1

Department of RadiologyHelsinki University Central Hospital

University of Helsinki, Finland

Pertti T. Karjalainen, M.D.

MAGNETIC RESONANCE IMAGING

OF ACHILLES TENDON

with special reference to

normal appearance,chronic disorders and

postoperative total ruptures

Academic Dissertation

to be presented with the permission ofthe Faculty of Medicine of the University of Helsinki,

for public discussion in Auditorium XII.

On May 31st, 2000, at 2 p.m.

2

Supervised by

Professor Hannu Aronen, M.D.Department of RadiologyHelsinki University Central Hospital, HelsinkiDepartment of Clinical RadiologyKuopio University

Reviewed by

Docent Sakari Orava, M.D.Department of Orthopaedics and TraumatologyUniversity of Oulu

Docent Timo Paakkala, M.D.Department of RadiologyUniversity of Tampere

ISBN 952-91-2133-4 (nid.)ISBN 952-91-2134-2 (PDF version)Helsingin yliopiston verkkojulkaisutHelsinki 2000

3

Contents

List of original papers ___________________________________ 6

Abbreviations and definitions _____________________________ 7

Introduction __________________________________________ 8

Review of the literature__________________________________ 9

Magnetic resonance (MR) imaging ________________________ 9Basic principles ____________________________________________ 9MR field strength and coils ___________________________________ 9Sequences________________________________________________ 9MR tissue characteristics of tendons___________________________ 10Magic angle phenomenon ___________________________________ 11Chemical shift artifact ______________________________________ 11Musculoskeletal MR imaging _________________________________ 11Foot and ankle MR imaging__________________________________ 11

Ultrasonography in tendons ____________________________ 12

X-ray and CT in Achilles tendons ________________________ 13

Achilles tendon anatomy_______________________________ 14Functional anatomy________________________________________ 14Normal MR appearance_____________________________________ 14

Achilles tendon rupture________________________________ 16Incidence and pathophysiology_______________________________ 16Diagnosis and treatment____________________________________ 16Rehabilitation ____________________________________________ 17MR appearance ___________________________________________ 17

Achilles tendon overuse injuries_________________________ 18Tendinosis _______________________________________________ 18Insertional disorders _______________________________________ 19Peritendinitis _____________________________________________ 19MR appearance ___________________________________________ 20

Other causes of Achilles tendinopathy ____________________ 21

The aims of the present study were _______________________ 22

Subjects, materials and methods _________________________ 23

Subjects ___________________________________________ 23

Clinical evaluation____________________________________ 24

Conservative treatment _______________________________ 24

4

Surgical treatment ___________________________________ 24Indications ______________________________________________ 24Surgical techniques________________________________________ 24Surgical evaluation ________________________________________ 25Postoperative rehabilitation _________________________________ 25Complications ____________________________________________ 25

Clinical follow-up and scoring___________________________ 25

Magnetic resonance imaging ___________________________ 26MR protocols _____________________________________________ 26MR image analysis ________________________________________ 27

Ultrasonography studies_______________________________ 28

Histopathological studies ______________________________ 29

Statistical methods ___________________________________ 29

Results______________________________________________ 30

Postoperative MRI findings in patients with Achilles tendonrupture (Paper I) ____________________________________ 30

3 weeks_________________________________________________ 306 weeks_________________________________________________ 303 months________________________________________________ 326 months________________________________________________ 351 to 3 years (Paper II) _____________________________________ 36Correlation between dimensions of MRI and US __________________ 36Postoperative ultrasonography _______________________________ 37

MR imaging of asymptomatic Achilles tendons (Paper III) ____ 38Dimensions ______________________________________________ 38Shape __________________________________________________ 39Plantaris tendon __________________________________________ 39Signal intensity ___________________________________________ 40Insertion to calcaneus______________________________________ 40Peritendinous tissues ______________________________________ 41

MR imaging of overuse injuries of the Achilles tendon (Paper IV)42Antero-posterior diameter___________________________________ 43Intratendinous lesions______________________________________ 43Tendon insertion and bursa _________________________________ 44Peritendinous tissues ______________________________________ 45MRI and clinical findings ____________________________________ 47MRI and surgical findings ___________________________________ 47MRI and histological findings ________________________________ 47Long-term follow-up _______________________________________ 49

5

Discussion ___________________________________________ 50

Postoperative follow-up of surgically repaired Achilles tendonruptures ___________________________________________ 50

Patient material___________________________________________ 50Cross-sectional area _______________________________________ 50Intratendinous lesions______________________________________ 50Return to sports __________________________________________ 51Reoperations _____________________________________________ 51Functional tests and MRI____________________________________ 52Miscallenous findings ______________________________________ 52Ultrasonography __________________________________________ 53

MR imaging of asymptomatic Achilles tendon ______________ 53Diameter ________________________________________________ 53Shape __________________________________________________ 53Signal intensity ___________________________________________ 54Plantaris tendon __________________________________________ 54

Intratendinous lesions of Achilles tendon _________________ 55Asymptomatic subjects _____________________________________ 55Symptomatic subjects______________________________________ 55

Peritendinous tissues _________________________________ 56Normal appearance________________________________________ 56Abnormal appearance ______________________________________ 57

Tendon insertion and retrocalcaneal bursa_________________ 57Normal appearance________________________________________ 57Abnormal appearance ______________________________________ 58

Multiple findings _____________________________________ 58

MR imaging as prognostic method _______________________ 58

Sequences__________________________________________ 59

High vs. low field MR imaging___________________________ 59

Limitations _________________________________________ 59

Conclusions and summary_______________________________ 60

Acknowledgements ____________________________________ 61

References___________________________________________ 63

6

LIST OF ORIGINAL PAPERS

This study is based on the following papers, which are referred to in thetext with the Roman numerals (I-IV).

I Karjalainen PT, Aronen HJ, Pihlajamäki HK, Soila K, Paavonen T,Böstman O. Magnetic resonance imaging during the healing ofsurgically repaired Achilles tendon rupture. Am J Sports Med1997;25:164-171

II Karjalainen PT, Ahovuo J, Pihlajamäki HK, Soila K, Aronen HJ.Postoperative MRI and ultrasonography of a surgically repairedAchilles tendon ruptures. Acta Radiol 1996;37:639-646

III Soila K, Karjalainen PT, Aronen HJ, Pihlajamäki HK, Tirman PFJ. Highresolution MR imaging of asymptomatic Achilles tendon: Newobservations. Am J Roentgenol 1999;173:323-328

IV Karjalainen PT, Soila K, Aronen HJ, Pihlajamäki H, Tynninen O,Paavonen T, Tirman PFJ. MR imaging of overuse injuries of theAchilles tendon. Am J Roentgenol 2000;175:000-000

7

ABBREVIATIONS AND DEFINITIONS

AP = anteroposterior

CP = circular polarized

CT = computerized tomography

FLASH = fast low angle shot T1-weighted spoiled gradient echo

FOV = field of view

MR = magnetic resonance

MRI = magnetic resonance imaging

ms = millisecond

SD = standard deviation

SE = spin echo

SI = signal intensity

STIR = fast short-inversion-time inversion-recovery or

short tau inversion recovery

T1 = longitudinal relaxation time

T2 = transverse relaxation time

TE = time to echo

TI = inversion time

TR = repetition time

US = ultrasonography

8

INTRODUCTION

The Achilles tendon is the largest andstrongest tendon in man. Also, it is oneof the most frequently torn and one ofthe most common sites of overuseinjuries among athletes (Galloway etal. 1992, Leppilahti et al. 1991).Among runners the occurrence ofAchilles tendon disorders varies fromabout 5 to 18% (Kvist 1994). Achillestendon rupture is a common traumaaffecting most often active, earlymiddle-aged men with tendondegeneration, which is considered to berequisite to have a rupture of Achillestendon (Jozsa et al. 1989). Operativetreatment of Achilles tendon rupture isfavored by many surgeons because ofits lower risk of rerupture comparedwith non-surgical treatment (Wills etal. 1986). The importance of thepostoperative evaluation of the reunionprocess of the operated Achilles tendonhas been emphasized in order to giveguidelines in pacing the rehabilitation(Marcus et al. 1989, Quinn et al.1987).

The spectrum of Achilles tendonoveruse injuries ranges frominflammation of the peritendinoustissue (peritendinitis), structuraldegeneration of the tendon(tendinosis), to partial or completetendon rupture (Kvist 1994). Theseconditions may co-exist (e.g.,peritendinitis with tendinosis)(Schepsis et al. 1994). Clinically, it isoften difficult to distinguish tendinosisfrom peritendinitis (Kvist 1994), andfrom partial tearing of the tendon(Allenmark 1992). Treatment andprognosis vary depending on thepathology (Allenmark 1992, Gallowayet al. 1992, Kvist 1994, Schepsis et al.1994).

Imaging method of the Achilles tendoninclude plain radiography, ultra-sonography (US) and magneticresonance imaging (MRI). State-of-artMR imaging offers an excellent softtissue contrast and spatial resolution.

Interpreting radiologists must be awareof the imaging appearance of a normaltendon, expected postoperativechanges and complications of thesurgically repaired ruptured Achillestendon because the patients with poorclinical outcome are often re-evaluatedwith US or MRI.

The Achilles tendon has classicallybeen described to possess uniform lowsignal intensity in all commonly usedMR sequences. Recently, two groups(Åström et al. 1996, Rollandi et al.1995) of authors have stated that thenormal Achilles tendon can haveincreased intratendinous signalintensity spots on axial T1-weightedand proton density - weighted images.As technical quality in musculoskeletalMR imaging has improved (Erickson1997), and experience in imageinterpretation increases, newobservations can be made regardingthe Achilles tendon and its surroundingtissues.

This study first evaluates thepostoperative appearance of surgicallytreated Achilles tendon ruptures in low-field MR unit. Then asymptomaticsubjects and symptomatic patientswith overuse injuries of the Achillestendon are imaged in a modern high-field MR unit.

9

REVIEW OF THE LITERATURE

Magnetic resonance (MR)imaging

Basic principles

A strong, homogeneous magnetic fieldis required for nuclear magneticresonance phenomenon to occurr at asufficient energy level to emit a signalwhich is strong enough for imaging.Powerful radiofrequency transmitter isused to give radio frequency pulseswith a radiofrequency (RF) antenna, orRF coil. The pulses are given in asequence which creates specificcontrast in the images. The energy isabsorbed by atomic nuclei andsubsequently emitted through aprocess called relaxation. The energy isemitted as radiofrequency wawes whichare detected with a sensitiveradiofrequency receiver via receivingcoils. The detected signal originatesfrom a slice of tissue at a time when agradient magnetic field in Z direction isapplied during rf stimuli. The signal iscoded for localization within the slice byapplying rapidly switching gradientfields in X, and Y directions. The signalis transferred from receiving coils tothe computer. This data is convertedinto an image through mathematicalfunction called Fourier transform. Theimages are displayed throughappropriate media such as film or acomputer workstation (Harms 1997).

MR field strength and coils

Most musculoskeletal studies today areperformed with high field MR units(1.5T). The use of low field units israpidly growing due to technicalimprovements. The most unfavourablequality of low field MRI is a lowerspatial resolution. Low field systemsare not capable of rapidly producing

high-resolution images, such as can beobtained with 1.5-T magnet (Beltran etal. 1987). However, new low-field (0.2T) MR units have shown goodagreement with pathological findings inimaging ankle injuries when comparedto 1.0-T MR units (Merhemic et al.1999). When compared to sonography,even low-field MRI investigation allowsmore accurate staging of tendinouschanges than sonography. It is morereproducible and includes theadvantages of the combined evaluationof bones, ligaments, and soft tissue(Rand et al. 1998).

Early studies in musculoskeletal MRimaging were performed with body andhead coils. There was a majorimprovement in image quality whencircular polarized extremity and surfacecoils were developed (Lucas et al.1997, Maurer et al. 1996, Maurer et al.1996).

Sequences

Spin echo imaging was the firsttechnique developed for clinical imagingand still is most widely used inmusculoskeletal imaging (Evancho etal. 1990, Kalmar et al. 1988). In spinecho imaging, some time (echo timedivided by two) after 90° pulse, a 180°pulse is applied. This rephases theprotons that are getting out of phase.

Gradient echo sequences weredeveloped for rapid imaging, they useshorter repetition time, low excitationpulse angles and have shorter echotimes than spin echo sequences. Theappearance of lesions on gradient echoimages can be different from that inspin echo images. This is due to thesignal being influenced by T2*relaxation, in which the dephasing

10

process of protons is not compensatedlike in spin echo sequences with 180°rephasing pulse. Also, artifacts can beintroduced in the images for differentreasons than in spin echo imaging,such as signal decrease from magneticsusceptibility effects. However, noconsiderable signal dephasing due tosusceptibility effects are found intendons (Schick et al. 1995).

Short inversion time or short tauinversion recovery sequences (STIR)was introduced to eliminate high signalemanating from fatty tissues. STIRsequence consists of a 180° inversionpulse followed typically by spin-echo orturbo spin echo sequence (tSTIR) foracquisition of the signal for imaging.After the inversion pulse, themagnetization recovers exponentiallyfrom the maximal negative value to amaximal positive value through theinversion time null point. The timeinterval between the inversion pulseand the excitation pulse is inversiontime. Inversion of magnetizationincreases sensitivity to tissue T1differences. By selecting the inversiontime relatively short, protons in fattytissues are at null point in recoverywhen imaging portion of the sequenceis initiated. Therefore, signal from fattytissues can be reduced thus increasingcontrast with lesions within it forimproved detectability.

MR tissue characteristics oftendons

The signal intensity of normal tendonexhibited by spin-echo and gradient-echo sequences with common echotimes (TE) >10ms is very low (Quinn etal. 1987, Schweitzer 1993). This is dueto characteristically long T1 and shortT2 relaxation times of tendons in whichhydrogen nuclei of water molecules(protons) are strongly associated withthe collagen matrix (Gold et al. 1995).T2 relaxation time for intact tendon

tissue is very short, approximately 0,25ms with the tendon aligned with themagnetic field (Fullerton et al. 1985). Itis practically independent of fieldstrengths commonly used (Koblik andFreeman 1993). Recognition of tendonpathology is based on the detection ofareas of increased signal within tendon.These increased signal lesionsrepresent areas of T2 prolongationassociated with disruption of organizedcollagen structure and edema (Cheunget al. 1992, Erickson et al. 1992).

Sensitivity of MR imaging to earlystages of tendon pathology can beimproved by application of sequenceswith very short echo times because thegradient echo methods allow shorterecho times than spin echo techniquesfor a given gradient system of theimager and given spatial resolution(Schick et al. 1995). Gradient echo andSTIR sequences have been found moresensitive in detecting focal signalchanges in patellar tendon than spinecho sequences alone (Davies et al.1991, Khan et al. 1996). Minimum echotime gradient echo sequences shouldbe used for sensitive imaging of tendonalterations, because no considerablesignal dephasing due to susceptibilityeffects that might be detrimental havebeen found in tendons (Koblik andFreeman 1993).

As modern MR equipment with high-performance gradient coil systems hasbecome available for clinical imagingsystems, an expanding role for MRI inthe evaluation of tendon disease ispossible.

The soft tissues surrounding theAchilles tendon are rich in fat (pre-Achilles fat pad, subcutaneous tissues,bone marrow). Fat suppressionsequences increase diagnosticcapabilities of MRI by being moresensitive for detection of lesions inmusculoskeletal imaging (Masciocchi et

11

al. 1998, Nakamura et al. 1999,Yoshioka et al. 1994).

Signal from fat may be suppressedtaking advantage of the difference inresonance frequency of lipid protonsfrom water protons by means offrequency selective pulses or phasecontrast techniques, or based on theshorter T1 relaxation time in fattytissues than in other soft tissues byutilizing inversion recovery sequences(de Kerviler et al. 1998).

Magic angle phenomenon

The "magic angle" effect in MR imagingis caused by changes in the dipolarinteractions between water hydrogenprotons that are loosely bound alongcollagen fibrils in organized tissue suchas tendon. When tendons are aligned at55 degrees to the main magnetic field,the T2 relaxation time is lengthened,causing focal increased signal on shortecho time MR images. Tendons in theankle, wrist, and rotator cuff of theshoulder are common sites to observethis effect (Erickson et al. 1993,Fullerton et al. 1985, Haygood 1997,Peto and Gillis 1990). However, Achillestendon is parallel or perpendicular tothe main magnetic field in clinicalimaging, and no increased signalintensity due to magic anglephenomenon will occurr within anormal tendon.

Chemical shift artifact

Chemical shift artifact is seen as a thin,low-intensity line, which partiallysurrounds many structures on MRimaging. This artifact should not bemistaken for a true morphologicstructure. This artifact can berecognized by its characteristicappearance along the direction of thefrequency-encoding gradient at theinterface of fatty and non-fatty tissues.Water and lipid protons differ in their

Larmor frequencies and the location offat is misregistered two to three pixelsin relation to water at the interface(Soila et al. 1984, Weinreb et al.1985). Around the Achilles tendon suchan interface is between thesubcutaneous fat and paratenon.

At low-field the chemical shift artifact isless problematic in clinical imaging(Weinreb et al. 1985).

Musculoskeletal MR imaging

Pettersson et al stated at 1985 that“the definite place of MRI withinmusculoskeletal diagnostic imaging isnot yet settled, but its potential isgreat, and it will have an important rolein the future” (Pettersson et al. 1985).At 2000 there is a broad range ofclinical applications for MRI in thediagnosis of musculoskeletal disease(Burk et al. 1988, Erickson 1997). MRimager with high-resolution andmultislice, multiecho techniqueprovides detailed information on alljoints, muscles, ligaments and spinewith an ever shortening examinationtime (Erickson 1997).

Foot and ankle MR imaging

The foot and ankle are among thehardest of all areas to image becauseof the complex three-dimensionalanatomy. MR imaging, with itsmultiplanar capabilities, excellent soft-tissue contrast, ability to image bonemarrow, noninvasiveness, and lack ofionizing radiation, has become avaluable tool for the diagnosis andstaging as well as the surgical planningof multiple disorders (Ferkel et al.1991, Mammone 1997, Schweitzer andKarasick 1994).

Avascular necrosis is common in thefoot, usually seen after talus fracturesor spontaneously in the metatarsalheads. Other causes of a marrow

12

edema pattern include stress fractureand occult fractures, which appear likebone bruises but behave more likefractures. Arthritic disorders, reactionsto altered biomechanics, osteomyelitis,and regional migratory osteoporosistypically cause marrow edema.

Characteristic signs of plantar fasciitison MRI are thickening of the plantarfascia and intratendinous signalintensity increase with contrastenhancement to some extent. Bonemarrow edema of the calcaneus andperitendinous edema close to theplantar fascia can be also detected(Steinborn et al. 1999).

MRI has also been utilized to assessankle sprains with complications, suchas the ankle impingement syndrome,the sinus tarsi syndrome, and chronicinstability, and to diagnoseosteomyelitis. MRI is promising for theevaluation of reflex sympatheticdystrophy and is as useful for theevaluation of bone and soft tissuetumors as it is elsewhere in theskeleton. MR imaging helps tocharacterize the biologicalaggressiveness of the tumor as well asits extent and therefore aids in surgicalplanning (Beltran et al. 1987,Chandnani and Bradley 1994, Cheunget al. 1992, Ferkel et al. 1991,Panageas et al. 1990, Schweitzer 1993,Schweitzer and Karasick 1994).

MRI is more specific than bonescintigraphy and provides moreinformation than ultrasound andcomputed tomography. Arthroscopy ofthe ankle is limited to the articularsurface and joint space. MRI allows aglobal evaluation of the bones,tendons, ligaments, and otherstructures with a single examinationthat exceeds the capabilities of all otheravailable techniques (Lucas et al.1997).

The most frequently diseased tendonsin the ankle are the Achilles, posteriortibial, and peroneal (Ferkel et al. 1991,Schweitzer and Karasick 1994). MRIcan be used to diagnose most disordersof these tendons, as well as stagethese disorders to allow appropriatetherapy. The major decisions to bemade when performing MR images ofthe tendons of the ankle are field ofview to be used and the planes ofimaging. Recommended field of view is12 to 16 cm, and imaging planes forAchilles tendon are sagittal and axial(Haygood 1997). For the diagnosis ofposterior tibialis tendon tears andinjuries, MRI is an important tool insurgical planning. Often, MRI is helpfulfor the diagnosis of peroneal tendoninjury, including dislocations andperoneal splits, two entities that areseen to effect the peroneal tendons.

Ultrasonography in tendons

Achilles tendon rupture has beendiagnosed by ultrasonography (US)since early 1980´s (Bruce et al. 1982).Since then the quality of the high-resolution real-time US scanners hasimproved remarkably. US is capable ofdetecting Achilles tendon ruptures,partial ruptures, tendinosis andpostoperative findings (Fornage 1986).Tendinosis is characterized byenlargement and decreased echogenityof the tendon (Fornage 1986). US maybe valuable in the diagnosis ofabnormalities in surrounding Achillestendon structures such as peritenon(Kainberger et al. 1990). Pre-operativeUS is said to be useful in thedifferentiation between partial ruptureand other tendinous changes (Lehtinen1996). However, US is not completelyreliable for diagnosing peritendinitisand tendinitis, and it cannot be used todifferentiate partial tendon rupturesfrom focal degenerative lesions(Paavola et al. 1998). Also, it is noteasy to differentiate partial Achilles

13

tendon rupture from total rupture(O'Reilly and Massouh 1993). In acomparison study, MR imaging wasmore sensitive to the presence ofintratendinous lesion than US (Åströmet al. 1996). According to Movin et al.,intratendinous lesions seen on MRimages are larger than correspondinghypoechoid areas in US. Hence, thenon-pathologic echo is not equivalentwith a healthy tendon structure inpatients with chronic achillodynia. Bothmethods equally measured thediameter and the shape of the tendon(Movin et al. 1998). US, unlike-MRimaging, is able to detectmicrocalcifications in the Achillestendon (Fornage 1986, Maffulli et al.1987).

