www.bjorl.org
Brazilian
Journal
of
OTORHINOLARYNGOLOGY
ORIGINAL
ARTICLE
Sonographic
scoring
of
solid
thyroid
nodules:
effects
of
nodule
size
and
suspicious
cervical
lymph
node
夽
Ozlem
Unsal
a,∗,
Meltem
Akpinar
a,
Bilge
Turk
a,
Irmak
Ucak
a,
Alper
Ozel
b,
Semra
Kayaoglu
c,
Berna
Uslu
Coskun
aaSisliEtfalTeachingandResearchHospital,HeadandNeckSurgery,ClinicofOtolaryngology,Istanbul,Turkey bSisliEtfalTeachingandResearchHospital,ClinicofRadiology,Istanbul,Turkey
cNisantasiFamilyHealthCenter,FamilyMedicine,Istanbul,Turkey
Received7November2015;accepted21January2016 Availableonline19April2016
KEYWORDS Thyroidmalignancy; Thyroidnodule; Ultrasound characteristics; Scoring; Suspicious
Abstract
Introduction:Ultrasoundisthemostfrequentlyusedimagingmethodtoevaluatethyroid nod-ules.Sonographiccharacteristicsofthyroidnoduleswhichareconcerningformalignancyare importanttodefinetheneedforfineneedleaspirationbiopsyoropensurgery.
Objective: Toevaluatemalignancyriskofsolidthyroidnodulesthroughsonographicscoring. Theeffectsofnodulesize≥2cmandassociatedpathologiccervicallymphnodeinscoringwere examinedinadditiontogenerallyexceptedsuspiciousfeatures.
Methods:Medical data of 123 patients underwent thyroid surgery were reviewed, and 89 patients(58females,31males)wereincludedinthestudy.Thepresenceandabsenceofeach suspicioussonographicfeatureofthyroidnoduleswerescoredas1and0,respectively.Total ultrasoundscorewasobtainedbyaddingthepositiveultrasoundfindings.Differentlyfromthe literature, nodulesize≥2cmandassociatedpathologiccervicalnodewereaddedinscoring criteria.Thediagnosticperformanceofnodulecharacteristicsformalignancyandtheeffectof totalUSscoretodiscriminatemalignantandbenigndiseasewerecalculated.
Results:Asignificantrelationshipwasfoundbetweenmalignancyandhypoechogenity,border irregularity,intranodularvascularity,andmicrocalcification(p<0.05).Pathologiccervicalnode wasobservedpredominantlyinassociationwithmalignantnodules.Positivepredictivevalue ofsuspiciouscervicalnodeformalignancywas67%,similartomicrocalcification.Nodulesize
≥2cmwasnotdistinctivefordiagnosisofmalignancy.Thenumberofsuspicioussonographic featuresobtainedwithreceiveroperatingcharacteristicanalysistodiscriminatebetween malig-nantandbenigndiseasewasthree.
夽 Pleasecitethisarticleas:UnsalO,AkpinarM,TurkB,UcakI,OzelA,KayaogluS,etal.Sonographicscoringofsolidthyroidnodules:
effectsofnodulesizeandsuspiciouscervicallymphnode.BrazJOtorhinolaryngol.2017;83:73---9.
∗Correspondingauthor.
E-mail:[email protected](O.Unsal).
PeerReviewundertheresponsibilityofAssociac¸ãoBrasileiradeOtorrinolaringologiaeCirurgiaCérvico-Facial.
http://dx.doi.org/10.1016/j.bjorl.2016.01.013
Conclusion:Sonographicscoringofthyroidnodulesisaneffectivemethodforpredicting malig-nancy.Theauthorssuggestincludingassociatedpathologicnodeinthescoringcriteria.Further studieswith largercohortswillprovidemore evidenceaboutitsimportanceinsonographic scoring.
© 2016 Associac¸˜ao Brasileira de Otorrinolaringologia e Cirurgia C´ervico-Facial. Published by Elsevier Editora Ltda. This is an open access article under the CC BY license (http:// creativecommons.org/licenses/by/4.0/).
