The value of MRI STIR signal intensity on return to play prognosis and reinjury risk estimation in athletes with acute hamstring injuries

(1)

ContentslistsavailableatScienceDirect

Journal

of

Science

and

Medicine

in

Sport

j o u r n al ho me p ag e:w w w . e l s e v i e r . c o m / l o c a t e / j s a m s

The

value

of

MRI

STIR

signal

intensity

on

return

to

play

prognosis

and

reinjury

risk

estimation

in

athletes

with

acute

hamstring

injuries

R.A.

van

der

Horst

a,∗

_,

_J.L.

_Tol

d,e,i

_,

_A.

_Weir

f,i

_,

_J.M.

_den

_Harder

b

_,

_M.H.

_Moen

h

_,

_M.

_Maas

b

_,

G. Reurink

c,d,e,g

a_Department_of_Sports_Medicine,_Amsterdam_University_Medical_Centers,_The_Netherlands

b_Department_of_Radiology_and_Nuclear_Medicine,_Amsterdam_University_Medical_Centers,_The_Netherlands

c_Department_of_Orthopaedic_Surgery,_Amsterdam_Movement_Sciences,_Amsterdam_University_Medical_Centers,_The_Netherlands

d_Academic_Center_of_Evidence_Based_Sports_Medicine_(ACES),_Amsterdam_University_Medical_Centers,_The_Netherlands

e_Amsterdam_{Collaboration}_for_Health_and_Safety_in_Sports_(ACHSS),_Amsterdam_University_Medical_Centers,_The_Netherlands

f_Department_of_{Orthopaedics,}_Erasmus_Medical_Centre,_The_Netherlands

g_Department_of_Sports_Medicine,_OLVG,_The_Netherlands

h_Department_of_Sports_Medicine,_Bergman_Clinics,_The_Netherlands

i_Aspetar_Sports_Groin_Pain_Centre,_Aspetar_Orthopaedic_and_Sports_Hospital,_Qatar

a

r

t

i

c

l

e

i

n

f

o

Articlehistory:

Received10November2020

Receivedinrevisedform26January2021

Accepted14February2021 Availableonlinexxx Keywords: Hamstringmuscles MRI Reinjury Returntoplay Injury

Sportsandexercisemedicine

a

b

s

t

r

a

c

t

Objectives:Previousstudieshaveshownlowtomoderateevidenceforavarietyofmagneticresonance imaging(MRI)featuresasprognosticfactorsinathleteswithhamstringinjuries.Short-tauinversion recovery(STIR)signalintensityhasnotyetbeeninvestigatedforassessingtheprognosisofacutemuscle injuries.OuraimwastoexploretherelationshipbetweenMRISTIRsignalintensityandtimetoreturnto play(RTP)andtoinvestigatetheassociationbetweenMRISTIRandreinjuryriskinathleteswithacute hamstringinjuries.

Studydesign:Case-controlstudy.

Methods:WeusedMRISTIRtomeasureintramuscularsignalintensityinpatientswithclinicallydiagnosed hamstringinjuriesattwotimepoints:atinjuryandRTP.Atinjury,wecalculatedtheassociationofMRI STIRsignalintensitywiththetimetoRTPandreinjuryrisk.AtRTP,theassociationofMRISTIRsignal intensityandreinjuryriskandthechangeinMRISTIRsignalintensityovertimeonreinjuryriskwas evaluated.

Results:51patientswereincluded.WefoundincreasedMRISTIRsignalintensity:(1)attimeofinjury nottobeassociatedwithtimetoRTP,(2)attimeofinjurytobeassociatedwithaslightlylowerriskfor reinjury:odds0.986(0.975–0.998,p=0.02)and(3)atRTPnottobeassociatedwithreinjuryrisk.(4)We foundnoassociationbetweenthechangeinMRISTIRsignalintensityovertimeandreinjuryrisk. Conclusion:IncreasedMRISTIRsignalintensityatinjuryhasnovalueintimetoRTPprognosis,butis associatedwithareducedreinjuryrisk.

Practicalimplications

• Intra-muscularsignalintensityafterinjuryhasnovaluein pre-dictingtimetoRTP.

• Intra-muscularsignalintensityafterinjuryisassociatedwitha slightlyreducedriskonreinjury.

Abbreviations: MRI,magneticresonanceimaging;STIR,short-tauinversion

recovery;RTP,returntoplay;PRP,plateletrichplasma.

∗ Correspondingauthor.