US has been used in postoperativeassessment of the ruptured Achillestendon (Draghi et al. 1999, Maffulli etal. 1990, Merk et al. 1997, Thermannet al. 1992). Maffulli et al. performedUS at different postoperative times ona total of 22 patients with total ruptureof the Achilles tendon. In US, theoperated tendon remained of increasedthickness 9 months after surgery(Maffulli et al. 1990). Merck et al.reported no correlation between the USimaging findings and unsatisfactoryclinical results following Achilles tendonrepairs (Merk et al. 1997).

X-ray and CT in Achilles tendons

Plain radiographs with specialtechniques and xeroradiography areable to visualize soft tissue alterationsaround Achilles tendon (Denstad andRoaas 1979, Lehtonen et al. 1981).Intratendinous calcifications could bedetected with these methods. Denstadand Roaas suggested that partialruptures can be visualized by thethickening of the tendon tissue onlateral soft tissue radiographs (Denstadand Roaas 1979). Plain radiographs canbe used to indicate the rare occurence

of ossification of the Achilles tendon (Yuet al. 1994). A fracture of an ossifiedAchilles tendon diagnosed byradiograph has been reported (Aksoyand Surat 1998). Patients withrheumatoid arthritis, ankylosingspondylitis, psoriasis, and Reitersyndrome have shown to haveincreased thickness of the Achillestendon on lateral radiographs (Resnicket al. 1977).

An excessive prominence of the bursalprojection in the posterosuperior aspectof the calcaneous constitutes Haglund'sdeformity. Swelling in this areaconstitutes Haglund's disease and isassociated with retrocalcaneal bursitis.Rigid and prominent heel counters withhigh heels impinge on the soft tissuesoverlying the prominence and give riseto symptoms of pain and swelling. Cavovarus deformities exacerbate thisproblem. Bursal projection can bedemonstrated radiologically by asuperior calcaneal angle of more than75 degrees, a combination of calcanealinclination and a posterior calcanealangle of more than 90 degrees, andexcessive bone above the upperparallel pitch line (Stephens 1994).Pre- and postoperative radiographs areused to determine the amount of thecalcaneal osteotomy in cases ofHaglund's syndrome (Sella et al. 1998).

Computer tomography (CT) hasrelatively low diagnostic value ofAchilles tendon disorders (Reiser et al.1985, Ulrich et al. 1991). CT issensitive in detection of intratendinouscalcifications, and fractures ofossifications can also be diagnosed. CThas been used to measure volume ofcalf muscles after Achilles tendonrupture (Leppilahti 1996). CT appearsto be the imaging method of choice fordemonstrating monosodium uratedeposits in entheses and tendons ingout (Gerster et al. 1996).

14

Achilles tendon anatomy

Functional anatomy

The Achilles tendon constitutes thedistal insertion of the gastrocnemiusand soleus muscles. The tendon insertsinto a rectangular area on the middlepart of the posterior surface of thecalcaneus. The space between thetendon and calcaneal tuberosity is filledwith retrocalcaneal bursa. From thepoint of confluence, the tendon fiberstake a slightly spiral course, where theposterior fibers go from medial tolateral and anterior fibers from lateralto medial. The degree of rotation isvariable (Cummins et al. 1946). Thetendon is composed of 30% collagenand 2% elastin embedded in anextracellular proteoglycan matrixcontaining 58-70% water. Collagenfibres are grouped together in primary,secondary and tertiary bundles (Józsaand Kannus 1997). The tendon bundlesare surrounded by a mesh of looseconnective tissue, the endotenon,which holds the bundles together andalso carries the blood vessels,lymphatics and nerves (Kvist 1991).The Achilles tendon is covered by athin, smooth epitenon, and by theparatenon on dorsomediolateral sides.Paratenon consists of several thingliding membranes and forms a thinspace between the tendon and thecrural fascia. Crural fascia is thencovered by subcutaneous tissue andskin (Kvist and Kvist 1980). On ventralside, the paratenon consists of fattyareolar tissue and contains bloodvessels and connective tissue. Ventralto the Achilles tendon is a triangularpre-Achilles fat pad, also known asKager´s fat pad (Kager 1939).

An accessory soleus muscle can form asoft-tissue mass bulging mediallybetween the distal part of the tibia andthe Achilles tendon. It usually inserts

with a separate tendon on thecalcaneus anteromedial to the Achillesinsertion, and may be a cause of painon exercise (Romanus et al. 1986).

Achilles tendon receives blood vesselsin three regions: 1) at themusculotendinous junction, 2) throughthe paratenon surrounding the tendonand 3) at the osteotendinous junction.The paratenon has abundant bloodsupply (Carr and Norris 1989). Åströmand Westlin evaluated microvascularperfusion in the human Achilles tendonby laser Doppler flowmetry. Blood flowwas significantly lower near thecalcaneal insertion but otherwise wasdistributed evenly in the tendon(Åström and Westlin 1994). Naito andOgata studied the blood supply to thecentral third of the Achilles tendon inadult rabbits using the hydrogenwashout technique (Naito and Ogata1983). Their results indicated that thecentral third of the tendon with aparatenon receives its blood supplyfrom the extrinsic vascular system by35% and from the intrinsic vascularsystem by 65%.

Immobilization causes tendon atrophy,but because of a low metabolic rate ofthe tendon tissue the effects are slowand not as dramatic as in calf muscle(Karpakka et al. 1990). The rate ofmetabolism of collagen is relativelyslow, and there is normally a balancebetween synthesis and breakdown.During growth and following injury, thesynthesis exceeds degradation(Leadbetter 1992).

Normal MR appearance

The MR appearance of the intactAchilles tendon has been described ashomogenous, low signal intensitystructure on all sequences (Åström etal. 1996, Marcus et al. 1989, Neuholdet al. 1992, Quinn et al. 1987). Theimage of the Achilles tendon in the

15

sagittal plane appears as thin structuresurrounded by anterior high signalintensity fat pad on T1- and T2-weighted images. The antero-posteriorthickness measurements, signalintensity of the normal Achilles tendonand MR technical data are summarizedon Table 1. The cross-sectional shapeof the Achilles tendon has beendescribed as oval or ellipse (Marcus etal. 1989, Neuhold et al. 1992,Weinstabl et al. 1991) and the anteriormargin as flattened or mildly concave(Ferkel et al. 1991, Quinn et al. 1987).The posterior margin has always beendescribed as convex. All intratendinoussignals are usually regarded aspathological (Beltran and Mosure 1990,Cheung et al. 1992, Ferkel et al. 1991,Quinn et al. 1987). Recently, however,several authors have observedintratendinous signals in the Achillestendons of asymptomatic subjects(Åström et al. 1996, Movin et al. 1998,

Movin et al. 1998, Rollandi et al. 1995).Movin et al. reported high intensity,patchy signal alterations in the distalAchilles tendon in 3 of 25asymptomatic tendons on T1-weightedimages (Movin et al. 1998). Åström etal. had signal alterations in 2 of 14asymptomatic subjects (Åström et al.1996). Rollandi et al. showedintratendinous signals consisted oflongitudinally oriented straight lines(sagittal images) and spots (axialimages) on T1-weighted and protondensity –images. The signals werevisible in the distal portion of thetendon in normal, asymptomaticvolunteers (Rollandi et al. 1995).Accessory soleus muscle is a rarecondition which presents as a soft-tissue mass medial to the calcaneumand distal Achilles tendon. MRI of theankle shows characteristic findings of anormal muscle in an abnormal location(Palaniappan et al. 1999).

TABLE 1. SUMMARY OF TECHNICAL DATA AND MRI FINDINGS OF STUDIES WITH NORMAL

ACHILLES TENDONS.Signal intensity

Author Year N Pixel size(mm)

Slicethick(mm)

ShortestTE (ms)

AP-thickness(mm)

T1-WI T2-WI

Quinn etal.

1986 20 0.63 x 0.63 5 20 SE NA low low

Marcus etal.

1989 30 variable0.63 to 1.25

NA 39 SE NA low low

SartorisandResnick

1989 20 NA NA 35 SE NA low low

Weinstablet al.

1991 NA NA 5 30 SE 5 - 6 low low

Neuholdet al.

1992 7 0.63 x 0.63 5 30 SE 5.7-6.2 low NA

Rollandiet al.

1995 11 0.63 x 0.89 2 30 SE(PD) a

NA in-creasedb

low

Åström etal.

1996 14 0.78 x 0.78 5 30 SE 6 ± 1 12 low2 high

low

Movin etal.

1998 25 0.78 x 0.39 3 15 SE 6 22 low3 highc

low

FOV = field of view, TE = echo time, NA = not available, SE = spin echo, PD = proton densitya proton density -images only on axial planeb on proton density images in the distal tendonc on T-weighted gadolinium enhanced images

16

Achilles tendon rupture

Incidence and pathophysiology

Leppilahti et al. determined theincidence of a total Achilles tendonrupture in the city of Oulu (Finland) andchanges over the 16-year period 1979-1994. The incidence increased from 2ruptures/105 inhabitants in 1979-1986to 12 in 1987-1994. The peak annualincidence, 18, was recorded in 1994.Eighty-one percent of the ruptureswere related to sports, mainly ballgames (Leppilahti 1996). Similarincrease have been reported inDenmark and Sweden (Levi 1997,Moller et al. 1996).

The exact pathogenesis of the Achillestendon rupture remains obscure. Mostfrequently discussed theories involvechronic degeneration of the tendon andfailure of the inhibitory mechanism ofthe musculotendinous unit (Arner andLindholm 1959, Barfred 1971, Barfred1971, Jozsa et al. 1990). The repetitivemicrotrauma may producemicroruptures, focal or diffuse tendondegeneration and inflammation.Regeneration is unable to keep pacewith the recurring microtrauma due tothe relatively poor vascular supply inthe midportion of the tendon, andfinally the tendon is weakenedsufficiently for total rupture to occur(Lagergren and Lindholm 1958).

Diagnosis and treatment

Diagnosis of the total Achilles tendonrupture is missed about 20% of thetime, thereby leading to delay intreatment (Carden et al. 1987, Inglis etal. 1976). The history of the trauma isoften inconsequential, and both thephysician and the patient tend todisregard it. Insignificant pain and theability to weakly plantar flex the foot

are additional reasons for misdiagnosis(Carden et al. 1987, Inglis et al. 1976).A delay in diagnosis and surgicaltreatment of longer than one monthwill downgrade the result of surgicaltherapy by at least 20%. Therefore,diagnosis and treatment should bestarted as soon as possible (Inglis et al.1976).

Nistor concluded that non-surgicaltreatment offers advantages, such aslower complication rate, over surgicaltreatment (Nistor 1981). Thecomplications of conservative treatmentinclude mostly reruptures and residuallengthening of the tendon, which mayresult in significant calf muscleweakness. In a review, covering 3245surgically and 437 non-surgicallytreated patients the rerupture rate wassignificantly higher in non-surgicalgroup (1.6% vs. 11%, respectively)(Leppilahti and Orava 1998). Thecomplication rate of major surgicalcomplications (deep wound infection,skin necrosis, deep vein thrombosisetc.) has varied from 1.9 % to 5.4%(Cetti et al. 1993, Leppilahti and Orava1998). Almost always after operativetreatment of Achilles tendon rupture asmall group of patients feel subjectiveunsatisfaction and report minor clinicalsymptoms at the Achilles tendon region(Arner and Lindholm 1959, Carden etal. 1987). Although conservativetreatment has its own supporters,surgical treatment seems to have beenthe method of choice in the late 1980’sand the 1990’s in athletes and youngpeople, and in cases of delayedruptures (Leppilahti and Orava 1998).

There are over 20 different surgicaltechniques described in the literature toreunite the Achilles tendon rupture(Cetti et al. 1993). No prospective,randomized clinical study comparingthe simple suture technique and repairwith augmentation could be found in

17

the literature (Cetti et al. 1993). Simpleend-to-end, gastrocnemius flap andplantaris tendon reinforcementtechniques described by Bunnell,Lindholm and Lynn are widely used inFinland (Bunnell 1940, Lindholm 1959,Lynn 1966).

Rehabilitation

Open surgical repair is followed by aperiod of cast immobilization generallylasting 6-8 weeks in a below-knee cast(Cetti et al. 1993, Waterston et al.1997). The foot is immobilized in fullequinus for 3-4 weeks and in thesemiequinus or neutral position foranother 3-4 weeks. Active range ofmotion exercises, stretching indorsiflexion, walking, cycling andswimming can be started after castremoval. Jogging is allowed after 4 to 6months, and return to competitivesports is allowed after 6 months(Carden et al. 1987, Landvater andRenstrom 1992). The average time forreturn to sport activities after a surgicalrepair of the Achilles tendon rupturehas been reported to vary from 6.5 to9.1 months (Beskin et al. 1987, Kellamet al. 1985).

Limited immobilization and earlyfunctional rehabilitation has beensuggested recently by manyresearchers (Carter et al. 1992,Mandelbaum et al. 1995, Motta et al.1997, Troop et al. 1995). Immediatemovement of the foot and ankle helpsthe remodeling of scar tissue, at thesame time inhibiting the formation ofskin adherences that can later interferewith full movement of the joint. (Mottaet al. 1997). Sölveborn and Mobergreported 100% excellent or goodresults after immediate free anklemotion after surgical repair of theAchilles tendon (Solveborn and Moberg1994). They used a 6-weekpostoperative plaster cast with aprotecting frame under the foot making

weightbearing possible (Solveborn andMoberg 1994). Troop et al. achievedgood return of plantar flexion strength,power, and endurance afterrehabilitated with early motion startingan average of 10 days after surgery.Active range of motion begun at anaverage of 23 days and weightbearingin a walking boot started at an averageof 3.5 weeks after surgery (Troop et al.1995). Mandelbaum et al. had noreruptures and excellent recovery ofisokinetic strength and endurance ofthe calf muscles in 29 operativelytreated ruptures (Mandelbaum et al.1995). Patients began range-of-motionexercises 72 hours after surgery, useda posterior splint for 2 weeks, and thenbegan ambulation in a hinged orthosis.Six weeks after surgery, use of theorthosis was discontinued, fullweightbearing was allowed, andprogressive resistance exercises wereinitiated. All patients returned topreinjury activity levels at a mean of 4months (range, 3 to 7) after repair(Mandelbaum et al. 1995).

MR appearance

Reinig et al. (1985) first published acase report in which MR imaging wasused to diagnose Achilles tendonrupture (Reinig et al. 1985). Sincethen, several authors have describedthe rupture site as an inhomogenous,increased signal intensity area on bothproton density and T2-weighted images(Keene et al. 1989, Quinn et al. 1987).As a sign of acute hemorrhage,intermediate signal intensity is typicallyseen at the rupture site on T1-weightedimages. Marked soft tissue swelling andthickening of the tendon is also present(Daffner et al. 1986, Weinstabl et al.1991). The level of the tear issometimes indicated by the large areaof moderate signal intensity on T1-weighted images in the pre-Achilles fatpad (Keene et al. 1989). The proximaltendon end of the ruptured Achilles

18

tendon is often more frayed than thedistal end, and high signal intensitystripes on T2-weighted images aredetected in-between the tendon ends(Keene et al. 1989). In subacuterupture some high signal intensitybands can be detected on T1-weightedimages due to old hemorrhage.Generally, T1-weighted images do notclearly demonstrate the definite rupturesite (Keene et al. 1989, Quinn et al.1987). Sometimes a localized(subacute) hematoma with high signalintensity on T1- and T2-weightedimages is seen surrounded by lowsignal halo of hemosiderin (Neuhold etal. 1992).

Achilles tendon overuseinjuries

Tendinosis

Achillodynia is often used as a generalterm for all painful conditions of theAchilles tendon (Schmitgen andHaasters 1990). Occasionally this termmeans Achilles tendon pain syndromesfor which no objective findings oretiology have been found. Typically,symptoms prohibit activity and clinicalexamination reveals swelling andtenderness 1.5-6 cm proximal to theAchilles tendon insertion. Tendinosis(focal degeneration), sometimescomplicated by partial rupture, appearsto be the major finding in chronicAchilles tendinopathy (Åström andRausing 1995). Important features area lack of inflammatory cells and a poorhealing response (Åström 1998, Åströmand Rausing 1995, Puddu et al. 1976).Åström et at studied 342 Achillestendons in 298 patients which wereoperated on for painful chronic Achillestendinopathy (81% men; mean age 35years; 79% athletes). A partial rupturewas found in 23%, tendinosis(degeneration) in 49% and no

macroscopic pathology in 28% of thetendons. In partial ruptures, ascompared with non-ruptured tendons,the lesion was more common in thedistal part of the tendon and morefrequent in physically active menslightly below middle age who hadreceived local steroid injections beforesurgery (Åström 1998). It ishistologically documented thatdegeneration of the Achilles tendon canexist in the absence of clinicalsymptoms but can becomesymptomatic with heavy trainingleading to localized pain andtenderness (Puddu et al. 1976).

Treatment of Achilles tendon overuseinjuries is initially conservative. Thebasic conservative treatment strategiesare (modified) rest, cryotherapy,rehabilitation of triceps surae muscle-tendon unit (stretching andstrengthening) and control ofbiomechanical variables (Clement et al.1984). Alfredson et al. reported verygood results after heavy-load eccentriccalf muscle training on athletes in theirearly forties (Alfredson et al. 1998).The results of 8-year follow-up byPaavola et al. showed that the long-term prognosis of patients with acute tosubchronic (symptoms < 6 months)Achilles tendinopathy was favorable asdetermined by subjective andfunctional assessments. In the clinicaland ultrasonographic examinations,mild to moderate changes wereobserved rather frequently in both theinvolved and initially uninvolvedAchilles tendons, but the occurrence ofthese changes was not clearly relatedto patients’ symptoms (Paavola et al.2000). However, chronic overuseinjuries do not always respond toconservative treatment. Surgery isoften needed in the final stage of anoveruse injury. In a large series byOrava et al., chronic Achilles tendonproblems formed 23% of all operativelytreated chronic symptomatic overuse

19

injuries (Orava et al. 1991). Leppilahtiet al. reported good results in 69% ofthe operatively treated Achilles tendonoveruse injuries (Leppilahti et al.1991). In a large series by Paavola etal., every 10th patient treated surgicallyfor chronic Achilles tendon overuseinjury suffered from a postoperativecomplication that clearly delayedrecovery (Paavola et al. 2000).

Insertional disorders

Retrocalcaneal bursitis is a distinctentity denoted by pain that is anteriorto the tendon, just superior to itsinsertion on the os calcis. The bursa,which lies between the anterior aspectof the tendon and posterior aspect ofthe os calcis, becomes inflamed,hypertrophied, and adherent to theunderlying tendon (Clain and Baxter1992, Lehto et al. 1994, Schepsis et al.1994).

Patients with insertional tendinitis havedirect tenderness over the Achillestendon insertion. In some cases, theinflammatory changes within thetendon may be seen in conjunction withretrocalcaneal bursatis (Schepsis et al.1994). According to Clain and Baxter,insertional tendinitis involves theadjacent bursa along with changes inthe tendon, including thickening,calcification, and fraying (Clain andBaxter 1992). In insertional disorders,conservative treatment is usuallysuccessful. In cases of continuedsymptoms (typically a period over sixmonths from initial onset of symptoms)surgical intervention is warranted(Schepsis et al. 1994). In surgery, theinflamed bursa is completely excised,and the postero-superior angle of theos calcis is removed (Schepsis et al.1994).

The pain syndromes in the insertionarea are occasionally called “Haglund´ssyndrome” (Pavlov et al. 1982). An

excessive prominence of the bursalprojection in the posterosuperior aspectof the os calcis constitutes Haglund'sdeformity. Rigid and prominent heelcounters with high heels impinge onthe soft tissues overlying theprominence and give rise to symptomsof pain and swelling. The results ofsurgery are satisfactory, providedadequate bone has been resected(Stephens 1994). However, also poorclinical outcomes have been publishedafter osteotomy of the posterosuperiorcorner of os calcis (Nesse and Finsen1994).

Peritendinitis

Peritendinitis is called paratenonitis bymany authors. A crepitating conditionin the paratenon is called “peritendinitiscrepitans”. With acute Achillesperitendinitis, an inflammatory cellreaction, edema, extravasation ofplasma proteins, and accumulation offibrin are seen in the paratenon. Inchronic cases, thickening of theparatenon, areas of proliferatingconnective tissue, formation ofadhesions, and obliterative changes inblood vessels are found. Paratenonbecomes adherent to underlying tendon(Kvist et al. 1985, Kvist et al. 1987,Kvist et al. 1988). Pain may beexperienced anywhere around Achillestendon, but most often in the middlethird. Palpable tenderness is typicallyfound at the sides of the tendon (Lehtoet al. 1994). Frequently there aretender nodules around the Achillestendon with chronic peritendinitis aswell as diffuse or focal thickening ofsubcutaneous tissue (Kvist and Kvist1980). Typically peritendinitis co-existswith tendinosis (Clement et al. 1984,Galloway et al. 1992, Kvist 1994,Schepsis et al. 1994). Clinically, it isvery difficult to distinguish tendinosisfrom paratenonitis unless palpationreveals nodules characteristic fortendinosis (Kvist 1994). Acute

20

peritendinitis is treated basically astendinosis, but heparin and non-steroidal anti-inflammatory drugs havebetter effect on acute than on chronicphase (Kvist and Kvist 1980, Nichols1989). In chronic cases operativetreatment include crural fasciotomy,removal of adhesions, and liberation ofclearly hypertrophied portions of theparatenon (Kvist and Kvist 1980,Schepsis et al. 1994). The results havebeen better in the operative treatmentof chronic peritendinitis than in otherAchilles tendon overuse injuries (86%vs. 69%, respectively) (Leppilahti, etal. 1991).