PALAVRAS-CHAVE Malignidadede tireoide;
Nódulotireoidiano; Características ultrassonográficas; Escore;
Suspeito
Escoreultrassonográficodenódulossólidosdetireoide:efeitosdotamanhodo nóduloedelinfonodocervicalsuspeito
Resumo
Introduc¸ão:A ultrassonografia é o método imagiológico mais frequentemente utilizado na avaliac¸ão denódulos tireoidianos.As característicasultrassonográficasdos nódulos tireoidi-anosquedizemrespeito àmalignidadesãoimportantesparaadefinic¸ãodanecessidadede umabiópsiaporaspirac¸ãocomagulhafinaouumacirurgiaaberta.
Objetivo:Avaliaroriscodemalignidade denódulos tireoidianossólidos pormeiodeescore ultra-sonográfico,verificandoosefeitosdenódulos≥2cm,emassociac¸ãocomlinfonodo cer-vicalpatológico,alémdecaracterísticassuspeitasgeralmenteomitidas.
Método: Foramrevisadosdadosmédicosde123pacientestratados comcirurgiadatireoide. Foramincluídosnoestudo89pacientes(58mulheres,31homens).Presenc¸aeausênciadecada característicaultrassonográficasuspeitadenódulotireoidianoreceberampontuac¸õesde1e0, respectivamente.Oescoreultrassonográficototalfoiobtidopelasomadosachados ultrassono-gráficospositivos. Diferentementedaliteratura, nódulos ≥2cmenodo cervicalpatológico associadoforamacrescentadosnoscritériosdepontuac¸ão.Foramcalculadosovalor diagnós-ticodascaracterísticasdosnódulos para malignidadeeoefeitodoescoreultrassonográfico totalnadiferenciac¸ãoentredoenc¸amalignavs.benigna.
Resultados: Foiencontrada uma associac¸ãosignificante entremalignidade e hipoecogenici-dade,irregularidadedasmargens,vascularidadeintranodularemicrocalcificac¸ão(p<0,05). Nodocervicalpatológicofoiobservadopredominantementeemassociac¸ãocomnódulos malig-nos.Ovalorpreditivopositivodenodocervicalsuspeitoparamalignidadefoide67%,similarao achadoparamicrocalcificac¸ão.Diâmetrodenódulo≥2cmnãofoifatordiferenciadorpara diag-nósticodemalignidade.Onúmerodecaracterísticasultrassonográficassuspeitasobtidocoma análisedacurvadecaracterísticasdeoperac¸ãodoreceptor(receiveroperatingcharacteristic, ROC)paradiscriminac¸ãoentredoenc¸amalignavs.benignafoiiguala3.
Conclusão:Oescoreultra-sonográficodosnódulostireoidianosémétodoefetivoparapredic¸ão demalignidade.Sugerimosainclusãodenódulopatológicoassociadoaoscritériosdepontuac¸ão. Futurosestudoscomcoortesmaioresproporcionarãomaisevidênciassobresuaimportânciano escoreultrassonográfico.
© 2016 Associac¸˜ao Brasileira de Otorrinolaringologia e Cirurgia C´ervico-Facial. Publicado por Elsevier Editora Ltda. Este ´e um artigo Open Access sob uma licenc¸a CC BY (http:// creativecommons.org/licenses/by/4.0/).
Introduction
Thyroidnodules arecommonly seen withmalignancyrate of10---15%.1 Althoughfine needleaspirationbiopsy(FNAB)
stillisthemostreliabletestformalignancydetectionwith >95%ofaccuracy,theindeterminatecytologywasobserved in15---30%ofFNABs.1,2 Mostlesionsshowingindeterminate
cytologyhappenedtobebenignafterhistopathological eval-uationofsurgicalspecimens.3
Ultrasonography(US)wasproventobeareliableand eas-ilyavailable diagnostic methodwithhigh sensitivity(90%) andspecificity(85%)forthyroidnodules.4 Thesonographic
characteristicsof thyroidnodulesthat suggestmalignancy
includehypoechogenicity,solidstructure,irregularmargin, microcalcification, andregionallymphnode metastasis.5---8
Intranodular hypervascularityand nodulesize≥2cmwere alsoconsidered asindicators ofthyroid malignancyin the literature.9---11
US-basedclassificationmethodshavebeenusedby sev-eralresearcherstoevaluatethyroidnodules.12However,the
rolesofsuspiciouscervicallymphnode(LN)andnodulesize
Methods
Patients
The records of 123 patients (58 females, 31 males) who underwent thyroid surgery (lobectomy, isthmolobectomy, or total thyroidectomy) between February 2013 and July 2014 in Otolaryngology Clinic at Sisli Etfal Teaching and ResearchHospitalwerereviewed.Fivepatientswith hyper-thyroidism, fourpatients withincompleteUS images,five patientswithinconclusivecytopathologyresult,11patients with revision surgery, and nine patients with pure cystic nodule were excluded from the study. The age, gender, pre-operative USreports, FNABresultsandhistopathology resultsofsurgicalspecimensof89patientswererecorded. The institutional researchethics committee approved the presentstudyunderNo.783,andaninformedconsentwas obtainedfromallpatients.