E-mailaddress:Robertvdhorst@hotmail.com(R.A.vanderHorst).

• ThepresentﬁndingsmayencourageclinicianstoleavetheMRI asideindecisionmakingwithregardtoRTP.

1. Introduction

Hamstringstraininjuries(HSI)areknowntobeoneofthemost commoninjuriesinsports.1,2_They_often_result_in_lengthy_absence

fromsport,rangingfromacoupleofweekstoseveralmonthsand highrecurrencerates,rangingfrom12 to33%.1 _There_is_a_clear

clinicalneedforeitherprimarypreventionorpropercounselingto minimizetheriskofrecurrence.3

Systematicreviewshaveshownthatthereisnostrongevidence forcommonlyusedMagneticresonanceimaging(MRI)parameters

https://doi.org/10.1016/j.jsams.2021.02.008

(2)

inpredictingreturntoplay(RTP)orreinjuryrisk.4,5_These_MRI

ﬁnd-ingsincludebutarenotlimitedto:radiologicalgrade,11_radiological

size,6,7,8_,_absence_of_{hyperintensity}6,8,9 _and_radiological _location

andtendoninvolvement.7,10_It_was_shown_that_the_majority_of_the

correlationswerefoundbyunivariateanalysis,hadahighriskof biasandwereoftenconﬂicting.4,5

ApotentiallyrelevantMRIfeaturethathasnotyetbeen investi-gatedisthesignalintensityonﬂuidsensitivesequencesandtheir changesovertime.Theuseofﬂuidsensitivesequencesisnotnewin HSIresearch.10,12_Some_lesions_may_appear_brighter_white_to

clini-ciansthanotherlesions,thebrightness(i.e.signalintensity)canbe measuredonfluidsensitivesequences.Thishasnotyetbeen inves-tigated.Commonlyadoptedmethodstovisualizeedemaonfluid sensitivesequencesare(1)Short-tauinversionrecovery(STIR),(2), frequencyselectivefatsuppressionand(3)Dixon.13_{Quantification}

ofsignalintensityonﬂuidsensitivesequenceshasshownpromise forassessingtheseverityofavarietyofmyopathiescharacterized byintra-muscularedema.14_{Quantiﬁcation}_of_signal_intensity_has

alsobeenusedintheassessmentofwristinjuriesofgymnasts,15

andtoassessdiseaseactivityinjuvenileidiopathicarthritis.16_As

edemaandhemorrhageareabsorbedovertime,changesinsignal intensitymaybeassociatedwithmusclerecovery.We hypothe-sizedthattheextentofincreasedsignalintensityonﬂuidsensitive sequences is associated witha longer period until RTP and an increasedreinjuryrisk.

Thepurposeofthispaperistoexploretherelationshipbetween MRISTIRsignalintensityandtimetoreturntoplayandto investi-gatetheassociationbetweenMRISTIRandreinjuryriskinathletes withacutehamstringinjuries.Inordertodoso,wecalculatethe associationsofMRISTIRsignalintensity:(1)atinjuryontimeto RTP,(2)atinjuryonreinjuryrisk,(3)atRTPonreinjuryriskand(4) thechangeinMRISTIRsignalintensityonreinjuryrisk.

2. Methods

The patients in this study consist of a cohort that partic-ipated in a double blind randomized controlled trial on the effect of platelet-rich plasma (PRP): Dutch trial register num-ber2771.ParticipantswereenrolledbetweenFebruary2011and November2012.Atinclusion,informedconsentwasobtainedfrom allpatients.ApprovalwasobtainedfromtheRegionalEthical Com-mitteeofSouthWestHollandandtheEthicalCommittee.Similar studiesconcerningMRIatinjurywiththiscohort(n=165,n=64, n=70)andRTPwiththiscohort(n=108&n=53)wereperformed byourstudygroup.17–21_These_studies_{speciﬁcally}_looked_at

intra-musculartendoninjury,radiologicalsizeofthelesion,radiological gradeoftheinjuryandﬁbrosis,absenceofincreasedintra-muscular signal intensity at RTP and the association with re-injury. The prognosticvalueofquantiﬁedMRISTIRsignalintensitywasnot investigatedaspartofthesestudies.Themethodsandresultsof thisstudyhavebeencomprehensivelydescribedpreviously.21_In

brief,patientswererandomlyallocatedtotheplaceboorPRPgroup. Thecontrolgroupreceived3mlsalineinjections,theintervention groupreceived3mlPRPinjections.Bothgroupsreceivedtwo injec-tionsatthesiteofmaximalmuscleinjury:theﬁrstwithin5days ofinjury, thesecond5–7dayslater.Participantsofbothgroups completedanidenticalstandardizedrehabilitationprogramwhich haspreviously beendescribed indetail.21 _In_brief: _both_groups

performed an identical daily progressive phased,criteria-based rehabilitationprogramconsistingofdailyhomeexercisesandtwice weeklyphysiotherapistsupervisedtrainingsessions.Patientswere instructedtokeepdailylogstoimproveandmonitoradherence. Thephysiotherapistsupervisedprogramwasmodiﬁedfrom Hei-derscheitetal.,2010.3

