohort Industry-FundedStatisticianInvolved

Country

USA OtherPositive N(percent)

200 (63.5) 45 (78.9)* 29 (65.9) 120 (65.9) 59 (63.4)

Negative N(percent)

101 (32.1) 12 (21.1) 15 (34.1) 62 (34.1) 34 (36.6)


RECOMMENDED READING

Blumenthal D, Campbell EG, AndersonMS, Causino N, Louis KS: Withholdingresearch results in academic life science:Evidence from a national survey offaculty. JAMA 277: 1224-1228, 1997.

Boyd K: Premature discontinuation ofclinical trial for reasons not related toefficacy, safety, or feasability(Commentary: Early discontinuationviolates Helsinki principles). BMJ 322:605-606, 2001.

Callaham ML, Wears RL, Weber EJ,Barton C, Young G: Positive-outcomebias and other limitations in theoutcome of research abstractssubmitted to a scientific meeting. JAMA280: 254-257, 1998.

King, R: Bitter pill: How a drugcompany paid for university study, thenundermined it, in Wall Street Journal.1996: New York. p. 1.

Mahoney MJ: Publication prejudices:An experimental study of confirmatorybias in the peer review system. Cog TherRes 1: 161-175, 1977.

Rennie D: Editorial: Thyroid storm.JAMA 277: 1238-1243, 1997.

Topically-related website:nofreelunch.org


Age and menopause related boneloss markedly compromiseskeletal integrity. The

accompanying incidence of skeletalfractures substantially increase healthcare expenditures, erode quality of lifeand raise levels of morbidity. Physicalexercise, an attractive lifestyle choice formany, holds promise as a non-invasive,non-pharmacologic interventionagainst age-related bone loss. However,the mild and moderate types of exercisemost successfully implemented in theelderly have proved ineffective inenhancing bone mass. Theineffectiveness of mild and moderateloading is consistent with the agerelated degradation in bonemechanotransduction (i.e., in theability to perceive, initiate and sustainan osteogenic response). Interestingly,we recently observed in both young andadult animals that insertion of a 10-sunloaded rest interval between loadcycles transformed low-magnitude,repetitive loading incapable ofinfluencing bone cell populations intoregimens that substantially enhancedosteoblastic activation (nearly 6-fold)and bone formation (over 4-fold).Given the surprising osteogenicpotency of this unique stimulus, wehypothesized that rest-inserted loadingcould potentially overcome age-relateddeficits in mechanotransductionfunction. Here, we test our hypothesisby examining whether insertion of restbetween load cycles enhances boneformation in the aged skeleton.

METHODS

Aged female C57BL/6 mice (21Months, n=30) underwent mechanicalloading utilizing the non-invasivemurine tibia model. The right tibiaewere subject to a 1-Hz waveform andreceived 50 load cycles/day over 2-weeks with animals being randomlyassigned to one of three protocols: i)repetitive low-magnitude loading(low), ii) repetitive high-magnitudeloading (high), iii) rest-inserted low-magnitude loading wherein a 10-s

unloaded rest interval was insertedbetween each load cycle (low - 10s rest).The low- and high magnitude protocolswere calibrated to induce 1200 µε and

2400 µε at the tibia mid-shaft. Animalswere allowed normal cage activitybetween loading protocols. Calcein wasadministered on days 3 and 12 and

Augmenting Skeletal Integrity at Senescence via “Restful”Physical Activity

SUNDAR SRINIVASAN, PH.D., STEVEN C. AGANS, B.S., KATY KING, B.S., NICHOLAS Y. MOY, B.S.,SANDRA L. POLIACHIK, PH.D., AND TED S. GROSS, PH.D.

(A)

(B)

(C)

Figure 1: Composite fluorescent images of mid-shaft cross-sections from control tibia (A) and tibiaeexposed to low-magnitude (B) and low-magnitude rest-inserted loading (C). While low-magnitudeloading only minimally influenced periosteal bone formation compared to controls, insertion of 10-s restbetween low-magnitude cycles substantially enhanced bone formation (C).


animals sacrificed on day 15. Uponsacrifice, the right (experimental) andleft (intact contralateral control) tibiaewere disarticulated and dissected of softtissue. Mid-shaft cross sections (100µm) were obtained and bone formationassessed via measures of dynamichistomorphometry.

RESULTS

Consistent with the response of thesenescent skeleton to mild types ofactivity, exposure of the aged murinetibia to 50 cycles/d of low-magnitudeloading (low, 1200 µε) did notsignificantly enhance periosteal boneformation rates compared to pooledcontralateral controls (0.05 ± 0.03 vs0.03 ± 0.01 µm3/µm2/d, p= 0.12, Figure1). In contrast, both high-magnitudeloading (high, 2400 µε, 0.12 ± 0.02µm3/µm2/d, p= 0.03), and low-magnitude rest-inserted loading (0.13± 0.03 µm3/µm2/d, p= 0.01)significantly enhanced periosteal boneformation (Figure 1 and Figure 2).

DISCUSSION

Similar to previous reports, low-magnitude mechanical loading (low,comparable in magnitude to thatencountered during mild functionalactivity) was not sufficient tosignificantly enhance periosteal boneformation in the aged skeleton.

Figure 2: When ‘standard’ or repetitive loading protocols were utilized, the aged skeleton required high-loading magnitudes (High) in order to initiatesignificant (*) bone formation (i.e., low magnitude loading was not osteogenic). Instead, insertion of a 10-s rest between low-magnitude load cycles (Low-10s Rest) was potently osteogenic in the senescent skeleton and significantly increased bone formation over 4-fold compared to controls.

Increasing strain magnitudes by 2-fold(high, similar in magnitude to thatattained during strenuous activities)was required to induce significant boneformation. However, implementationof this approach (i.e., high-magnitudeloading) is of limited feasibility as theelderly are unable to consistentlycomply with strenuous types of loadingevents capable of building bone mass.As an alternate strategy, we found thatinserting a 10-s rest interval betweenload cycles transformed a non-stimulatory low-magnitude regimeninto a potent osteogenic stimulus in theaged skeleton.

In terms of mechanisms underlyingthe observed responses, the benefit ofrest-inserted loading in the agedskeleton is surprising (particularly, thepotency of low-magnitude rest-insertedloading). This is especially soconsidering the limited availability ofosteoprogenitor cells and the increasedpropensity for fully differentiatedosteoblasts in aged bone tissue toundergo apoptosis. Within this context,while rest-inserted loading transformsa subtle stimulus and enables sustainedactivation of the aged periosteum, therest interval may also provide a criticalnon-stimulatory window that facilitatesimproved osteoblast survival. Given theease of implementation of this novelstimulus (inserting an unloaded rest

interval between load cycles), ourfuture investigations of underlyingmechanisms and optimization of thisconcept, including exploitation ofsynergy with select growth factors,should allow consideration of low-magnitude, rest-inserted exerciseregimens for building bone mass andaugmenting skeletal integrity in the frailelderly.

RECOMMENDED READING

Prince, R. L., Smith, M., Dick, I. M.,Price, R. I., Webb, P. G., Henderson, N.K. & Harris, M. M. Prevention ofpostmenopausal osteoporosis. Acomparative study of exercise, calciumsupplementation, and hormone-replacement therapy. N Engl J Med 325,1189-95 (1991).

Rubin, C. T., Bain, S. D. & McLeod, K. J.Suppression of the osteogenic responsein the aging skeleton. Calcif Tissue Int50, 306-13 (1992).

Srinivasan, S., Weimer, D. A., Agans, S.C., Bain, S. D. & Gross, T. S. Lowmagnitude mechanical loadingbecomes osteogenic when rest isinserted between each load cycle. J BoneMiner Res 17, 1613-1620 (2002).


Preisinger, E., Alacamlioglu, Y., Pils, K.,Saradeth, T. & Schneider, B.Therapeutic exercise in the preventionof bone loss. A controlled trial withwomen after menopause. Am J PhysMed Rehabil 74, 120-3 (1995).

Jilka, R. L., Weinstein, R. S., Takahashi,K., Parfitt, A. M. & Manolagas, S. C.Linkage of decreased bone mass withimpaired osteoblastogenesis in amurine model of acceleratedsenescence. J Clin Invest 97, 1732-40(1996).

Chan, G. K. & Duque, G. Age-relatedbone loss: old bone, new facts.Gerontology 48, 62-71. (2002).


I n certain instances, highervertebrates including man respondto injury with regeneration of

functional tissues rather than with theproduction of scar. Digit tipamputation is one such instance. Thecellular and molecular events involvedin this phenomenon are not wellunderstood. The most common animalmodel of this process is presently themouse, whose small size makes certaintypes of study technically challenging.In this study, digit tip amputations were

performed on the middle digit of theleft forepaw of ten New Zealand Whiterabbits. Amputation of the distalphalanx immediately distal to the distalinterphalangeal (DIP) joint wasperformed under 3.5x loupemagnification. Eight weeks later thedigits were removed for study.Regrowth of multiple tissue types(bone, nail, epithelium, subcutaneoustissue) was seen in all cases where theplane of amputation passed distal to theDIP joint, suggesting that the rabbit

may be useful in the study of thisphenomenon, particularly for workwhere the small size of the mouse digitmakes the manipulation of regrowingtissues difficult.

INTRODUCTION

The observation that children canregrow amputated fingertips hasinterested many researchers since thiswas first noted in the 1970s. Our groupworks in the field of hand traumasurgery and is highly motivated todevelop improved treatments for ourseverely injured patients. We havewished to learn more about the cellularand molecular processes involved indigit tip regrowth, believing that thisinformation alone or combined withtissue-engineering techniques may helpus improve present treatment options.

Finding work with mouse digit tipsquite technically challenging due totheir small size, we decided to assess theresponse to digit tip amputation in alarger animal.


MethodsTen immature New Zealand White

rabbits were anesthetized and themiddle digit of the left forepawamputated in a region distal to the distalinterphalangeal joint. Amputated digittips were preserved to document levelof amputation. The amputation sitewas dressed and animals given painmedication. Daily dressing changeswere performed and dressingsdiscontinued when the wound was dry.Eight weeks after surgery the operateddigit was removed from the left forepawfor histological evaluation. At the sametime, the contralateral (right forepaw)middle digit was removed forcomparison.

HistologyAmputated digit tips (Day 0) and

operated digits (eight weeks) as well ascontrol digits (eight weeks) were treatedin formalin and decalcified for twenty-four hours, then embedded in paraffinfor sectioning. Specimens were stainedwith hematoxylin and eosin andsectioned for standard light

Regrowth of Amputated Rabbit Digit Tips

CHRISTOPHER H. ALLAN, M.D., TRACI BARTHEL, M.D., AND BRIAN P. RUBIN, M.D., PH.D.

Figure 1a: Amputated and regrowing digit tip, 8 weeks; line = plane of amputation.

Figure 1b: Second specimen; note preparation artifact with nail overlying distal tip.


microscopy. Day zero amputated digittips were compared with eight weekspecimens to confirm level ofamputation and to evaluate theresponse to tip amputation.

RESULTS

Of the ten specimens evaluated,four passed through the DIP joint,rather than distal to it. In these casesno regrowth of digit tip tissues was seen,and the wound healed with a stump. Inthe remaining six cases where the planeof amputation appeared to pass distalto the DIP joint and through the baseof the proximal phalanx, regrowth ofthe digit tip to varying degrees was seen(Figure 1a and b), with the formationof several tissue types including bone,nail, epidermis, dermis, andsubcutaneous tissue, and withoutevidence of scarring. The tissues regrewincompletely, as compared with theextent of the tissue removed at the timeof amputation (Figure 2) and as shownin an image of a normal rabbit digit DIPjoint without tip amputation (Figure3).

DISCUSSION

Millions of persons are affected bytraumatic limb loss and severeextremity injury. In 1996, (NationalHealth Interview Survey), there were1,285,000 amputees in the US alone.Present treatments includereattachment, repair, revision Figure 3: Rabbit digit DIP joint, no amputation: note volume of bone of distal phalanx compared

with regrowing specimens.

Figure 2: Amputated rabbit digit tip; plane of amputation distal to DIP joint.

amputation, use of a prosthesis,transplantation, and variousreconstructive procedures. It isextremely rare for any of thesetreatment options to restore normalfunction, and several have potentiallysevere side effects.

The features of the regenerativeresponse to injury are the same as thoserequired for successful tissueengineering: cells, scaffolds andsignaling molecules are all involved. Weplan to identify cells and signaling

pathways involved in this regrowth. Wemay either stimulate enhancedregeneration in situ or better engineerreplacement tissues for patients withsevere limb injuries, since the cells andsignaling molecules involved inregenerative healing form new andfunctional tissue so successfully.

Evaluation of healing at differenttime periods and in animals of differentages should provide interestinginformation. More stringent analysis ofthe healing tip is needed, as one of ourpriorities is to identify the firstpopulation of cells to appear afterwounding. These cells may be stem-cell-like in function, and thereforeuseful in enhancing in-situregeneration and tissue engineeringstrategies.

SUMMARY

At eight weeks after tip amputationthrough the digit just distal to the DIPjoint, the rabbit digit tip demonstratesregrowth of multiple tissue typesincluding bone, epithelium,subcutaneous tissue, and nail, with noevidence for scar tissue formation.These findings are consistent withtissue regeneration. The rabbit digit tipamputation may represent a usefulmodel of this phenomenon of tissueregeneration in higher vertebrates.


ACKNOWLEDGEMENTS

The authors gratefully acknowledgeresearch funding support from theRoberts Helping Hand Fund.

RECOMMENDED READING

Illingworth, C.M., Trapped fingers andamputated finger tips in children. JPediatr Surg, 1974. 9(6): 853-58.

Borgens, R.B., Mice regrow the tips oftheir foretoes. Science, 1982. 217(4561):747-50.

Reginelli, A.D., et al., Digit tipregeneration correlates with regions ofMsx1 (Hox 7) expression in fetal andnewborn mice. Development, 1995.121(4): 1065-76.

Muller, T.L., et al., Regeneration inhigher vertebrates: limb buds and digittips. Semin Cell Dev Biol, 1999. 10(4):405-13.

Singer, M., et al., Open finger tip healingand replacement after distalamputation in rhesus monkey withcomparison to limb regeneration inlower vertebrates. Anat Embryol, 1987.177(1): 29-36.