MR appearance

In patients with chronic Achilles tendondisorders, the MR imaging is oftendescribed in terms of enlargement andswelling on sagittal images, and alteredsignal appearance within the tendontissue (Åström et al. 1996, Marcus etal. 1989, Neuhold et al. 1992,Weinstabl et al. 1991). On cross-sectional images, the tendon loses itsnormal lenticular shape and becomesovoid and, with severe disease,rounded (Schweitzer 1993). Weinstablet al. introduced a classification ofAchilles tendon disorders of 20 patientson MR images as follows: 1)inflammatory reaction; thickenedtendon without structural changes oftendon tissue, 2) degenerativechanges; thickened tendon withlongitudinal and centrally locatedchanges that did not reach tendonsurface, and 3) incomplete rupture;thickening of the tendon with structuralchanges longitudinally and horizontallyextending to paratenon (Weinstabl etal. 1991). Neuhold et al. published thesame 20 patient material, but did notdifferentiate between inflammatoryreaction and degenerative changes,instead the term achillodynia was usedin cases of centrally locatedintratendinous lesions (Neuhold et al.

1992). Incomplete rupture was judgedin five cases of intratendinous lesionsthat extended to tendon surface. Theseintratendinous lesions had intermediatesignal intensity on T1-weighted imagesand moderate signal increase on T2*-gradient echo images (Neuhold et al.1992). Some case reports of chronicAchilles tendon disorders have alsoshown similar thickening of the tendonwith or without intratendinousabnormalities (Bonner et al. 1990,Marcus et al. 1989, Quinn et al. 1987).

Åström et al. took histological samplesduring surgery from the intratendinousabnormalities. A large and severeintratendinous lesion with sagittaldiameter >10mm suggested partialrupture. The location of the lesion wasno criteria for partial rupture (Åströmet al. 1996). Similarly, Khan et al.showed that intratendinous lesions ofthe patellar tendon had loss of cleardemarcation of collagen bundles,increased noncollagenous extracellularmatrix and capillary proliferation (Khanet al. 1996). Movin et al. comparedgadolinium enhancement of T1-weighted images with conventional T1-,proton density, and T2-weightedimages. Gradient echo images were notused in this study. Gadoliniumenhancement improved the imaging ofintratendinous signal abnormality onT1-weighted images. Histologicalsamples showed an increased non-collagenous extracellular matrix andaltered fiber structure in the lesions(Movin et al. 1998). It has beensuggested by many authors thatlocalized degenerative changes mayprogress to partial ruptures (Åström1998, Åström and Rausing 1995,Panageas et al. 1990, Schweitzer 1993,Schweitzer and Karasick 1994). Highsignal intensity of the intratendinouslesion seen on T2-weighted images issuggested to indicate interstitial tear(Schweitzer and Karasick 1994).

21

On MR imaging, the asymptomaticretrocalcaneal bursa normally containsdetectable high signal intensity fluid orsynovium or both. A bursa larger than1 mm anteroposteriorly, 11 mmtransversely, or 7 mm craniocaudally isabnormal (Bottger et al. 1998). Thereare no prospective MR studies ofinsertional tendinosis or peritendinitis inthe literature. Schweitzer hassuggested that peritendinitis isdiagnosed on MR images of the Achillestendon if high signal is seen around thetendon on T2-weighted images. Thisabnormality is medially predominateddue to common overpronation of thefoot (Schweitzer 1993). Edema in thepre-Achilles fat pad may be seen onfat-suppressed images (Beltran et al.1987, Schweitzer 1993). Evaluation ofthe paratenon is not possible onconventional MR images (Åström et al.1996).

There are no sequential or longitudinalstudies concerning the intratendinousMR changes. In a case report, anintratendinous lesion seen on MRimages had been disappeared in afollow-up MR study eight months later(Nicolaisen et al. 1997).

Other causes of Achillestendinopathy

Rheumatoid tendinopathy can bedistinguished from degenerativetendinopathy in patients with chronicpain of the heel with MR imaging.Inflammation of the retrocalcanealbursa and the absence of enlargementof the tendon combined with the

presence of intratendinous signalalterations are characteristic findings ofrheumatoid tendinopathy (Stiskal et al.1997).

MR imaging is a more sensitive methodthan physical examination for detectingabnormalities in Achilles tendons ofpatients with hyperlipidemia. MRpattern of xanthomas is often differentfrom that of partial tendon tears, ortendinosis, although the tendons aretypically diffusely enlarged as intendinosis. Abnormal MR signal ofxanthomas has a diffuse stippledpattern with many low-signal roundstructures of equal size surrounded byhigh-signal material on all pulsesequences. Also, xanthomas are seenbilaterally. MR imaging of patients withfamilial hyperlipidemia shows signalpattern with or without enlargement orabnormal configuration of the tendon.Although the MR signal pattern ofxanthomas is often different from thatof partial tendon tears, tendondegeneration, or tendinitis, a significantoverlap in appearance can be observedand the MR appearance of a xanthomais not always pathognomonic (Dussaultet al. 1995). US offers a sensitive andcost effective method in detectingAchilles tendon xanthomas (Koivunen-Niemelä 1995).

Achilles tendon pain or rupture afterfluoroquionolone treatment has beendescribed as an uncommon adverseeffect. Fluoroquinolone-inducedtendinopathy appears more commonlyin tendons under high stress such asAchilles tendon (Movin et al. 1997).

22

THE AIMS OF THE PRESENT STUDY WERE

1. To monitor MR imaging findings during the healing process of thesurgically repaired Achilles tendon ruptures and to correlate the findingswith clinical restoration (Paper I).

2. To evaluate and to compare MRI and US findings on an unselectedgroup of patients with one to three years old surgically repairedcomplete Achilles tendon ruptures (Paper II).

3. To describe the normal appearance of the Achilles tendon andperitendinous tissues in asymptomatic active volunteers using high-resolution MR imaging (Paper III).

4. To describe and classify MR imaging findings related to overuse injuriesmanifesting as a painful Achilles tendon (Paper IV).

5. To compare MR imaging findings with clinical, surgical, histopathologicalfindings, and long-term outcomes (Paper IV).

23

SUBJECTS, MATERIALS AND METHODS

The research was performed at theDepartment of Radiology (I-IV) atHelsinki University Central Hospital,Helsinki, Finland in co-operation withthe Department of Orthopaedics andTraumatology (I-II), the Department ofPathology (I, IV) and the DeaconessHospital in Helsinki (IV).

Subjects

Paper I consisted of twenty consecutivepatients (16 men and 4 women; meanage 37 years; range 33-56 years) witha complete Achilles tendon rupture.They were treated surgically at theDepartment of Orthopaedics andTraumatology, Helsinki UniversityCentral Hospital between January andMay 1994. Five men (20%) hadpreviously had an Achilles tendonrupture on the contralateral side andonly three patients (15%) had previousknowledge of Achilles tendon symptomson affected side. The total number oftendons were 21 because one patienthad a rupture on contralateral side fivemonths apart.

Paper II consisted of thirteenunselected male patients with asurgically repaired, complete Achillestendon ruptures. They were reviewed ina follow-up study on the average 17.7months (range, 12 to 36) post-operatively including clinical, MRI andUS examinations. The mean age of thepatients at the time of the injury was36 years (range 24 to 49 years). Inaddition, 4 of the 13 patients also had asubsequent Achilles tendon rupture onthe contralateral side, but at the timeof the follow-up examination the age ofthat rupture was less than 1 year.Therefore, only 9 patients areconsidered to have a healthy uninjuredcontralateral tendon for comparison.

Paper III consisted of 100asymptomatic Achilles tendons. Eighty-one volunteers (61 men and 20women) underwent high-resolution MRIof their asymptomatic Achilles tendon.The average age was 32 years (range,15 to 56). All volunteers wereparticipants in competitive orrecreational sports, which includedrunning or track and field (n=64),soccer or basketball (n=8), ballet ordancing (n=7), and tennis (n=2). Allpatients with any systemic disease thatmight affect the Achilles tendon such ashypercholesterolemia were excluded(Dussault et al. 1995). Of the 81volunteers, 62 asymptomatic tendonswere imaged in patients who hadclinically manifest Achilles tendonoveruse injury on the contralateral side(they were included in Paper IV).Additionally, 38 tendons in 19volunteers who had never had Achillestendon symptoms were imaged.

In Paper IV, 100 patients (75 men and25 women) with 118 painful Achillestendons were prospectively evaluated.The time period from the initial onset ofsymptoms until the MR imaging of theAchilles tendon was, on average, 18weeks (range, 6 to 160 weeks).Patients with acute injury or suddenonset of symptoms were excluded fromthe study. The average age was 33years (range, 15 to 58). All patientswere involved in competitive orrecreational sports, and were referredfor MR imaging from sports medicineclinics by orthopaedic surgeons. 75patients participated in running or trackand field events. Other sports includedsoccer (n=9), ballet or dancing (n=6),tennis (n=4), walking (n=2), skating(n=2), bicycling (n=1), and wrestling(n=1). The duration of symptoms andany previous history of Achilles tendon

24

disorders were registered. Of the 82patients with unilateral Achilles tendondisorder, 20 were excluded because ofprevious history of Achilles tendonsurgery or disorder on the contralateralankle, and therefore imaging of theasymptomatic side was performed in 62available cases. These 62 cases wereincluded in Paper III as asymptomaticsubjects.

Clinical evaluation

In Papers I - IV, clinical examinationincluded careful palpation of theAchilles tendon region and registrationof the level of maximal pain,tenderness, and focal thickening. InPaper III, all cases with anytenderness, nodularity or thickeningwere excluded from the study. In allstudies the clinical examination wasperformed on the same day as MRstudy.

Conservative treatment

All patients in Paper IV were firsttreated conservatively. Conservativetreatment included modified orcomplete rest for 3-12 weeks, heel liftsof 10-15 mm, cryotherapy, non-steroidal anti-inflammatory drugs,static stretching, and eccentric andconcentric strength training.Corticosteroid injections were rarelyused (n=7 patients in Paper IV after MRimaging).

Surgical treatment

Indications

In Papers I and II, the preoperativediagnosis of total Achilles tendonrupture was based on positiveThompson test, palpable defect, local

tenderness and swelling. In twouncertain cases (Paper I) apreoperative MRI was performed toconfirm the total rupture. All patientswere operated on within 48 hours fromthe injury as an emergency procedureby surgeon on duty, and the diagnosiswas verified at surgery.

In Paper IV, the indication for surgerywas long-standing, persistent pain ofthe Achilles tendon, which did notrespond to conservative treatment. Thetime interval from the initial onset ofsymptoms to the surgery was onaverage 10 months (range 6 to 25).The detection of the thickenedparatenon, as well as the largeintratendinous lesion on MR imagescontributed towards surgicalintervention.

Surgical techniques

In Papers I and II, the surgicalmethods included Lindholm,tendorrhaphy and Lynn techniquesdepending on the preferences of thesurgeon (Tables 3 and 4) (Lindholm1959, Lynn 1966). In all operationsabsorbable polyglyconate (typically No.0 Maxon; Davis & Geck, Gosport,United Kingdom) intratendinous sutureswere used.

In Paper IV, 28 patients did notrespond to conservative treatment, andeventually underwent surgery. Thesurgeons (all specialized in sportsmedicine) were preoperatively informedof the MR imaging findings. Operativetreatment included crural fasciotomy,liberation of tissue elements adherentto one another, removal ofhypertrophic paratenon, excision oftendon lesions, and when necessaryexcision of inflamed retrocalcanealbursae and prominent superiorcalcaneal tuberosity.

25

Surgical evaluation

In Paper IV, thickening of theparatenon; thickening, consistency, andfiber structure of the tendon; andpresence of intratendinous lesions wererecorded during surgery. Signs ofbursitis were recorded when theretrocalcaneal bursa had to beexposed. Paratenon was consideredthickened if redundant and hyperemictissue was detected.

Postoperative rehabilitation

In Papers I and II, post-operatively theankle was immobilized for three weeksin equinus cast without weight bearingand for another three weeks in a shortwalking cast in neutral position duringwhich weight bearing was graduallyallowed. After cast removal the patientsstarted active range of motionexercises of the ankle and walkingexercises. The patients were instructedto gradually resume activities, but theywere not allowed unlimited activity untilsix months after the surgery. In PaperIV, our postoperative program was agraduate return to sports, usually notuntil 4 to 6 months. In none of theoperated patients we used castimmobilization. During the firstpostoperative weeks patients typicallyhad limited weight-bearing, and theywere encouraged to range of motionexercises, especially in dorsiflexion.

Complications

In Paper I, three re-operations wereperformed because of persistent painand limping, and large intratendinouslesion in MRI (two cases) or anaccidental re-rupture (one cases).

Clinical follow-up andscoring

In Paper I, all patients were seen atthree and six weeks and underwent acomplete objective and functionalevaluation at three and six monthsfollow-up by an orthopaedic surgeon.To assess the subjective recovery,patients were asked for stiffness, pain,weakness and ability to walk and run.Clinical examination included theassessment of the appearance of thewound scar, measurement of range ofmotion in active dorsiflexion andplantarflexion of the ankle, patient’sability to perform 3 cm heel raises ontoes (pacing with a metronome 60 perminute). At six months the adhesiontendency between tendon and skin wasassessed by palpation. Walking abilitywas evaluated by physiotherapeutistwithout knowledge of the imagingfindings. He classified the patients intotwo groups: normal and abnormal.Based on clinical evaluation the resultswere classified into three categories atsix months using a scale presentedoriginally by Arner and Lindholm andrecently by Sölveborn et al. (Arnerand Lindholm 1959, Solveborn andMoberg 1994). We modified the scaleby adding the rise-on-toes test resultsin the scale, as follows:

1. Excellent: normal clinical findings,i.e. normal walking, ability to riseon toes over 30 times or the equalamount as contralateral leg, anklemobility that was normal ordecreased by at most 5° indorsiflexion and/or plantarflexion.

2. Good: mild discomfort, normal orslightly hampered walking, abilityto do at least 10 toe-risings,restricted mobility of less than 10°in dorsal and/or 15° in plantardirection.

3. Poor: patient was dissatisfied orhad marked discomfort, incomplete

26

wound healing, clearly abnormalwalking (i.e. limp), patient haddifficulty in rising on toes and hadrestricted ankle mobility of morethan 10° in dorsal and 15° inplantar direction.

In Paper II, patients underwent similarsubjective and objective evaluation atthe time of MR study. At classificationexcellent and good categories werecombined as follows:

1. Good: patient declared himselfsatisfied, had no or mild discomfort,normal or slightly hamperedwalking, normal or only slightlydecreased power of calf muscles,ability to rise repeatedly on toes,normal or restricted mobility of lessthan 10° in dorsiflexion and/or 15°in plantarflexion.

2. Poor: patient was dissatisfied orhad marked discomfort, incompletewound healing, clearly abnormalwalking (i.e. limp), could not rise ontoes or repeat rising on toes atleast once, had restricted anklemobility of more than 10° in dorsaland/or 15° in plantar direction.

In Paper IV, for long-term follow-up, allpatients were interviewed using acomprehensive questionnaireconcerning functional activity, painassociated with athletic activity, painduring daily activities, and ability toparticipate in sports. Outcome wasclassified according to the modifiedevaluation method of Schepsis et al.(1994). Full return to sportscomparable with the preinjury statuswas considered an excellent result. Agood result was considered to be returnto sports with only intermittent or milddiscomfort, and persisting mildlimitation of range of motion. A fairresult was scored if discomfort did notallow return to preinjury level, and thusdictated cessation of competitive sportsactivity or mandated a change in the

form of the recreational sport activity.Result was considered poor if patientshad given up all sports activities andhad pain during activities of daily living.

Magnetic resonanceimaging

MR protocols

In Papers I, all MR images wereobtained with a 0.1-T resistive Merit(Picker Nordstar Inc., Helsinki, Finland)scanner at three and six weeks, andthree and six months post-operatively.In Paper II, the MR images wereacquired 17.7 months (range, 12 to 36)post-operatively with the 0.1T scanner.At three weeks the MR study wasperformed supine without cast removaland thereafter without cast. Separatetransmitter and surface receiver coilswere used. In all images a 192 x 256matrix and field of view of 192 x 256mm were used which gives 1 mm inplane resolution. Gradient echo T1-(repetition time (TR) of 225 ms, andecho time (TE) of 18 ms) and T2-weighted (TR = 1500, TE = 55) sagittalimages with 3 mm slice thickness wereacquired. In axial plane, double-echoproton density- (PD-) and T2-weightedimages were obtained (TR = 2000, TE= 18 and 80, respectively) with 8 mmslice thickness and without an interslicegap. The total scanning time was 30minutes. In all studies the tibiotalarjoint space was used as an easilydetectable reference for slicepositioning.

In Papers III-IV, MR imaging of theAchilles tendon was performed with a1.5-T imager (Vision, Siemens MedicalSystems, Erlangen, Germany) using astandard circular polarized (CP)-extremity coil supplied by Siemens. Inthis prospective study we obtainedsagittal T1-weighted spin-echo images

27

(460/14) and fast short inversion timeinversion recovery (STIR) images(4000/30). The sagittal images wereplaced along the long axis of thetendon utilizing a coronal scout viewwhen necessary. Sagittal sectionthickness was 3 mm with a gap of 0.3mm, and the field of view (FOV) wasrectangular 113 x 180 mm with a 182 x256 matrix. In the axial plane weacquired high-resolution Tl-weightedspoiled gradient echo (FLASH) imageswith a short echo time (600/10), flipangle 90°, bandwidth 78 kHz, 113 x180 mm rectangular FOV and 211 x512 matrix (pixel size, 0.54 x 0.35mm). We utilized the postero-anteriordirection in frequency-encoding in orderto place the low intensity component ofthe chemical shift artifact adjacent tothe paratenon on the dorsal aspect ofthe tendon and in order to keepsymmetry of the artifact in themediolateral direction (Soila et al.1984). This enabled us to acquirebetter images of the thin layersbetween the Achilles tendon and theposterior skin. We also obtained axialfast STIR- (4700/30) and dualconventional spin-echo (2100/20/80)(Paper IV) –images with 113 x 180 mmrectangular FOV and 154 x 256 matrix.All axial sequences had 17-19 sections,4-mm section thickness, 1.0-mmintersection gap. Total imaging time forthis protocol was 12 (without dual spin-echo sequence) to 17 minutes.

MR image analysis

In Papers I and II, the thickest antero-posterior and transverse dimensions ofthe Achilles tendon 2, 6 and 10 cmabove the postero-superior corner ofthe calcaneus (not the insertion) weremeasured. All dimensions werecompared with the contralateral tendonexcluding the patients previouslyoperated. We also measured thelargest cross-sectional area of the

tendon and the size of intratendinouslesion visible within the opposed tendonends on PD- and T2-weighted imagesusing the computer of the imagingsystem. The percentage of this regionof the total cross-sectional area of theAchilles tendon was calculated. Theintensity of the intratendinous lesionwas visually evaluated into twocategories, intermediate or high. Wealso evaluated the contour and theover-all signal intensity of the tendonand the appearance of the surroundingperitendinous tissues. The generalsignal pattern of the contralateraltendon was evaluated as well.

The high field (1.5T) MR images wereindependently interpreted by twomusculoskeletal readers. Images werefirst evaluated independently, if the twointerpreters did not fully agree, aconsensus was reached. Thesymptomatic and asymptomatic legswere examined.

The anteroposterior diameter of theAchilles tendon on axial FLASH imagesand its distance from the postero-superior corner of os calcis on sagittalimages was measured. Inasymptomatic cases discussed in PaperIII, the width was recorded at astandard level of 3-cm proximal to thelevel of the posterosuperior corner ofos calcis, because there was a greatvariance in width related to the level ofmeasurement. The length of thetendon, from insertion to the level ofthe distal end of the soleus muscle wasassessed on sagittal images.

The anterior margin of the tendon wasevaluated on axial FLASH images as flat(=0), moderately (=1) or severely (=2)convex, and the anterior contour onsagittal images as normal or bulging. Inasymptomatic cases presented in PaperIII, the anterior contour on axialimages was evaluated on a workstation in “cine” mode. The tendon was

28

diagnosed as abnormal if the tendonwas thickened (>6 mm in antero-posterior diameter or ≥2-mm thickerthan the asymptomatic side), had aconvex anterior margin on axial imagesor anterior bulging on sagittal images,or showed a focal intratendinous lesionlarger than 3 mm in size on axialFLASH images.

The overall signal intensity of theAchilles tendon was evaluated on axialFLASH images as homogeneous orinhomogeneous. If the signal intensityof the Achilles tendon wasinhomogeneous, the signalabnormalities were further classified asfollows:

1. Intratendinous intermediate signalintensity strands typically located inthe distal part of the tendon,

2. Punctate or patchy intermediatesignal intensity intratendinous focialong the tendon,

3. Diffuse ground-glass typeintermediate signal intensityintratendinous lesions that couldnot be judged as punctates.

The level, size (percentage of thecross-sectional area and height), andlocation (deep versus surface) of thediffuse intratendinous lesion wasregistered on axial FLASH images. Thesignal intensity of the lesion wasfurther analyzed on PD-, T2-, and STIRimages as having low (or no),intermediate, or high signal intensity.

Other recorded findings were the sizeand intensity of the retrocalcanealbursae and the signal intensity of thecalcaneal marrow deep to at theAchilles insertion on STIR images.Bursae larger than 11 mm transversallyor 7 mm craniocaudally was consideredabnormal (Bottger et al. 1998).

The peritendinous soft tissuesevaluated included the pre-Achilles or

Kager´s fat pad (anterior to theAchilles tendon and posterior to deepflexor tendons) (Kager 1939).Thickening of the paratenon on axialFLASH images and signal intensity ofthe paratenon and Kager´s fat pad onSTIR images were subjectivelyevaluated. Signal intensity andthickness of the paratenon were gradedusing a scale of 0 (normal), 1(moderate) or 2 (severe).

In asymptomatic cases discussed inPaper III, the plantaris tendon in theanteromedial margin of the Achillestendon was identified.