Table1

Eligibilitycriteria.

Inclusioncriteria Exclusioncriteria *Age18–50years *ContraindicationtoMRI *Clinicaldiagnosisofacute

hamstringinjury

*Chronichamstringinjury *PresentingandMRIwithin5days

ofinjury

*Causeofinjuryisanextrinsictrauma *MRIconﬁrmedgradesIorII

hamstringlesion

*Notcapableofperforming rehabilitation

*SecondMRIavailablewithin7 daysofRTP

*Nointentiontoreturntofullsports activity

*Presenceofincreasedsignal intensityoninitialscan

*Unwillingtoreceivethe intramuscularinjections *NoRTPMRIavailable

*Morethan7daysbetweenRTPand secondMRI

*NoReinjurydataavailable RTP,Returntoplay;MRI,magneticresonanceimaging.

Clearanceforsportresumptionwasgivenbyasportsmedicine physician once the athlete successfully and asymptomatically fulﬁlledtherehabilitationprogramsupervisedbyasports phys-iotherapist. The athletehad tobesymptom-free (e.g. pain and stiffness)during: full range ofmotion, full speed sprinting and duringsport-speciﬁcmovementssuchasjumpingandcutting.

Participantswerefollowedoverthecourseofoneyearafter ini-tialinjurytoregisterreinjuries.Reinjurywasdeﬁnedasanacute onsetofposteriorthighpainatthesamesideastheinitialinjury resultinginabsencefromplay.Athleteswererequestedtocontact thecoordinatingresearcher immediatelyin theevent ofa sus-pectedreinjury,andreinjuryoccurrencewasmonitoredat4,8,16, 26,and52weekswithtelephonecallstotheparticipants.This orig-inalstudyfoundnodifferencesbetweenPRPandplaceboinjections onthetimetoRTPandreinjuryratewithinoneyear,neitheronany ofthehamstringsfunctionrelatedsecondaryoutcomemeasures (e.g.strength,ﬂexibility,subjectiverecovery).Eligibilitycriteriaof thepresentstudyarepresentedinTable1.Thesecriteriaconsists oftheeligibilitycriteriaofthePRPstudyandthepresenceofboth abaselineandRTPMRI.

AllparticipantsunderwentMRItwice.ThefirstMRIwas per-formedwithin5daysoftheinitialinjuryandthesecondwithin 7daysofRTP.InallcasestheinitialMRIprecededtheinjection procedure.Participantsofthecurrentstudywerebothfromthe PRPandSalinegroup.SincethebaselineMRIismadepriorsaline andPRPadministration,theinjectionwillnothaveeffectonthe measurementstakenofthebaselineMRI.Tocorrectforapossible confoundingeffectoftheSalineorPRPinjectionontheRTPMRI measurementswewilladjustfortheinterventionofaPRPorSaline injection.Effectoftheseinjectionsseemunlikely as histopatho-logicalrodentstudieswithsalineinjectionsshowedonlyminimal edematouschangesforthefirsttwodaysandRTPMRIwas con-ductedatamedianof14daysafterthefinalinjection.22

MRIsettingswerekeptconstantacrosspatients.Pleasesee Sup-plementaryAppendixforMRIprotocoldetails.