Stocum D. Wound Repair andRegeneration. 1995, Landes Co, Austin.206-207.


A cute traumaticosseoligamentous instability atthe craniocervical junction is

usually fatal. However, improvedpatient retrieval and emergentmanagement techniques have increasedpatient survival to hospital. Delays indiagnosis are thought to havepotentially devastating consequences.Causes of missed diagnosis aremanifold, and include low clinicalsuspicion, presence of multiple lifethreatening injuries, and difficultyvisualizing relationships at thecraniocervical junction onconventional radiographs. Our goalwas to evaluate potentially correctablecauses of delayed diagnosis of CCDwhile comprehensively evaluating aseries of patients treated with posteriorinstrumented occipitocervical fusion.


A retrospective study wasconducted of 17 consecutive survivorsof CCD at Harborview Medical Center(HMC) between 1994 and 2002. Datacollected from spine and traumaregistries, medical records and spinalimaging studies were retrospectivelyreviewed to identify (i) the timing and

manner of diagnosis; (ii) the effect ofdelayed diagnosis on neurologicfunction; (iii) the diagnostic reliabilityof the lateral cervical spine radiograph(C-spine lateral); (iv) surgical and non-surgical complications, and (v)neurologic outcome. A delay indiagnosis was defined as the failure toidentify or suspect CCD uponcompletion of the ATLS-mediatedspinal evaluation, typically defined ascompletion of a lateral plain radiographand CT scan of the cervical spine.

Two patients (12%) had normalsensorimotor function of theextremities, and 15 patients (88%) hadspinal cord injuries (SCI) (2 complete,13 incomplete). All patients hadassociated injuries.

A lateral cervical spine radiographwas obtained as part of the initialtrauma evaluation. CT scans of thehead and cervical spine were obtainedin all patients. Radiographic screeningof the craniocervical junction was doneusing Harris’ lines, Wackenheim’s line,and Powers’ ratio. Nine patients wereseen at other hospitals first and mayhave been initially treated according toa different algorithm.

Once CCD was identified or

suspected, provisional stabilization wassecured through manual reduction andHALO application under fluoroscopicguidance (Figure 1). MRI of thecervical spine was obtained todetermine the extent of ligamentousdisruption and SCI. Patients with MRIfindings of craniocervical osseo-ligamentous injury who had grosslypreserved alignment (< 2mmdisplacement) and whosecraniocervical stability remainedunclear were evaluated with tractiontests (Figure 2).

Definitive management of CCDentailed rigid occipitocervicalsegmental instrumentation andarthrodesis with structural bone graftsecured to the occiput and cervicalspine (Figure 3). Postoperativeimmobilization was achieved with abrace or HALO. Postoperative CT scansassessed alignment and hardwareplacement. Healing was confirmed 3months postoperatively with uprightflexion-extension lateral radiographs.Clinical follow-up was providedthrough the Northwest Regional SpinalCord Injury System.

Diagnosis and Treatment of Craniocervical Dissociation: A Seriesof 17 Consecutive Survivors Over Seven Years

CARLO BELLABARBA, M.D., SOHAIL K. MIRZA, M.D., G. ALEXANDER WEST, M.D., PH.D.,FREDERICK A. MANN, M.D., DAVID W. NEWELL, M.D., AND JENS R. CHAPMAN, M.D.

Figure 1: Widely displaced Stage 3 CCD. From left to right: C-spine lateral x-rays after injury, after provisional closed reduction with a HALO, and afterinternal fixation in a 35 year-old woman who sustained a highly displaced stage 3 CCD when she was struck by a motor vehicle. Even with open reduction,the atlanto-occipital articulation remained distracted and anteriorly translated (black lines).


RESULTS

DiagnosisFour of 17 patients (23%) were

diagnosed with CCD during initialtrauma evaluation. Two of theseinjuries (12%) were detected on theinitial C-spine lateral and the other 2were diagnosed on the cervical spineCT scan. None of these 4 patientsworsened neurologically before surgery.

Thirteen of 17 patients (76%) hada delay in diagnosis, which averaged 2days in length (range 1 to 15 days). Fiveof these 13 patients (38%) hadworsening of their neurologiccondition before CCD was clinicallyrecognized.

Operative stabilization was

undertaken in all patients. Averagefollow-up was 15 months (range 6 to45 months).

One patient had postoperativeworsening of sensorimotor functioncaused by a malreduction that resultedin hydrocephalus and cerebellarinfarction, presumably fromvertebrobasilar insufficiency. Heunderwent urgent occipitocervicalrealignment, decompression andstabilization, and placement of aventriculoperitoneal shunt. At 16-month follow-up, the quadriplegia hadresolved, with a mildly broad-based gaitas his only sign of neurologicdysfunction.

No patient had occipitocervicalpseudarthrosis or hardware failure.

One (6%) superficial woundinfection required antibiotics and localwound care. One (6%) patient who hadrepair of a traumatic dural tear requireda lumbar subarachnoid drain to resolveCSF wound drainage.

The average ASIA motor scoreimproved from 50 preoperatively to 79at follow-up. Eleven of 13 (85%)patients with incomplete SCI improvedby one ASIA grade. Neither of the twopatients without SCI worsenedneurologically. One of the two patientswith complete SCI improved from ahigh cervical to a T5 level. The secondpatient, who presented with widecraniocervical displacement (Figure 1)remained ventilator-dependent with amotor score of zero 16 monthspostoperatively. She was the onlypatient without motor improvement.

The average Basion-Dens Interval(BDI) improved from 16 to 10 mm, andaverage Basion-Axial Interval (BAI)decreased from 10 mm to 8 mm.

DISCUSSION

Autopsy studies have ascribed up to90% of traumatic fatalities to injuriesof the upper cervical spine. Bucholzestimated a survival chance of 0.65 to1% for patients with CCD based on aprospective morgue study. However,dramatic improvements in theemergent management of traumavictims appear to have improvedsurvival. The burden of appropriatelyanticipating, identifying and treatingsurvivors is thus shifting to theclinician.

Previously recognized difficulties indiagnosing craniocervical instabilitywere highlighted in this series. Thehighly unstable nature of these injurieswas either entirely unrecognized ordrastically underappreciated in overthree-fourths of our patients. Thisproblem can be attributed tomisleading clinical and radiographicfeatures, and the lack of a systematicapproach to evaluating thecraniocervical junction. Most patientshad cognitive impairment due to lossof consciousness (13 patients, 76%) ora closed head injury with intracranialhemorrhage (13 patients, 76%), whichcontributed to diagnostic difficulties.Since the trauma C-spine lateralremains key in the initial evaluation ofa patient with suspected cervical spine

Figure 2: Mildly displaced Stage 2 CCD. Clockwise from top left: C-spine lateral, sagittal MRI throughleft atlanto-occipital joint, sagittal MRI through right atlanto-occipital joint, and traction C-spine lateralin a 39 year-old woman who sustained a mildly displaced stage 2 CCD with a basion-dens relationshipwithin 2 mm of normal and only mild, unilateral right-sided loss of atlanto-occipital congruence (whitearrows). The traction radiograph was used to distinguish this as a highly unstable stage 2 injury (blacklines) requiring internal fixation, rather than a stage 1 injury in which closed treatment would havesufficed.


trauma the most alarming finding ofthis study is how infrequentlycraniocervical instability was suspectedon standard radiographic evaluation(2/17 patients, 12%). Retrospectivereview of these x-rays showed both theBDI and BAI to be within normalparameters in only 1 patient.

Even a short delay in diagnosis andtreatment carried potential life-threatening consequences. Five of 13(38%) undetected injuries wererecognized after a decline in neurologicfunction, which occurred even withcervical collar immobilization and fullspine precautions, highlighting theinadequacy of standard bracing instabilizing CCD.

Other pitfalls in the radiographicdiagnosis of craniocervical instabilitycan be attributed to associated fracturesof the upper cervical spine . This tendedto lead to insufficient immobilizationfor craniocervical stability and a loss ofurgency in further evaluating thecraniocervical junction. In the currentstudy, ligamentous avulsion fracturesoccurred in 9 patients (53%), withmore than one fracture type seen in 6patients (35%). The frequency withwhich avulsion fractures were identifiedsuggests that their presence shouldheighten suspicion for CCD and triggerearly follow-up investigation ofligamentous integrity with MRI.

Traynelis has identified threeoccipitocervical injury patterns basedon direction of displacement. However,directional criteria, may be misleadingsince complete ligamentous disruptionrenders the position of the headarbitrary. Furthermore, thisclassification does not address injuryseverity or spontaneously repositioned

injuries. Based on our clinicalexperience, we propose a 3-stageclassification system with therapeuticimplications. A Stage 1 injury is a stableminimally- or non-displacedcraniocervical injury with preservationof sufficient ligamentous integrity fornonoperative treatment. Stage 2represents a spontaneously reducedbilateral craniocervical disruption withminimal displacement (BAI/BDI < 2mm beyond normal) in which atraction test confirms complete loss ofcraniocervical stability, requiringinternal fixation. Stage 3 represents ahighly unstable injury defined by grosscraniocervical malalignment (BAI/BDI> 2mm beyond acceptable limits), alsorequiring internal fixation. The term“craniocervical dissociation” is reservedfor Stage 2 and 3 injuries, whereligamentous instability is complete.

Increased awareness is importantfor prompt diagnosis and stabilizationof craniocervical instability. Thirteenof 17 survivors had a delay in diagnosiseither at our institution or thetransferring hospital. A disciplined,systematic approach to the evaluationof screening cervical spine imagingstudies offers the potential to identifymost injuries. In addition to itstherapeutic implications, aclassification system that includes acategory for spontaneously reduced yethighly unstable injuries may heightenawareness that even a life-threateninginjury such as CCD may present with amisleadingly subtle radiographicappearance. Delayed diagnosis placesthe patient at risk for neurologicworsening. With timely diagnosis andinternal occipitocervical fixation, evenpatients with severe neurologic deficits

may experience a great deal offunctional improvement.

RECOMMENDED READING

Alker GJ, Oh YS, Leslie EV. Postmortemradiology of head and neck injuries infatal traffic accidents. Radiology1975;114:611-7.

Anderson PA, Montesano PX.Morphology and treatment of occipitalcondyle fractures. Spine 1988;13:731-6.

Bell HS. Paralysis of both arms from theinjury of the upper portion of thepyramidal decussation: “Cruciateparalysis”. J Neurosurg 1970;33: 376-80.

Bucholz RW, Burkhead WZ. Thepathologic anatomy of fatal atlanto-occipital dislocations. J Bone Joint Surg1979;61:248-9.

Chapman JR, Bellabarba C, Newell DW,et al. Craniocervical injuries: Atlanto-occipital dissociation and occipitalcondyle fractures. Seminars in SpineSurgery 2001;13:90-105.

Deliganis AV, Baxter AB, Hanson JA, etal. Radiologic spectrum ofcraniocervical distraction injuries.Radiographics 2000;20:S237-50

Harris JH, Carson GC, Wagner LK.Radiologic diagnosis of traumaticoccipitovertebral dissociation: 1.Normal occipitovertebral relationshipson lateral radiographs of supinesubjects. AJR Am J Roentgenol1994;162:881-6.

Figure 3: Operative Technique. Postoperative sagittal CT (left) and lateral C-spine x-ray (right) illustrating congruent reduction of atlanto-occipital and atlanto-axial joints and arthrodesis using segmental occipitocervical instrumentation, C1-2 transarticular screws, structural bone graft with suboccipital and C2-sublaminar fixation (solid arrow) and autologous cancellous bone graft in the C1-2 joints and at the craniocervical junction (open arrows).


Each year, the number of personsliving past the age of 65 yearsincreases. It is estimated that by

the year 2020, the population ofindividuals older than 65 years willclimb to 52 million from its currentlevel of 30 million. Similarly and notsurprisingly, the incidence of acetabularfractures in this age group is alsoincreasing.

Different opinions exist regardingthe optimal treatment for seniorpatients with acetabular fractures.Treatment remains a controversial issuesince these patients have morecomorbid medical conditions whichincrease their risk of complications.These medical comorbidities causesome authors to propose conservativetreatment of acetabular fractures,usually consisting of bed rest with orwithout skeletal traction. Thecomplications of prolonged bed restinclude skin breakdown, urinaryinfections, pulmonary problems,among others. Closed management ofdisplaced acetabular fractures results inarticular malunion with resultantarthritis. Hip arthroplasty in patientswith acetabular malunion is associatedwith increased rates of acetabular

component loosening. In order toavoid such problems, surgical treatmentis advocated by some authors in anattempt to provide accurate fracturereduction and stable fixation.Unfortunately for these patients,osteopenia allows more extensivearticular impaction injuries andsimilarly complicates stable fracturefixation. Certain surgical modificationssuch as cable fixation, allograft corticalbone supplementation, or bone cementaugmentation have been described toaddress the problem of poor bonequality in senior patients withacetabular fractures. Few authorsadvocate acute total hip arthroplasty inthese patients as treatment for theacetabular fracture. Hip arthroplastyin such situations is associated withpredictable complications such asacetabular component loosening andpostoperative dislocation. Acute hiparthroplasty requires restoration of theacetabular fracture osseous anatomyprior to implantation of the componentwhich further prolongs the surgicalprocedure.

We evaluated our senior patientswith displaced acetabular fracturestreated operatively using routine

surgical techniques and implants. Wereport the epidemiology of this formof acetabular fracture treatment inpatients over the age of 65, with specialattention to hospital course andperioperative complications.


Between January 1, 1996 andDecember 31, 1999, we treated 352patients with displaced acetabularfractures operatively usingconventional techniques of openreduction and internal fixation. Only 27(7.67%) of these patients were 65 yearsof age or older. The average age of these27 senior patients was 76.3 years,ranging from 65 to 89 years. There were18 male and 9 female patients. Fifteenfractures were left-sided and 12 wereright-sided injuries. The average timefrom injury until surgery was 3.8 days(range 1 to 8 days). The mechanismsof injury included falls in 14 patients,motor vehicle accidents in 11 patients,and sports injuries in 2 others.Fourteen of the 27 senior patients hadassociated injuries, while 13 hadisolated acetabular fractures. Eightpatients had fracture-dislocationswhich required manipulativereductions acutely in the emergencydepartment. All of the patients wereplaced in skeletal traction prior tosurgery.