Ultrasonography studies

In study II, in US examination, thepatient was placed in the prone positionwith feet hanging free over the edge ofthe table and the ankles in neutralposition. A real-time US scanner(Toshiba SS 270A, Tokyo, Japan) witha linear-array transducer of 7,5 MHzwas used. A standoff gel pad wasplaced on the skin in order to betterimage the Achilles tendon in theoptimal focal zone of the transducer.Perpendicular transverse andlongitudinal scans were taken from thesurgically repaired Achilles tendon, aswell as from the contralateral tendon.Delineation and echogenic properties ofthe tendon were assessed. The greatestantero-posterior dimension andtransversal width of the Achilles tendonat levels of 2 and 6 cm cranial from thepostero-superior corner of thecalcaneus were measured.

29

Histopathological studies

In three re-operations in Paper I, thehistological samples of theintratendinous lesions were obtained.At surgery, it was possible to obtainspecimens for histologic examinationfrom 13 intratendinous lesion in PaperIV. Hematoxylin-eosin and van Giesonstains were used. Pathologic changesincluded derangement of collagenfibers, lack of collagen stainability,round-shape of tenocyte nuclei, andincrease in vascularity (Åström andRausing 1995). Disturbance of parallelfiber bundles and discontinuity of fiberswere noted when evaluating fiberarrangement. Regarding collagenstainability, the deep red staining oftightly packed, mature collagen fibersin van Gieson staining was consideredas normal. The red staining wasreduced in areas of scarring and newlyproduced collagen. Amount of round-shaped nuclei and any increase in thenumber of nuclei were noted andclassified as pathological changes.Deviation from normal vascular densityof tendon tissue and capillaryproliferation were estimated andgraded.

Statistical methods

In statistical analysis the mean ± SD oftendon dimensions were calculated. Forinterobserver variability (III, IV), allestimated categories were calculated withthe use of Cohen kappa values. κ valuesgreater than 0.60 indicated goodagreement. κ values were also calculated foragreement between MR imaging andmacroscopic evaluation of the operativesurgeon (IV).

A paired T-test was used to assess thedifference between Achilles tendondimensions of the left and right legs in the19 healthy volunteers (III).

The size of high intensity lesion inside thetendon and patients with normal and

abnormal walking, as well as the size of thetendon in patients operated with Lindholmplasty and tendorrhaphy were compared byusing the Mann-Whitney u-test (I).

The operated and non-operated tendonareas in the 9 patients with uninjuredcontralateral tendon, as well as the size ofthe intratendinous lesion on PD- and T2-weighted images were compared usingWilcoxon signed rank test (II).

The correlation between MRI and US studies(II), and between two observers (III) inmeasuring anterior-posterior dimensionswere calculated with regression analysis(III). Regression analysis was also used toinvestigate the correlation between the sizeof intratendinous lesion on PD- and T2-weighted images and the percentage ofscore on toe-risings test compared tounaffected side of the same patient, andbetween the size of lesion and the range ofmotion tests (I). For correlation between thelevel of peritendinous abnormal signalintensity on STIR images and the location ofclinically-observed maximal tenderness, aswell as between the intratendinous lesionand the clinically-observed area ofthickening we used regression analysis (IV).

Analysis of variance was used to comparethe size of the largest cross-sectional area ofAchilles tendon between each follow-upperiod (I), the size of the intratendinouslesion detected only on FLASH images and inall other sequences (IV), and the AP-dimensions of symptomatic andasymptomatic Achilles tendons (IV).

A Fisher’s exact test was performed toassess MR imaging findings with respect tolong-term outcome analysis (IV). Thesignificant P value was 0.05 or less.

30

RESULTS

Postoperative MRI findings in patients with Achillestendon rupture (Paper I)

3 weeksThe postoperative dimensions of theAchilles tendon are presented in Table2. On MRI scans, at the level of therupture site the normally low intensityAchilles tendon was replaced by highintensity substance (Figs. 1A and 2A).At levels above and below, some lowintensity elements persisted. Usually,the tendon as a whole was bestvisualized on T2-weighted sagittalimages (Fig. 3A).

6 weeksThe rejoined rupture site was easilydetected in all cases. Eight of 21tendons showed early formation of anintratendinous high-intensity signalareas at the rupture site, which wastypically in the center of the tendonand was located at the level of therejoined tendon ends (Figs. 1C and

2C). This area had high intensity onboth PD- and T2-weighted images, buton the T2-weighted images the areawas slightly smaller. The rest of thetendons (13 of 21) showed diffuselyheterogeneous signal in the Achillestendon at the rupture site. The overallintensity of the tendon was still highand almost as heterogeneous as atthree weeks (Figs. 1C and 2C). On T2-weighted images the outer margins ofthe tendon were better visualized thanon PD-images (Fig. 2C). On the sagittalT2-weighted images, the tendon at thelevel of the rupture site appearedintermediate to high in intensity (Figs.3B and 4B). One patient (case 10) hadan accidental total re-rupture (verifiedby MRI) three days after cast removaland had surgical repair by the Lindholmtechnique.

TABLE 2. POSTOPERATIVE DIMENSIONS OF ACHILLES TENDON RUPTURE ON MRIDimension on MRI

b (mm)Follow-up Achilles tendon area

(mm2)Ratio a Thickness Width

Uninjured 75 (60 to 100) 5,9 13,7

3 weeks 218 (150 to 330) 2,9 (range, 2.3 - 4.7, p<0.001) 11,8 16,7

6 weeks 255 (170 to 370) 3,4 (range, 2.4 - 5.7, ns c) 13,4 18,1

3 months 454 (260 to 600) 6,1 (range, 4.0 – 6.8, p<0.001) 18,4 25,7

6 months 418 (260 to 510) 5,6 (range, 3.5 – 7.5, ns) 17,5 26,1

1 to 3 years d 300 (240 to 380) 4.2 (range, 3.2 to 5.1, p<0.01) 12,7 22,0

a average area compared to uninjured sideb average ap- (thickness) and width dimensions at 6 cm cranial to the posterosuperior calcaneal cornerc ns = non-significantd based on results of Paper II

31

Fig. 1.Axial proton densityMR images (TR =2000, TE = 20) ofnormal reunionprocess of rupturedand surgicallyrepaired Achillestendon. Samemagnification factor isused.

A, Affected andB, Unaffected sides three

weeks post-operatively. Onaffected side, tendon ishyperintense andpoorly demarcated.

Ruptured tendon 6 weeks, 3 months and 6 months post-operatively.C, At 6 weeks, high signal intensity area (✻) is visualized.D, At 3 months, lesion inside tendon is best seen, when tendon is at its largest. Low intensity periphery of

healing tendon is thicker and shows high intensity strands (arrow).E, At 6 months, tendon has slightly decreased in size and lesion has almost disappeared.

Fig. 2

Matching T2-weightedimages from samepatient (case 14)shown in Fig. 1(TR = 2000, TE = 80).

A,Operated tendon showshigh intensity withperipheral thin rim oflow intensity and alsolow intensity elementscentrally.

B, Unaffected side.

C, At 6 weeks, intratendinous lesion and margin of Achilles tendon are better visualized (arrowhead) onT2-weighted images than on PD-images (Fig. 1-C).

D, At 3 months, periphery of healing tendon has returned to normal low intensity level. Intratendinouslesion inside tendon is rather small in size (arrow). Note seven times as large cross-sectional area ofhealing tendon as on unaffected side (B and D).

E, At 6 months, scar is barely seen and also edema around tendon has decreased compared to previousfollow-ups.

32

3 months

At 3 months, the average cross-sectional area on MRI had reached itsmaximum (Table 2), and hadsignificantly increased from the 6-weekvalue (P<0.001). In 19 of 21 tendons, ahigh intensity, intratendinous signalwithin the Achilles tendon on PD-images(Table 3) emerged as healingproceeded. On T2-weighted images thisarea was smaller (p<0.001) and theintensity was classified as high (10 of19) or intermediate (9 of 19; Table 3).In 15 of the 19 cases, the lesion wascentrally located (Fig. 2D); in 4 cases itwas located adjacent to ventral or dorsalsurface of the tendon. The lesionappeared in one to five adjacent axialslices (height, 8 to 40 mm, Figs. 3C and4C). In all cases the intensity of thetendon above and below the rupture siteon T2-weighted images was low, but stillslightly higher than in an intact Achillestendon.

At 3 months, three patients showed aclinically poor preliminary recovery.They all had large (more than 50% ofthe tendon) high intensity lesions in theAchilles tendon on PD-images. Theselesions appeared as compact, highsignal intensity and were rounded withthin, low signal intensity peripheraltendon fibers (Fig. 5A). This was unlikethe more common form of high signalintensity lesion, which typically spreadlongitudinally along the rupture site andhad rather irregular margins (Figs. 3Cand 4C). Two of these patients (cases 2and 20) were reoperated on because ofthe clinically poor function, including anobvious limp, and the MRI lesion. Duringsurgery, when the tendon was incised,the center of the lesion was found to bevery soft with some hemorrhagicelements. It was removed with acurette. The histological specimen of themore peripheral portion of the lesionshowed incomplete healing with ongoingactive granulation tissue. The sample

from the most peripheral zone of thelesion consisted of more maturerepairing tissue with some collagenbundles. One patient (case 3) continuedwith elastic dressing. He had very slowrecovery even though the size of thehigh intensity lesion decreased from 39to 21% at six months (Table 3). On thispatient, we also performed an additionalMRI at 9 months, and the lesion haddecreased to 15%. This patient did notreturn to previous level of activities, buthe resumed normal walking.

Clinically, in 5 of 21 tendons, themargins of the wound had not fusedproperly. Two of them had red scarformation. However, none of thepatients had wound infection. All exceptone of the patients complained ofstiffness, either in the morning or after along day. Ten of 21 cases felt painaround the tendon during physicalactivity; this pain was commonly at theinsertion area. The range of motion indorsiflexion of the affected side was onaverage 2.7° (range, -12 to 30) lessthan that of the unaffected leg. Inplantar flexion, the affected leg had 6.5°(range, 0 to 20) less motion than theunaffected leg. Walking was consideredabnormal in five patients, which all hadintratendinous lesion of larger than 33%of the cross-sectional area on PD-images. This was statistically larger thanin normally walking patients on both PD-and T2-weighted images (P=0.003).Statistically significant inversecorrelations between the percentage ofheel raises and the size of theintratendinous lesion on PD-images (r=-0.66 and P=0.002) and on T2-weightedimages (r=-0.62 and P=0.004) werefound. The range of motion indorsiflexion and the size of theintratendinous lesion on PD- and T2-weighted images revealed correlation:r=0.50, p=0.02 and r=0.57, p=0.008,respectively.

33

Fig. 3. -- Sagittal T2-weighted MR images (TR = 1500, TE = 55) of sensitive patient (case16) with subjectively delayed recovery during rehabilitation of ruptured and surgicallyrepaired Achilles tendon.A, At 3 weeks, level of reunion area of Achilles tendon is seen (black arrow).B, At 6 weeks, tendon has increased only slightly in size. Diffuse intratendinous lesion formation is visible

and it has intermediate signal intensity.C, At 3 months, lesion is better-visualized (white arrow). Clinically, her ankle had restricted range of motion

in dorsiflexion and probably therefore, her tendon was one of the smallest in cross-sectional area (4.8times the normal).

D, At 6 months, lesion has almost disappeared although contour of Achilles tendon remains abnormal.

Fig. 4. -- Sagittal T2-weighted MR images (TR = 1500, TE = 55) of patient (case 12) withmore active rehabilitation than patient in Fig 3. Generally tendon is thicker and has largerintratendinous lesion within it.A, At 6 weeks, the rejoined tendon ends are clearly seen (black arrow) showing misleading similarity with

acute tendon rupture.B, At 3 months, the Achilles tendon is at its thickest (23 mm) and the anterior margin of the tendon bulges

anteriorly above the calcaneal corner. The intratendinous, medium sized lesion has spread longitudinallyand has irregular margins (white arrow).

C, At 6 months, the size of the intratendinous lesion has decreased significantly as seen in all cases.D, At 9 months extra MR images the ap-diameter of the tendon, and the size of the lesion have still

decreased.

34TABLE 3. POSTOPERATIVE MRI AND CLINICAL EVALUATION FINDINGS IN SURGICALLY REPAIRED ACHILLES TENDON RUPTURES.

Clinical data Evaluation at 3 months Evaluation at 6 months

Rupture Tendon Lesion size c Lesion Tendon Scar size

No. Age/sex

Activity sitea

(cm)Plastytype

size b

(mm2)on PD(%)

on T2(%)

intensityon T2e Walk size b

(mm2)on PD(%)

on T2(%)

Clinicalcategory

1 44/M Running 7,5 L 490 22 14 Int normal 360 NL NL Excellent2 32/M Badminton 6,5 T 550 51 42 High abnormal - - - -2 e L 510 18 10 High normal 480 5 4 Excellent3/R 34/M Volley ball 5 L 530 56 39 High abnormal 490 32 21 Poor3/L 34/M Tennis 6 L 410 29 13 Int normal 330 NL NL Good4 26/F Volley ball 5 T 475 21 12 Int normal 475 NL NL Excellent5 45/M Tennis 5 L 330 33 20 Int abnormal 300 NL NL Good6 34/M Badminton 6,5 T 460 14 12 Int normal 440 NL NL Excellent7 32/M Badminton 5 T 450 27 12 Int normal 430 NL NL Excellent8 36/M Bandy 4,5 T 340 21 14 Int normal 320 NL NL Excellent9 32/M Badminton 7 T 530 NL NL NL normal 500 NL NL Excellent10 37/M Tennis 5 T - - - - - - - - -10 e L 260 NL NL NL abnormal 350 NL NL Good11 35/M Basket ball 7,5 L 270 11 7 Int normal 260 NL NL Excellent12 28/M Badminton 7 T 490 36 30 High normal 420 23 16 Good13 35/M Badminton 7 T 600 26 18 High normal 510 17 12 Good14 32/M Soccer 6,5 Lynn 500 14 12 High normal 430 6 4 Good15 40/F Aerobic 5 T 560 9 7 High normal 420 8 4 Good16 42/F Badminton 6 Lynn 330 16 11 Int normal 330 12 3 Good17 54/M Tennis 6 T 520 16 9 High normal 470 6 3 Excellent18 32/F Hand ball 7,5 L 540 33 22 High abnormal 510 15 9 Good19 37/M Soccer 1,5 Lynn 570 33 25 High normal 450 22 18 Good20 51/M Dancing 6,5 L 320 65 58 High abnormal - - - -20 e L 320 48 31 High abnormal 320 26 20 PoorMean 36.8 5.9 456 28.5 19.9 418 15.6 11,4SD 7.1 1.4 103 15.5 13.3 80 9,2 7,3

L = Lindholm, T= tenorrhaphy, NL = no lesion, Int = intermediate.a Measured from the calcaneal insertion at 6 weeks after surgery on MRI scans.b Normal Achilles tendon has an area of 75 mm2, on average.c Size of intratendinous lesion (at rejoined tendon ends) on proton density- and T2-weighted images, % of total area.d Signal intensity of the intratendinous lesion on T2-weighted images.e Reoperated patients during follow-up

35

6 months

Eight of 9 patients with an intermediatesignal intensity area on T2-weightedimages at 3 months no longer had thislesion at 6 months (Table 3).Correspondingly, 9 of 10 patients withhigh signal intensity areas on T2-weighted images at 3 months still hadthem at 6 months, although the areaswere reduced in size (Fig. 2E). In allcases the signal intensity of the tendonon T2-weighted images had returned tonormal, except the high signal intensityarea at the rupture site (Figs. 3D and4C). On PD-images the Achilles tendonstill showed a slightly higher signalintensity than the unaffected tendon.On sagittal T1-weighted images somevertically oriented thin intermediateintensity stripes were seenintratendinously. On sagittal T2-weighted images the tendon appearedhomogeneous, except high signalintensity areas (Figs. 3D and 4D) in 10cases.

At 6 months, two patients (cases 3 and20) were not able to walk normally;these patients had the largest intra-tendinous lesions on MR images.Clinically they had poor outcome andthe rest of the patients had good orexcellent scores on the scale used(Table 3). Six of 20 patients either feltuncomfortable with running or had nottried it. Stiffness or soreness was stillfelt by seven patients. In two patientsthe wound had not fused properly andthey had numerous adhesions. Fiveother patients had adhesions betweenthe skin and tendon. One patientcomplained pain at calcaneal insertion.The range of motion in dorsiflexion wason average 1.6° (range, -8 to 18) lessthan on uninjured side. In plantarflexion, minor limitation averaged 3.3°(range, 0 to 13). Statistically significantinverse correlations were foundbetween the percentage of heel raisesand the size of the intratendinous

lesion on PD-images (r=-0.53 andP=0.02) and on T2-weighted images(r=-0.50 and P=0.04). The range ofmotion in plantar flexion and the size ofthe intratendinous lesion on PD- andT2-weighted images showedcorrelation: r=0.60, P=0.01 andr=0.55, P=0.02, respectively. Therewas no statistically significantdifference (P>0.11) between the size ofAchilles tendons operated withLindholm technique (n=7) andtendorrhaphy (n=10) at any follow-ups.

Fig 5. -- Sagittal T2-weighted MR images ofpatient with poorly healed Achilles tendonrupture subsequently requiring reoperation.A, At 3 months after initial injury tendon has large

intratendinous lesion (✻) formation. Note largerounded and compact appearance of centralscar and thin peripheral low intensity fibers.Patient had pain and abnormal walk. He hadsecond surgery, and lesion was removed.

B, At 6 months after reoperation lesion (arrow)had decreased significantly.

C, Below, histopathology of lesion. Activegranulation tissue with proliferating fibroblastsand capillaries (arrow) are seen (HE;magnification x100).

36

1 to 3 years (Paper II)

The mean postoperative dimensions ofthe ruptured Achilles tendons at 1 to 3years are presented in Table 4. Theshape of the affected tendon appeared

more rounded and irregular in contourthan the unaffected tendon thattypically had ellipsoid shape.

TABLE 4. POSTOPERATIVE DIMENSIONS OF RUPTURED ACHILLES TENDON ON MRI AND USMRI (n=13) US (n=13)

thickness(mm)

width(mm)

thickness(mm)

width(mm)

Ruptured tendon at 2 cm 12.2 ± 3.7 22.8 ± 2.7 11.4 ± 2.6 20.0 ± 3.0

Ruptured tendon at 6 cm 15.9 ± 2.4 22.2 ± 3.1 12.7 ± 2.2 22.0 ± 4.6

Uninjured tendon at 6 cm 6.0 ± 1.1 13.2 ± 1.8 6.0 ± 1.7 17.8 ± 4.0

On axial images the affected tendonhad generally low signal intensity onPD- and T2-weighted images. However,on axial PD-images 9 of 13 affectedtendons showed intermediate signalintensity stripes of small size within thetendon. On T2-weighted images 4 of 13tendons showed an intratendinous areawith intermediate (1 cases) to high (3cases) signal intensity (Table 5). Thesignal intensity of these areas were inall 4 cases higher on PD-images, andthese areas were typically located inthe middle of the tendon. On sagittalT1-weighted images the intratendinousareas showed intermediate signalintensity in all 4 cases. It’s size rangevaried from 4 to 18% (mean 9.5%,Table 5) of the total cross-sectionalarea on PD-images and it was smallerin all cases on T2-weighted images(p=0.06, n=4). The intratendinous areawas typically located at the level ofmost thickness of the tendon.

Clinically 2 of 13 patients wereclassified to have poor outcome and therest had good score on the scale used(Table 5). These two patients with pooroutcome could not run, felt weaknesson ruptured side in rising on toes,complained marked discomfort andwere unsatisfied. On MR images, theyhad the two largest intratendinousareas of increased signal intensity.

Concerning the other dimensionsmeasured, there was no differencebetween the patients with poor andgood outcome. Two patients in goodclinical category complained of stiffnessof the ankle with normal imagingfindings in both MRI and in USexamination.

On sagittal MR images the anteriormargin of the operated Achilles tendonshowed remarkable bulging, which wasseen right above the level of thecalcaneal corner. In addition, theoverall contour of the tendon was wavyin 7 of 13 cases and smooth in 6 cases.At the level of the calcaneal cornerthere was increased signal intensity inthe tendon in 6 of 13 cases. Also thedorsal surface of the tendon on axialMR images was typically uneven, andthe surgical wound scar was identifiablein all cases and it’s size varied. It hadlow signal intensity similar to thetendon.

Correlation between dimensions ofMRI and US

There was a significant correlation in APdimension of the Achilles tendon at 2and 6 cm levels above the calcanealcorner between MRI and US (r=0.61,p=0.05 and r=0.87, P=0.001,respectively). However, in transversal

37

width there was no statisticallysignificant correlation at 2 or 6 cmlevels, although the average width wasapproximately the same in US and MRI(Table 4). When the MR and US

dimensions of the opposite intact sidewere compared, there was a significantcorrelation in AP dimension (r=0.83,P=0.01) and in transversal width(r=0.78, P=0.04, n=9).

TABLE 5. PATIENTS AND POSTOPERATIVE MRI FINDINGS OF ACHILLES TENDON RUPTURES.Intratendinous lesion

Signal intensity on SizePat/Age

Event Plastytype

IntervalRupture-MR

Clinicaloutcome

Overallsignalintensity T2-WI PD-WI T1-WI %*

1/35 volley ball Lindh 15 mo good homogen - - - -

2/24 long jumping Lynn 14 mo good homogen - - - -

3/28 volley ball Lynn 14 mo good homogen - - - -

4/49 motocross Lindh 16 mo poor inhomogen high high inter 18

5/47 tennis Lindh 15 mo good homogen - - - -

6/37 indoors Lindh 16 mo good homogen - - - -

7/35 basket ball Lindh 36 mo good inhomogen inter high inter 4

8/33 badminton Lynn 36 mo good homogen - - - -

9/34 volley ball Lindh 13 mo poor inhomogen high high inter 10

10/36 tennis T 12 mo good inhomogen high high inter 6

11/39 badminton Lindh 13 mo good homogen - - - -

12/33 volley ball Lynn 14 mo good homogen - - - -

13/40 running Lindh 16 mo good homogen - - - -T =tendorrhaphy,* The size of the intratendinous lesion, percentage of the cross-sectional tendon area

Postoperative ultrasonography

The echogenic properties of the Achillestendon were similar in all 13 patients.As compared with the 9 contralateraluninjured Achilles tendons, theoperated tendon delineated lesssharply. A linear subcutaneous fat layerwas disturbed in all cases. In addition,the border of the operated Achillestendon was more uneven than in thenormal Achilles tendon. No fluidaccumulation around the Achillestendon was seen in any patient. Incontrast to thicker and longerechogenic bands seen in USexamination of the normal Achillestendon, there were thin and shorterechogenic bands in the Achilles tendonafter rupture. In addition, an Achilles

tendon after a rupture was more roundand irregular, whereas the normalAchilles tendon was ellipsoid in shape.Both increased and decreasedechogenic irregular areas were seen inthe tendon after operation, whereas thehealthy tendon was uniform in itsechogenic properties. In US,intratendinous calcifications were seenin 2 patients. A patient with poorclinical outcome showed more irregularhypoechoid areas and had the largestintratendinous lesion on MR images.The other patient with poor outcomedid not show any difference in UScompared with good-outcome patients.The site of the rupture was not visiblein US in any patient.