Image acquisition is described in detail in Supplementary Appendix,in brief: two MRI imagesper patientwere recorded usingtheImageJsoftwareprogram(NationalInstitutesofHealth, Bethesda,Maryland,USA):oneT0image(MRIattimeofinjury) andthecorrespondingRTPimage.Aswewerespecifically inter-estedinintra-muscularedemaandhemorrhage,weonlymeasured increased signal intensity within the muscle. Increased signal intensitywasdefinedasanabnormallyincreasedsignalinthe intra-musculartissuecomparedwiththeunaffectedsurroundingmuscle tissue.Imageswerematchedusinganatomicallandmarks.Signal intensitywasquantifiedusingthegrayvaluemeasurementtoolof ImageJ.SignalintensityisexpressedinGrayvalue.Grayvalueis

(3)

Fig.1. (A)Shorttau-inversionrecovery(STIR)imageoftheinitialinjuryshowingincreasedintra-muscularsignalintensityofthemusculusbicepsfemoris,indicatingedema

andhemorrhage.Theareaofincreasedsignalintensityisencircledinordertomeasurethemeanintramuscularsignalintensity.(B)STIRimageatRTPshowingareduction

ofedemaandhemorrhagecomparedtotheinitialinjury.Themeasurementtemplateusedfortheinitialinjurywaspositionedinsuchawaythatitwouldcorrespondtothe

imageofinitialinjury.Measurementsweretakenandrecordedforfurtheranalysis.(C)STIRimageatinjurydepictinganexampleofminorincreasedsignalintensity.(D)

STIRimageatRTPshowinganalmostcompleteresolutionofedemaandhemorrhagecomparedtotheinitialinjury.

deﬁnedasthesumofthegrayvaluesofallthepixelsdividedby thenumberofpixelswithinaselection(Fig.1).

Asinprinciple,absolutesignalintensitiescannotbecompared directlybetweendifferentscans,signalintensitywasnormalized beforeanalysis,pleaseseeSupplementaryAppendixforadetailed descriptionofthisprocedure.Inbrief:weperformednormalization bytaking3referencemeasurementsofthevastusintermediusin thesameimage,anddividingthemeansignalintensitywithinthe ROIintheinjurybythemeansignalintensityinthesereference measurementsasillustratedinFig.2inSupplementaryAppendix. StatisticalanalysiswasperformedusingSPSSsoftware(V.20.0; SPSS,Chicago,Illinois,USA).Weusedtheindependentt-testto ana-lyzedifferencesinintra-muscularsignalintensitybetweengroups. Toassesstheassociationofincreasedintra-muscularsignal inten-sityontimetoRTPandreinjuryrisk,linearregressionandlogistic regressionweredeployedrespectively.Weadjustedforipsilateral hamstringinjuries,SalineandPRPinjectionsandagebyadding thesevariablesascovariatestothelinearandlogisticregression analysis.Intra-andinterobserverreproducibilitywasdetermined bycomputingtheintraclasscorrelationcoefﬁcients(ICC)from two-waymixed-effectsANOVA.PleaseseeSupplementaryAppendixfor detailsonthereliabilityanalysisanddetailsaboutstatistics.

3. Results

Weincluded51patientsthatmettheeligibilitycriteria(Table1) intheanalysis.Intotal29of80patientswereexcludedofwhichthe majority(n=16)wereexcludedduenofollow-upMRIwas avail-able.Intheoriginalstudy6of80patientsdidnotcompletethe1 yearfollow-up,4werecensoredastheysustainedanotherinjury beforetheyreturnedtoplay.1patientwaslosttofollow-upafter RTPand1patientdidnotreturntoplaywithinthestudyperiod.All ofthesepatientswereexcludedinthecurrentstudyeitherdueno RTPMRIwasavailableorthelackofreinjurydata.Theflowdiagram (Fig.3)inSupplementaryAppendixgivesanoverviewon exclu-sioncriteria.PatientcharacteristicsaresummarizedinTable2.The mediantimetoRTPwas42days(range14–99days).Themedian timebetweeninjuryandthefirstMRIwas2days(range1–5days). ThemediantimebetweenRTPandsecondMRIwas4days(range0 daysbefore–7afterRTP).Themediantimebetweenfinalinjection andsecondMRIwas23days(range5–71days).

Intra-observervariabilityoftheoutcomemeasurement RTPN

andT0,Nafternormalizationwereexcellent:ICC=0.98(0.96–0.98).

Inter-observervariabilityofthesemeasurementswereexcellentas well:ICC=0.98(0.94–0.97).

(4)

Fig.2.(A,B&C)STIRimagesoftheinitialinjuryshowingincreasedintra-muscularsignalintensityofthebicepsfemorismuscleindicatingedemaandhemorrhage.

Normal-izationmeasurementsweretakenatthevastusintermediusmuscleinsuchawaythatnovesselsorﬁbroustissuewasincorporated.Generally,onemeasurementwastaken

atthe(A)lateralsectionofthevastusintermediusmuscle,oneatthe(B)middlesection)andoneatthe(C)medialsection.Inrareoccasions,presenceofvesselsorthesize

oftheintermediusmuscledidnotallowformeasurementsaccordingtothedescribedmethod.Inthiscasetwomeasurementswouldbetakenateitherthelateral,medial

orinnersectiondependingonwhichsectionwasnotmeasurable.