Ultrasound
Gray scale ultrasound and color Doppler study were per-formed by radiologists experienced on neck US with a high-resolutionultrasound(SiemensS2000---Erlangen, Ger-many)equippedwitha5---14MHzlinearprobe.
Onlysolidnoduleswereincludedinthestudy.Asolid thy-roidnodulewasdescribedasmerelysolidorpredominantly solid when 10% or less of the total volume was cys-tic.Suspicioussonographicnodulecharacteristicsincluding marked hypoechogenicity, irregular margins, microcalcifi-cations,intranodalvascularity,associatedcervicalLNwith intranodal cystic components, or microcalcifications and nodulesize≥2cminthelongestdiameterwereevaluated. Inthepresenceofmultiplethyroidnodules,thenodulewith thelargestsizeand/orwiththehighestmalignancy poten-tialatsonographicfindingswasselected.
Marked hypoechogenicity was described as having the sameordecreasedechogenicityregardingstrapmuscles.A nodulemarginincludingmicrolobulationand/oraspiculated bordermayberelatedtothyroidmalignancy. Microlobula-tionwasdescribedasthepresenceofsmalllobulesonthe surfaceofanodule;aspiculatedborderwasdescribedasthe presenceofirregularspiculationlocatedonthesurfaceofa nodule.Microcalcificationswereobservedassmall, hyperec-hoicfoci(<1mminsize),presentingnocomet-tailartifacts orposteriorshadow.Intranodalvascularitydeterminedwith colorDopplerUSwasdescribedasflowinthecentralpart orinbothcentralandperipheralpartofthenodule.
Scoringofsuspiciouscharacteristics
InUSimagesandreports,the presenceofeach suspicious feature of evaluated nodule-hypoechogenity, microcal-cification, border irregularity, size ≥2cm, intranodal vascularity, and associated pathologic cervical LN, were scoredas1,while theabsencewasscoredas0.The total US scorewasobtained by addingeach individual scoreof suspicioussonographicfeatures.ThetotalUSscorewas sep-aratedintotwogroups(lowandhigh)basedontheoptimal
cut-offvaluecalculatedusingreceiveroperating character-istic(ROC)analysis.
Fineneedleaspirationbiopsy
US-guidedFNABs wereperformed by experienced radiolo-gistsusinga27-gaugeneedleattachedtoa10-mLdisposable plastic syringe and aspirator. No local anesthesia was applied.Eachnodulewascarefullyaspiratedatleasttwice. AfterFNAB,thecollected specimenwassmeared onglass slidesandfixedwith95%alcoholforPapanicolaoustaining.
Histopathologicalevaluation
All thyroidectomy specimens were made available for histopathological examination. Tissue samples were immersedintoparaffinblocks,fromwhich3---5multrathin sections were obtained. The sections were stained with hematoxylin and eosin. Special stains were performed whenrequired.Allsmearsandsectionswereexaminedby experiencedpathologists.Thefinaldiagnosisofthenodules wasreachedthroughthespecimenhistopathology.
Statisticalanalysis
Statisticalanalysiswasperformed withSPSS15.0software package for Windows (SPSS, Inc., Chicago, IL). Descrip-tive statistics were given as numeric and percentage for categorical variables. The dependent ratio of categorical variables among the groups was tested using MC Nemar analysiswhereasindependentratiowasanalyzedwith chi-squared test. When conditions of comparisons could not provide the proportions of independent groups, a Monte Carlo simulation was applied. The Kappa coefficient was usedasameasureofconcordanceintheappliedprotocol, considering: <0, no concordance; 0.0---0.20, low concord-ance; 0.21---0.40, moderate concordance; 0.41---0.60, high concordance,0.81---1.00;veryhighconcordance.The statis-ticalsignificancewasdefinedasp<0.05.