38

MR imaging of asymptomatic Achilles tendons (PaperIII)

DimensionsCorrelation between themeasurements of the twoobservers was excellent (r =.92, P<.001). The Cohen Kvalues for interobserveragreement varied frommoderate (.43) to good (.96).There was little variation in thesize of the asymptomatic, intactAchilles tendons, averaging 5.2mm in ap-diameter (Table 6).The difference in ap-diameterbetween the left and the rightside was on the average 0.30 ±0.30 mm (±SD; range 0-0.8mm). The thickest ap- diameterwas on the average 32 ± 10.5mm (±SD; range 0-50 mm)above the level of the calcanealcorner on axial images. Thelength of the Achilles tendonwas quite variable, rangingfrom 20 to 120 mm (Figs. 6and 7 and Table 6). There wasno statistically significantdifference in ap- or otherdimensions between the leftand right sided tendons in the19 entirely symptom-freevolunteers.

Fig. 6. -- Exceptionally long(12 cm) Achilles tendon in32-year-old asymptomaticmale volunteer. Sagittal T1-weighted spin-echo MR image(TR/TE 460/14) showshomogeneous, low intensityAchilles tendon. Distal end ofsoleus muscle (arrow). Noteinsertion of Achilles tendon inmiddle of posterior os calcis(✻).

Fig. 7. -- Short Achillestendon in 26-year-oldasymptomatic male volunteer.Sagittal T1-weighted spin-echo MR image (TR/TE460/14) demonstrates thesoleus muscle extendingdistally (arrow) and occupingspace in the fat pad anteriorto the Achilles tendon (Kager'sfat pad).

TABLE 6. DIMENSIONS OF ASYMPTOMATIC ACHILLES TENDONS MEASURED ON MR IMAGES

Dimension Mean (mm) SD Min (mm) Max (mm)

Antero-posterior thickness a 5.2 ± 0.73 3.5 6.8Width b 14.7 ± 2.06 10.0 19.1Length or height c 52 ± 18.8 20 120Min = minimum, Max = maximum.a Measured at the level of greatest thickness.b Measured at 3-cm level above the calcaneal corner.c Measured from insertion to the level of most distal part of the soleus muscle.

39

Shape

The cross-sectional anatomy was bestshown on axial FLASH images (Fig. 8).In 10 tendons the anterior margin wasconsidered to be mildly convex. In theremainder it appeared flat or concave.Mild anterior bulging on sagittal imageswas found in seven cases. Werecognized a more complex variation inthe cross sectional shape of the tendonin 56 of the 100 tendons. In thesecases, the tendon had a localizedanterior bulge that shifted in a wave-like fashion from the lateral to medialmargin across the anterior surface ofthe tendon on sequential axial imagesfrom proximal to distal (Fig. 9).

Another variant of the anterior contourwas seen centered at the level of the

musculotendinous junction in 29 cases.The anterior margin of the Achillestendon showed a longitudinal, thicker,cordlike component. This tendon bundleis formed within the most distal muscletissue of the soleus, proximal to thelevel where gradual merging into themain body of the Achilles tendon takesplace (Fig. 10).

Plantaris tendon

The plantaris tendon was seen as aseparate structure in 54 cases. In 20cases it blended into the medial marginof the Achilles tendon, forming a high-intensity longitudinal cleft just proximalto this level of merging (Figs. 8 and11).

Fig. 8. -- Asymptomatic Achilles tendon in34-year-old male volunteer. Axial high-resolution FLASH MR image 4 cm abovecalcaneal corner shows paratenon whichappears as intermediate signal intensitylayer on dorsomediolateral aspect ofAchilles tendon (arrowhead).Outline of paratenon is marked by low intensityband that represents chemical shift artifactbetween paratenon and subcutaneous adiposetissue. Intratendinous vessel has round spot-likehigh intensity appearance.Note wave-like segment of anterior contour oftendon, a normal variant, and plantaris tendon(white arrow) seen anterior to medial margin ofAchilles tendon.

Fig. 9. -- Achilles tendon in 29-year-oldasymptomatic male volunteer. Serial axialFLASH MR images of right Achilles tendonfrom proximal to distal, showing wave-likecrescent (arrow) in anterior contour oftendon shifting lateral to medial. This iscaused by fibers originating in soleusmerging with fibers originating ingastrocnemius and spiraling, as they extendtowards insertion in calcaneus.Also demonstrated is normally seen loss ofdefinition between paratenon and skin as level ofimages approaches calcaneal corner.

40

Signal intensity

The normal Achilles tendon was ahomogeneous, low signal-intensitystructure on STIR images. In somecases, round foci of increased signalintensity most likely representing smallvessels were detected in the tendonsubstance on STIR images. These hadcharacteristic tubular curving shape onsequential images. On axial FLASHimages, the intratendinous signalintensity was mildly inhomogeneous in45 of 100 tendons. Thirty-eight of 100tendons had some very thin,intermediate signal-intensity stripesdistally (Fig. 12), and 30 of 100tendons had patchy or punctateintratendinous intermediate to highsignal intensity foci, typically slightlymore proximal than the distal stripes.Four of 100 asymptomatic tendonsdemonstrated small areas ofintratendinous ground-glass inter-mediate signal-intensity (average 10%in cross-sectional area, 3 mm indiameter, and 10 mm in height) onaxial FLASH images (Fig. 13). None of

these intratendinous regions wasdetectable on STIR images (Fig. 14).

Insertion to calcaneus

The calcaneal bone marrow had ahomogeneous low signal intensity in allbut one of 100 asymptomatic tendonson STIR images. This one volunteerhad a well demarcated round highsignal intensity osseous lesion, 8 mm insize, close to the insertion of theAchilles tendon.

The retrocalcaneal bursae had variousforms typically appearing as a thinhigh-signal-intensity band on STIRimages between the Achilles tendonand the posterosuperior corner of oscalcis. Fifteen percent of theasymptomatic cases showed aprominent fluid collection in theretrocalcaneal bursae. The averagebursal dimensions in these cases were11 mm (range, 7-16 mm) incraniocaudal dimension, and 13 mm(range, 8-22) in transverse dimension.

Fig. 10. -- Achilles tendon in 24-year-oldasymptomatic volunteer. Axial FLASH MR imagedemonstrating prominent cord-like, low intensitytendon fibers (arrow) originating in soleus merginginto anterior aspect of Achilles tendon.

Fig. 11. -- Achilles and plantaris tendons 2 cmabove calcaneal corner. Plantaris tendon may insertupon calcaneus directly, or merge with medialmargin of Achilles tendon. Just proximal tomerging, axial images show cleft (arrow), givingfalse impression of medial tear of Achilles tendon.

41

Fig. 12. -- Achillestendon in 17-year-oldasymptomatic malevolunteer. Axial FLASHMR image of Achillestendon just abovecalcaneal cornerillustrates intratendinousstripes commonly seenon this sequence at thislevel.

Fig. 13. -- Achilles tendon in25-year-old asymptomaticfemale volunteer. AxialFLASH MR image shows focalground-glass area (arrow) ofincreased signal intensitysuggesting early tendondegeneration.

Fig. 14. -- Musculotendinousjunction. Axial STIR MRimage at level ofmusculotendinous junctionshows relatively high signalintensity in distal soleusmuscle as a normal variant(arrow), likely due to richvascular network normallypresent. Also, note relativelyhigh signal intensity ofproximal paratenon(arrowhead).

Peritendinous tissues

The cross-sectional anatomy of theperitendinous tissues was best seen onaxial FLASH images (Fig. 8). Thesuperficial component of the cruralfascia merged posteriorly with theparatenon. Outline of the paratenonwas accentuated by the low-intensitycomponent of the chemical shiftartifact, seen as a signal void bandbetween the subcutaneous adiposetissue and the paratenon of the Achillestendon (Figs. 8 and 9). The cruralfascia continued anteriorly on bothsides of the pre-Achilles fat pad tomerge near the medial and lateralmalleolus with the deep component ofthe fascia covering the deep flexormuscles.

The paratenon appeared as a thin,homogeneous layer of signal intensity,higher than the tendon in intensity butsimilar to muscle on axial FLASH

images (Figs. 8 and 9). It wasconsistently detected along theposterior, the medial and the lateralaspects of the Achilles tendon. At levelscloser to the musculotendinousjunction, the paratenon was clearlyseparated from tissues posterior to theAchilles tendon (Fig. 9). Towards thecalcaneal insertion of the Achillestendon, the paratenon and crural fasciawere demarcated less clearly, andseemed to fuse with the posteriorsubcutaneous structures (Fig. 9). Nocases of thickening of the paratenonwere detected on axial FLASH images.

On axial STIR images the paratenonhad generally uniform intermediate-signal-intensity which appearedmoderately bright in contrast to the lowintensity of the tendon and suppressedfat. In 40 of the 100 cases, there was ahigher intensity segment of theparatenon proximally (Fig. 14) anddistally, while the middle third

42

appeared less intense. Also, in somecases 1 to 2 axial slices demonstratedhigh-signal-intensity veins drapingposteriorly along the paratenon. Pre-Achilles fat pad had homogeneous, lowsignal intensity in all except fiveasymptomatic cases in which someabnormal increased signal intensity onSTIR images was noted. In almost allcases, there were high-intensity veinstraversing the pre-Achilles fat pad.Eleven asymptomatic tendons hadabnormal, high signal intensity in the

paratenon on axial STIR images: Sixwere seen in healthy volunteers, andfive in asymptomatic tendons inpatients with contralateral tendondisease.Mild symmetrically increased signalintensity on STIR images in 19 cases atthe musculotendinous junction at thelevel of fusion of the distal fibers of thesoleus muscle within the Achillestendon (Fig. 14).

MR imaging of overuse injuries of the Achilles tendon(Paper IV)

Of 118 cases of Achillodynia, abnormalMR imaging findings were present in111. A summary of various MR imagingfindings is presented in Table 7. In 62asymptomatic contralateral tendons wefound 12 abnormal findings: four smallintratendinous lesions, three slightlyenlarged bursae and five moderateperitendinous abnormalities. Theoverall sensitivity of MR imaging in the

detection of abnormalities in cases ofpainful Achilles tendon was 94%,specificity 81%, positive predictivevalue 90%, negative predictive value88% and overall accuracy 89%. Theinterobserver agreement for thedifferent MR imaging findings evaluatedwere good in all categories (κ=0.60-0.87).

TABLE 7. SUMMARY OF MRI FINDINGS IN 118 ACHILLES TENDON OVERUSE INJURIES.

MRI Findings Tendonsa

n (%)Abnormal tendon

No focal intratendinous lesion 36 (31 %)Focal intratendinous lesion 37 (31 %)Focal lesion with high intensity centerb 17 (14 %)

Enlarged retrocalcaneal bursa 23 (19 %)Tendon insertion

Abnormal tendon at insertion 18 (15 %)Abnormal calcaneal bone marrow 10 (8 %)

Musculotendinous junctionAbnormal SIb in distal soleus muscle 3 (3 %)

Peritendinous tissuesIncreased SIb with tendon abnormalitiesIncreased SIb with normal tendonThickening of paratenon

611432

(52 %)(12 %)(27 %)

Finding not related to Achilles tendon 2 (2 %)Achilles tendon with normal MR appearance 7 (6 %)SI = signal intensitya The total number exceeds 118 because several MR imaging findings co-existb Detected on axial STIR images

43

Antero-posterior diameter

The most common abnormal findingwas thickening of the Achilles tendon inthe ap-direction (Fig. 15) seen in 87cases. The average ap-diameters of thesymptomatic and asymptomaticAchilles tendons were 7.6 mm (±2.25)and 5.2 mm (±0.77), respectively(P<.001). Anterior bulging (Fig. 16)was seen in 67 tendons, and convexityof the anterior margin (Fig. 15) in 77tendons (29 mild and 48 severe).

Fig 15. -- 23-year-old female with long-standing pain in left Achilles tendon.A, Right asymptomatic Achilles tendon shows

normal appearance on axial FLASH image.B, Left leg with anterior convexity (arrow) and

thickening of Achilles tendon as sign oftendinosis. Note lack of intratendinous signalchanges.

Fig. 16.33-year-oldmale withsymptomatic Achillestendinosis.

Sagittal T1-weighted SEimage showsanteriorbulging andthickening(arrows) oftendon asevidence oftendinosis.

Intratendinous lesions

In 54 tendons (45%) an intratendinouslesion was detected. The level of thecenter of the lesion varied frominsertion distally to 8 cm moreproximally. Fourteen of theintratendinous lesions were located at alevel close to the tendon insertion. Thearea of the lesion varied from 5 to 90%(average, 28%) of the cross-sectionalarea on axial FLASH images (Figs. 17Aand 18A). The height of the lesionvaried from 5 to 100 mm (average, 28mm). The center of the lesion wasusually in the central substance of thetendon (Fig. 18A). In 35 of 54 Achillestendons, the lesion reached the tendonsurface (Fig. 17A). In ap-direction, thecenter of the lesion was located towardanterior margin in 20 tendons, andtoward posterior margin in seven. Inthe mediolateral direction, the lesionwas situated medially in 13 tendonsand laterally in seven. In all tendons,the area of the intratendinous lesionwas seen most clearly and at its largeston axial FLASH images (Figs. 17A and18A). In 19 tendons, the lesion wasclearly detected on axial FLASH imagesonly (Fig. 17A). Significantly smallerlesions were visualized with FLASHsequence than on other sequences(P=.04). On T2-weighted images, thelesions were least visible (Fig. 17D) andsmaller than on PD- and STIR imagesin all except three cases, in which theywere equal in size.

In 17 cases, a focal, higher signalintensity area was detected centrally inthe intratendinous lesion on axial STIRimages (Fig. 18B). In nine of thesetendons, this area had equally highsignal intensity on T2-weighted images;however, in the other eight, this areawas less intense, and in both instancesthe lesions appeared smaller.

44

Fig. 17. -- Intratendinous lesion ofpainful Achilles tendon in 38-year-oldmale.A, Axial FLASH image shows large

intermediate intensity intratendinouslesion in thickened Achilles tendon.Paratenon (arrowhead) is thickenedposterior to Achilles tendon.

B, Corresponding STIR image demonstrateslow signal intensity within lesion, andmoderately increased intensity withinposterior paratenon (open arrow).

C, PD- and, D, T2-weighted SE imagesexhibit very mild intratendinous changes.Paratenon is poorly demarcatedcompared to high-resolution image in A.

Fig. 18. -- 39-year-old female withlong-standing Achilles tendon pain.A, Axial FLASH image shows large

intratendinous lesion (arrowhead) withradiating stripes to surface of tendon.

B, Corresponding STIR image shows lesionwith high intensity center (arrow).Paratenon posterior to intratendinouslesion has increased intensity suggestingparatenonitis.

Tendon insertion and bursa

In 28 symptomatic tendons,abnormalities were detected at theAchilles tendon insertion. Ten tendoninsertions (8%) had an area ofincreased signal intensity on STIRimages in the calcaneal bone marrow(Figs. 20A and C). The area varied from3 to 20 mm in diameter. In 18 tendons(15%), the tendon was abnormal at thelevel of 0-2 cm from tendon insertionwith (n=14, Fig. 20B), or without (n=4)a focal intratendinous lesions. Thelesions were typically located along theanterior surface of the tendon at thesame level (Fig. 20B).

Twenty-three cases (19%) had anenlarged retrocalcaneal bursa, and ineight tendons this was the onlyabnormality (Fig. 19). Of the 18tendons with abnormal tendon

appearance at the Achilles tendoninsertion (Fig. 20B), 10 had increasedsignal intensity in the calcaneal bonemarrow (Fig. 20C), and 15 had anenlarged retrocalcaneal bursa on STIRimages (Fig. 20D).

Fig. 19.Retro-calcanealbursitis in20-year-oldfemale.

Sagittal STIRimage showsenlargement ofbursa (arrow)due toexcessive fluid.Achilles tendoninsertion tocalcaneus hasnormalappearance.

45

Fig. 20. -- 22-year-old female with clinical diagnosis of Achillesinsertitis.A, Sagittal STIR image shows increased signal intensity (arrows) in

calcaneal bone marrow as sign of reactive edema. Also, retrocalcanealbursae and adjacent Kager’s fat pad shows increased signal intensity.

B, Axial T1-weighted FLASH image at level of calcaneal corner showsanteriorly located intratendinous lesion.

C, Axial STIR image at lower level, shows increased signal intensity incalcaneal bone marrow at insertion of Achilles tendon (solid arrow).

D, Axial STIR image (corresponding to B) shows retrocalcaneal bursitis(✼) typically associated with other insertional abnormalities.

Peritendinous tissues

Eighty-one tendons (69%) hadabnormalities in the peritendinoustissues. Anterior fat pad was abnormalin 12 tendons (10%) (Figs. 21A-B). Theparatenon had increased signalintensity along the posterior, medial orlateral aspect of the Achilles tendon onSTIR images in 48 tendons (41%).Both sides were involved in 18 tendons(15%) (Fig. 22C). Three tendons hadthickening of the paratenon bestdetected on FLASH images. In 45tendons (36%), the peritendinouschanges seen on STIR images wereclassified as severe, and in 36 tendons(31%) as moderate.

Thickening of the paratenon, best seenon axial FLASH images, was found in32 tendons. In 15 tendons, thickeningof the paratenon was classified asmoderate and in 17 as severe (Figs.23B-C). Of the 32 tendons withthickened paratenon, 29 also showedincreased signal intensity of theparatenon on STIR images.

Both the tendon and peritendinoustissue were abnormal in 75 tendons(64%). Of 90 tendons with anabnormality, 61 had increased signalintensity in the peritendinous tissues onSTIR images (Figs. 17B and 18B).Regional thickening of the paratenonwas associated in 23 (43%) of the 54intratendinous lesions (Fig. 17A).

Fig. 21. --Anterior peritendinitis in Kager’sfat pad in 20-year-old male.A, Sagittal T1-weighted SE image shows abnormal

low signal intensity strands in Kager’s fat padanterior to normal Achilles tendon.

B, Abnormal high signal intensity changes (∗ ) areseen on sagittal STIR image.

46

Fig. 22. -- Various patterns of peritendinitis on axial STIR images.

A, 25-year-old male with small intratendinous lesion (arrowhead) and increased signal intensity posteriorlyas sign of paratenonitis.

B, 32-year-old female with increased signal intensity in Kager's fat pad anterior to Achilles tendon (✼), assign of anterior peritendinitis.

C, 20-year-old male with both anterior and posterior peritendinitis (solid arrow), posteriorly also calledparatenonitis.

Fig. 23. -- Various patterns of thickening of paratenon on axial high-resolution FLASH images.

A, 32-year-old asymptomatic male volunteer shows normal, intermediate signal intensity paratenon deep tolow signal intensity chemical shift artifact between paratenon and subcutaneous fat.

B, 24-year-old male with Achilles tendinosis shows moderately thickened paratenon (between arrowheads)and thickened Achilles tendon.

C, 48-year old male with long-standing Achillodynia. Posterior paratenon (between solid arrows) is severelythickened. Note also large intratendinous lesion extending to surface of abnormally thickened tendonindicating tendinosis.

47

MRI and clinical findings

Only 30 (56%) of the 54 cases with anintratendinous lesion on MR imaginghad palpable nodular thickening. Goodcorrelation was found between theanatomic level of the thickening andthe lesion detected on MR imaging(r=.76, P<.001). Four tendons werethickened on palpation and a thickenedappearance on MR imaging but no focallesion.

The level of the palpable tendernesscorrelated well with the level ofincreased signal intensity of theparatenon as seen on STIR images(r=.79, P<.001). All 28 patients withabnormal MR imaging findings at thelevel of the insertion of the Achillestendon also had maximal pain andtenderness at that level.

A significant difference (P=.02) wasfound in the duration of symptomsbetween patients with tendon insertionand calcaneal bone marrowabnormalities (mean duration, 10.2 ±7.8 months), and patients withenlarged retrocalcaneal bursae alone(mean, 3.2 ± 3.0 months). The samewas true for cases with thickening ofparatenon (mean, 12.1 ± 8.9 months)or without it (mean, 6.0 ± 5.0 months)(P=.001).

MRI and surgical findings

Twenty-eight patients had surgery, onthe average 5.6 weeks (range 1 to 15weeks, in 3 cases > 8 weeks) after theMR imaging. The main MRI findingswith surgical correlation are listed inTable 8. Of 21 surgically-detectedintratendinous lesions, 20 were seen onMR images (sensitivity 95%). Thelesions were best detected on axialFLASH images (Table 8). However, inthree cases, axial FLASH imagesdisplayed the lesion as somewhat

larger than was found at surgery. Inone MR negative case, a small,surgically detected lesion was notidentified on MR imaging. In all MRpositive cases the lesion appeared tohave disorganized tendon fiberstructure. In six cases, a lack ofcontinuity of some of the tendon fiberswas found at surgery, suggesting apartial rupture. Of these six cases, fourhad a high signal intensity focus in theintratendinous lesion on STIR imagesand three had that focus on T2-weighted images. Fifteen of the 21surgically detected intratendinouslesions was described as soft or hard,nodular areas with deranged fiberstructure, indicating degenerativechanges or healed partial tears.Surgical findings in three cases withswollen appearing Achilles tendons didnot reveal focal disturbances in fiberstructure. On MR images, thesetendons were thickened tendon withouta focal lesion. In surgery, 19 tendonswere described to have thickenedparatenon. Twelve of these (63%) weredetected on axial FLASH images.