Fig.3.Flowdiagramdepictingpatientsincludedinthestudy.

Table2

Patientcharacteristicsn=51.

Medianage(range) 27(18–45) Gender,male/female 50/1 Sports Football 39 Fieldhockey 7 Americanfootball 1 Tennis 1 Athletics 3 Levelofsports Competitive 38 Recreational 13

AKolmogorov-SmirnovtestindicatesthattheT0andRTPsignal intensity measurementswerenormally distributedD(53)=0.06, p>0.20andD(53)=0.09,p>0.20.Linearregressionshowedno sig-niﬁcantassociationbetweenbaselineSTIRvaluesandtimetoRTP: betacoefﬁcient−0.001(−0.008–0.007,p=0.98).

Theparticipantswithoutareinjurywithin1yearafterprimary injury,hadsigniﬁcantlyincreasedintra-muscularsignalintensity onthebaselineMRIcomparedtotheparticipantswithareinjury; 241(95%CI,221–261)vs.198(95%CI,169–227),p=0.02,Cohen’s d=0.74indicatingamediumeffectsize.Thesigniﬁcantdifference between groupsat baseline wasfurtheranalyzedusing logistic regression showinganoddsratioof0.986(0.975–0.998,p=.02),

indicatinganincreasedriskofreinjuryinpatientswithlower intra-muscularsignalintensityattimeofinjury.Noevidencewasfound ofaconfoundingeffectofthevariables:previoushamstringinjury, PRP-,Salineinjectionorage.

No signiﬁcant difference was detected at the MRI at RTPN

betweenthegroupwithoutareinjuryandwithreinjury;139(95% CI,125–153)vs.122(95%CI,104–140),odds0.987(0.970–1.005, p=0.15)

Nosigniﬁcantdifferencewasdetectedin‘intra-muscularsignal intensity’reductionbetweentheinjuryMRIandtheRTPMRI,odds 0.986(0.972–1.001,p=0.06)(Table3).

4. Discussion

ThisisthefirststudythatevaluatedthevalueofMRISTIRsignal intensityforacutehamstringinjuries.Therearetwomain find-ings.Firstly,intra-muscularsignalintensityonfluidsensitiveMRI imagesattimeofinjuryis notassociatedwiththetimetoRTP. Secondly, lowersignal intensityat injury is associated withan increasedriskforreinjury.

WedidnotﬁndanassociationbetweenMRISTIRsignal inten-sityattimetoRTPMRIandreinjuryrisk.Therewasnoassociation betweentime-courseMRIchangesofintra-muscularsignal inten-sityandreinjuryrisk.

(5)

Table3

Meanincreasedintra-muscularsignalintensityvaluesonﬂuidsensitiveMRISequencesinpatientswithandwithoutreinjury.

Reinjury(95%CI∨) Noreinjury(95%CI∨) Total(95%CI) IndependentT-test

(n=17) (n=34) (n=51) p

T0,N* 198(169–227) 241(221–261) 227(210–244) 0.02*

RTPN 122(104–140) 139(124–153) 133(122–144) 0.15

RTPN·T0,N◦ 76(55–98) 102(86–118) 94(81–107) 0.06

T0,Ntimeofinitialinjury;RTPN,Returntoplay;MRI,magneticresonanceimaging.

∨95%CIintervalofthemean.

*Statisticallysigniﬁcantdifferencebetweenreinjuryandnoreinjuryp=0.01.

◦_Mean_difference_in_increased_{intra-muscular}_signal_intensity_between_RTP_and_T0_relative_to_T0.

Fig.4.ScatterplotdepictingtherelationshipbetweentimetoRTPand‘intra-muscularsignalintensity’attimeofinjury.

We found no association between increased intra-muscular signalintensityonMRIatinjuryandthetimetoRTP.The associa-tionbetweensignalintensityatinjuryandtimetoRTPhasnever beeninvestigatedbefore.Beforethestudywehypothesizedthat increasedsignalintensitycouldbeassociatedwithincreasedtime toRTPasitmayreflecttheextentofinflammation.Basedonour findings,cliniciansshouldnotusesignalintensityonfluid sensi-tiveMRIsequencesforRTPprediction.Thelackofcorrelationis depictedinFig.4.