Each patient was fully evaluated bythe geriatric medicine and/or generalsurgery consultants prior to surgery. Allpatients were screened for venousthromboses using duplex ultrasoundevaluations. The patients were treatedperioperatively with thromboembolicstockings, sequential compressivedevices, and subcutaneous heparin.

Radiographic evaluations includedanteroposterior pelvic and Judetoblique acetabular plain radiographs,and pelvic computed tomographypreoperatively for all patients. Twopatients were noted to have pelvic ringinvolvement in addition to theiracetabular fractures. The femoral headand acetabular dome were carefullyevaluated for impaction injuries usingthe preoperative pelvic imaging studies.The fractures were then classified

Acetabular Fractures in Senior Patients

ALYSANDRA SCHWARZ, M.D. AND M.L. CHIP ROUTT, JR., M.D.

Figure 1A: This 72 year old female was injured in an automobile accident. She sustained a fracture-dislocation of her right hip. The obturator-oblique pelvic radiograph demonstrates the posterior wallacetabular fracture pattern.


according to Letournel based on theimaging studies. There were 10 patientswith elementary fracture patterns: 5posterior wall and 5 anterior columnfractures. Seventeen patients hadassociated fracture patterns: 9associated both column, 3 anteriorcolumn with associated posteriorhemitransverse, 3 T-type, 1 transversewith associated posterior wall, and 1posterior column with associatedposterior wall.

Operative treatment included openreduction with internal fixation, and

percutaneous fixations. All 27 patientswere positioned on a radiolucentoperating table and receivedintravenous antibiotic prophylaxis.Twenty-five patients had significantlydisplaced acetabular fracturesmandating open reduction and internalfixation. An ilioinguinal anteriorsurgical exposure was chosen for 15patients: 8 with associated bothcolumn, 5 with anterior column, and 2with anterior column with associatedposterior hemitransverse fractures. Alateral Kocher Langenbeck surgical

exposure was selected for 6 patients: 5posterior wall, and 1 posterior columnwith associated posterior wall fractures.A prone Kocher-Langenbeck approachwas used in 4 patients: 3 T-type, and 1transverse with associated posteriorwall fractures .

Standard fixation techniques wereused for all 27 patients. Twenty-threepatients were treated with openreduction and internal fixation using atleast one reconstruction plate, with anaverage of 1.5 plates per patient. Twoother patients underwent openreduction and internal fixation usinglag screws only. Two patients hadminimally displaced acetabularfractures which demonstratedinstability on examination of the hipusing fluoroscopic imaging. These twopatients had closed reduction andpercutaneous fixation of theiracetabular fractures using cannulated7mm cancellous screws only. Onepatient had an associated both columnfracture, and the other patient had ananterior column with associatedposterior hemitransverse fracture.

Seven patients had impactionfractures of the femoral head. Sixpatients had acetabular domeimpaction injuries. Three of the fiveposterior wall fractures had associatedmarginal impaction injuries. Three ofthe six patients with dome impactionrequired cancellous allograft bonegrafting, and all three patients withmarginal impaction underwentautogenous bone grafting from thegreater trochanter.

Postoperative duplex scans wereobtained on all patients immediatelyafter surgery and at one week intervalsuntil discharge.

RESULTS

Twenty-one patients had significantassociated medical conditions atpresentation.

Eight patients had a prior operationat the proposed site for acetabularfixation. The previous surgical site wasincorporated into the acetabularexposure in seven patients. Onepatient’s surgical plan was changed dueto the preexisting surgical incision.

The average estimated blood loss forthe patients treated with openreduction and internal fixation was 817cc. Eight of the 25 patients treated withopen reduction received cell saverblood, with the average amount

Figure 1C: The postoperative pelvic CT scan shows the fracture reduction and implant location. Theosteochondral impaction fragment has been elevated and its defect filled with cancellous bone graft.

Figure 1B: This pelvic computed tomography (CT) axial image of her acetabular dome reveals theposterior displacement, fracture comminution and osteochondral impaction components.


returned being 422 cc. Six of the 25patients treated with open reductionreceived intraoperative bloodtransfusions, with an average of 2.3units transfused. The two patients withpercutaneous fixation had minimalblood loss.

Twenty patients (74%) developed42 documented complications (210%).Only seven patients (26%) had anuncomplicated hospital course.

Multisystem organ failureaccounted for the only postoperativedeath. Ten patients were admitted to theintensive care unit immediatelyfollowing surgery, one patient wastransferred to the ICU on the firstpostoperative day, one patient wasadmitted to a cardiac monitoring ward,and the other fifteen patients wereadmitted directly to a routineorthopedic hospital ward. Sixteenpatients underwent an additional 30invasive diagnostic or therapeuticprocedures while hospitalized.

Fourteen patients (51.9%) haddiagnosed deep venous thromboses(DVT). Two DVTs were notedpreoperatively, and treated with a venacaval filter. Twelve patients haddocumented DVTs postoperatively yetprior to hospital discharge. Three ofthese patients required placement of aninferior vena caval filter due to medicalconditions which prohibited the use ofsystemic anticoagulation. Nocomplications resulted from the venacaval filters. The other nine patients hadroutine systemic anticoagulation.There were no diagnosed incidents ofgastrointestinal bleeding as a result oftherapeutic anticoagulation in thesepatients. There was a single woundhematoma which required evacuationtwo weeks postoperatively.

Cardiopulmonary complicationsoccurred in 14 patients (51.9%). Fivepatients developed pneumonia.Another patient developed asymptomatic pulmonary embolusconfirmed by a ventilation perfusionlung scan. Six patients requiredintubation and mechanical ventilationwhile hospitalized, one requiringconversion to a tracheostomy prior todischarge. Respiratory depressionresulted from excessive narcoticsedation and required chemical reversalin one patient. Another patient hadcongestive heart failure postoperatively.

Neurological complications

resulted in seven patients (25.9%). Fivepatients had transient altered mentalstatus. Two patients sufferedcerebrovascular accidents. Only onepatient suffered delerium tremenssecondary to alcohol withdrawal. Therewas one postoperative sciatic nervepalsy.

Gastrointestinal complications werelimited to postoperative symptomaticileus in only two patients. Two patientsdeveloped bacteremia without knownetiology. Only one urinary tractinfection was identified.

Unusual complications includedpostoperative blindness in one patient,great toe embolus in one patient, andsubdiaphragmatic placement of a chesttube in one patient. Another patienthad thrombocytopenia requiringtransfusion, while one patientdeveloped acute renal failure requiringhemodialysis. One patient died frommultisystem organ failure.

The average duration ofhospitalization was 15.3 days (range 5to 42 days). Twenty three patients wereindependent with activities of dailyliving prior to injury. Of these, 16 weredischarged to a skilled nursing facility,4 were transferred to inpatientrehabilitation, 2 were discharged home,and one expired. Two patients residedin assisted living quarters prior toadmission. Of these, one returned tothe same location and the other wasadmitted to a skilled nursing facility.Both patients who were residing in askilled nursing facility prior to surgerywere readmitted to the same facility.

SUMMARY

Displaced and unstable acetabularfractures in senior patients are difficultto manage effectively for a variety ofreasons. Poor bone quality complicatesthe surgeon’s attempts to achieve stableinternal fixation. The patient’s medicalcomorbidities complicate theirtreatment. Spencer (1989) reported onthe conservative treatment of 25patients over the age of 65 withacetabular fractures. He found a 30%rate of unacceptable functional result.He noted that fractures had latedisplacement, despite sustainedtreatment with traction. He definedpoor prognostic indicators asosteoporosis, femoral head fractures,delayed diagnosis, inadequateradiographs, inappropriate traction,and early weight bearing. Somesurgeons believe that if conservativetreatment does not lead to anacceptable outcome, the elderly patientmay be a candidate for a delayed totalhip arthroplasty. Unfortunately, theosseous deformity of the malreducedacetabulum makes these reconstructiveprocedures more difficult. Clinicaloutcomes after delayed arthroplasty areinferior to those for degenerativeconditions. Romness and Lewallen(1990) found the loosening rates ofacetabular components after posttraumatic reconstructions are four tofive times greater than in surgeriesperformed for routine hip arthroplasty.

Letournel defined osteopenia of theinnominate bone the most importantcontraindication to operative

Figure 1D: She was asymptomatic at her one year follow up evaluation. The oblique acetabularradiographs confirm uneventful healing without arthritis or ectopic bone formation. The osteochondralfragment did not collapse.


treatment, yet he treated 103 acetabularfractures in patients over the age of 60with open reduction and internalfixation. Advantages of a reconstructedacetabulum include possiblerestoration of function, preservation ofnative joint, and, if future arthroplastyis required, more normal osseousanatomy for implant placement.Potential disadvantages areencountered if conversion to a total hiparthroplasty is required and include thepresence of scar tissue and an increasedinfection rate.

In our series, 27 senior patients withacetabular fractures were treatedoperatively and, 74% of these patientshad a complicated clinical course.Twenty of our patients had a 210%complication rate.

CONCLUSION

The surgical treatment of displacedacetabular fractures is especiallychallenging in senior patients. Thosewith medical comorbidities haveincreased postoperative complicationrates. Senior patients and their familymembers should be aware of suchassociated risk prior to operativeintervention.

RECOMMENDED READING

Helfet DL, Borrelli J Jr, DiPasquale T,Sanders R. Stabilization of acetabularfractures in elderly patients. J BoneJoint Surg Am. 1992 Jun;74(5):753-65.

Mears DC, Velyvis JH, Chang CP.Displaced acetabular fracturesmanaged operatively: indicators ofoutcome. Clin Orthop. 2003Feb;(407):173-86.

Kang CS, Min BW. Cable fixation indisplaced fractures of the acetabulum:21 patients followed for 2-8 years. ActaOrthop Scand. 2002 Dec;73(6):619-24.


H emiarthroplasty has beenshown to be effective inimproving the comfort and

function of shoulders in which thehumeral joint surface has been lost,specifically in patients with primaryglenohumeral arthritis, secondarydegenerative joint disease, avascularnecrosis of the humeral head, and thosewith combined loss of theglenohumeral joint surface and rotatorcuff. However, the published reportsindicate large variability in theeffectiveness of this procedure. Forexample, in two recent series ofhemiarthroplasties, the rate ofunsatisfactory results were 25%(Sanchez-Sotelo, 2001) and 47%(Sperling, 1998). The factorscontributing to the variability ineffectiveness of hemiarthroplasty havenot been well studied.

The authors suggest thatpreoperative decision-making bypatients and their surgeons could bebetter informed if the factors associatedwith the effectiveness of the procedurewere identified. The purpose of thisstudy was to test the hypothesis that theeffectiveness of humeralhemiarthroplasty is significantly relatedto factors that can be identified beforesurgery.

METHODS

Excluding procedures for themanagement of acute fractures andnon-unions, the senior authorperformed eighty-six primaryhemiarthroplasties between January 1,1990 and October 1, 1999. Each ofthese patients was invited to participatein follow-up study using follow-up self-assessment questionnaires. This reportconcerns the seventy-one shoulders insixty-eight patients who providedfollow-up data at least 24 months aftertheir surgery.

The preoperative characteristics ofeach shoulder were characterizedincluding: (1) patient age, (2) patientgender, (3) diagnosis (primarydegenerative joint disease, secondary

degenerative joint disease, avascularnecrosis, post-traumatic arthritis, cufftear arthropathy, rheumatoid arthritis,and capsulorrhaphy arthropathy), (4)the integrity of the rotator cuff, (5) thepresence of prior surgery, (6) astandardized shoulder functionquestionnaire, and (7) general healthstatus as assessed by the SF-36.

The preoperative radiographicappearance of the glenoid onstandardized anteroposterior andaxillary views was characterized basedon methods previously described, butmodified to use plain films rather thancomputerized tomographic scans. Theclassification was initially conductedindependently by three of the authorswho were asked to classify eachshoulder as uneroded, superiorlyeroded, posteriorly eroded, mediallyeroded, or anteriorly eroded accordingto the dominant direction of theerosion. All three of the authors agreedin 86% of the cases. One disagreed in11% and all three disagreed in 3% ofthe cases. The disagreement in these14% was resolved by discussion amongthe three reviewers so that eachshoulder could be included in thestatistical analysis.

RESULTS

Patients undergoinghemiarthroplasty had significantimprovements overall in shoulderfunction with the average number ofshoulder functions performableincreasing from 3.3 to 5.9 (p <.00000003) with an averageimprovement of 2.6 ± 3.6. Theimprovement in total number offunctions was not significantlycorrelated with age (R2 = 0.04).

Ten of the twelve individualshoulder functions also weresignificantly improved for the entirestudy population: (1) arm comfortableat side (p < .001), (2) sleep comfortably(p < .001), (3) tuck in shirt (p < .001),(4) hand behind head (p < .01), (5) onepound to shoulder height (p < .001),(6) eight pounds onto a shelf (p < .001),

(7) carry 20 pounds (p < 0.05), (8) tossa softball underhand (p < .01), (9)throw overhand (p < .05), and (10)wash back (p < .001). The ability toplace a coin on the shelf (p < .07) andto do usual work (p < .11) were notsignificantly improved.

The SF-36 comfort score for theentire study population improved from31 to 52 (p < .001). The remainingscores did not improve significantly.

The pre- and postoperativeshoulder function and SF-36 painscores were compared for each of theseven diagnoses (see Table 1). Thediagnosis had a significant effect on theeffectiveness of the procedure (ANOVAp < .01). Hemiarthroplasty was mosteffective in improving the function andcomfort in shoulders with avascularnecrosis and posttraumatic arthritis. Itwas effective with secondarydegenerative joint disease and primarydegenerative joint disease. It may havebeen effective in rheumatoid arthritis,but the improvement was notsignificant with the number of patientsin this study. It was least effective incuff tear arthropathy andcapsulorrhaphy arthropathy; theseeffects were not significant.

The effectiveness of

Preoperative Factors Associated with Greater ImprovementFollowing Humeral Hemiarthroplasty

CAROLYN M. HETTRICH, B.S., EDWARD J. WELDON III, M.D., RICHARD S. BOORMAN, M.D.,AND FREDERICK A. MATSEN III, M.D.