MRI and histological findings

Histological specimens of 13 intra-tendinous lesions detected on FLASHimages were obtained at surgery.Pathologic changes were found in allspecimens. The most prominentobservations were tendon fiberdisturbances and lack of continuity ofthe collagen fibers (Fig. 24B). Lightcollagen staining and roundness oftenocyte nuclei were also present in allspecimens (Fig 24C). In some tendons,active capillary proliferation was noted(Fig. 24D) while in three specimensvessel density was considered normal.Inflammatory cells were not found inany speciman. The histological changeswere most severe in tendons with ahigh signal intensity lesion on STIRimages (n=4).

48

TABLE 8. COMPARISON OF MRI AND SURGICAL FINDINGS OF 28 PATIENTS WITH ACHILLES

TENDON DISORDERSNormaltendona

___________

Intratendinouslesion

__________________Bursitis

___________

Thickenedparatenon

___________

MR Surg MR Surg MR Surg MR Surg

FLASHotherseq.

N:o of cases 4 4 20 15 21 11 10 12 19

False pos. 0 - 0 0 - 1c - 0 -

False neg. 1 - 1 6 - 0 - 7 -

κ-valueb .89 - .91 .56 - .92 - .52 -

Surg=surgery, seq=sequences, FLASH=T1-weighted fast low-angle shot gradient echo imagesa No thickening or disturbances in tendon fiber structure of the Achilles tendon at surgeryb Surgical findings are considered as golden standardc Long time interval between MR and surgery (15 months)

Fig. 24. -- Various patternsof histological finding ofAchilles tendon lesions.

A, Normal collagen structure oftendon. Tenocytes are barelyvisible and spindle-shaped, andstaining is deep (vanGiesonstain, original magnification, x300.)

B, Histopathology ofintratendinous lesion showsweak collagen staining andderangement of tendon fibers.(vanGieson stain; originalmagnification, x 300.)

C, Histopathology of lesion showspresence of rounded tenocytenuclei (arrows) in Achillestendon tissue. (H&E stain;original magnification, x 400.)

D, Damaged Achilles tendontissue with high capillaryproliferation (arrowheadsindicate some single vessels).Note also high number andround morphology of tenocytenuclei. (H&E stain; originalmagnification, x 400.)

Light photomicrograph images

49

Long-term follow-up

The time interval between MR imagingand the follow-up interview was 1.5 to4.5 years (mean 3.4). Ninety patients(92%) with 108 painful Achillestendons were available for theinterview. Twenty-seven of thesetendons were treated surgically.Overall, 63 tendons (58%) hadexcellent, 21 (19%) good, 22 (20%)fair, and 2 (2%) poor results. Thepercentage of satisfactory (excellent orgood) results of operatively andconservatively treated patients were78% (n=21) and 80% (n=65),respectively. All patients with normalMR imaging had fully recovered. Thesummary of the long-term follow-up ispresented in Table 9.

In the conservatively treated subgroup,patients without an intratendinouslesion had better results than patientswith such a center (Table 9). However,when subjected to statistical analysis,the results did not show significantvariation (P=.09). Patients with anintratendinous lesion showing a highsignal intensity center on STIR images

had significantly less satisfactoryrecovery (P=.02) than patients withoutsuch a lesion. Surgically treatedpatients showed similar results (Table9).

The long-term results for bothconservatively and surgically treatedinsertional abnormalities weresignificantly less satisfactory (P=.04 forboth subgroups) than for patients withnon-insertional MR imaging findings.There were two patients with poorresults, and they both had surgicallytreated insertional abnormalities.

The group of conservatively treatedpatients in which findings were limitedto the peritendinous tissues hadsignificantly (P=.02) better long-termresults than the patients with combinedperitendinous and tendinousabnormalities. Similarly, patients with afluid-filled retrocalcaneal bursa but noother abnormalities had significantly(P= .04) better outcome afterconservative treatment than patientswith abnormal findings at tendoninsertion.

TABLE 9. SUMMARY OF LONG-TERM FOLLOW-UP IN PATIENTS WITH DIFFERENT MRIFINDINGS.

Satisfactory (excellent or good) results

MRI FindingsConservative

treatmentn=81 (%)

Operativetreatmentn=27 (%)

Abnormal tendonNo focal intratendinous lesionFocal intratendinous lesionFocal lesion with high intensity center a

25135

(83%)(66%)(55%)

3124

(100%)(85%)(57%)

Enlarged retrocalcaneal bursaWith normal Achilles tendon

87

(67%)(100%)

6- c

(60%)

Insertional tendinosis 5 (57%) 5 (56%)Peritendinous tissues

Increased SIb with tendon abnormalitiesIncreased SIb with normal tendonThickening of paratenon

361115

(78%)(100%)(78%)

19- c

8

(79%)

(80%)Achilles tendon with normal MR appearance 6 (100%) - c

Note. Total number exceeds 108 because several MR imaging findings coexist.a on STIR images b SI = signal intensityc n ≤ 1

50

DISCUSSION

Postoperative follow-up ofsurgically repairedAchilles tendon ruptures

Patient material

Achilles tendon rupture is known tooccur at a much younger age thanother degenerative tendon ruptures,the mean age being 35 years (Jozsa etal. 1989). The typical patient is arecreational athlete, usually a male,and in 85% of cases there is nopreceding symptoms (Kannus andJozsa 1991). Our group of patientsconfirms these observations. On theother hand, our patients did have anuncommonly high incidence of previouscontralateral Achilles tendon rupture.One patient first ruptured his righttendon. He had three follow-up scansand then went on to rupture his lefttendon. Therefore, we had preinjury MRimages of the left Achilles tendon, thelatest one 7 weeks before the rupture.In none of these images did we see anyobvious indications of intratendinousdegenerative lesions that typicallyappear as signal intensity changes(Weinstabl et al. 1992). This suggeststhat at least low field MRI may not beable to predict the risk of Achillestendon rupture.

Cross-sectional area

During the reunion process aftersurgical treatment of the Achillestendon rupture, the cross-sectional areaof the tendon increases greatly. Thefastest increase occurs during the earlymobilization phase after cast removal,i.e. between the 6 weeks and 3 months.For example, in antero-posteriordimension this resulted in an average6-mm increase in the middle of the

tendon during the first 6 weeks ofmobilization. This increase alone equalsthe normal tendon thickness (Weinstablet al. 1992). In width the increase wasonly moderate and the explanationcould be that there is less spaceavailable in this direction because of thecrural fascia. The abrupt increase in thearea of Achilles tendon after six weekscasting observed in our study might beminimized by using the free anklemotion rehabilitation programs thathave been suggested by many authors(Beskin et al. 1987, Carter et al. 1992,Solveborn and Moberg 1994). Therewas also a considerable variation in thecross-sectional area between theindividual patients during the healing.

The cross-sectional area of theruptured Achilles tendon decreasesslowly after 3 months, but neverreturns to normal. The cross-sectionalarea of the ruptured Achilles tendonremained enlarged 1 to 3 yearspostoperatively in all of our patients.

Intratendinous lesions

The main finding of the Paper I was thehigh signal intensity area at therejoined tendon ends on PD-imageswhich appeared in nearly all surgicallyrepaired Achilles tendons. At 6 weeks,intratendinous and continuous highintensity areas were identified for thefirst time. During the earlyrehabilitation period, within 3 months ofthe rupture, a variable-sized highintensity area was detected in all excepttwo cases. Thereafter, by 6 months thisarea had reduced greatly in size ordisappeared. The intratendinous lesionspersisted significantly longer if T2-weighted images at 3 months hadshown large high signal intensity areas.The present study demonstrated thatan intratendinous lesion is part of the

51

reunion process of the ruptured Achillestendon.

Intratendinous lesions were alsodetected in our group of 1 to 3 oldAchilles tendon ruptures. Two patientswith poor outcome had the largestintratendinous lesions. In patients withdelayed healing the lesions may persistlonger.

Characteristically, all intratendinouslesions showed larger size on protondensity than on T2-weighted images.The inner part showing high signalintensity on both images may representfibrous scar formation with high mobileproton density, whereas the outer layershowing high signal intensity on PD-images and low on T2-weighted imageshad shorter T2-relaxation timesuggesting difference in proteincomposition between the inner andouter layers in the lesion.

Intratendinous increased signalintensity lesions of the Achilles tendonhave been described previously in somecase reports (Beskin et al. 1987,Marcus et al. 1989, Neuhold et al.1992, Reinig et al. 1985). A thickened,complex Achilles tendon with onlypartially continuos tendon fibers wasdetected by MRI at fourteen monthspostoperatively (Quinn et al. 1987). Inthe study by Marcus et al. (Marcus etal. 1989) one patient was treated non-surgically and MRI at four monthsrevealed moderate signal intensitysubstance between the tendonfragments on PD-images suggestingincomplete healing with a fibrous ratherthan direct union. Liem et al. (Liem etal. 1991) in their paper describing apolylactid acid implant in repair ofAchilles tendon rupture found therepaired tendon area to be 4.4 to 6.4times larger than the unaffected side at6 to 13 months postoperatively. Theyreported only one conventionally,without the use of polylactide implant,

operated patient. In this single patientthe reunited Achilles tendon revealedless thickening and had a compactcentral zone of intermediate signalintensity on T1-weighted images.

Return to sports

The average return to normal activitiesafter Achilles tendon rupture has beenreported at 6.5 to 9.1 months (Beskinet al. 1987, Kellam et al. 1985). In thepresent prospective study, the lastfollow-up was at 6 months; obviouslyhealing still continued in patients whohad poor or good clinical results. Asstated by many authors, there isalmost always a small group of patientswith poor clinical outcomes (Arner andLindholm 1959, Carter et al. 1992,Kellam et al. 1985, Kiviluoto et al.1985). A rather unusual cause for pooroutcome could be a delay in treatmentthat can lead to the development of alarge gap between the ends of aruptured Achilles tendon (Mann et al.1991). According to the present study,another reason for unsatisfactoryhealing could be an extensiveintratendinous lesion, which can bedetected by MRI.

Reoperations

Three of our patients had reoperations.However, only one was surgicallytreated for a typical traumaticrerupture. This rate is similar to thatreported previously (Nistor 1981). Twopatients had second surgeries becauseof poor clinical outcomes, includingpersistent pain, abnormal walk, totallack of ability to perform heel raises,and large high signal intensity lesion onMR images. At surgery, the MRIfindings matched well with the soft,incompletely healed area inside thetendon. This type of unorganized tissuehas a higher mobile proton contentthan highly arranged type I maturetendon collagen (Peto et al. 1990).

52

These findings suggest that the highintensity areas seen on PD- and T2-weighted images inside the Achillestendon may represent active scarformation.

After the second surgery, the firstpatient recovered well and returned toprevious activities at 6 months fromreoperation. The second is still slowlyrecovering at 6 months afterreoperation. Despite the lesion seen on3-month images, one patient continuedwith conventional rehabilitationincluding temporary use of elasticdressing. The intratendinous lesiongradually reduced in size and heresumed normal walking, but he failedto return to his previous levels ofactivity, mainly because of the longrecovery period.

Functional tests and MRI

The correlation between the size of thehigh signal intensity area on MR imagesand functional recovery became evidentas five patients with abnormal walk at3 months had statistically significantlarger intratendinous lesion thannormally walking patients. The score inheel raises and the size of theintratendinous lesion showed inversecorrelation, indicating that the largerthe lesion the less the ability to do heelraises. This correlation was foundparticularly at 3 months but also at 6months. A standardized heel raisingtest as a measurement of the functionalwork capacity was chosen instead ofisokinetic evaluation with dynamometerbecause it is easily performed and hasfound to correlate well with reducedmuscle capacity (Haggmark et al.1986). However, in clinical evaluation,ability to walk remains as an importantfactor during the early rehabilitationperiod.

In Paper I, a significant correlation wasfound between the intratendinous

lesions on PD- and T2-weighted imagesand increased dorsiflexion at 3 months,which may reflect the lengthening ofthe Achilles tendon after surgery. Inprevious studies, Nistor suggested thattendon lengthening results in increasedrange of motion in dorsiflexion (Nistor1981). Mortenson et al. described howradiographic markers placed on theruptured Achilles tendon ends beforesurgery separated an average of 10.5mm during the first 7 weeks aftersurgery (Mortensen et al. 1992). Thisview is also supported by the fact thattwo of our patients with no visibleintratendinous lesion at 3 months hadmore restricted dorsiflexion of the ankle(>15°) than the others.

Miscallenous findings

The rupture site in our material (PaperI) was located an average of 5.8 cmabove the calcaneal insertion, and onlyone patient had a rupture lower than4.5 cm, which has been considered asthe upper limit of an area with fewvessels in the Achilles tendon (Carr andNorris 1989). Therefore, our results arein accordance with Schmidt et al.(Schmidt-Rohlfing et al. 1992) whorecently have stated that the area ofpoor vascularization and the rupturesite of the Achilles tendon do notnecessarily correlate with each other.

One of the interesting findings in PaperII was the increased intratendinoussignal intensity on T1-weighted imagesat the level of calcaneal corner.However, this finding did not seem tohave any clinical relevance, because all6 patients with this phenomenon hadgood outcome and did not complainany discomfort at calcaneal corner.

Adhesions between the tendon and theskin have previously been reported inas many as 40% of the patients withAchilles tendon rupture after 1-yearfollow-up (Solveborn and Moberg

53

1994). The surgical wound scar wasclearly detectable on MR images. At thesite of the skin scar, there was no highsignal intensity subcutaneous fat tissuebetween the skin and the tendon onaxial images. Therefore, free motion ofthe tendon may be prevented.

Ultrasonography

In US, our observations andmeasurements of the postoperativeAchilles tendons were in concordancewith those previously presented (Blei etal. 1986, Fornage 1986). Afteroperation the tendon was thickened(Fornage 1986) and sausage shaped(Blei et al. 1986). One yearpostoperatively, the reunited Achillestendon has been reported to graduallydecrease in thickness (Blei et al. 1986).In the present study both increasedand decreased echogenic areas wereseen in mixed fashion. The orientationof tendon fibers was not properlyorganized.

Only one of 4 patients withintratendinous lesions on MR imageshad abnormal (hypoechoid) appearancein US. MRI provided better details ofintratendinous changes of surgicallyrepaired Achilles tendon ruptures. Bothimaging modalities revealed the shapeof the tendon and the irregularity of theaffected tendon surface. US revealedcalcifications which was not seen onMRI.

The variable correlation between MRIand US in Achilles tendon dimensionswas somewhat expected. However, theoverall increased size of the operatedAchilles tendon was well visualized bothwith MRI and US.

MR imaging ofasymptomatic Achillestendon

Diameter

The MR appearance of the intactAchilles tendon has been describedpreviously (Åström et al. 1996, Marcuset al. 1989, Neuhold et al. 1992, Quinnet al. 1987). However, in ourevaluation we found additional details,which are important to recognize asnormal anatomy when interpreting MRstudies in the clinical setting. In ourcontrol patients and volunteers, theaverage antero-posterior diameter ofthe Achilles tendon was 5.2 mm, whichis slightly less than in previous MRstudies (Åström et al. 1996, Neuhold etal. 1992, Weinstabl et al. 1991).Åström et al. and Neuhold et al.reported 6 mm as the average of themaximum antero-posterior diameter offourteen and seven asymptomaticAchilles tendons, respectively (Åströmet al. 1996, Neuhold et al. 1992). Thesmall number of cases studied, largerpixel size and lack of interobserveraveraging in previous studies couldpartially explain the differencesdiscovered between previous studiesand our study.

Shape

Most of the tendons in our study hadflat or concave anterior margins andonly 10% had mild convexity (Table10). Similarly, Koivunen-Niemelä andParkkola in an ultrasound study, foundthat in 13% of the volunteers, theanterior margin is somewhat convex(Koivunen-Niemela and Parkkola 1995).In previous MR studies the cross-sectional shape of the Achilles tendonhas been described as oval or ellipse(Marcus et al. 1989, Neuhold et al.1992, Weinstabl et al. 1991) and theanterior margin as flattened or mildly

54

concave (Ferkel et al. 1991, Quinn etal. 1987). However, the small numberof subjects and low resolution ofimages may partly account thedifference to our results.

The wave-like crescent of the anteriorcontour of the tendon, seen as anormal variant, swept from the lateralside to the medial side on successivecranio-caudal axial images. The “wave”was most recognizable when paging theaxial images in rapid succession on theviewing console. This is a normalvariation seen in asymptomaticvolunteers and should not to bemistaken as a focal thickening due totendon pathology, which it mayresemble in the most pronouncedcases. We believe this appearance isdue to the intertwining ofgastrocnemius and soleus componentsof the tendon which rotate relative toeach other as they approach thecalcaneal corner. The anterior bundle atthe level of the musculotendinousjunction, seen in 10 tendons, is likely amanifestation of the two componentsremaining only partially fused at thetendon surface.

Another feature of the shape of theAchilles tendon is anterior bulging onsagittal images, which can also bemisleading because the course of thetendon is not always parallel to thesagittal plane of the MR scanner.Therefore, we recommend usage of acoronal scout view of the Achillestendon before placing sagittal imagesalong the long axis of the tendon.

Signal intensity

According to several previous reports,normal tendons, including the Achillestendon, have a low signal intensityappearance on all sequences

(Chandnani and Bradley 1994, Marcuset al. 1989, Neuhold et al. 1992, Quinnet al. 1987, Weinstabl et al. 1991).Schweitzer has stated that normalAchilles tendon has no internal signalon any sequence except at its insertion(Schweitzer 1993). In our study, 45%of tendons had signal intensity changeswithin their Achilles tendons on axialFLASH images with a short echo time.The increased signal intensity stripesand punctate foci were small in size,had longitudinal orientation and weretypically located in distal tendon, andnot only at the level of the insertion. Inagreement with our findings, somerecent studies have describedinhomogeneous signal of the normalAchilles tendon on certain pulsesequences (Mantel et al. 1996, Rollandiet al. 1995). Rollandi et al. reportedthat intratendinous signal intensityspots or straight lines could be seen onT1- and proton density -weightedimages (Rollandi et al. 1995). In thestudy by Mantel et al. MR images withhistologic micro-anatomic correlationshowed connective tissue septa withintratendinous vessels betweencollagen bundles resulting in punctatefoci and short linear high signalintensity foci on T1-weighted axialimages (Mantel et al. 1996).

Plantaris tendon

In cases where the plantaris tendonmerges with the medial margin of theAchilles tendon, the cleft seen justproximal can resemble a smallsuperficial tear. This is analogous to theinsertion of the meniscal-femoralligament on the posterior horn of thelateral meniscus reported as sometimesmimicking a tear (Vahey et al. 1990).Once one is familiar with this variant, itis easily recognized.

55

TABLE 10. PITFALLS IN ACHILLES TENDON MR IMAGING

Normal variant Percentage Sequence

Anterior bulging 6% sagittalAnterior convexity 9% axialAnterior wave 56% axialIntratendinous distal stripes 30% axial gradient echo with short TE*Intratendinous punctate foci(diameter < 2mm)

28% axial gradient echo with short TE*

Intratendinous lesions(diameter >2mm)

4% axial gradient echo with short TE*

Prominent retrocalcaneal fluid 12% axial or sagittal STIR/fat sat T2-WI

Increased peritendinous signal intensity 14% axial STIR or fat sat T2-WITE = echo time, WI = weighted images*on other sequences to lesser extent

Intratendinous lesions ofAchilles tendon

Asymptomatic subjects

The actual normal collageninfrastructure of the Achilles tendoncannot be imaged under the TEs andFOVs routinely used (Koblik andFreeman 1993). The small ground glassareas seen in four of the asymptomatictendons probably represent the earliestsign of tendon degeneration. Anotherexplanation is that a small percentageof tendon fibers are not parallel to themain axis and thus could reach obliqueenough orientation to show magicangle phenomenon (Jozsa and Kannus1997, Peto and Gillis 1990).

Symptomatic subjects

The FLASH sequence proved effectiveand accurate in detection ofintratendinous lesions. Small lesionswere found, even when not evident onpalpation. This effectiveness is inagreement with recent reports statingthat the highest intrinsic signal intendinous tissue is seen with short echotime gradient echo sequences (Koblikand Freeman 1993, Schick et al. 1995).Of the lesions found at surgery, 25%would have been missed without theuse of the FLASH sequence. The

increased detection of the lesions bythe FLASH sequence is likely a result ofhigh-resolution matrix and the shortTE. Only one patient had a smallintratendinous lesion not detected inthe preoperative MR imaging. This wasprobably caused by volume averagingbetween two slices. In three tendons,the intratendinous lesion appearedsomewhat larger than the lesiondiscovered at surgery. This appearancewas likely due to the sensitivity of theFLASH sequence for subtle changes notvisible macroscopically in the tendonstructure. In these tendons, the STIR-and PD–images more accuratelydelineated the actual macroscopic sizeof the lesion.

Histological specimens obtained in ourstudy confirmed that the lesions seenon FLASH images representedabnormal tendon tissue with a lack ofcontinuity of collagen fibers and anongoing repair process in the Achillestendon. This was seen in all 13specimens available from the operatedtendons. The histological findings alsocorrelated with the severity of the MRimaging findings; the high signalintensity center of the intratendinouslesion on STIR images (or T2-weightedimages) represented the areas withmore severe disruption of the normalcollagen structure of the tendon.

56

The shape, location or intensity of theintratendinous lesion did not allow us toclassify the lesion as degenerativeversus traumatic (partial tear), asreported by other researchers(Weinstabl et al. 1991). We found thatthe signal intensity and histology of thelesions deep in the tendon substanceappeared similar to the intensity andhistology of lesions extending to thesurface. Most often the intratendinouslesion had a longitudinal orientation,but several did appear in a single slice.The lesions were predominantly locatedantero-medially in the tendon.