Wefoundincreasedintra-muscularsignalintensityafterinjury tobeassociatedwithalowerriskforreinjury.Wehadhypothesized thatincreasedsignalintensityafterinjurywouldbeassociatedwith agreaterriskofreinjury.Theassociationwefoundis counterintu-itiveasitshowsincreasedsignalintensitywasassociatedwitha betteroutcome.Thedifferenceinsignalintensitybetweengroups suggestsapossibleprotectiveassociationofincreased-intra mus-cular signalintensityonreinjury risk. Increasedintra-muscular

signalintensity on MRI ﬂuid sensitivesequences is commonly consideredtoreﬂect increasedintracellularorextracellular free water,typically described asmuscleedema.23,24_At _present_the

pathophysiologicalsigniﬁcance of increased signal intensityon MRIafterinjuryremainsunclear.25_There_is_no_research_to

com-pareourresultswithasthishasnotbeeninvestigatedinhumans before.Cutlipetal.usedacontraction-inducedhamstringinjury rat model to study the effect of increasing stretch-shortening contraction(SSC)repetitiononMRIsignalintensityand histopatho-logical findings. They found that increasing the SSC repetition (i.e.,increasing stresson thehamstring muscles)corresponded withincreasedsignalintensityonfluid-sensitiveMRIsequences and increased inflammatory cells and degenerative myofibers on histologic analysis.26 _This _finding _shows _that _signal

inten-sitycanrepresentdamagebut alsotheextentofinﬂammation, which is one of the ﬁrst important steps in muscle recov-ery.

(6)

Fig.5.Scatterplotdepictingtheoverlapofnormalizedmeansignalintensityatinjurybetweensubjectswithandwithoutreinjury.Thevaluesshownareallmeasurements

ofthestudy.

Inourcohort17ofthe51athletes(33%)sustainedareinjury within12monthsafterprimaryinjuryyieldingapretest proba-bilityof33%.Thiscorrespondswithapre-testoddsforsustaining areinjuryof: _ProbabilityProbability_noreinjury(33%)_{reinjury(66%)} ≈0.5

.Thepatientsthat sus-tainedareinjuryhadameansignalintensityatinjuryof198(gray value)whereaspatientsthatdidnotsustainareinjuryhadamean signalintensityof241(grayvalue).Toillustratethemagnitudeof theassociationwefound,wepresenttwodummyathletes. Ath-lete Aand Bhadatinjury anintra-muscularsignalintensityof 200and 250(grayvalue)respectively. We foundinourcohort anodds ratioof0.986(0.975–0.998,p=.02).Thisimpliesthata changeof1unitinintra-muscularsignalintensityafterinjuryis associated witha changeof oddsof0.986onsustaininga rein-jury.FordummyathleteBwithasignalintensityof50unitshigher thandummyathleteA,thechangeofoddsis0.98650_≈_0.5,

result-inginapost-testoddsforsustainingareinjuryof0.5*0.5=0.25. Astheprobabilityisequaltooddsdividedby(1+odds),wecan calculatethepost-testprobabilityofsustainingare-injuryis20% (95%CI, 12–31%).Thisfindingmeansthatthespecific MRI find-ing of an increased intra muscular signal intensity of 50 (gray value)reducesthepretestprobabilityof33%toaposttest prob-abilityof20%(95%CI,12–31%),whichwillnotmakeiteasierfor theclinicianintermsofreinjuryriskestimationduetothevery smalldecreaseinprobabilityandtherelativelywide confidence interval.Also,duetotheconsiderableoverlapinincreased intra-muscularsignalintensitybetweengroups,thisfindingislikelyto beoflimitedvaluefor theindividualathleteinclinicalpractice (Fig.5).

Participants receivedtwo injections of PRPor Salinewhich mayinfluencethefindingsonRTPMRI’s.TheseRTPMRI’swere performedatamedianof23days(range:5–71)afterthelast injec-tions.Histopathologicalstudiesinlaboratoryrodentsandrabbits showedonlyminimaledematouschangesforthefirsttwo days afteradministrationofsalineinjections.Thismakesitunlikelythat theinjectionssignificantlyaffectedtheresultsthreeweeksafter theinjections.LessisknownaboutPRPinjections.Tsaietal.,used arodentmodelwithpartialtransverseincisionsinthe gastrocne-miusmuscletomimic muscleinjuryand subsequentlyassessed

theeffectofPRPinjectionsvs. a controlgroupwhere no injec-tionwasgiven.Histopathologyshowednosigniﬁcantdifference ininﬂammationafter5daysofinjurybetweenbothgroups.27_It

seemsunlikelythatPRPinjectionshadaneffectonedemaonthe MRIatRTPdueto:(1)themediantimeof23days(range5–71days) betweentheﬁnalPRPinjectionandMRIassessmentatRTPand(2) theabsenceofevidenceofPRPinjectionstobea confounderas shownbyourstudy.