Picture 1: X-ray of humeral prosthesis.


hemiarthroplasty in improving thenumber of functions performable andthe SF-36 comfort score was alsocorrelated with patient gender, cuffintegrity, lack of radiographic evidenceof glenoid bone erosion, and previoussurgery. These results are illustrated inTable 2. Shoulders withradiographically uneroded glenoidbone and those without previoussurgery demonstrated the greatest gainsin shoulder function. Shoulders withsuperior erosion demonstrated the leastimprovement, but with the numbers inthis study, this result was notsignificantly different than other

patterns of glenoid erosion.Rotator cuff status did not have a

significant effect on the effectiveness ofshoulder arthroplasty in improvingoverall function in this study, but thepower of this conclusion was low (.21).However, the effect of cuff integrity wassignificant on the improvement ofpatients’ ability to place a coin on a shelfat shoulder level (p .009), to lift a poundto the level of the shoulder (p .02), andto lift eight pounds to the level of theshoulder (p .01).

DISCUSSION

The results of this study support the

hypothesis that identifiablepreoperative characteristics arestatistically associated with greatereffectiveness in improving shouldercomfort and function afterhemiarthroplasty. Previous surgery, aradiographically eroded glenoid, anddiagnoses such as rheumatoid arthritisand capsulorrhaphy arthropathy wereassociated with less effectiveness.Defects in the rotator cuff wereassociated with less improvement in theability to lift different weights toshoulder level.

Little has been published regardingthe relationship of preoperativecharacteristics to the effectiveness ofshoulder hemiarthroplasty. Kay et alfound better results in patients withavascular necrosis compared to thosewith sequela of trauma (1988). Whilethere was pain relief on both categories,there was a better postoperative rangeof motion for avascular necrosis.Levine discovered that 86% of patientswith concentric glenoids achievedsatisfactory results followinghemiarthroplasty, whereas only 63% ofpatients with non-concentric glenoidsachieved the same result (1997). Theseresults are consistent with ourobservation that a diagnosis ofavascular necrosis and a radiographicappearance of uneroded glenoid boneare associated with greater effectiveness.

CONCLUSION

This study confirmed that theeffectiveness of hemiarthroplasty isstatistically associated withcharacteristics of the shoulder that canbe determined before surgery. Theseassociations may be of use inpreoperative decision making bysurgeons and patients consideringshoulder hemiarthroplasty.

RECOMMENDED READING

Levine, W. N.; Djurasovic, M.; Glasson,J. M.; Pollock, R. G.; Flatow, E. L.; andBigliani, L. U.: Hemiarthroplasty forglenohumeral osteoarthritis: resultscorrelated to degree of glenoid wear. J.Shoulder Elbow Surg., 6(5):449-454,1997.

Sanchez-Sotelo, J.; Cofield, R. H.; andRowland, C. M.: Shoulderhemiarthroplasty for glenohumeralarthritis associated with severe rotatorcuff deficiency. J. Bone Joint Surg.,

Diagnosis #Shoulders

PreopSST

PostopSST

Significance(p)

PreopComfort

PostopComfort

Significance(p)

Post-TraumaticArthritis (PT)

7 1.7 6.8 .041 26.7 51.9 .26

Avascular Necrosis(AVN)

11 3.9 9.1 .00004 27.8 61.1 .0004

Cuff Tear Arthropathy(CTA)

23 3.1 4.6 .09 33 54.4 .027

Degenerative JointDisease (DJD)

8 3.3 5.4 .02 26.3 37.4 .005

SecondaryDegenerative JointDisease (SDJD)

12 3.9 6.2 .03 30.5 63.2 .007

CapsulorrhaphyArthropathy (CA)

3 3.7 4.7 .09 25.9 38.9 .55

Rheumatoid Arthritis(RA)

7 3 5 .23 42.2 53.3 .26

Table 1: Changes in function an comfort for different diagnoses.

Table 2: Differences between patients in shoulder function pre- and postoperatively looking at previoussurgery, cuff tear, and glenoid surface. (1) Preoperative radiographs were not available on 15 shoulders;(2) Data on cuff integrity was not available on 10 patients; (3) Data on previous surgery was notavailable on 12 patients.

Characteristic # Shoulders Preop SST Postop SST Significance ofChange (p)

Significance ofgroup (p)

Female Gender 40 2.6 5.1 .0002 NS

Male Gender 31 4.1 6.7 .0001

Cuff Intact1 34 3.2 6.4 .00002 .26

Cuff Tear 27 3.0 5.2 .0019

Uneroded GlenoidBone2

16 3.3 8.4 .0001 .001

Eroded GlenoidBone

40 3.0 4.7 .0017

Previous Surgery3 16 3.4 4.0 .3783 .015

No PreviousSurgery

43 2.9 6.6 .0000001


83A(12):1814-1822, 2001.

Sperling, J. W.; Cofield, R. H.; andRowland, C. M.: Neer hemiarthroplastyand Neer total shoulder arthroplasty inpatients fifty years old or less. Long-term results. J. Bone Joint Surg.,80A(4):464-473, 1998.

Matsen, F. A., III; Antoniou, J.;Rozencwaig, R.; Campbell, B.; andSmith, K. L.: Correlates with comfortand function after total shoulderarthroplasty for degenerative jointdisease. J. Shoulder Elbow Surg.,9(6):465-469, 2000.

Goldberg, B. A.; Smith, K.; Jackins, S.;Campbell, B.; and Matsen, F. A., III: Themagnitude and durability of functionalimprovement after total shoulderarthroplasty for degenerative jointdisease. J. Shoulder Elbow Surg.,10(5):464-469, 2001.


Lower limb injuries are secondonly to head trauma as theprimary consequence of motor

vehicle accidents and may be the mostfrequent cause of permanent disability.In a frontal crash, a commonmechanism of foot injury involves theoccupant’s knee pocketing against thedashboard due to his or her continuedforward motion as the vehicledecelerates. This effectively locks thetibia in position momentarily. Forceapplied to the foot from either thefloorboard deforming into the footwellspace, and/or the load against the footfrom the occupant’s own weight anddeceleration can result in significantfoot injuries. An example, from ourCIREN (Crash Injury Research andEngineering Network, NationalHighway Traffic Safety Administration)crash investigation data, is shown inFigure 1. In this example, the floorpan

remained intact, yet a calcaneusfracture was sustained by the driver.

Understanding the basicmechanisms and the magnitude offorces required to cause injury is a firststep to modifying vehicle structures toprevent these injuries. A driver with his/her foot on the brake during the crashmay experience forced dorsiflexion, ordorsiflexion combined with eitherinversion or eversion, along with thecompressive load applied to the footduring the collision. We found thatwhen impact forces are applied to thefoot in these different positions, theforce required to create injury with thefoot everted and dorsiflexed was 40%lower than with the foot dorsiflexed.From field studies we found a numberof cases with no footwell intrusion butsignificant foot injuries, as shown inFigure 1, implying that mechanismsother than trapping the foot against the

crushed floorpan may be causing footinjuries. These observations, along withepidemiological data, demonstrate thatthe brake pedal, specifically how italigns the foot during a frontal collisionwith the driver braking, is a source offoot injuries. New technologies, such as“drive by wire”, that is withoutmechanical linkages, will allowplacement of driving controls,including the brake pedal, in positionswhich could reduce the potential forfoot injury.

METHODS

Field studiesThe motor vehicle crash

information included in this study wascollected from our CIREN Center. Eachcrash site had scaled documentation ofthe roadway, traffic controls, roadsurface type, conditions, and road gradeat both pre- and post- impact locations.A scaled drawing with impact and finalresting positions was completed toassist in calculation of the speed andforce at impact. Exterior inspections ofthe vehicle were performed, a damagecrush profile was collected and enteredinto crash analysis software (WinSMASH, U.S. Dept of Transportation)to calculate the change in velocity(Delta V) of the vehicle during impactand the energy dissipated during thecrash event. An inspection of theinterior of the vehicle from which theinjured person had been removed wasperformed to determine points ofcontact and restraint system use. Thisinspection also included assessingintrusion into the passengercompartment especially the toe panand side panels forward of the A pillaror kick panel. With Institutional ReviewBoard approval, the injuries wereassessed by examining the patient’smedical records and imaging studies.

Biomechanical studiesThe IRB approved cadaveric

protocol utilized 15 pairs of fresh frozenfeet, with the tibia of each specimensectioned about 20 cm above the ankle.The soft tissues around the foot andankle remained intact. Each specimenwas potted and mounted, with about

The Effect of Foot Position on Mechanism of Fracture inAutomobile Frontal CollisionsM. ASSAL, M.D., P. HUBER, B.ENG., ALLAN F. TENCER, PH.D., E. ROHR, C. MOCK, M.D.,AND R. KAUFMAN, M.SC.

Figure 1: Example crash from CIREN data showing (upper) the frontal collison of the vehicle, (middle)lack of any significant intrusion into the footwell area, (lower) the resulting calcaneus fracture.


10cm of tibial shaft exposed, into animpact apparatus. This consisted of aframe with a high capacity uniaxial loadcell. Attached to the load cell was acylinder into which the pottedspecimen was placed such that impactforce was transmitted through the footto the cylinder and the load cell. TheAchilles tendon was inserted into anelectrical cable trap, connected to apneumatic cylinder, and loaded with atensile force of 1100N. High speedvideo was taken of each specimenduring loading. The impactor consistedof a 4 bar pendulum and an impactsurface which contacted the plantarsurface of the foot. The impact surfaceof the pendulum was an aluminumplate which could be angled either to20o of dorsiflexion, or 20o ofdorsiflexion combined with 30o ofeversion as the compressive load wasapplied. Impact force was generated byraising the 110kg impactor to a heightof 38 cm and releasing it. Post impact,specimens were radiographed anddissected to determine the injuryproduced. A paired t test was used toassess differences in peak load betweenthe two groups of paired specimens.

RESULTS

Field studiesIn a total of 17 frontal crashes

selected from the Harborview CIRENdata, an example of which is shown in

Figure 1, significant tibial shaft, ankleor foot injuries were reported. Themean principal direction of force(PDOF) was 0.6o or just about head-on, with a mean speed change (delta V)of 52 kph (32.1 mph) resulting in amean toe pan intrusion of 15.1 cm.Almost all victims were drivers (16 of17) and 10 were female. The majoritywere belted, with 4 restrained by bothairbags and lap/shoulder belts, 4 onlyhad lap/shoulder belts, one wasrestrained by an airbag and a shoulderbelt, and 8 had only a deployed airbagfor protection. The majority (13 of 17)injured their right lower extremity.Fibula fracture occurred in 8 cases,calcaneus fracture in 3 cases, malleolusfracture in 3 cases, and talus fracture in4 cases. Ten had additional lowerextremity or pelvic injuries.

Biomechanical testingBased on results from 9 of the 15

paired specimens, Figure 2, the meanload to failure for feet that were loadedin compression with dorsiflexion was6468N, (sd = 2435N). When feet wereloaded in dorsiflexion and eversion, themean load was 4107N, (sd = 1630N),which was significantly lower (p =0.001). There were notable differencesin the types of fractures sustained in thetwo groups. With the foot dorsiflexed,only one malleolar fracture was noted.In contrast, with eversion anddorsiflexion, 5 malleolar fractures

resulted. In the example of Figure 3, acalcaneus fracture resulted from a footloaded in dorsiflexion. In the sequence,the gap between the posterior facet ofthe talus and its articulation with thecalcaneus decreases, then the talar facetacts as a fulcrum with the compressiveload applied anterior to it and theAchilles load posterior. This creates 3point bending and a crack opens, notat the point of contact with the talarfacet, but on the opposite (plantar) side.The crack rapidly progresses and thecalcaneus fractures into twocomponents. With the foot evertedunder load, the load path is medial tothe longitudinal axis of the tibia causingthe foot to continue to evert andfracturing the malleolus as it dislocates.

CONCLUSION

The results of testing showed thatthere was a difference in the mean loadto failure of specimens forced intoeversion along with dorsiflexion whensubjected to axial compression,compared with feet positioned just indorsiflexion. There were differences inthe types of fractures created withconsiderably more malleolar fracturesin specimens which were everted as wellas dorsiflexed. The load paths appeareddifferent as well with the compressiveload being offset to the tibial shaft withthe foot everted.

These observations along with

Figure 2: Peak load to failure of the dorsiflexed and everted foot (dorsi+everted), the foot dorsiflexed only (dorsi only) and data from two other studies wherethe specimens were loaded only in dorsiflexion (MCW - Medical College of Wisconsin, WSU - Wayne State University).


epidemiological data, not presentedhere, showing that drivers who werebraking at the time of the crash weremuch more likely to sustain footfracture, indicates that the brake pedal,even in vehicles where the floor pandoes not intrude, can be a mechanismof injury. New designs, such as ayielding brake pedal, devised by oneauto manufacturer, may reduce theinjury potential to the foot in frontalcrashes.

ACKNOWLEDGEMENT

This study was supported by CIREN(Crash Injury Research andEngineering Network, NationalHighway Traffic SafetyAdministration).

RECOMMENDED READING

Bedewi PG, Digges KH, Investigatingthe ankle injury mechanisms in offsetfrontal collisions utilizing computermodeling and case study data,Proceedings of the 43rd STAPP CarCrash Conference, 99SC14, 1999.

Dischinger PC, Cushing BM, O’QuinnTD, Ho SM, Burgess AR, SchmidhauserCB, Juliano PJ, Bents FD, Lowerextremity trauma in vehicular front-seat occupants: patients admitted to aLevel 1 trauma center, SAETransactions, 940710, 1994.

Otte, D, von Rheinbaben H, Zwipp H,Biomechanics of injuries to the foot andankle joint of car drivers andimprovements for an optimal car floordevelopment, SAE Transactions,922514, 1992.

Pattimore D, Ward E, Thomas P,Bradford M, The nature and cause oflower limb injuries in car crashes, SAETransactions, 912901, 1991.

Pilkey WD, Sieveka EM, Crandall JR,Klopp G, The influence of footplacement and vehicular intrusion onoccupant lower limb injury in fullfrontal and frontal-offset crashes, In :Biomechanics of impact injury andinjury tolerances of the extremities, edby SA Backaitis, SAE PT-56, 1996,199-206.

Figure 3: (upper left) dorsiflexed specimen before impact, (upper right) initial contact, (lower left) with loss of joint space crack forms on plantar surface ofcalcaneus (arrow), (lower left) calcaneus fractured.