Chronic tendon disorders are commonlyreferred to as tendinitis or partialruptures. This nomenclature has norelevance: tendinitis indicates aninflammatory reaction, which was,absent in all our chronic cases; andpartial rupture indicates a clearlydiscernible partial discontinuity, whichis seldom found in chronic cases. Asymptom-related diagnosis such asachillodynia is preferable until ahistopathological examination has beenmade. In patients in whom MR imagesreveal a tendon disorder (thickeningwith or without intratendinous lesion),the term tendinosis should be used.

Peritendinous tissues

Normal appearance

In the literature we found nodescription of the MR anatomy ofperitendinous layers. Galloway et al.have suggested that MR imaging maydifferentiate paratendinous adhesionsand swelling from true tendinousdisease (Galloway et al. 1992). Otherauthors have stated that MR imaging isnot able to demonstrate the paratenonor any abnormal changes within it(Åström et al. 1996, Schepsis et al.

1994). In our opinion, this may be dueto the spatial resolution of MR imagesused in the previous studies. Åström etal. used a pixel size of 0.78 x 0.78 mmand Neuhold et al. of 0.63 x 0.63 with5-mm slice thickness (Åström et al.1996, Neuhold et al. 1992). Our pixelsize of 0.54 x 0.35 with 4-mm slicethickness on axial FLASH images wassmaller than in previously publishedstudies. This increased spatialresolution was likely the reason that wewere able to detect finer anatomicalstructures such as paratenon. Theparatenon forms a distinct layer overthe dorsal aspect of the tendon, butnormally becomes indistinct towardsthe calcaneal insertion, as the cruralfascia and paratenon fuse with theposterior subcutaneous structures.

The higher signal intensity of theparatenon seen on STIR images at theproximally and distally portions of theAchilles tendon may be a reflection ofdecreased vascularity at the level of“critical zone” (Åström and Westlin1994, Carr and Norris 1989, Lagergrenand Lindholm 1958). High signalintensity subcutaneous venousstructures crossing posteriorly withrespect to the Achilles tendon maysuggest posterior peritendinitis on STIRimages and caution should be used ininterpreting high signal in this region.Pre-Achilles fat pad consistentlydemonstrated low intensity on STIR-and high intensity on T1-weightedimages with thin septa and vesselstraversing it, except in five cases wheremild increased signal intensity wasseen on STIR images. Any signalabnormality in this tissue must beviewed with a high degree of suspicionfor significant pathology.

The increased signal intensity detectedin the distal soleus muscle on STIRimages was another normal variant andnot related to Achilles tendon orperitendinous disease. This is likely due

57

to rich vascularity present in themusculotendinous junction.

Some studies have reported that MRimaging does not demonstrate theposterior paratenon, or any abnormalchanges related to it (Åström et al.1996, Schweitzer and Karasick 1994).High-resolution FLASH sequenceallowed us to consistently evaluate theappearance of the paratenon. On theSTIR sequence it is only faintly visible,when normal in thickness and signalintensity. The Achilles tendon lacks atrue synovial sheath, but instead relieson the thin membranes on the dorsal,medial and lateral sides that form theparatenon between the skin,subcutaneous fascia cruris and theAchilles tendon for gliding function(Jozsa and Kannus 1997, Kvist 1994).With the frequency encoding gradientof the scanner oriented antero-posteriorly, normal paratenon is seenas a distinct layer of soft tissue againstthe low intensity tendon. The paratenonis normally so thin that actualmeasurements of this structure couldnot be performed, but with experiencein the use of FLASH sequence theappearance of the normal paratenon isquite characteristic.

Abnormal appearance

Peritendinitis involves inflammation andedema in the peritendineal tissues,which are rich in vascularity (Jozsa andKannus 1997, Kvist 1994, Scioli 1994).The STIR sequence, which suppressesthe high signal emanating from fat, wasmost important in overall evaluation ofperitendinous signal changes. On STIRimages, a zone of increased signalintensity limited to the posteriorparatenon indicated paratenonitis.

Peritendinitis may also appear in thepre-Achilles (Kager’s) fat pad anteriorto the tendon (Schweitzer 1993,Schweitzer and Karasick 1994). It is

seen as increased signal intensity onSTIR images and as low intensitystrands on T1-weighted sagittal spin-echo images. As many as 38% of thepatients with peritendinitis had thisfinding in pre-Achilles fat pad, and in11 tendons this was the only finding.We suggest term “anteriorperitendinitis” to describe cases withabnormal signal intensity in this fatpad, and “posterior peritendinitis” and“paratenonitis” to describeinflammation and thickening of theparatenon, respectively.

The surgical appearance of chronicperitendinitis includes thickening andirregularity of the paratenon, whichcauses adhesions around the Achillestendon (Kvist and Kvist 1980, Kvist1994, Saxena 1995, Schepsis et al.1994). We found that patients with athickened paratenon had an associatedlonger duration of symptoms thanpatients without the thickenedparatenon. This finding is in agreementwith a histological and histochemicalstudy by Kvist et al. (Kvist et al. 1987),who found that chronic peritendinitisleads to hypertrophy of the paratenon.At surgery, the most commonperitendinous finding is the chronicallythickened paratenon posteriorly (Kvistand Kvist 1980). In our follow-up,conservatively treated patients withisolated anterior or posteriorperitendinitis, had a significantly betterprognosis than patients with combinedperitendinous and intratendinousabnormalities.

Tendon insertion andretrocalcaneal bursa

Normal appearance

At the level of the insertion, the mostnotable variation was identified withinthe retrocalcaneal bursae where 15

58

asymptomatic cases had additional fluidpotentially confused for an appearanceof bursitis. This is in agreement with astudy on normal and abnormalappearance of retrocalcaneal bursae(Bottger et al. 1998). However, in ourmaterial the average size of theenlarged bursae was larger. This maybe due to higher level of athleticactivity in our subjects.

Abnormal appearance

All patients with an MR diagnosis ofinsertional tendinosis (abnormal tendonappearance at insertion with or withoutcalcaneal bone marrow edema on STIRimages) had a clinical diagnosis ofinsertitis (pain at the level of thetendon insertion). Differentiatingbetween the underlying causes was notpossible on clinical examination. Thetendon tapers at the level of thecalcaneus, and at this levelintratendinous lesions were seenwithout the tendon being abnormallythick (>6 mm). This anteroposteriormeasurement should be used onlyproximal to the level of the calcanealcorner as a criterion to diagnoseabnormal tendon thickening related totendinosis.

Retrocalcaneal bursitis is considered adistinct entity from insertionaltendinosis (Allenmark 1992, Bottger etal. 1998, Kvist 1994, Scioli 1994).However, they may often coexist(Schepsis et al. 1994). In our study,increased fluid signal on STIR- and T2-weighted images indicatingretrocalcaneal bursitis was seen in 83%of the patients with insertionaltendinosis. MR imaging enabled us todifferentiate retrocalcaneal bursitisfrom an abnormality in the distalAchilles tendon or calcaneal bonemarrow edema. This is importantbecause patients with isolatedretrocalcaneal bursitis had excellent orgood long-term results. One false

positive case of bursitis hadexceptionally long time intervalbetween imaging and surgery. Theinflammatory process in the bursa mayhave subsided in the interim.

Multiple findings

The clinical diagnosis of pain in thehindfoot region can be challenging. Inthis study we found that similar clinicalsymptoms were caused by the diseasedtendon itself or peritendinous processesor both. In most cases, more than onetype of pathology was identified on MRimages. As many as 68% of thepatients with an intratendinous lesionhad peritendinitis, which was typicallycontiguous with the area of tendoninvolvement. Schepsis et al. (Schepsiset al. 1994) found a similar associationin their study of surgical treatment ofAchilles tendon overuse injuries.

MR imaging as prognosticmethod

Surgically treated patient with a focalintratendinous lesion had better long-term results than conservativelytreated patients. Surgical interventionseems to be warranted in cases ofprolonged Achillodynia caused by sucha lesion. The number of surgicallytreated patients in which no focal lesionwas present was too small for statisticalevaluation.

Our long-term follow-up indicates, thatthe prognosis is worse in cases ofinsertional tendinosis or when highintensity intratendinous lesions aredetected. Excluding patients withinsertional tendinosis, the patients withan intratendinous lesion have lesssatisfactory results than patientswithout such a lesion. Patients withisolated retrocalcaneal bursitis, isolatedperitendinitis and a normal MR imagingstudy had a good prognosis.

59

In the present study, surgically andconservatively treated patients withinsertional tendinosis (with or withoutretrocalcaneal bursitis) had a worseprognosis than any other subgroup.This is in concordance with the resultsof a previous study, in which distalpartial ruptures had less satisfactoryresults after surgery than proximalrupture (Morberg et al. 1997).However, Schepsis et al. (Schepsis etal. 1994) reported better results ininsertional tendinosis than in patientswith tendinosis more proximally.

Sequences

We evaluated the ability of MR imagingin diagnosing the causes of Achillodyniain athletic adults. High-resolution T1-weighted FLASH gradient echo, andfluid-sensitive STIR sequence werecompared with conventional spin-echoproton density and T2-weightedsequence in diagnosing these disorders.Histopathologic correlation of detectedtendon lesions was obtained.

In our protocol, the T2-weighted dualecho spin-echo sequence did not givesignificant additional information. Thecombination of sagittal T1-weightedimages, axial high-resolution FLASH,and axial STIR images provideddiagnostic information. The sagittalSTIR images are helpful in patients withpain at the level of the tendon insertionor musculotendinous junction. The totalimaging time with the protocolconsisting of the three sequencesmentioned above is approximately 15minutes, including patient positioning.In some instances, imaging of thecontralateral unaffected leg givesadditional information, but experiencewith the normal and abnormal MR

anatomy of the Achilles tendondecreases the need for this.

High vs. low field MRimaging

Lower spatial resolution of the low-fieldMR unit was sufficient to demonstratethe intratendinous changes thatoccured over time in the postoperativeAchilles tendons. MR imaging ofpostoperative tendon at high-field mayactually prove problematic due topresence of metallic artifacts. We haveobserved in occasional cases suchartifacts on gradient echo sequences(Karjalainen and Soila, unpublisheddata).

In evaluation of overuse injuries, thesmallest intra- and peritendinouschanges may not be detectable at low-field MR unit. Spatial resolution, lowersignal-to-noise and lack of short TEsequences have been limiting factorsin MR imaging of Achilles tendondisorders with low-field units.

Limitations

A limitation in the design of study IIIwas the select group of volunteers, notrepresentative of a patient populationof all ages and activity levels. Furtherstudies are needed in sedentaryvolunteers of varying age groups, inorder to establish imaging criteria forall patients. Our study was designed toestablish a baseline appearance for thetype of active patient commonlyreferred for MRI examination of theAchilles tendon. We wanted to illustratewhat is detected in an asymptomatictendon on high-resolution images,which are now becoming more widelyavailable.

60

CONCLUSIONS AND SUMMARY

1. MR imaging provides precise andvaluable tool to reveal the internalstructure of the surgically repairedAchilles tendon rupture. MRimaging shows remarkablepostoperative thickening seen at itslargest 3 months postoperatively,and thickening persists at 3 yearsas well. Typically, an intra-tendinous lesion detected onproton density- and T2-weightedMR images belongs to normalhealing during the reunion process.Patients with poor clinical outcomehave the largest intratendinouslesions postoperatively.

2. One to three years postoperatively,occasional intratendinous lesionson MR imaging as well as mixedechogenity in US seem to belong tothe normal stage of the surgicallyrepaired Achilles tendon rupture.

3. High-resolution MR imaging depictsthe Achilles tendon andperitendinous soft tissues in greatdetail. Normal anatomy of theAchilles tendon in asymptomatic,active population is variable whichone should be familiar with, asthey are a potential source ofmisinterpretation. The most typicalanatomical variations of theAchilles tendon are related toanterior margin of the tendonproper, inhomogenous intra-tendinous signal intensity changes(stripes and punctates) on high-resolution T1-weighted gradientecho sequence (FLASH), plantaris

tendon, length of the soleusmuscle, form and size of theretrocalcaneal bursa, andsegmented signal intensity of theposterior paratenon on STIRimages.

4. High-resolution MR imaging ofpatients with Achilles tendonoveruse injuries precisely detectsabnormalities in the entirelocomotor unit, including thetendon proper, the calcanealmarrow, the insertion, theretrocalcaneal bursa, theperitendinous tissues and themusculotendinous junction.Findings in MR imaging correlatewell with findings in surgery andwith histopathology. Histologicspecimen taken from intra-tendinous lesions showed anongoing repair process in theAchilles tendon. High-resolutionT1-weighted gradient echosequence (FLASH) proved accuratein detection of smallestintratendinous lesions. Peri-tendinous changes were bestdetected with STIR sequence,except the thickening of theparatenon, which was only seenwith FLASH. MR imaging also haspromising prognostic value.

We expect high-resolution MRimaging to have an increasinglysignificant role as an accurate andeffective diagnostic method in themanagement of patients withAchillodynia.

61

ACKNOWLEDGEMENTS

This study was carried out during 1993 – 1999 at the Department of Radiology,Helsinki University Hospital, in co-operation with the Department of Orthopedicsand Traumatology and Helsinki Deaconess Hospital.

I would like to thank the head of the Diagnostic Radiology, Professor Carl-GustafStandertskjöld-Nordenstam, M.D., who provided optimum conditions and anencouraging atmosphere during the course of the work.

Professor Hannu Aronen, M.D. has been my supervisor during the “long-acting”years of 1993 – 2000. His devotion and endless drive to magnetic resonanceimaging and scientific work has also inspired me to carry out this work during thelong hours in the hospital.

I am also indebted to Kalevi Soila, M.D., who has been my personal teacher inmusculoskeletal MR imaging. I thank him for careful pre-examination of mymanuscripts and reading hundreds of MR images with me. He has also taught me“re-thinking” of medical issues, and I admire his attitude for accurate, effective andevidence based MR imaging.

I wish to thank the official reviewers, Docent Ilkka Paakkala, M.D. and DocentSakari Orava, M.D. for their constructive criticism and advice during the revision ofthe manuscript.

I have had the privilege to collaborate with some of the most brilliant scientists:Docent, orthopedic surgeon Harri Pihlajamäki, M.D., Docent, pathologist TimoPaavonen, M.D., Docent, radiologist Juhani Ahovuo, M.D. and Docent, orthopedicsurgeon Ole Böstman, M.D. Olli Tynninen, M.D. has done major work in preparingpathological samples. Also, I wish to thank Professor Pekka Soila, M.D. (✝) havingtime to sit down and encourage me to continue my work.

My sincere thanks to Philip J. Tirman, M.D., San Francisco, for pre-reading my lasttwo major manuscripts in a very tight schedule.

I also had the honour to work with orthopedic surgeon Ilkka Tulikoura, M.D. andProfessor Pekka Peltokallio, M.D. They both have introduced and given me personalguidance to the field of sports medicine and Achilles tendon injuries.

In addition to my supervisor, I would like to thank the following senior scientists forthe helpful discussions over the years: Physicist Veli-Pekka Poutanen, Docent AnttiLamminen, M.D., Docent Sören Bondenstam, M.D. and Docent Ossi Korhola, M.D.

When I started this project at October 1993, Hannu Aronen had just returned fromHarward. At that time, lots of things had to be done alone, but during followingyears, “Hannu´s team” got many talented personalities, and also the facilitiesbecame remarkably better. Today, the power of his team consists of youngscientists with versatility, broad-mindiness and industriousness. Sami Martinkauppi,

62

medical student, has provided me additional support in computer technology. TimoTukeva, M.D., has been a big friend during this journey. Also, I want to thank allthe rest of the scientists of the Hannu’s team.

The staff of HUS-Röntgen MRI section has always been interested in my work, andmost of all; they have not criticized my working hours or my poor cleaning skills.Thank You.

My sincere thanks to my life time colleagues and friends Johanna Aaltonen, M.D.,Tuomo Karila, Lasse Lönnqvist, M.D., Karri Norio, M.D., Jarkko Pajarinen, M.D.,M.D., Mikko Sillanpää M.D. for both academic and non-academic support.

Then comes all the Achilles patients which were motivated to arrive for MR imagingas late as 3 a.m! In sports medicine the traditional doctor-patient relationship hasmore friendship, and the doctor has to spend more time to explain the clinical andradiographic situation. It is sometimes hard to be objective when it comes to bigdecisions such as long and full rests or operations. My colleagues in the field ofsports medicine, as well as physioterapeutists and coaches, have encouraged me tocontinue this work.

I want to thank all my friends for support in my “outside the scientific” life. MarkoJärvi has taken care of the nutritional aspects during long night-hours in thehospital. Science, and especially writing-part, needs well rested brains, therefore, Iwish to thank my golf-colleagues Gary, Fuzzy, Jesper and Tiger.

I wish to express my warmest thanks to my parents Erkki and Ritva Karjalainen,my brother and sisters and their families, and to my personal favorite, JohannaLassi for all their patience and support during the study.

P.O. Klingendahl foundation gave me the first grant in 1994. Since then this workhas been supported by Foundations of Pekka Peltokallio, the Radiological Society ofFinland, the Helsinki University Hospital Research Funds and Duodecim. Finally, Iwish to thank Jaana Lyyra and Jukka Vento for remarkable financial support duringthese years.

Helsinki, May 2000

63

References

Aksoy MC , Surat A. Fracture of the ossified Achilles tendon. Acta Orthop Belg 1998;4:418-421.

Alfredson H, Pietila T, Jonsson P , Lorentzon R. Heavy-load eccentric calf muscle training forthe treatment of chronic Achilles tendinosis. Am J Sports Med 1998;3:360-366.

Allenmark C. Partial Achilles tendon tears. Clin Sports Med 1992;4:759-769.Arner O , Lindholm Å. Subcutaneous rupture of the Achilles tendon: A study of 92 cases. Acta

Chir Scand (Suppl) 1959;1-51.Barfred T. Experimental rupture of the Achilles tendon. Comparison of experimental ruptures in

rats of different ages and living under different conditions. Acta Orthop Scand1971;5:406-428.

Barfred T. Kinesiological comments on subcutaneous ruptures of the Achilles tendon. ActaOrthop Scand 1971;5:397-405.

Beltran J , Mosure JC. Magnetic resonance imaging of tendons. Crit Rev Diagn Imaging1990;2:111-182.

Beltran J, Noto AM, Herman LJ , Lubbers LM. Tendons: High-field-strength, surface coil MRimaging. Radiology 1987;735-740.

Beskin JL, Sanders RA, Hunter SC , Hughston JC. Surgical repair of Achilles tendon ruptures.Am J Sports Med 1987;1:1-8.

Blei CL, Nirschl RP , Grant EG. Achilles tendon: US diagnosis of pathologic conditions. Work inprogress. Radiology 1986;3:765-767.

Bonner AC, Cloutier J, Econopouly J , Feitz D. Diagnosis of partial rupture of tendo Achillis withmagnetic resonance imaging. J Foot Surg 1990;3:212-217.

Bottger BA, Schweitzer ME, El-Noueam KI , Desai M. MR imaging of the normal and abnormalretrocalcaneal bursae. Am J Roentgenol 1998;5:1239-1241.

Bruce RK, Hale TL , Gilbert SK. Ultrasonography evaluation for ruptured Achilles tendon. J AmPodiatry Assoc 1982;1:15-17.

Bunnell S. Primary repair of severe tendons. Am J Surg 1940;502-516.Burk DL, Jr., Dalinka MK, Schiebler ML, Cohen EK , Kressel HY. Strategies for musculoskeletal

magnetic resonance imaging. Radiol Clin North Am 1988;3:653-672.Carden DG, Noble J, Chalmers J, Lunn P , Ellis J. Rupture of the calcaneal tendon. The early

and late management. J Bone Joint Surg Br 1987;3:416-420.Carr AJ , Norris SH. The blood supply of the calcaneal tendon. J Bone Joint Surg Br

1989;1:100-101.Carter TR, Fowler PJ , Blokker C. Functional postoperative treatment of Achilles tendon repair.

Am J Sports Med 1992;4:459-462.Cetti R, Christensen SE, Ejsted R, Jensen NM , Jorgensen U. Operative versus nonoperative

treatment of Achilles tendon rupture. A prospective randomized study and review of theliterature. Am J Sports Med 1993;6:791-799.

Chandnani VP , Bradley YC. Achilles tendon and miscellaneous tendon lesions. Magn ResonImaging Clin N Am 1994;1:89-96.

Cheung Y, Rosenberg ZS, Magee T , Chinitz L. Normal anatomy and pathologic conditions ofankle tendons: current imaging techniques. Radiographics 1992;3:429-444.

Clain MR , Baxter DE. Achilles tendinitis. Foot Ankle 1992;8:482-487.Clement DB, Taunton JE , Smart GW. Achilles tendinitis and peritendinitis: etiology and

treatment. Am J Sports Med 1984;3:179-184.Cummins EJ, Anson BJ , Carr BW. The structure of the calcaneal tendon in relation to

orthopaedic surgery with additional observations on the plantaris muscle. Surg GynecolObstet 1946;107-116.

Daffner RH, Riemer BL, Lupetin AR , Dash N. Magnetic resonance imaging in acute tendonruptures. Skeletal Radiol 1986;8:619-621.

Davies SG, Baudouin CJ, King JB , Perry JD. Ultrasound, computed tomography and magneticresonance imaging in patellar tendinitis. Clin Radiol 1991;1:52-56.

de Kerviler E, Leroy-Willig A, Clement O , Frija J. Fat suppression techniques in MRI: anupdate. Biomed Pharmacother 1998;2:69-75.

64

Denstad TF , Roaas A. Surgical treatment of partial Achilles tendon rupture. Am J Sports Med1979;1:15-17.

Draghi F, Calliada F, Fulle I, Madonia L, Bottinelli O , Campani R. Valutazione dei risultati dellaricostruzione dei tendini della gamba. Aspetti ecografici. Radiol Med (Torino)1999;5:337-340.