Theoptimal cut-offvaluefor thetotalUSscoreto dif-ferentiatemalignantandbenignsolidnoduleswasobtained withROCanalysisatthemaximumsensitivityandspecificity.
Results
Thirty-onemale(34.8%)and58female(65.2%)patientsaged between21and 82(median 51.9±13.1)wereincluded in thestudy.Thefrequencyofeachsuspiciousfeatureandtotal USscoresofnodulesaresummarizedinTable1.Size≥2cm, hypoechogenity,andmicrocalcificationwerethemost com-monlyencounteredsuspicioussonographicfeaturesamong allnodules,withincidenceratesof68.5%,33.7%,and20.2%, respectively.
Table1 Suspicioussonographicfeatures,totalultrasound (US)scores,andhistopathologyresultsofevaluatedthyroid noduleswithincidencerates.
n %
Hypoechogenity 30 33.7
Size≥2cm 61 68.5
Borderirregularity 11 12.4
Intranodalvascularity 16 18.0
Microcalcification 18 20.2
Suspiciouslymphnode 3 3.4
TotalUSscore
0 7 7.9
1 45 50.6
2 20 22.5
3 12 13.5
4 5 5.6
HighUSscore
≥3 17 19.1
LowUSscore
<3 72 80.9
Histopathology
Malignant 29 32.6
Benign 60 67.4
benignin60(67.4%)ofthenodulesaftersurgicalremoval. Histopathologicexaminationofmalignantnodulesrevealed 28(96.6%)papillarycarcinomasandone(3.4%)lymphoma.
Themalignancyratesinhighandlowscorednodulesare presented in Table 2. Seventeen nodules were located in thehighscoregroup,and15(82.2%)ofthemrevealed malig-nancy,whereas14nodulesoutof72(19.4%)inthelowscore group presented malignancy. The malignancy risk of high scorednoduleswassignificantlyhighercomparedtonodules withlowscores(p=0.004).
Whensuspiciousnodulecharacteristicsincluding hypoe-chogenity, border irregularity, intranodular vascularity, andmicrocalcificationwerecomparedwithhistopathology results, the malignancy risk wasfound tobe significantly higherinthenoduleswiththesesonographicfeatures,when comparedwiththosewithoutthem(p<0.05;Table3).
Nodule size ≥2cm was not distinctive for the diagno-sisofmalignancyinthisstudy(p>0.05).Contrarily,itwas
detected that the nodules ≥2cm were associated with benign histopathology. However, the distribution of the nodules sized ≥2cm among high and low scored nodules wassimilar. Nosignificant differences weredemonstrated (p>0.05)
Three cervical LNs with suspicious features were reportedsonographically in thepresent study,and twoof them(66.6%)wereassociatedwiththyroidmalignancy.
Foreachsuspicioussonographicfeature,thecalculated diagnostic performances including sensitivity, specificity, positive predictive value (PPV), negativepredictive value (NPV)andaccuracy,whicharedepictedinTable4. Micro-calcificationandpresenceofsuspiciousLNhadthehighest PPV(0.667).
Discussion
In adults,theprevalence ofthyroid noduleswasreported as50%.13 Ultrasonographicexamination,whichis themain
diagnostic modality for the detection of thyroid nodules, increasedtheidentificationrateofclinicallyasymptomatic thyroid nodules,withaprevalence of 19---67%.14 However,
its ability todifferentiate between malignant and benign nodulesisstillunclearandaconsensushasnotbeenreached yet.
US-guided FNAB is the diagnostic method for the histopathological evaluation of thyroid nodules. However, therearesomedisadvantages,suchasinadequatesampling, indeterminate cytology, cost, invasiveness, and operator dependency.12 At this point, the use of ultrasonographic
featurestopredictmalignancymaydecreasetheneedfor FNAB,buttodate,noneofthesonographicnodullary fea-tureswasfoundtobe100%sensitiveorspecificforthyroid malignancy.