Schueler F, Mattern R, Zeidler F,Scheunert D, Injuries of the lower legs-Foot, ankle, joint, tibia; Mechanisms,Tolerance limits, injury criteriaevaluation of a recent biomechanicexperiment series, In : Biomechanics ofimpact injury and injury tolerances ofthe extremities, ed by SA Backaitis, SAEPT-56, 1996, 551-566.


Many people with diabetesexperience foot ulcerations,and subsequently lower-

limb amputations. Footwear has beenimplicated as a primary cause of footulcers, yet research is limited on theefficacy of shoe and insertcombinations to prevent reulceration.Our main objective was to determinewhether extra-depth and extra-widththerapeutic shoes used with twodifferent types of inserts reducereulceration in diabetic individualswith a history of a previous footulceration. A secondary objective wasto evaluate the pivotal events leading tolower limb amputation in thesepatients.

DESIGN, SETTING, AND PARTICIPANTS

This study was a randomizedclinical trial of 400 diabetes patientswith history of foot ulcer in twoWashington State health careorganizations. Study eligibility criteriawere diagnosed diabetes; ages 45-84;males from either the VeteransAdministration or Group Health

Cooperative (GHC) and women fromGHC (as there were very few femaleveterans who could meet the eligibilitycriteria); history of a full thickness footlesion, paronychia or a foot infectionrequiring antibiotic treatment; no footdeformities requiring a custom shoe;shoe size 8-12 for men and shoe size 7-10 for women (sizes were limited tocontain manufacturing costs); ability towalk one block and climb one flight ofstairs per day (to insure the footwearwould be worn); and willingness toconsent to randomization and studyfootwear provisions. Exclusion criteriawere a prior lower extremityamputation of more than one digit; alesion either unhealed or healed for lessthan 1 month; requirement of boots,custom shoes, or non-traditionalfootwear for daily activities; a historyof or active Charcot foot deformity;non-ambulatory status; or a terminalillness that would make two yearsurvival unlikely.

Subjects were enrolled betweenAugust 1997 and December 1998 andfollowed closely for 2 years. Data

collected at regular intervalsdocumented physical, foot, anddiabetes characteristics; footwear use;foot lesions; and ulcers.

INTERVENTIONS

Participants were randomlyassigned one of three study arms.Group 1 received 3 pairs of therapeuticshoes and 3 pairs of customizedmedium-density cork inserts with aneoprene closed-cell cover (n = 121).Group 2 received 3 pairs of therapeuticshoes and 3 pairs of prefabricated,tapered polyurethane inserts with abrushed nylon cover (n = 119). Group3 did not receive therapeutic shoes orinserts, and were instructed to weartheir usual footwear (controls; n = 160).

OUTCOME MEASURES

The main outcome measure wasfoot reulceration, compared among the3 groups. A secondary outcome waslower limb amputation.

RESULTS

Two-year cumulative reulcerationincidence across the 3 groups was lowin all three groups: 15% in the studyshoe plus custom customized medium-density cork inserts group, 14% in thestudy shoe plus prefabricated,polyurethane insert group, and 17% incontrols. These rates, for all threegroups, were much less than wepredicted using previously publisheddata that indicated reulceration rates of30 to 50% in similar high risk diabeticindividuals.

In the intent-to-treat analysis, wewere unable to show a significantlylower risk of reulceration in patientsassigned to therapeutic shoes comparedwith controls (risk ratio [RR] for thecork-insert group, 0.88; 95%confidence interval [CI], 0.51-1.52 andRR the for prefabricated-insert group,0.85; 95% CI, 0.48-1.48). All ulcerepisodes in patients assigned totherapeutic shoes and 88% wearing

Effect of Therapeutic Footwear on Foot Reulceration in Patientswith Diabetes: A Randomized Controlled TrialDOUGLAS G. SMITH, M.D., GAYLE E. REIBER, M.P.H., PH.D., CAROLYN WALLACE, PH.D.,KATRINA SULLIVAN, D.P.M., SHANE HAYES, C.PED., CHRISTY VATH, B.S.,MATTHEW L. MACIEJEWSKI, PH.D., ONCHEE YU, M.S., PATRICK J. HEAGERTY, PH.D.,AND JOSEPH LEMASTER, M.D.

Table 1: Pivotal Events and Foot Insensitivity for Ulcer Episodes in the Clinical Trial of Footwear, fromJAMA - 2002;287:2552-58.

Pivotal Events for 84

Ulcer Episodes

Number By Group

Shoe related Ulcers Group 1

N = 121

Group 2

N = 119

Group 3

N = 160

Percent

of Total

Total % Insensate

Study shoe 7 2 0 10.7 9 100

Non study shoe 5 0 19 28.6 24 88

Non Shoe related Ulcers

External trauma 8 10 8 30.9 26 96

Self care 2 4 0 7.1 6 83

Critical ischemia 0 3 1 4.8 4 75

Paronychia 3 1 1 6.0 5 40

Decubitus 0 1 5 7.1 6 67

Other 0 1 3 4.8 4 75

Total Episodes 25 22 37 100 84 86


nonstudy shoes did occur in patientswith foot insensitivity.

Table 1 shows the two pivotal eventsthat contributed most to ulcer episodeswere shoes 39% (10.7% due to studyshoes and 28.6% due to nonstudyshoes), and external trauma 31%.Insensate patients comprised themajority of patients reporting ulcers:100% insensate among ulcers reportedto have occurred in study shoes; 88%in non-study shoes; and 96% attributedto external trauma.

Amputation was not the primarystudy endpoint as few were expected.During the two-year follow-up, elevenparticipants underwent lower limbamputation. Incidence of amputationwas 13.8 per 1000 persons-year.Analysis of the pivotal events is detailedin Table 2. This revealed: sixamputations directly related to minorexternal trauma, two due to progressionof vascular disease with dry gangrene,one due to decubitus ulcer, one due to

self-care injury, and only one due to afootwear related ulcer. The amputationlevels were: four transfemoral, twotranstibial, one hindfoot, one fifth ray,and three at the toe level.

CONCLUSION

This study of persons without severefoot deformity does not provideevidence to support widespreaddispensing of therapeutic shoes andinserts to diabetic patients with ahistory of foot ulcer. Study shoes andcustom cork or preformedpolyurethane inserts conferred nosignificant ulcer reduction comparedwith control footwear. This study doessuggest that careful attention to footcare by health care professionals maybe more important than therapeuticfootwear. The study does not negatethe possibility that special footwear isbeneficial in persons with diabetes whodo not receive such close attention tofoot care by their health care providers

or in individuals with severe footdeformities.

This prospective study also showedthat amputations in high risk patientswith diabetes were mostly related tohousehold accidents, and onlyinfrequently related to footwear.Progression of severe vascular diseasewas a major factor that lead to higheranatomic level of amputation.Strategies to reduce external traumaunrelated to shoes appear indicated.

The highest risk patients, those whoare insensate, have proven once againthey are the most vulnerable tocomplications. Limited health careresources might be better utilized todevelop a comprehensive approach tocare for these individuals. Thesepatients and their providers shouldjointly explore individualized strategiesto decrease the events that give rise tofoot ulcers and to amputations.

Table 2: Description of Pivotal Events Leading to Lower Limb Amputation, from: Smith DG, et al, Accepted by Foot and Ankle International. M: male; F:female; PAD: peripheral arterial disease; DM: diabetes mellitus; CHF: chronic heart failure; IHD: ischemic heart disease; HTN: hypertension; COPD: chronicobstructive pulmonary disease; CLF: chronic liver failure; AAA: abdominal aortic aneurysm; TIA: transient ischemic attack; CRF: chronic renal failure; PY:pack-year; TcPO

2: transcutaneous oxygen pressure.

Pat ien ts AverageA g e

S e x P i v o t a lEven t

Description ofPivotal Event

Co-Morb id i ty Amputat ionLevel

1 . 51 M Shoe related Fissure and infection onlarge callus located lat.

to 5th metatarsal head.

DM, CHF, IHD R 5th ray

2 . 74 M Non shoerelated, minorenvironmentaltrauma

Patient's foot ran intobed at night. Cut onlateral side of heel.

DM, CHF, IHD, LungCA, Prior LL bypass

R Transfemoral


Caregiver at nursinghome stepped onpatient's foot. Wound

on dorsum 3rd toe.

DM, CHF, HTN,COPD, CLF

R Transfemoral


Patient bumped hislateral foot against wall.

DM, IHD, Quintupleheart bypass, BilateralLL bypass, AAA repair

R Transtibial


Minor trauma, patientbumped his toe whilevacuuming barefoot.

DM L Transtibial


Patient's foot bumpedinto cart leading to

traumatic wound 2nd

toe.

DM, CRF requiringdialysis R distal 2nd toe

7 . 66 F Non shoerelated, minorenvironmentaltrauma

Bumped 5th toe againstfurniture.

DM L 5th toe

8 . 69 M Criticalischemia

No history of trauma."Spontaneous"appearance of necroticarea on dorsum of foot.

DM, CHF, HTN, COPDrequiring O2, Prostatemetastatic CA

R Transfemoral

9 . 68 M Criticalischemia

No history of trauma."Spontaneous"appearance of necroticarea on great toe.

DM, HTN L Hindfoot

10. 71 M Self-care trauma Accidentally cut lateralborder of toe nail withhis scissors.

DM, bilateral carotidsendarterectomy, TIA,HTN, CHF, IHD

R hallux

11. 50 M Decubitus Decubitus ulcer lateralmalleolus afterprolonged bedrest.

DM, Repaired rupturedAAA, COPD, chronicpancreatitis, Prior LLbypass

L Transfemoral


RECOMMENDED READING

Reiber GE, Smith DG, Wallace C,Sullivan, K, Hayes S, Vath C,Maciejewski M, Yu O, Heagerty PJ:Effect of Theraputic Footwear on FootRe-Ulceration in patients withDiabetes: A Randomized Clinical Trial.JAMA 287(19): 2552-2558, 2002.

Smith DG, Assa M, Reiber GE, Vath C,LeMaster J, Wallace C: MinorEnvironmental Trauma and LowerExtremity Amputation in High-RiskPatients with Diabetes: Incidence,Pivotal Events, Etiology andAmputation Level in a ProspectivelyFollowed Cohort. Accepted Foot andAnkle International, March 2003.

Reiber GE, Smith DG, Boone DA, et al.Design and pilot testing of the DVA/Seattle footwear system for diabeticpatients with foot insensitivity. JRehabil Res Dev. 1997; 34: 1-8.

Wallace C, Reiber GE, LeMaster J, SmithDG, Sullivan K, Hayes S, Vath C: Falls:Incidence, Risk Factors, and Fall RelatedFractures in Persons with Diabetes anda Prior Foot Ulcer. Diabetes Care, Vol25, No 11, November 2002, 1983-86.

Reiber GE, Smith DG, Wallace C,Sullivan K, Hayes S, Vath C, Yu O,Martin D, Maciejewski M. Footwearused by individuals with diabetes anda history of foot ulcers. J ofRehabilitation, Research &Development 2002, 39(5): 615-621.

LeMaster J, Reiber GE, Smith DG,Heagerty P. Foot ulcers among peoplewith diabetes: Daily ambulatory activitydoes not increase the risk. Accepted -Medicine And Science In Sports AndExercise. January 2003.

ACKNOWLEDGEMENTS

Funding/Support: This research wassupported by VA RehabilitationResearch and Development, HealthServices Research and Development,the Epidemiology.

Figure 1:Superficial diabetic ulceration into the dermis located under the right 2nd metatarsal head.

Figure 2: Deep diabetic ulceration through the dermis into the subcutancous tissue located under the4th metatarsal head.


C ancers arise when geneticchanges, such as chromosomaltranslocations and point

mutations, enable cells to escape thecontrol mechanisms that usually serveto limit cell division to periods ofgrowth or normal replenishment ofcells. A better understanding of thespecific pathways that lead from agenetic change to a cancerous cell mayyield targets for molecular treatments.In addition, by studying the aberrantprocesses of tumor cells one can oftengain a window of insight into normalcellular processes.

Ewing’s sarcoma is a highlyaggressive bone and soft tissue tumorthat occurs in adolescents and is ofunknown tissue origin. Most Ewing’ssarcomas exhibit a specific t(11;22)chromosomal translocation that resultsin the fusion of EWS to the ETS-familymember Fli-1. In a subset of patients, aspecific t(21;22) chromosomaltranslocation involving the Ewing’ssarcoma gene (EWS) and the ETS-related gene (ERG) generates instead anEWS/ERG fusion gene. In the resultanthybrid EWS/ERG fusion protein, aportion of the ETS family protein ERGis replaced by a piece of the RNA-binding protein EWS.

CADO-ES1 is an established humanEwing’s sarcoma cell line that candifferentiate into neural andmesenchymal cell lineages. Because

expression of mouse Col11a2 has beenlinked to neural and mesenchymal cellphenotypes, we analyzed this cell linefor expression of the COL11A2 collagengene. We confirmed that CADO-ES1cells express both COL11A2 mRNAand EWS/ERG fusion protein. Since theERG transcription factor is known toparticipate in cartilage differentiationfrom mesenchyme (chondrogenesis), amarker of which is the activation ofCOL11A2, we investigated whetherERG and/or EWS/ERG participate inregulating COL11A2 expression.

RESULTS

We characterized the COL11A2promoter and tested the ability of bothwild-type (normal) ERG and the EWS/ERG sarcoma fusion protein tostimulate the COL11A2 promoter usinga luciferase assay. Various proteins(transcription factors) interact with thepromoter region of genes to regulategene expression. Luciferase is theprotein that enables fireflies to glow.The gene for luciferase can beintroduced into living cells under thecontrol of the COL11A2 promoter andluciferase expression can then bequantified using a luminometer, thusgiving an indication of how efficientEWS and EWS/ERG are at promotingsynthesis of COL11A2 protein. Wefound that expression of EWS/ERG, butnot wild-type ERG, activated the

COL11A2 promoter (Figure 1) and thisactivation required not only the N-terminal region of EWS but also anintact DNA-binding domain fromERG.