Ourstudyhasanumberofstrengthsandlimitations.Pleasesee SupplementaryAppendix.

5. Conclusion

Insummarythis istheﬁrststudytoinvestigatethevalueof intra-muscularsignalintensityontimetoRTPprognosisand rein-juryriskestimationafteracutehamstringinjury.Intra-muscular signalintensityatinjurywasnotassociatedwiththetimetoRTP. HavinghighersignalintensitypresentonthebaselineMRIwas associatedwithaslightlylowerriskofreinjury.Dueto consider-ableoverlapinincreasedintra-muscularsignalintensitybetween groups,thisﬁndingisoflimitedvalueforreinjuryriskestimation intheindividualathlete.

Datasharingstatement

Additionalunpublisheddataareavailableonrequestby con-tactingthecorrespondingauthor.Theunpublisheddatacontains thedetailedmeasurementsoftheincreasedintra-muscularsignal intensityand the intra and inter-observer variability measure-ments.

Funding

TheDutchrandomizedcontrolledtrialwassupportedbythe RoyalDutchFootballAssociationandArthrexMedizinische Instru-menteGmbH.Nofundingwasreceivedforthecurrentstudy.

(7)

Declarationsofinterest

Nonedeclared.

Acknowledgements

RAHwasinvolvedinthestudydesign,analysisand interpre-tationofdataanddraftingofthemanuscript.GRwasinvolvedin datacollection,studydesignanddraftingofthemanuscript.JLT, AW,MHM,JMHandMMwereinvolvedindatainterpretationand draftingofthemanuscript.Theauthorswouldliketothank Ams-terdamUniversityMedicalCentersforgrantingresearchapproval. ThisstudywasconductedusingdataoftheHamstringInjection Therapy(Dutch)study.Thehamstringinjectiontherapystudywas supportedby ArthrexMedizinischeInstrumenteGmbHand the RoyalDutchFootballAssociation.Thisresearchdidnotreceiveany speciﬁcgrantfromfundingagenciesinthepublic,commercial,or not-for-proﬁtsectors.

AppendixA. Supplementarydata

Supplementarymaterialrelatedtothisarticlecanbefound,in theonlineversion,athttps://doi.org/10.1016/j.jsams.2021.02.008.

References

1.ErnlundL,deAlmeidaVieiraL.Hamstringinjuries:updatearticle.RevBrasOrtop 2017;52(4):373–382.

2.EkstrandJ,HägglundM,WaldénM.Epidemiologyofmuscleinjuriesin profes-sionalfootball(soccer).AmJSportsMed2011;39(6):1226–1232.

3.HeiderscheitBC,SherryMA,SilderAetal.Hamstringstraininjuries: recommen-dationsfordiagnosis,rehabilitationandinjuryprevention.JOrthopSportsPhys Ther2010;40(2):67–81.

4.vanHeumenM,TolJL,deVosR-Jetal.TheprognosticvalueofMRIindetermining reinjuryriskfollowingacutehamstringinjury:asystematicreview.BrJSports Med2017;51(18):1355–1363.

5.ReurinkG,BrilmanEG,deVosR-Jetal.Magneticresonanceimaginginacute hamstringinjury:canweprovideareturntoplayprognosis?SportsMed2015; 45(1):133–146.

6.GibbsNJ,CrossTM,CameronMetal.TheaccuracyofMRIinpredictingrecovery andrecurrenceofacutegradeonehamstringmusclestrainswithinthesame seasoninAustralianRulesfootballplayers.JSciMedSport2004;7(2):248–258.

7.SilderA,SherryMA,SanﬁlippoJetal.Clinicalandmorphologicalchanges follow-ing2rehabilitationprogramsforacutehamstringstraininjuries:arandomized clinicaltrial.JOrthopSportsPhysTher2013;43(5):284–299.

8.SlavotinekJP,VerrallGM,FonGT.Hamstringinjuryinathletes:usingMR imag-ingmeasurementstocompareextentofmuscleinjurywithamountoftimelost fromcompetition.AmJRoentgenol2002;179(6):1621–1628.