The sonographic suspicious features considered in this studytodifferentiatethepotentialmalignantsolidnodules werebasedonsomepreviousstudies.12,14,15Thepresenceof
hypoechogenity,microcalcification,borderirregularity,and intranodalhypervascularitywasassessed.Differentlyfrom the literature,the presence ofsuspicious cervical LN and nodulesize≥2cmwereincluded inthescoring criteriain thisstudy.
Prelaryngeal,pretracheal,andparatrachealLNsare com-monly observed sites of thyroid carcinoma metastasis. Cervical LN features, includinghyperechoic punctuations, cystic appearance,peripheral vascularization, and loss of
Table2 Distributionofmalignantandbenignnodulesinthelowandhighscoregroups.
Histopathologyofnodules
Malignant Benign p Kappa
n % n %
TotalUSscore 0.004 0.542
High(≥3) 15 88.2 2 11.8
51.7 3.3
Low(<3) 14 19.4 58 80.6
48.3 96.7
Table3 Malignancyratesofthenoduleswithandwithoutsuspicioussonographicfeatures.
Histopathologyofnodules p
Malignant Benign
n % n %
Hypoechogenity 0.003
+ 16 53.3 14 46.7
− 13 22.0 46 78.0
Borderirregularity 0.019
+ 7 63.6 4 36.4
− 22 28.2 56 71.8
Hypervascularity 0.026
+ 9 56.3 7 43.8
− 20 27.4 53 72.6
Microcalcification 0.001
+ 12 66.7 6 33.3
− 17 23.9 54 76.1
hilum, have been associated with malignancy.16 A
thy-roid nodule co-existing with a cervical LN presenting the aforementionedcharacteristicsisaremarkablesonographic findingregardingmalignancy. Inthisstudy,suspicious cer-vical LN wasencountered in three patients.Two of them accompanied thyroid carcinoma. A PPV of 66.7% for can-cerwasobservedinthepresenceofsuspiciouscervicalLN, similarlyinmicrocalcifiednodules. Therefore,theauthors believe that the inclusion of suspicious LN in the scoring criteriahasan importantrolein predictingthyroid malig-nancy. However, further studies are required due to the inadequatenumberofsuspiciouscervicalLNstounderstand itsimportanceintermsofsonographicscoringand predict-ingthyroidmalignancy.
Although nodulesize is considered a non-specific find-ing,apreviousstudy revealedthattherewasanon-linear relationshipbetweensizeandmalignancyrates.The approx-imatethresholdfor increasedmalignancywasfoundtobe 2cm.10 In addition,Smith-Bindman etal.11 reported that
threeUScharacteristics---nodulesize≥2cm, microcalcifi-cationandsolidcomposition---werecloselyassociatedwith thyroidcarcinoma risk.Therefore, theauthorsdecidedto includethenodulesize≥2cminscoringcriteria.However, it wasobserved that thedistribution of nodules≥2cmin bothhighandlowscoregroups weresimilar,andthatthis
distributionwasnotstatisticallysignificant(p>0.05).Thus, nodulesize ≥2cm wasconsidered not tohave an impor-tantroleinscoring.Althoughnodulesize≥2cmwasfound tobestatisticallyinsignificant,in practicethereis a ten-dency to sample from the largest nodules in the thyroid gland.Moreover,therevised2009AmericanThyroid Associ-ationGuidelinessuggeststosamplefromthelargestnodule inthepresenceoftwoormorenodulesifnonehassuspicious sonographicappearanceandthenodulesaresonographically similar.1
Predictors of thyroid malignancy including microcalci-fication, marked hypoechogenicity, and irregular margins havebeenalsoreported.15,16 Microcalcification,despiteof
lowsensitivity,isarelativelyspecificultrasonographic indi-cator of thyroid carcinoma. In a previous study, a total of45---60%ofmalignantnodulesshowedmicrocalcifications versus7---14%ofbenignnodules.16Ahnetal.17reportedaPPV
pf85.1%forcancerwiththepresenceofmicrocalcifications. In anotherstudy, 60% of patients withmicrocalcifications werefoundtohavemalignantdisease.18Thepresentstudy
observedaPPVof 66.7%for cancer inmicrocalcified nod-ules,thehighestPPV,similarlyinthepresenceofsuspicious cervicalLN.