Electrophoretic mobility shift assay(EMSA) is used to detect whether aspecific protein binds to the regulatory(promotor) sequence that controlsexpression of a specific gene. In essence,the bound protein slows down themigration of the regulatory genesequence through a gel. The complexis then detected by a specific anti-protein antibody and the distancetraveled through the gel can bevisualized. Using the COL11A2promoter sequence, EMSA showedinvolvement of EWS/ERG and ERG inthe formation of DNA-proteincomplexes.

Based upon our previous studies ofsarcoma fusion proteins, we wereinterested in whether both EWS/ERGand wild-type ERG would bind to RNApolymerase II (Pol II). This was testedusing western blotting andimmunoprecipitation with antibodiesspecifically targeting RNA Pol II andEWS/ERG or ERG. Interestingly, EWS/ERG but not wild-type ERG wouldbind RNA Pol II.

Finally, we have previouslypublished our discovery of an ERG-associated histone methyltransferasewith a SET domain that we named

COL11A2 Collagen Gene Transcription is DifferentiallyRegulated by EWS/ERG Sarcoma Fusion Protein andWild-Type ERGYOSHITO MATSUI, M.D., PH.D., HOWARD A. CHANSKY, M.D., FARIBA BARAHMAND-POUR, PH.D.,ANNA ZIELINSKA-KWIATKOWSKA, M.D., NORIYUKI TSUMAKI, AKIRA MYOUI, HIDEKI YOSHIKAWA,LIU YANG, PH.D., AND DAVID R. EYRE, PH.D.

Figure 1: Differential effects of ERG and EWS/ERG on the COL11A2 promoter. a) Schematic of EWS, ERG, and EWS/ERG. In the EWS/ERG fusion protein thetransactivation domain of ERG is replaced by the N-terminal domain of EWS. b) Effect of ERG and EWS/ERG expression constructs on the activity of wild-typeCOL11A2 promoter in NIH3T3 cells. EWS/ERG but not ERG activates the COL11A2 promoter thus resulting in a greater induction of luciferase expression asmeasured by the luminometer (horizontal axis).


ESET. Histones are proteins that controlthe physical structure and thusultimately the expression of genes inchromosomes. Histonemethyltransferases in turn bind otherproteins such as histone deacetylasesthat together determine theconformation of histones and thusregulate the physical accessibility oftranscription factors to gene promoters.Histone deacetylases in fact inhibit genetranscription. We hypothesized that ahistone methyltransferase (such asESET) bound to both ERG and histonedeacetylase may in part account for theinability of ERG to activate theCOL11A2 promoter. To evaluate thispossibility we treated CADO-ES1 cellswith the histone deacetylase inhibitortrichostatin A. This in fact enabled ERGto activate the COL11A2 promotertherefore abolishing the differentialeffects of EWS/ERG and ERG (Figure2).

DISCUSSION

Our findings indicate that ERG andthe sarcoma fusion protein EWS/ERGdifferentially regulate expression of theCOL11A2 gene. Even though bothwild-type ERG and the EWS/ERGfusion protein share the same DNA-binding domain and are expected tobind to the same promoter sequence,the human COL11A2 promoterconstruct is transactivated by EWS/ERG, but not by ERG. Our resultstogether with a recent report oninduction of the Id2 gene by EWS-ETS,reveal that an EWS fusion protein isable to activate an authentic promoterthat is not activated by its wild-type

counterpart. Furthermore,electrophoretic mobility shift assayconfirms the interaction of EWS/ERGand/or ERG with the COL11A2promoter. Thus, activation ofCOL11A2 is at least partially mediatedby the interaction between theCOL11A2 promoter and the DNAbinding domain of EWS/ERG.

In this study we also investigatedwhy EWS/ERG, but not ERG, activatesCOL11A2. Our study points to at leasttwo potential mechanisms. First, sinceinhibition of histone deacetylasesignificantly enhanced the effects ofERG on COL11A2 promoter activity,transactivation by ERG appears to beepigenetically (a change in the patternof gene expression without a change inthe DNA sequence) suppressed.Second, as evidenced byimmunoprecipitation experiments,EWS/ERG fusion protein recruits RNAPol II in vivo whereas wild-type ERGdoes not. By fusing the N-terminus ofEWS to the DNA-binding domain ofERG, the EWS/ERG oncogenic proteinis able to recruit RNA Pol II directly tothe site of transcription and this couldserve to bypass suppression by histonedeacetylase complexes. These findingstherefore provide new insights into themechanisms underlying collagen geneexpression in general, and COL11A2expression in particular.

In addition to transcriptionalderegulation, EWS sarcoma fusionproteins are reported to interfere withbasic cellular processes such as RNAsplicing. Unlike wild-type EWSprotein, the C-terminal domains ofEWS fusion products are unable to

recruit certain factors that regulatespilicing of RNA, which may partlyexplain the abnormal splicing patternsand abnormal protein expression inEwing’s sarcoma cells. The finding thatEWS/ERG binds RNA Pol II alsosuggests a coupled mechanism forstimulated gene expression and alteredRNA splicing. Since gene transcriptionand RNA splicing are physically linkedvia RNA Pol II and its associatedproteins, EWS/ERG has the potential toderegulate gene activation, epigeneticgene repression, and RNA-splicing. Allof these processes are integral to cellgrowth and differentiation.

While the cell of origin (benigncounterpart) of Ewing’s sarcomaremains unknown, the expression ofCOL11A2 in CADO-ES1 cells supportsa neural/mesenchymal origin.Interestingly, the regulatory elements ofthe COL11A2 active in CADO-ES1 cellsdiffer from those active in cells ofcartilage and neural tissues. It is likelythat other genes with promoterproperties in common with COL11A2are also differentially regulated by EWS/ERG and ERG, and the implications webelieve are therefore fundamental to anunderstanding of Ewing’s tumorpathobiology.

ACKNOWLEDGEMENTS

The authors thank Ken Kodama forthe CADO-ES1 cells. This work issupported by a VA Merit Review Award(to H.A.C.), the Orthopaedic Researchand Education Foundation (to H.A.C.),and by NIH grants 1R01CA90941 (toL.Y.), 5R01AR37318 and 5R37AR36794(to D.R.E.).

RECOMMENDED READING

Ginsberg, J. P., de Alava, E., Ladanyi, M.et al. (1999) J. Clin. Oncol. 17, 1809-1814.

Delattre, O., Zucman, J., Plougastel, B.et al. (1992) Nature 359, 162-165.

Sorensen, P. H. B., Lessnick, S. L., Lopez-Terrada, D. et al. (1994) Nat. Genet. 6,146-151.

Yang, L., Chansky, H.A., andHickstein,D.D. (2000) J. Biol. Chem.273, 37612-37618.

Yang, L., Xia, L., Wu, D. Y. et al. (2002)Oncogene 21, 148-152.

Figure 2: Derepression of ERG after treatment with Trichostatin A (TSA). NIH3T3 cells were co-transfectedwith the wild-type COL11A2 promoter and ERG or EWS/ERG expression plasmid. Transfected cellswere assayed for luciferase 18 hours after treatment with TSA and compared with untreated cells.Luciferase activities relative to the untreated cells are shown as fold of derepression (stimulation). ERGis able to activate the COL11A2 promoter after treatment with the histone deacetylase inhibitor TSA.


Cartilage contains three tissue-specific collagens, types II, IXand XI. In the extracellular

matrix all three co-polymerize to forma fibrillar framework that is stabilizedby inter-molecular covalent cross-links.Trivalent hydroxypyridinoline (HP)cross-links between the amino- (N) andcarboxy- (C) telopeptides of type IIcollagen molecules and the helicalregion of adjacent type II collagenmolecules are the most prevalent.Pyridinoline and divalent cross-linksalso bond type IX collagen moleculesto the N- and C- telopeptides of type IIcollagen. Type XI collagen is linked tothe N- and C- telopeptides of type IIcollagen by divalent ketoamine cross-links. This co-polymeric fibrillarnetwork is the basis of the tensilestrength of cartilage.

In vivo, chondrocytes synthesize anddeposit collagen molecules in thecartilage matrix. In vitro, chondrocytesare difficult to culture as theydedifferentiate and synthesize matrixmolecules not characteristic of cartilage(e.g. type I collagen). Over the pastdecade, a better understanding ofchondrocyte biology has led toimproved cell culture techniques and anew field of research, cartilage tissueengineering. Under suitable cultureconditions chondrocytes can beinduced to synthesize and form amatrix based on type II collagen. Towhat degree the fibrillar matrix isnormal, in terms of the coassembly ofthe minor collagens (types IX and XI)is not well characterized.

This study investigated the ability ofchondrocytes, cultured under differentconditions, to assemble collagen typesII, IX and XI into the co-polymericcross-linked network that typifiescartilage matrix in vivo.

METHODS

AntibodiesThe monoclonal antibody 1C10

recognizes an epitope near the C-terminus of the helix (CB9,7) of typeII collagen chains. The monoclonalantibody 10F2 recognizes a cleavage-

site (neoepitope) in a sequence in theC-telopeptide cross-linking domain oftype II collagen. This antibody detectsthe C-telopeptides of type II collagen(even as short fragments) if cross-linked to collagen chains.

Cell CultureChondrocytes were cultured under

conditions that would favor thedifferentiated phenotype. The Swarmrat chondrosarcoma chondrocyte cellline RCS-LTC, was maintained as a highdensity monolayer culture in DMEMwith BCS and ascorbate for a period of2 weeks. Human fetal epiphysealchondrocytes were maintained as highdensity pellet cultures in the presenceof FBS and ascorbate for 3 weeks.Chondrocytes from adult human taluscartilage were maintained in highdensity micromass cultures in DMEMcontaining ascorbate, with or without100ng/ml human recombinant FGF-18for a period of 5 weeks. Alginaterecovered chondrocyte (ARC)engineered tissue, using human adultarticular chondrocytes (2 and 4 weekspost alginate culture), was a gift fromKoichi Masuda, M.D. (Department ofOrthopaedics, Rush Presbyterian St.Luke’s Medical Center, Chicago).

Collagen extractionThe collagen network laid down by

the cultured cells was depolymerized

using pepsin. Pepsin cleaves in thetelopeptide domains of type II collagenand in the non-collagenous domains ofthe minor collagens, type XI and IX, butleaves the triple helical domains intact.The short stubs of cleaved telopeptidesremain cross-linked to the triple-helicalsites to which they were attached in thematrix and the antibody 10F2 detectsthe α1(II) C-telopeptide wherever it iscross-linked to collagen chains or chainfragments (Figure 1).

RESULTS

The Swarm rat chondrosarcoma cellline RCS-LTC synthesizes and depositscollagen types II, IX and XI in theextracellular matrix. As seen in Figure2, lane 1, the antibody 10F2 reacted withthe α1(II) collagen chain as expectedfor a cross-linked type II collagenpolymer. (C-telopeptide of type IIcollagen cross-linked to α1(II) collagenchains). The antibody also reacted withα1(XI) collagen chains implying thischain was cross-linked to the C-telopeptide of type II collagen. A thirdimmunoreactive band is seen only afterreduction with DTT. This band fromits properties is the α3(IX) collagenchain cross-linked to a C-telopeptide oftype II collagen. As shown in Figure 1these cross-linking sites have all beenreported to occur in normal articular

Cross-Linking of Collagens II, IX and XI in ChondrocyteCultures As a Monitor of Normal Matrix AssemblyRUSSELL J. FERNANDES, PH.D. AND DAVID R. EYRE, PH.D.

Figure 1: C-telopeptides of type II collagen are cross-linked to chains of type XI, II and IX in normalcartilage.


cartilage.The fetal human chondrocytes

elaborated an extensive extracellularmatrix containing type II collagen. Theantibody 1C10 detected the α1(II) andthe β1(II) chains of type II collagen(Figure 3). The antibody 10F2 reactedwith the α1(II) chain indicating a cross-linked collagen network within a weekin culture. 10F2 also reacted with the

α1(XI) chain after 2 weeks in culture,showing that type XI collagen wascopolymerized and cross-linked to C-telopeptides of type II collagen.

FGF-18 is a trophic factor formature chondrocytes. Cultured inmicromass, the human adultchondrocytes elaborated a matrix lesscopious than that of fetal chondrocytes.The antibody 1C10 was able to detect

type II collagen in the matrix of bothFGF-18 treated and untreatedchondrocytes indicating adifferentiated phenotype (Figure 4).The antibody 10F2 showed a slightlystronger reaction with the α1(II) chainsin FGF-18 treated cultures than incontrol. The α1(XI) collagen chainswere faintly immunoreactive. Theresults indicate that a copolymericcollagen type II/XI network hadformed.

When cultured by the alginaterecovered chondrocyte method, theextracellular matrix of the engineeredtissue contained type II collagen asdetected by antibody 1C10 (Figure 5).The antibody 10F2 reacted stronglywith the α1(II) collagen chains showingthat the type II collagen waspolymerized and cross-linked. Theα1(XI) collagen chains were alsoimmunoreactive showing the presenceof a type XI-type II collagen co-polymer.

Figure 2: The collagen heteropolymer formed in RCS-LTC chondrocyte monolayers.

Figure 3: The collagen network formed in human fetal chondrocyte pellet cultures.

Figure 4: Collagen from human adult chondrocyte micromass culture.


CONCLUSION

Under appropriate cultureconditions we can demonstrate thathuman and rat chondrocytes remaindifferentiated and form a fibrillarmatrix of the collagen II, IX, XIheteropolymer characteristic of hyalinecartilage.

Since it is clear that each of theminor collagens is essential inregulating the organization of thenetwork (e.g. fibril diametersmodulated by type XI), fingerprintingthe pattern of inter-type cross-linkingprovides a screen for normal matrixassembly. This could be useful inassessing the quality of cartilage indisease and when produced as a healingresponse or by tissue engineeringmethods.

RECOMMENDED READING

Wu, J-J and Eyre, D.R. (1984),Identification of hydroxypyridiniumcross-linking sites in type II collagen ofbovine articular cartilage. Biochemistry23:1850-1857.

Eyre, D.R. et al., (1987), Collagen typeIX: evidence for covalent linkages totype II collagen in cartilage. FEBSLetters 220:337-341.

Wu, J-J and Eyre, D.R. (1995),Structural analysis of cross-linkingdomains in cartilage type XI collagen.J. Biol. Chem. 270:18865-18870.