9.VerrallG,SlavotinekJ,BarnesPetal.Clinicalriskfactorsforhamstring mus-clestraininjury:aprospectivestudywithcorrelationofinjurybymagnetic resonanceimaging.BrJSportsMed2001;35(6):435–439.

10.GreenkyM,CohenSB.Magneticresonanceimagingforassessinghamstring injuries:clinicalbeneﬁtsandpitfalls–areviewofthecurrentliterature.Open AccessJSportsMed2017;8:167–170.

11.PollockN,PatelA,ChakravertyJetal.Timetoreturntofulltrainingisdelayed andrecurrencerateishigherinintratendinous(’c’)acutehamstringinjuryin elitetrackandﬁeldathletes:clinicalapplicationoftheBritishAthleticsMuscle InjuryClassiﬁcation.BrJSportsMed2016;50(5):305–310.

12.KoulourisG,ConnellD.Hamstringmusclecomplex:animagingreview. Radio-graphics2005;25(3):571–586.

13.BloemJL,ReijnierseM,HuizingaTWJetal.MRsignalintensity:stayingonthe brightsideinMRimageinterpretation.RMDOpen2018;4(1):e000728.

14.DeroideN,BoussonV,MambreLetal.MuscleMRISTIRsignalintensityand atrophyarecorrelatedtofocallowerlimbneuropathyseverity.EurRadiol2015; 25(3):644–651.

15.KoxLS,KraanRBJ,MazzoliVetal.It’sathinline:developmentandvalidationof DixonMRI-basedsemi-quantitativeassessmentofstress-relatedbonemarrow edemainthewristsofyounggymnastsandnon-gymnasts.EurRadiol2020; 30(3):1534–1543.

16.HemkeR,TzaribachevN,BarendregtAMetal.Imagingofthekneeinjuvenile idiopathicarthritis.PediatrRadiol2018;48(6):818–827.

17.vanderMadeAD,AlmusaE,ReurinkGetal.Intramusculartendoninjuryisnot associatedwithanincreasedhamstringreinjuryratewithin12monthsafter returntoplay.BrJSportsMed2018;52(19):1261–1266.

18.VanderMadeAD,AlmusaE,WhiteleyRetal.Intramusculartendoninvolvement onMRIhaslimitedvalueforpredictingtimetoreturntoplayfollowingacute hamstringinjury.BrJSportsMed2018;52(2):83–88.

19.DeVosR-J,ReurinkG,GoudswaardG-Jetal.Clinicalﬁndingsjustafterreturn toplaypredicthamstringre-injury,butbaselineMRIﬁndingsdonot.BrJSports Med2014;48(18):1377–1384.

20.ReurinkG,GoudswaardGJ,TolJLetal.MRIobservationsatreturntoplayof clinicallyrecoveredhamstringinjuries.BrJSportsMed2014;48(18):1370–1376.

21.ReurinkG,GoudswaardGJ,MoenMHetal.Platelet-richplasmainjectionsin acutemuscleinjury.NewEnglJMed2014;370(26):2546–2547.

22.ThuilliezC,Dorso L,HowroydPet al.Histopathologicallesionsfollowing

intramuscularadministrationofsalineinlaboratoryrodentsandrabbits.Exp

ToxicolPathol2009;61(1):13–21.http://dx.doi.org/10.1016/j.etp.2008.07.003.

Epub2008Oct1.PMID:18835765.

23.McMahonCJ, WuJS,EisenbergRL. Muscleedema.AmJRoentgenol2010; 194(4):W284–W292.

24.MayDA,DislerDG,JonesEAetal.Abnormalsignalintensityinskeletal mus-cleatMRimaging:patterns,pearls,andpitfalls.Radiographics2000;20Spec No:S295–S315.

25.Mueller-WohlfahrtH-W,HaenselL,MithoeferKetal.Terminologyand classiﬁ-cationofmuscleinjuriesinsport:theMunichconsensusstatement.BrJSports Med2013;47(6):342–350.

26.CutlipRG,HollanderMS,JohnsonGAetal.Magneticresonanceimagingof gradedskeletalmuscleinjuryinliverats.EnvironHealthInsights2014;8(Suppl. 1):31–39.

27.TsaiW-C,YuT-Y,ChangG-Jetal.Platelet-richplasmareleasatepromotes regen-erationanddecreasesinﬂammationandapoptosisofinjuredskeletalmuscle. AmJSportsMed2018;46(8):1980–1986.