Malignant nodules are frequently seen to be markedly hypoechoic.16 Cappelli et al.19 found a 3.8 odds ratio of
Table4 Diagnosticperformancesofsuspicioussonographiccharacteristicsofthyroidnodules.
Sensitivity(%) Specificity(%) PPV(%) NPV(%) Accuracy
Ultrasound 0.517 0.967 0.882 0.806 0.802
Hypoechogenity 0.552 0.767 0.533 0.780 0.764
Size≥2cm 0.483 0.217 0.230 0.464 0.303
Borderirregularity 0.241 0.933 0.636 0.718 0.708
Hypervascularity 0.310 0.883 0.563 0.726 0.697
Microcalcification 0.414 0.900 0.667 0.761 0.742
malignancy in solid hypoechoic nodules in a prospective studyof349surgicallyexcisedthyroidnodules.Thepresent studyfound a 2.8 oddsratioof malignancy in hypoechoic nodules.
Border irregularity of a thyroid nodule is considered to be notable sonographic feature. The increased malig-nancy risk related toirregular margins has been debated in the literature.16 In the present study, irregular border
wasdetected in 11 of 89 nodules. Seven of them (7/11; 63.6%)werefound tobemalignant,whereasfournodules (4/11;36.4%)werebenign.Borderirregularitywasfoundto bestatisticallyhigherin malignantnodulesthaninbenign (p=0.019).
Nodule vascularity is generally evaluated withthe use ofcolorDopplerUS.Ithasbeendemonstratedthat42---74% ofmalignantthyroidnodulespresent hypervascularityand aprominent central blood flow.20 Frates etal.21 reported
thatacentrallypredominantvascularitywasencounteredin malignantnodules(42%)moreoftenthaninbenignnodules (14%).Similarly, thepresent studyrevealed a hypervascu-larity rate significantly higher in malignant nodules when comparedwithbenignnodules(p=0.026).
Variousreportinganddatasystemsbasedonsonographic featureshave been discussed to identifymalignancy risk. Horvathetal.,22withamodifiedrecommendationfromJin
Kwak etal.,23 proposed the Thyroid Image Reportingand
DataSystem(TIRADS)inordertoimprovepatient manage-mentandcost-effectivenessbyavoidingunnecessaryFNAB ofthyroidnodules.However,itsclinicaluseisverylimited. AccordingtotheguidelinesoftheAmericanAssociation of Clinical Endocrinologists, the co-existence of at least twosuspiciousUS featuresincreasesthyroid cancerrisk.24
Ozeletal.25 suggestedthatat leastthreeUS featuresare
consideredformalignancyriskinnodulessized<1cmwith diagnosticaccuracyof89.9%,whereastwoUSfeaturesare consideredfornodulessized≥1cmwiththediagnostic accu-racyof 93.8%. In the present study, thecut-off value for suspicious sonographic features including size ≥2cm and presence of suspiciouscervical LN wasthree(Az=0.675), andthediagnosticaccuracyofUSwas80.2%.
Thesonographicscoringofthyroidnodulesmayincrease diagnosticaccuracyof USfor thyroidmalignancyand pre-ventunnecessaryinvasivetissuesamplingandopensurgical procedures, and also provides a quick method to decide whichnoduletobesampled.
There are some limitations in this study. The first is that only one suspicious thyroid nodule for each patient wasselected.Thesecond isthattheretrospective images might be limited, providing inadequateinformation com-paredtoadynamicstudywheremoreextendedanalysiscan beperformed.Furthermore,inthepresentseries,thehigh incidenceofpapillarythyroidcarcinoma(96.5%)maylead toabias,sincegenerallyaccepted suspicious USfeatures arenotenoughtodetectfollicularcarcinoma.
Conclusion
Ultrasoundisanon-invasiveandeffective toolfor evalua-tionof thyroidnodules.The scoring ofsuspiciousfeatures basedonUSispracticaltodecidewhichnoduleneedstobe sampled.Contrarytonodulesize,theauthorsbelievethat
the presenceof suspicious cervical LN shouldbeincluded ingenerallyacceptedscoringcriteriainanattemptto pre-dictthyroidmalignancyeventhoughanadequatenumberof suspiciousLN foranaccuratestatisticalanalysiscouldnot beachievedinthisstudy.Largercohortsmayprovidemore prominencetotheassociation betweensuspiciouscervical LNandthyroidmalignancyrisk.