Fernandes, R.J. et al., (2001),Procollagen II amino propeptideprocessing by ADAMTS-3. Insights onDermatosparaxis. J. Biol. Chem.276:31502-31509.

Ellsworth, J.L. et al., (2002), Fibroblastgrowth factor-18 is a trophic factor formature chondrocytes and theirprogenitors. Osteoarthritis andCartilage 10:308-320.

Masuda, K. et al., (2002), Human tissueengineered cartilage by the alginate-recovered-chondrocyte method afteran expansion in monolayer.Transactions of the 48th OrthopaedicResearch Society 27:467.

Figure 5: Demonstration of the collagen heteropolymer in tissue engineered from human adult chondrocytes.


Department of Orthopaedics and Sports Medicine Faculty

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

Christopher H. Allan, M.D.Assistant Professor

Dheera Ananthakrishnan, M.D.Assistant Professor

David P. Barei, M.D.Assistant Professor

Carlo Bellabarba, M.D.Assistant Professor

Stephen K. Benirschke, M.D.Professor

Richard J. Bransford, M.D.Assistant Professor

James D. Bruckner, M.D.Associate Professor

Howard A. Chansky, M.D.Associate Professor

Jens R. Chapman, M.D.Professor

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

Ernest U. Conrad III, M.D.Professor

David R. Eyre, Ph.D.Professor

John R. Green, M.D.Associate Professor

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

Ted S. Gross, Ph.D.Associate Professor

Douglas P. Hanel, M.D.Professor

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

M. Bradford Henley, M.D.Associate Professor

Todd Jarosz, M.D.Assistant Professor

Nancy J. Kadel, M.D.Assistant Professor

Roger V. Larson, M.D.Associate Professor

Seth S. Leopold, 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.Professor

Bruce J. Sangeorzan, M.D.Professor

Gregory A. Schmale, M.D.Assistant Professor

John A. Sidles, Ph.D.Professor

Douglas G. Smith, M.D.Associate Professor

Kevin L. Smith, M.D.Associate Professor

Kit M. Song, M.D.Associate Professor

Sundar Srinivasan, Ph.D.Research Assistant Professor

Lynn T. Staheli, M.D.Professor Emeritus

Lisa A. Taitsman, M.D., M.P.H.Assistant Professor

Carol C. Teitz, M.D.Professor

Allan F. Tencer, Ph.D.Professor

Thomas E. Trumble, M.D.Professor

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

Liu Yang, Ph.D.Research Assistant Professor

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

Associate Professor, PhysiologicalNursing

Jack W. Berryman, Ph.D.Professor, Medical History & Ethics

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

Randal P. Ching, Ph.D.Associate Professor, MechanicalEngineering

Eva M. Escobedo, M.D.Associate Professor, Radiology

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

Thurman Gillespy, M.D.Associate Professor, Radiology

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

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

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

David W. Newell, M.D.Professor, Neurosurgery

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

Wendy Raskind, M.D., Ph.D.Professor, General Internal Medicine

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

Christopher I. Shaffrey, M.D.Associate Professor, Neurosurgery

Robert B. Schoene, M.D.Professor, Medicine

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

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

Joint FacultyMark A. Harrast, M.D.

Assistant Professor, Rehab Medicine

John E. Olerud, M.D.Professor, Division of Dermatology

Nathan J. Smith, M.D.Professor Emeritus, Pediatrics

Michael D. Strong, Ph.D.Research Professor, Surgery

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

Clinical FacultySarah E. Jackins, R.P.T.

Assistant Professor, Rehab Medicine


1959 N.E. Pacific StreetBox 356500Seattle, Washington 98195-6500Phone: (206) 543-3690Fax: (206) 685-3139

Affiliated Institutions

University of Washington

University of Washington School of Medicine

Children’s Hospital andRegional Medical Center4800 Sand Point Way NESeattle, WA 98105(206) 987-1776

Harborview Medical Center325 Ninth AvenueSeattle, WA 98104(206) 731-3466

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

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

Department of Orthopaedics and Sports Medicine


Graduating ResidentsClass of 2003

Ben DuBois, M.D.: Ben will return tothe University of WashingtonDepartment of Orthopaedics as ashoulder and elbow fellow in ourAdvanced Clinical Experienceprogram.

Andrew Howlett, M.D.: Andrew willreturn next year as our Trauma Fellowat Harborview Medical Center. Thefollowing year he will join our SarcomaService as the fellow under Dr. ErnestConrad.

Emma Woodhouse, M.D.: Emma willbe an ACE next year in our AdvancedClinical Experience program with ourshoulder and elbow service.

Guy Schmidt, M.D.: Following hisresidency he will be doing a sportsmedicine fellowship at The Rosenberg/Cooley Clinic in Utah. Afterwards, hehopes to return to Montana.

Brian Shafer, M.D.: Next year Brian willbe doing a sports medicine fellowshipwith James Tibone at the University ofSouthern California Center for AthleticMedicine, in Los Angeles, afterwards hemight return to Arizona where he isfrom.


Incoming Residents

Jamie Antoine: Jamie attended UnionCollege in Schenectady, New Yorkwhere he received a B.S. in Biology. Hereceived his medical degree fromAlbany Medical College. He is currentlycompleting his surgical internship atthe University of Washington. Hisinterests outside of medicine include icehockey, weight training and basketball.

Jeremiah Clinton: Jeremiah attendedMontana State University in Bozeman,Montana where he received a B.S. inBiomedical Science. He earned his M.D.from the University of Washington. Heis finishing his surgical internship at theUniversity of Washington. His interestsinclude alpine skiing, rock climbing andwoodworking.

Mary R.A. Cunningham: Maryattended George Mason University inFairfax, Virginia where she received aB.S. in Biology. She received her medicaldegree from the University of Virginia.She is completing her surgicalinternship at the University ofWashington. Personal interests includetraveling, backpacking, hiking, andgardening.

Joseph Lynch: Joe received a B.A. inEnvironmental Science and PublicPolicy from Harvard. In addition, hecompleted four years of course work atMIT studying Military Science asrequired by the Navy ROTC program.He received his medical degree fromOregon Health Sciences University.Interests outside of medicine includeskiing and mountain biking.

Allison MacLennan: Allison attendedthe University of Michigan where shereceived a B.S. in Biology. She alsoreceived her medical degree from theUniversity of Michigan. She is currentlycompleting her surgical internship atthe University of Washington. Personalinterests include music, both singingand playing the viola, dance, films, andjogging.

Evan Ellis: Evan received his B.A. inPsychology from Johns HopkinsUniversity. He earned his medicaldegree from the University of Michigan.He is currently completing hisinternship at the University ofWashington. Outside of medicine heenjoys basketball, golf, volleyball, andjogging.


2003 Department of OrthopaedicsNew Faculty

Dheera Ananthakrishnan, M.D. Richard J. Bransford, M.D. Lisa A. Taitsman, M.D., M.P.H.

Dr. Ananthakrishnan completedher undergraduate work inMechanical Engineering at

Massachusetts Institute of Technologyin 1990, followed by a masters degreein Bioengineering from the Universityof Washington. She then attendedmedical school at the Albany MedicalCollege in New York, earning her degreein 1996. Her internship and residencywere completed at the University ofIllinois in Chicago in 2001. She has justfinished a two-year spinal deformityfellowship under Dr. David Bradford atthe University of California SanFrancisco. She joins our faculty as aspine surgeon based at the Universityof Washington Medical Center.

A lthough born in Omaha,Nebraska, Richard is fluent inKiswahili and French after

being raised in Kijabe, Kenya. Whileliving in Kenya, he attended a boardingschool for five hundred missionarychildren where Richard becameinvolved in sports, especially basketball,volleyball and soccer. During this timehe spent summers observing his fatherin the operating room. Richard leftAfrica to attend college in the UnitedStates where he received hisundergraduate B.S. degree in Biologyfrom Westmont College in 1992. Hegraduated summa cum laude as well asValedictorian of his graduating collegeclass. In 1995 he was awarded the MAP-Reader's Digest Internationalfellowship which was a sponsorship tocontinue his work at a mission hospitalin Indonesia. He proceeded to earn hisM.D. degree at Vanderbilt UniversitySchool of Medicine in 1996. Hecompleted his residency at theUniversity of Washington, where,afterwards, he was a fellow in the spineservice’s Advanced Clinical Experienceprogram. He has just completed apediatric fellowship in Australia andnow joins our faculty as a spine surgeonpracticing at the University ofWashington and the local VA hospital.

Dr. Taitsman received both herundergraduate and medicaldegrees from Brown

University. She then earned her Mastersin Public Health from Harvard Schoolof Public Health. She served as asurgery intern at Beth Israel MedicalCenter and then completed herresidency in Boston in the HarvardCombined Orthopaedic ResidencyProgram. After acting as chief residentat Massachusetts General Hospital, shecompleted a trauma fellowship atHarborview Medical Center/ Universityof Washington. Dr. Taitsman hasremained at Harborview where shejoins the orthopaedic trauma faculty.


RESEARCH GRANTS

DEPARTMENT OF ORTHOPAEDICS AND SPORTS MEDICINE

National Institutes of Health (NIH)

Augmentation of Peak Bone MassTed S. Gross, Ph.D.

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

Disuse Induced Osteocyte HypoxiaTed S. Gross, Ph.D.

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

Outcomes Following Limb Reconstruction vs. AmputationDouglas G. Smith, M.D.

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

Safety of Lumbar Fusion Surgery for Chronic Back PainSohail K. Mirza, M.D.

Skeletal DysplasiasDavid R. Eyre, Ph.D.

TLS and TLS Fusion Proteins in LeukemiaLiu Yang, Ph.D.

Veterans Affairs Medical Center Review Grants

Effect of Motor Imbalance on Bony Deformity and Plantar PressureBruce J. Sangeorzan, M.D.

Ewing’s Sarcoma Fusion Proteins and mRNA Splicing FactorsHoward A. Chansky, M.D.

Rehabilitation Research and Development Center of Excellence for Limb LossPrevention and Prosthetic Engineering

Bruce J. Sangeorzan, M.D.


RESEARCH GRANTS


Centers for Disease Control

Chest Injuries Due to Motor Vehicle Side ImpactsAllan F. Tencer, Ph.D.

Orthopaedic Research and Education Foundation (OREF)

Overcoming Nerve Defect by Growth Factor Stimulated Regeneration AlongIntact Nerves

Thomas E. Trumble, M.D.Ben DuBois, M.D.

Reduction of Total Knee Arthroplasty Risk in Morbidly Obese Patients UsingLaparoscopic Bariatic Surgery: A Prospective, Controlled Trial

Seth S. Leopold, M.D.

The Role of Wild-Type EWS and the EWS/FLI-1 Fusion Protein in the Genesisof Ewing’s Sarcoma

Howard A. Chansky, M.D.

Toward Clinical Application of the Intact Nerve Bridge Technique: Longer TermStudy in a Rabbit Model

Thomas E. Trumble, M.D.Wren V. McCallister, M.D.

American Federation for Aging Research

Rest-Inserted Loading: Examination of a Novel Osteogenic Stimulus For theAged Skeleton

Sundar Srinivasan, Ph.D.

American Society for Surgery of the Hand

Bridging Versus Non-Bridging External Fixation of Distal Radius Fractures:A Measure of Structural Rigidity in an Unstable Fracture Model

Thomas E. Trumble, M.D.


RESEARCH GRANTS


American Fracture Association

Central Screw Placement in Stimulated Scaphoid Waist Fractures:A Biomechanical Study

Wren V. McCallister, M.D.Thomas E. Trumble, M.D.

American Shoulder and Elbow Surgeons

Periosteal Arthroplasty of the Reamed Glenoid: A Histologic andHistomorphometric Study in a Rabbit Model of Non-ProstheticArthroplasty

Kevin L. Smith, M.D.

Defense Advanced Research Projects Agency

Achieving Molecular Observation in Four YearsJohn A. Sidles, Ph.D.

The Accelerated Development of MRFMJohn A. Sidles, Ph.D.

Florence and Marshall Schwid Memorial Foundation

The Role of Wild-Type TLS and the TLS/CHOP Sarcoma Fusion Protein in theGenesis of Myxoid Liposarcoma

Howard A. Chansky, M.D.

Genetics Institute

A Feasibility and Safety Study of rhBMP-2/ACS and Allograft Compared toAutogenous Bone Graft for Patients with Severe Tibial Shaft Fractures

Sohail K. Mirza, M.D.

Leukemia Research Foundation

Disruption of RNA Splicing by TLS/ERG Leukemia Fusion ProteinLiu Yang, Ph.D.


National Childhood Cancer Foundation

Childrens Cancer GroupErnest U. Conrad III, M.D.

National Science Foundation Engineering Research Center

Tissue-Engineered Digit ReplacementChristopher H. Allan, M.D.

Novartis Pharmaceuticals Corporation

Cartilage Collagen StudyDavid R. Eyre, Ph.D.

Orthopaedic Trauma Association

Immediate vs. Delayed Closure of Type II and IIIA Open Tibia FracturesM. Bradford Henley, M.D.

Ostex International, Inc.

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

Pfizer, Inc.

Guinea Pig OsteoarthritisDavid R. Eyre, Ph.D.

Sevrain, Christophe J-P

Biomechanical Testing of a Cervical Spine Fixation PlateAllan F. Tencer, Ph.D.

Skeletech, Inc.

Micro-CT Imaging of Mice VertebraeTed S. Gross, Ph.D.

RESEARCH GRANTS



RESEARCH GRANTS


The Boeing Company

Randomized Clinical Trial of Open versus Endoscopic Carpal Tunnel Releaseand Hand Therapy Comparing Patient Satisfaction: Functional Outcomeand Cost Effectiveness

Thomas E. Trumble, M.D.

Tyco Healthcare

Interdiscal Electrothermography (IDET) in Cervical DiscsAllan F. Tencer, Ph.D.

US Army

Achieving Molecular Observation in four yearsJohn A. Sidles, Ph.D.

Whitaker Foundation

Examining Processes Underlying the Dramatic Osteogenic Response Elicited byRest-Inserted Loading

Sundar Srinivasan, Ph.D.