Dussault RG, Kaplan PA , Roederer G. MR imaging of Achilles tendon in patients with familialhyperlipidemia: comparison with plain films, physical examination, and patients withtraumatic tendon lesions. Am J Roentgenol 1995;2:403-407.

Erickson SJ. High-resolution imaging of the musculoskeletal system. Radiology 1997;3:593-618.

Erickson SJ, Fitzgerald SW, Quinn SF, Carrera GF, Black KP , Lawson TL. Long bicipital tendonof the shoulder: normal anatomy and pathologic findings on MR imaging. Am JRoentgenol 1992;5:1091-1096.

Erickson SJ, Prost RW , Timins ME. The "magic angle" effect: background physics and clinicalrelevance [editorial]. Radiology 1993;1:23-25.

Evancho AM, Fajman WA, Stiles RG, Brunner MC, Peterson J , Fleming L. Magnetic resonanceimaging of knee menisci. Comparison of spin echo pulse sequences. Invest Radiol1990;11:1224-1227.

Ferkel RD, Flannigan BD , Elkins BS. Magnetic resonance imaging of the foot and ankle:correlation of normal anatomy with pathologic conditions. Foot Ankle 1991;5:289-305.

Fornage BD. Achilles tendon: US examination. Radiology 1986;3:759-764.Fullerton GD, Cameron IL , Ord VA. Orientation of tendons in the magnetic field and its effect

on T2 relaxation times. Radiology 1985;2:433-435.Galloway MT, Jokl P , Dayton OW. Achilles tendon overuse injuries. Clin Sports Med

1992;4:771-782.Gerster JC, Landry M, Rappoport G, Rivier G, Duvoisin B , Schnyder P. Enthesopathy and

tendinopathy in gout: computed tomographic assessment. Ann Rheum Dis1996;12:921-923.

Gold GE, Pauly JM, Macovski A , Herfkens RJ. MR spectroscopic imaging of collagen: tendonsand knee menisci. Magn Reson Med 1995;5:647-654.

Haggmark T, Liedberg H, Eriksson E , Wredmark T. Calf muscle atrophy and muscle functionafter non-operative vs operative treatment of achilles tendon ruptures. Orthopedics1986;2:160-164.

Harms SE. Generation and manipulation of MR imaging. 2nd Philadelphia: Lippinscott - Raven,1997:1-20.

Haygood TM. Magnetic resonance imaging of the musculoskeletal system: part 7. The ankle.1997;336:318-336.

Inglis AE, Scott WN, Sculco TP , Patterson AH. Ruptures of the tendo achillis. An objectiveassessment of surgical and non-surgical treatment. J Bone Joint Surg Am 1976;7:990-993.

Jozsa L , Kannus P. Histopathological findings in spontaneous tendon ruptures. Scand J Med SciSports 1997;2:113-118.

Józsa L , Kannus P. Human tendons. Anatomy, physiology and pathology. Champaign, IL:Human Kinetics, 1997: 46-96.

Jozsa L, Kvist M, Balint BJ, et al. The role of recreational sport activity in Achilles tendonrupture. A clinical, pathoanatomical, and sociological study of 292 cases. Am J SportsMed 1989;3:338-343.

Jozsa L, Reffy A, Kannus P, Demel S , Elek E. Pathological alterations in human tendons. ArchOrthop Trauma Surg 1990;1:15-21.

Kager H. Zur Klinik und Diagnostik des Achillessehnenrisses. Chirurg 1939;691-695.Kainberger FM, Engel A, Barton P, Huebsch P, Neuhold A , Salomonowitz E. Injury of the

Achilles tendon: diagnosis with sonography. Am J Roentgenol 1990;5:1031-1036.Kalmar JA, Eick JJ, Merritt CR, et al. A review of applications of MRI in soft tissue and bone

tumors. Orthopedics 1988;3:417-425.Kannus P , Jozsa L. Histopathological changes preceding spontaneous rupture of a tendon. A

controlled study of 891 patients. J Bone Joint Surg Am 1991;10:1507-1525.

65

Karpakka J, Vaananen K, Virtanen P, Savolainen J, Orava S , Takala TE. The effects ofremobilization and exercise on collagen biosynthesis in rat tendon. Acta Physiol Scand1990;1:139-145.

Keene JS, Lash EG, Fisher DR , De SA. Magnetic resonance imaging of Achilles tendonruptures. Am J Sports Med 1989;3:333-337.

Kellam JF, Hunter GA , McElwain JP. Review of the operative treatment of Achilles tendonrupture. Clin Orthop 1985;201:80-83.

Khan KM, Bonar F, Desmond PM, et al. Patellar tendinosis (jumper's knee): findings athistopathologic examination, US, and MR imaging. Victorian Institute of Sport TendonStudy Group. Radiology 1996;3:821-827.

Kiviluoto O, Santavirta S, Klossner O, Sandelin J , Hakkinen S. Surgical repair of subcutaneousrupture of the Achilles tendon. Arch Orthop Trauma Surg 1985;5:327-329.

Koblik PD , Freeman DM. Short echo time magnetic resonance imaging of tendon. InvestRadiol 1993;12:1095-1100.

Koivunen-Niemela T , Parkkola K. Anatomy of the Achilles tendon (tendo calcaneus) withrespect to tendon thickness measurements. Surg Radiol Anat 1995;3:263-268.

Koivunen-Niemelä T. Imaging of the Achilles tendon. Diagnostic evaluation of sonography anfmagnetic resonance imaging. Thesis. Annales universitatis Turkuensis, Series D. Medica- Odontologica 190. University of Turku, Turku, Finland, 1995: 1-54.

Kvist H , Kvist M. The operative treatment of chronic calcaneal paratenonitis. J Bone Joint SurgBr 1980;3:353-357.

Kvist M. Achilles tendon overuse injuries. A clinical and pathophysiological study in athleteswith special reference to Achilles paratenonitis. Thesis. Annales UniversitatisTurkuensis. Series D. Medica - Odontologica 87. University of Turku, Turku, Finland,1991: 1-121.

Kvist M. Achilles tendon injuries in athletes. Sports Med 1994;3:173-201.Kvist M, Jozsa L, Jarvinen M , Kvist H. Fine structural alterations in chronic Achilles

paratenonitis in athletes. Pathol Res Pract 1985;4:416-423.Kvist M, Jozsa L, Jarvinen MJ , Kvist H. Chronic Achilles paratenonitis in athletes: a histological

and histochemical study. Pathology 1987;1:1-11.Kvist MH, Lehto MU, Jozsa L, Jarvinen M , Kvist HT. Chronic achilles paratenonitis. An

immunohistologic study of fibronectin and fibrinogen. Am J Sports Med 1988;6:616-623.

Lagergren C , Lindholm Å. Vascular distribution in the Achilles tendon: an angiographic andmicroangiographic study. Acta Chir Scand 1958;491-495.

Landvater SJ , Renstrom PA. Complete Achilles tendon ruptures. Clin Sports Med 1992;4:741-758.

Leadbetter WB. Cell-matrix response in tendon injury. Clin Sports Med 1992;3:533-578.Lehtinen A. Ultrasound studies of tendons and joints. Thesis. Department of Radiology and

Institute of Clinical Medicine. Helsinki University, Helsinki, Finland, 1996: 1-70.Lehto MU, Jarvinen M , Suominen P. Chronic Achilles peritendinitis and retrocalcanear bursitis.

Long-term follow-up of surgically treated cases. Knee Surg Sports Traumatol Arthrosc1994;3:182-185.

Lehtonen A, Makela P, Viikari J , Virtama P. Achilles tendon thickness inhypercholesterolaemia. Ann Clin Res 1981;1:39-44.

Leppilahti J. Achilles tendon rupture, with special reference to epidemiology and results ofsurgery. Thesis. Acta Universitatis Ouluensis D Medica 383. University of Oulu, Oulu,Finland, 1996: 1-66.

Leppilahti J , Orava S. Total Achilles tendon rupture. A review. Sports Med 1998;2:79-100.Leppilahti J, Orava S, Karpakka J , Takala T. Overuse injuries of the Achilles tendon. Ann Chir

Gynaecol 1991;2:202-207.Levi N. The incidence of Achilles tendon rupture in Copenhagen. Injury 1997;4:311-313.Liem MD, Zegel HG, Balduini FC, Turner ML, Becker JM , Caballero SA. Repair of Achilles

tendon ruptures with a polylactic acid implant: assessment with MR imaging. Am JRoentgenol 1991;4:769-773.

Lindholm A. New Method of Operation in Subcutaneous Rupture of the Achilles tendon. ActaChir Scand 1959;261-265.

66

Lucas P, Kaplan P, Dussault R , Hurwitz S. MRI of the foot and ankle. Curr Probl Diagn Radiol1997;5:209-266.

Lynn TA. Repair of the torn achilles tendon, using the plantaris tendon as a reinforcingmembrane. J Bone Joint Surg Am 1966;2:268-272.

Maffulli N, Dymond NP , Regine R. Surgical repair of ruptured Achilles tendon in sportsmen andsedentary patients: a longitudinal ultrasound assessment. Int J Sports Med 1990;1:78-84.

Maffulli N, Regine R, Angelillo M, Capasso G , Filice S. Ultrasound diagnosis of Achilles tendonpathology in runners. Br J Sports Med 1987;4:158-162.

Mammone JF. MR imaging of tendon injuries about the foot and ankle. Clin Podiatr Med Surg1997;2:313-335.

Mandelbaum BR, Myerson MS , Forster R. Achilles tendon ruptures. A new method of repair,early range of motion, and functional rehabilitation. Am J Sports Med 1995;4:392-395.

Mann RA, Holmes GB, Jr., Seale KS , Collins DN. Chronic rupture of the Achilles tendon: a newtechnique of repair. J Bone Joint Surg Am 1991;2:214-219.

Mantel D, Flautre B, Bastian D, Delforge PM, Delvalle A , Leclet H. Structural MRI study of theAchilles tendon. Correlation with microanatomy and histology. Journal de Radiologie1996;4:261-265.

Marcus DS, Reicher MA , Kellerhouse LE. Achilles tendon injuries: the role of MR imaging. JComput Assist Tomogr 1989;3:480-486.

Masciocchi C, Catalucci A , Barile A. Ankle impingement syndromes. Eur J Radiol 1998;S70-73.Maurer J, Requardt H, Herrling T, et al. Theoretical principles and technical realization of high

resolution nuclear magnetic resonance tomography with the example of a dedicated coilsystem. Bildgebung 1996;1:40-46.

Maurer J, Requardt H, Sander B, et al. Applications of specialized coils for high-resolution MRIon a whole-body scanner. Magma 1996;1:27-33.

Merhemic Z, Breitenseher M, Trattnig S, et al. MRT des Sprunggelenks. Vergleich einesGanzkorper-Mittelfeldgerats (1.0 T) mit einem dedizierten Niederfeldgerat (0.2 T).Radiologe 1999;1:41-46.

Merk H, Wissel H , Merkel M. Sonographische Nachuntersuchung nach operativ versorgtenAchillessehnenrupturen. Ultrascahll Med. 1997;6:254-257.

Moller A, Astron M , Westlin N. Increasing incidence of Achilles tendon rupture. Acta OrthopScand 1996;5:479-481.

Morberg P, Jerre R, Sward L , Karlsson J. Long-term results after surgical management ofpartial Achilles tendon ruptures. Scand J Med Sci Sports 1997;5:299-303.

Mortensen NH, Saether J, Steinke MS, Staehr H , Mikkelsen SS. Separation of tendon endsafter Achilles tendon repair: a prospective, randomized, multicenter study. Orthopedics1992;8:899-903.

Motta P, Errichiello C , Pontini I. Achilles tendon rupture. A new technique for easy surgicalrepair and immediate movement of the ankle and foot. 1997;2:172-176.

Movin T, Gad A, Guntner P, Foldhazy Z , Rolf C. Pathology of the Achilles tendon in associationwith ciprofloxacin treatment. Foot Ankle Int 1997;5:297-299.

Movin T, Kristoffersen-Wiberg M, Rolf C , Aspelin P. MR imaging in chronic Achilles tendondisorder. Acta Radiol 1998;2:126-132.

Movin T, Kristoffersen-Wiberg M, Shalabi A, Gad A, Aspelin P , Rolf C. Intratendinousalterations as imaged by ultrasound and contrast medium-enhanced magneticresonance in chronic achillodynia. Foot Ankle Int 1998;5:311-317.

Naito M , Ogata K. The blood supply of the tendon with a paratenon. An experimental studyusing hydrogen washout technique. Hand 1983;1:9-14.

Nakamura T, Yabe Y , Horiuchi Y. Fat suppression magnetic resonance imaging of thetriangular fibrocartilage complex. Comparison with spin echo, gradient echo pulsesequences and histology. J Hand Surg [Br] 1999;1:22-26.

Nesse E , Finsen V. Poor results after resection for Haglund's heel. Analysis of 35 heels in 23patients after 3 years. Acta Orthop Scand 1994;1:107-109.

Neuhold A, Stiskal M, Kainberger F , Schwaighofer B. Degenerative Achilles tendon disease:assessment by magnetic resonance and ultrasonography. Eur J Radiol 1992;3:213-220.

Nichols AW. Achilles tendinitis in running athletes. J Am Board Fam Pract 1989;3:196-203.

67

Nicolaisen T, Skovgaard D , Kjaer M. Conservative treatment of a partial Achilles tendonrupture with an intratendinous lesion. Scand J Med Sci Sports 1997;3:191-193.

Nistor L. Surgical and non-surgical treatment of Achilles Tendon rupture. A prospectiverandomized study. J Bone Joint Surg Am 1981;3:394-399.

Orava S, Leppilahti J , Karpakka J. Operative treatment of typical overuse injuries in sport. AnnChir Gynaecol 1991;2:208-211.

O'Reilly MA , Massouh H. Pictorial review: the sonographic diagnosis of pathology in theAchilles tendon. Clin Radiol 1993;3:202-206.

Paavola M, Kannus P, Paakkala T, Pasanen M , Järvinen M. Long-term prognosis of patientswith Achilles tendinopathy. An observational 8-year follow-up study. accepted forpublication in Am J Sports Med.

Paavola M, Orava S, Leppilahti J, Kannus P , Järvinen M. Chronic Achilles tendon overuseinjury: Complications after surgical treatment. Am J Sports Med 2000;1:77-82.

Paavola M, Paakkala T, Kannus P , Jarvinen M. Ultrasonography in the differential diagnosis ofAchilles tendon injuries and related disorders. A comparison between pre-operativeultrasonography and surgical findings. Acta Radiol 1998;6:612-619.

Palaniappan M, Rajesh A, Rickett A , Kershaw CJ. Accessory soleus muscle: a case report andreview of the literature. Pediatr Radiol 1999;8:610-612.

Panageas E, Greenberg S, Franklin PD, Carter AP , Bloom D. Magnetic resonance imaging ofpathologic conditions of the Achilles tendon. Orthop Rev 1990;11:975-980.

Pavlov H, Heneghan MA, Hersh A, Goldman AB , Vigorita V. The Haglund syndrome: initial anddifferential diagnosis. Radiology 1982;1:83-88.

Peto S , Gillis P. Fiber-to-field angle dependence of proton nuclear magnetic relaxation incollagen. Magn Reson Imaging 1990;6:705-712.

Peto S, Gillis P , Henri VP. Structure and dynamics of water in tendon from NMR relaxationmeasurements. Biophys J 1990;1:71-84.

Pettersson H, Hamlin DJ, Mancuso A , Scott KN. Magnetic resonance imaging of themusculoskeletal system. Acta Radiol Diagn (Stockh) 1985;3:225-234.

Puddu G, Ippolito E , Postacchini F. A classification of Achilles tendon disease. Am J Sports Med1976;4:145-150.

Quinn SF, Murray WT, Clark RA , Cochran CF. Achilles tendon: MR imaging at 1.5 T. Radiology1987;3:767-770.

Rand T, Bindeus T, Alton K, et al. Low-field magnetic resonance imaging (0.2 T) of tendonswith sonographic and histologic correlation. Cadaveric study. Invest Radiol 1998;8:433-438.

Reinig JW, Dorwart RH , Roden WC. MR imaging of a ruptured Achilles tendon. J Comput AssistTomogr 1985;6:1131-1134.

Reiser VM, Rupp N, Lehner K, Paar O, Gradinger R , Karpf PM. Die Darstellung derAchillessehne im Computertomogramm. Normalbefunde und pathologischeVeranderungen. Rofo 1985;2:173-177.

Resnick D, Feingold ML, Curd J, Niwayama G , Goergen TG. Calcaneal abnormalities in articulardisorders. Rheumatoid arthritis, ankylosing spondylitis, psoriatic arthritis, and Reitersyndrome. Radiology 1977;2:355-366.

Rollandi GA, Bertolotto M, Perrone R, Garlaschi G , Derchi LE. MRI of normal Achilles tendon.Eur Radiol 1995;596-598.

Romanus B, Lindahl S , Stener B. Accessory soleus muscle. A clinical and radiographicpresentation of eleven cases. J Bone Joint Surg Am 1986;5:731-734.

Saxena A. Surgery for chronic Achilles tendon problems. J Foot Ankle Surg 1995;3:294-300.Schepsis AA, Wagner C , Leach RE. Surgical management of Achilles tendon overuse injuries. A

long-term follow-up study. Am J Sports Med 1994;5:611-619.Schick F, Dammann F, Lutz O , Claussen CD. Adapted techniques for clinical MR imaging of

tendons. Magma 1995;2:103-107.Schmidt-Rohlfing B, Graf J, Schneider U , Niethard FU. The blood supply of the Achilles tendon.

Int Orthop 1992;1:29-31.Schmitgen G , Haasters J. Bei "Achillodynie" auch an seltene Ursachen denken--Ein Fallbericht

uber ein Schwannom des Nervus tibialis. Sportverletz Sportschaden 1990;1:45-47.

68

Schweitzer ME. Magnetic resonance imaging of the foot and ankle. Magn Reson Q 1993;4:214-234.

Schweitzer ME , Karasick D. MRI of the ankle and hindfoot. Semin Ultrasound CT MR1994;5:410-422.

Scioli MW. Achilles tendinitis. Orthop Clin North Am 1994;1:177-182.Sella EJ, Caminear DS , McLarney EA. Haglund's syndrome. J Foot Ankle Surg 1998;2:110-

114; discussion 173.Soila KP, Viamonte MJ , Starewics PM. Chemical shift misregistration effect in magnetic

resonance imaging. Radiology 1984;819-820.Solveborn SA , Moberg A. Immediate free ankle motion after surgical repair of acute Achilles

tendon ruptures. Am J Sports Med 1994;5:607-610.Steinborn M, Heuck A, Maier M, Schnarkowski P, Scheidler J , Reiser M. MR-Tomographie der

Plantarfasciitis. Rofo 1999;1:41-46.Stephens MM. Haglund's deformity and retrocalcaneal bursitis. Orthop Clin North Am

1994;1:41-46.Stiskal M, Szolar DH, Stenzel I, et al. Magnetic resonance imaging of Achilles tendon in

patients with rheumatoid arthritis. Invest Radiol 1997;10:602-608.Thermann H, Reimer P, Milbradt H, Zwipp H, Hoffmann R , Wippermann B. Sonographische

Primardiagnostik und Verlaufskontrolle von Muskel- und Sehnenschaden der unterenExtremitat. Unfallchirurg 1992;8:412-418.

Troop RL, Losse GM, Lane JG, Robertson DB, Hastings PS , Howard ME. Early motion afterrepair of Achilles tendon ruptures. Foot Ankle Int 1995;11:705-709.

Ulrich I, Pedersen BH , Lindewald H. CT af akillessener. Ugeskr Laeger 1991;18:1287-1288.Vahey TN, Bennett HT, Arrington LE, Shelbourne KD , Ng J. MR imaging of the knee:

pseudotear of the lateral meniscus caused by the meniscofemoral ligament. Am JRoentgenol 1990;1237-1239.

Waterston SW, Maffulli N , Ewen SW. Subcutaneous rupture of the Achilles tendon: basicscience and some aspects of clinical practice. [Review] [124 refs]. 1997;4:285-298.

Weinreb JC, Brateman L, Babcock EE, Maravilla KR, Cohen JM , Horner SD. Chemical shiftartifact in clinical magnetic resonance images at 0.35 T. Am J Roentgenol 1985;1:183-185.

Weinstabl R, Stiskal M, Neuhold A, Aamlid B , Hertz H. Classifying calcaneal tendon injuryaccording to MRI findings. J Bone Joint Surg Br 1991;4:683-685.

Weinstabl R, Stiskal P, Neuhold A , Hertz H. MR and ultrasound study of Achilles tendon injury.Unfallchirurgie 1992;4:213-217.

Wills CA, Washburn S, Caiozzo V , Prietto CA. Achilles tendon rupture. A review of theliterature comparing surgical versus nonsurgical treatment. Clin Orthop 1986;156-163.

Yoshioka H, Anno I, Niitsu M, Takahashi H, Matsumoto K , Itai Y. MRI of muscle strain injuries.J Comput Assist Tomogr 1994;3:454-460.

Yu JS, Witte D, Resnick D , Pogue W. Ossification of the Achilles tendon: imaging abnormalitiesin 12 patients. Skeletal Radiol 1994;2:127-131.

Åström M. Partial rupture in chronic achilles tendinopathy. A retrospective analysis of 342cases. Acta Orthop Scand 1998;4:404-407.

Åström M, Gentz CF, Nilsson P, Rausing A, Sjoberg S , Westlin N. Imaging in chronic achillestendinopathy: a comparison of ultrasonography, magnetic resonance imaging andsurgical findings in 27 histologically verified cases. Skeletal Radiol 1996;7:615-620.

Åström M , Rausing A. Chronic Achilles tendinopathy. A survey of surgical and histopathologicfindings. Clin Orthop 1995;316:151-164.

Åström M , Westlin N. Blood flow in chronic Achilles tendinopathy. Clin Orthop 1994;308:166-172.