APRIL 2002 THROUGH MARCH 2003

Friends of Orthopaedics

John AberleAircastJeffrey AkesonMaria AlexanderChris AllanFrank AlvineMarilyn AlvineAmerican Society of BiomechanicsMarlene AngelLarry ArbourIra E. ArmadAscension OrthopaedicsAXYA Medical Inc.Samuel BakerBallard Orthopedic & Fracture ClinicDavid BareiWilliam BarrettThomas BarthelBayfield ChiropracticTimothy BealsDavid BelfieCarlo BellabarbaSteve BenirschkeGary BergmanInez BowmanThomas BoycePaul BrandAndrew BronsteinBrown Middle School StaffJim BrucknerCarolyn & Bob CampbellSusan CeroHoward ChanskyJens ChapmanChildren’s HospitalSandy ChristensonGary ClanceyJohn ClarkBurnet ClarkeMichael CoeErnest ConradMarjolijn ConradJay CraryKathryn Del BeccaroDavid DenekaDePuy

We express our appreciation to all who have contributed to the work of the Department of Orthopaedics and Sports Medicineover the past year. Your assistance makes possible special research activities, educational programs, and other projects thatwe could not offer without this extra support from our alumni, faculty, and friends in the community. We owe a specialthanks to the University of Washington Resident Alumni who have made significant contributions to help further the educationof our current residents. We have tried to include in this list all who contributed; if anyone was overlooked, please be sure tolet us know!

Contributors to Departmental Research and Education

Mohammad DiabRichard DimondOriente DitanoLee & Jeanne DobbsJoan DohertyMargaret DohertyAllen & Twila EaslerEBI Medical SystemsEly Community FoundationDonald EricksenEdward FarrarNancy & Jose FigueroaJohn & Marianne FishbackDaniel FlugstadJonathan FranklinBetty Jean FrantzFzio Med Inc.Jeff GarrDick & Marjorie GatkeRobert GieringerPark GloydJohn GreenThomas GreenTed GreenleeTed GrossDale HaakensonArlene HaaserDouglas HanelMargaret HanelSigvard HansenHarbor Orthopedic ClinicNorman & Neva HartyMark HarrastThomas HarrisJohn HendricksonHilsboro PharmacyJoel HoekemaHowmedica Osteonics/Stryker SpineFrederick HuangBurton & Margaret HuggettLarry HullThomas HutchinsonTodd JaroszDoug & Bev JohnsonJeanne JohnsonNancy Kadel


Friends of Orthopaedics

Allan Tuney KannapellTom KarnezisDavid KargesCarleton KeckKatayo KhakpourRichard KirbyHans KlassJonathan KnightWalter KrengelRoger LarsonSeth LeopoldGeorge LuckNeil MakiFrederick MannCarol MartinErma MartinsonFrederick MatsenGregory MayJanice McCafferyRobert McConnellDavid McGuireLaVerne & Shirley MeachamMichael MetcalfJohn MichelottiMidwest Orthopaedic CenterWilliam MillsPeter MitchellMitekJohn MiyanoVincent MoscaMarr MullenWilliam MunsonJohn MurrayFlorence NelsonKitty NewmanMary NofsingerSean NorkNorthwest BiometNorthwest Orthopaedic AssociatesSusan NosackaJohn O’KaneWilliam OppenheimPacific OrthopedicsBarbara PaetzholdMichael & Jennifer PatriarcheDon & Cindy PetersonPfizerPharmacia CorporationMary PlegerMichael PowerLila ProppGregory RafijahSteven RatcliffeStan ReedJim ReiffRandy & Debi ReischmanMark Remmington

Christine RobertsJohn RobertsonRaymond RobinsonCharles Rockwood Jr.Oranne Sally RoedigerJerry & Marilyn RossM.L. Routt and Janet JohnsonJohn SackBruce SangeorzanShanon SaraKenichi SatoGregory SchmaleRobert SchoeneScottsdale Orthopaedic SurgeonsSeattle Hand Surgery GroupRichard SemonAl & John ShiptonConstance SidlesPeter SimonianSimonian Sports ClinicCarla SmithDoug SmithJeffrey SmithKevin SmithSmith & NephewKit SongMuriel & Douglas StraughanMarc SwiontkowskiSynthes SpineCarol TeitzAllan TencerWilliam TheimeMichael ThorpeDavid ThullThomas TrumbleMartin TullusTyco Healthcare/Surgical DynamicsDale & Jo Anne TysonEric VanderhooftMary Stuart Van MeterNicholas VedderRobert VeithWilliam WagnerWashington Hand SurgeryWashington Orthopedic CenterEdward WeinbergerAmy WeisnerJohn WestTed & Gerry WillieRobert WinquistRobert WoodJiann-Jiu WuRobert YanceyLiu YangZimmerJoseph Zuckerman


Alumni

1952Park W. Gloyd, M.D. ★

1954Trygve Forland, M.D. ★

1955Robert W. Florence, M.D.

1956J. Michael Egglin, M.D. ★John E. Goeckler, M.D.Robert L. Romano, M.D.

1957John H. Aberle, M.D. ★John R. Beebe, M.D.

1958Harry H. Kretzler, Jr., M.D.James R. Friend, M.D. ★Kenneth L. Martin, M.D. ★Samuel L. Clifford, M.D.

1959James W. Tupper, M.D.

1960Irving Tobin, M.D. ★William V. Smith, M.D. ★★

1961Robert C. Colburn, M.D.

1962Arthur Ratcliffe, M.D.Marr P. Mullen, M.D. ★

1963Alfred I. Blue, M.D.Robert A. Kraft, M.D.

1964David E. Karges, M.D. ★★★Harold J. Forney, M.D. ★Theodore K. Greenlee II, M.D.★★★★★Thomas E. Soderberg, M.D.

1966F. Richard Convery, M.D. ★Joseph S. Mezistrano, M.D. ★William A. Reilly, Jr., M.D.

1967Ivar W. Birkeland, M.D.J. Conrad Clifford, M.D.Robert F. Smith, M.D.

1968Lynn T. Staheli, M.D. ★Stewart M. Scham, M.D. ★William T. Thieme, M.D. ★

1969Edward E. Almquist, M.D. ★★Edward L. Lester, M.D.Hugh E. Toomey, M.D. ★★★Sigvard T. Hansen, Jr., M.D. ★★★

1970John C. Brown, M.D. ★John M. Coletti, Jr., M.D. ★Malcolm B. Madenwald, M.D. ★Michael T. Phillips, M.D. ★Robert D Schrock, Jr., M.D.

1971Bruce E. Bradley, Jr., M.D.Franklin G. Alvine, M.D. ★★★Jerome H. Zechmann, M.D.Louis A. Roser, M.D. ★Nils Fauchald, Jr., M.D.

1972David J. LaGasse, M.D.David R. Nank, M.D. ★★Donald D. Hubbard, M.D. ★John A. Neufeld, M.D. ★Thomas L. Gritzka, M.D. ★

1973Frederick J. Davis, M.D. ★Larry D. Hull, M.D. ★Robert P. Watkins, Jr., M.D. ★Theodore A. Wagner, M.D. ★★★★

1974Richard A. Dimond, M.D. ★★Ronald B.H. Sandler, M.D. ★★★Samuel R. Baker, M.D. ★★Robert A. Winquist, M.D. ★★★★★

1975Donald L. Plowman, M.D. ★★★Frederick A. Matsen III, M.D.★★★★★Gunter Knittel, M.D.Larry R. Pedegana, M.D. ★Thomas M. Green, M.D. ★★★William M. Backlund, M.D., P.S. ★

1976Douglas K. Kehl, M.D.Douglas T. Davidson III, M.D. ★John F. Burns, M.D.Peter Melcher, M.D.Richard A. Zorn, M.D. ★

1977Carl A. Andrews, M.D. ★Geoffrey W. Sheridan, M.D. ★★Larry D. Iversen, M.D. ★Mark C. Olson, M.D. ★Steven T. Bramwell, M.D.

1978Arnold G. Peterson, M.D. ★★★★Gary J. Clancey, M.D. ★★★John W. Brantigan, M.D.Richard S. Westbrook, M.D. ★★Robert J. Strukel, M.D.William Oppenheim, M.D. ★

1979Allan W. Bach, M.D. ★★★★Gregory M. Engel, M.D. ★★Jonathan L. Knight, M.D. ★Richard L. Semon, M.D. ★★★

1980Carol C. Teitz, M.D. ★★Douglas G. Norquist, M.D.John M. Hendrickson, M.D. ★Michael A. Sousa, M.D. ★★Stuart R. Hutchinson, M.D. ★

1981Dennis J. Kvidera, M.D. ★John M. Clark, Jr., M.D., Ph.D. ★★★Martin S. Tullus, M.D. ★★★★Robert G. Veith, M.D. ★★★★★


1982John L. Thayer, M.D. ★Richard M. Kirby, M.D. ★★★★Steven S. Ratcliffe, M.D. ★★William D. Burman, M.D.

1983E. Anne O. Elliot, M.D. ★Edward L. Farrar III, M.D. ★★★Henry K. Yee, M.D.Joseph D. Zuckerman, M.D. ★★★★Keith A. Mayo, M.D. ★★★Robert M. Berry, M.D.

1984Jeffrey C. Parker, M.D. ★Jeffrey W. Akeson, M.D. ★★★Kevin P. Schoenfelder, M.D.Marc F. Swiontkowski, M.D.★★★★★Thomas J. Fischer, M.D. ★★★★

1985Daniel L. Flugstad, M.D. ★Jeffrey N. Hansen, M.D. ★★★Paul J. Abbott, M.D. ★★★Richard J. Barry, M.D. ★William P. Barrett, M.D. ★★★★

1986Carleton A. Keck, Jr., M.D. ★★Gary Bergman, M.D. ★★★★★Lawrence E. Holland, M.D. ★Michael E. Morris, M.D. ★★★

1987Craig T. Arntz, M.D. ★★Herbert R. Clark, M.D. ★★Michael K. Gannon, M.D. ★Steven L. Reed, M.D. ★

1988Jonathan L. Franklin, M.D. ★★★★Michael A. Thorpe, M.D. ★★★★★Richard V. Williamson, M.D. ★

1989James P. Crutcher, M.D. ★★★★Lawrence V. Page, D.O. ★Martin G. Mankey, M.D. ★★Nancy J. Ensley, M.D.Steve C. Thomas, M.D. ★★

1990David M. Kieras, M.D. ★J. Roberto R. Carreon, M.D.Jay A. Winzenried, M.D. ★★Ken Fujii, M.D. ★Walter F. Krengel III, M.D. ★★

1991David H. Bishop, M.D. ★★Kit M. Song, M.D.Mark Remington, M.D. ★★Mark E. Murphy, M.D., Ph.D. ★Tim P. Lovell, M.D. ★★

1992Curt Rodin, M.D.Don Striplin, M.D. ★★Eli Powell, M.D. ★Jeff Stickney, M.D. ★John D. West, M.D. ★Michael Sailer, M.D. ★★

1993J. Eric Vanderhooft, M.D. ★★★★Lyle S. Sorensen, M.D. ★★★★Philip J. Kregor, M.D. ★★Susan R. Cero, M.D. ★★★

1994Brodie Wood, M.D. ★★Eric Bowton, M.D. ★Jim Vahey, M.D. ★Sohail K. Mirza, M.D.William Obremskey, M.D. ★★

1995Ron Kristensen, M.D. ★Scott Hormel, M.D. ★Timothy Beals, M.D. ★Todd Clarke, M.D. ★William J. Mills III, M.D. ★

1996David Deneka, M.D. ★Peter Mitchell, M.D. ★★Peter T. Simonian, M.D. ★Vernon Cooley, M.D. ★William Wagner, M.D. ★★

1997Daniel Stechschulte, Jr., M.D.David Levinsohn, M.D. ★L. Anthony Agtarap, M.D. ★Mohammad Diab, M.D.Randall W. Viola, M.D.

1998Colin Poole, M.D. ★David Belfie, M.D. ★Don Ericksen, M.D. ★★Jay Crary, M.D. ★Oriente DiTano, M.D. ★

1999Craig Boatright, M.D.Jeffrey Garr, M.D.John Michelotti, M.D. ★Julie A. Switzer, M.D.Thomas D. Chi, M.D. ★

2000Brett Quigley, M.D. ★Cara Beth Lee, M.D.Daniel Jones, M.D. ★Joel Hoekema, M.D. ★Patrick McNair, M.D.

2001Eric Novack, M.D.Frederick Huang, M.D. ★Matthew Camuso, M.D.Michael Metcalf, M.D. ★Richard Bransford, M.D.

2002Timothy DuMontier, M.D.Scott Hacker, M.D.Timothy Rapp, M.D.William Sims, M.D. ★Carla Smith, M.D. ★

2003Ben DuBois, M.D.Andy Howlett, M.D.Guy Schmidt, M.D.Brian Shafer, M.D.Emma Woodhouse, M.D.

★★★★★ = $10,000 and over★★★★ = $7,500 - $9,999

★★★ = $5,000 - $7,499★★ = $2,500 - $4,999

★ = $1 - $2,499

STARS INDICATE DONATIONS IN SUPPORT

OF THE RESIDENCY


Endowments

Ernest M. Burgess Endowed Chair for Orthopaedics Investigation

Sigvard T. Hansen Jr. Endowed Chair in Orthopaedic Traumatology

Jerome H. Debs Endowed Chair in Orthopaedic Traumatology

Endowed Chair for Women’s Sports Medicine and Lifetime Fitness

Surgical Dynamics Endowed Chair for Spine Research

Douglas T. Harryman II/DePuy Endowed Professorship for Shoulder Research

Zimmer Fracture Fixation Biology Endowed Professorship

Ostex Bone and Joint Research Endowment

Orthopaedic Traumatology Endowed Lectureship

John F. LeCocq Lectureship in Orthopaedic Surgery

Don and Carol James Research Fund in Sports Medicine and Fitness

Victor H. Frankel Award

Esther Whiting Award

Ed Laurnen Award

Spine Research Endowment

We express our appreciation to all who have contributed to the endowments of the Department of Orthopaedics and SportsMedicine. Your assistance makes possible special research activities, educational programs, and other projects that we couldnot offer without this extra support from our alumni, faculty, and friends in the community. Additional Contributions tothese and new endowments are most welcome! If you have any questions, please contact our Chair, Rick Matsen, or ourAdministrator, Diana Jansen.

Documents

UNIVERSITY OF WASHINGTON Department of Orthopaedics · 2019. 7. 30. · Contents Foreword