Althoughnoneof thesuspicioussonographic character-isticofferssufficientspecificityandsensitivitytodiagnose malignancy, their cumulative effecthighlights a potential thyroidmalignancy.Inthisstudy,threeor moresuspicious sonographicfeaturesindicatedapotentialmalignantthyroid nodule.However,furtherprospectivecohortstudiesbased onrealtimeUSareneededformoreaccurateresults.
Conflicts
of
interest
Theauthorsdeclarenoconflictsofinterest.
References
1.American Thyroid Association (ATA) Guidelines Taskforce on ThyroidNodulesandDifferentiatedThyroidCancerCooperDS, DohertyGM,HaugenBR,KloosRT,LeeSL,etal.Revised Amer-icanThyroidAssociationmanagement guidelinesfor patients withthyroidnodulesanddifferentiatedthyroidcancer.Thyroid. 2009;19:1167---214.
2.CibasES,AliSZ.NCIthyroidfineneedleaspirationstateofthe ScienceConference.TheBethesdasystemforreportingthyroid cytopathology.AmJClinPathol.2009;132:658---65.
3.YoonJH,KwakJY,KimEK,MoonHJ,KimMJ,KimJY,etal.How toapproachthyroidnoduleswithindeterminatecytology.Ann SurgOncol.2010;17:2147---55.
4.BranderAE,ViikinkoskiVP,NickelsJI,KivisaariLM.Importance ofthyroid abnormalities detected at US screening: a 5year follow-up.Radiology.2000;215:801---6.
5.GinatDT,ButaniD,GiampoliEJ,PatelN,DograV.Pearlsand pitfallsofthyroidnodulesonographyandfine-needleaspiration. UltrasoundQ.2010;26:171---8.
6.HaberRS.Roleofultrasonographyinthediagnosisand manage-mentofthyroidcancer.EndocrPract.2000;6:396---400. 7.HegedüsL.Thyroidultrasound.Endocrinol MetabClinNAm.
2001;30:339---60.
8.WeberAL,RandolphG,AksoyFG.Thethyroidandparathyroid glands.CTandMRimagingandcorrelationwithpathologyand clinicalfindings.RadiolClinNAm.2000;38:1105---29.
9.Tay SY, Chen CY, Chan WP. Sonographic criteria predictive of benign thyroid nodules useful in avoiding unnecessary ultrasound-guidedfineneedleaspiration.JFormosMedAssoc. 2015;114:590---7.
10.KamranSC, Marqusee E,KimMI,Frates MC,RitnerJ,Peters H,etal.Thyroidnodulesizeandpredictionofcancer.JClin EndocrinolMetab.2013;98:564---70.
11.Smith-BindmanR,LebdaP,FeldsteinVA,SellamiD,Goldstein RB,BrasicN,etal.Riskofthyroidcanceronthyroidultrasound imagingcharacteristics: results of a population-based study. JAMAInternMed.2013;173:1788---96.
12.KimEK,ParkCS,ChungWY,OhKK,KimDI,LeeJT,etal.New sonographiccriteriafor recommendingfine-needleaspiration biopsyofnonpalpablesolidnodulesofthethyroid.AJRAmJ Roentgenol.2002;178:687---91.
14.DeanDS,GharibH.Epidemiologyofthyroidnodules.BestPract ResClinEndocrinolMetab.2008;22:901---11.
15.Frates MC,Benson CB, CharboneauJW, Cibas ES,ClarkOH, ColemanBG,etal.Managementofthyroidnodulesdetectedat US:SocietyofRadiologistsinUltrasoundconsensusconference statement.Radiology.2005;237:794---800.
16.KangelarisGT,KimTB,OrloffLA.Roleofultrasoundinthyroid disorders.OtolaryngolClinNAm.2010;43:1209---27.
17.AhnSS,KimEK,KangDR,LimSK,KwakJY,KimMJ.Biopsyof thyroidnodules:comparisonofthreesetsofguidelines.AJRAm JRoentgenol.2010;194:31---7.
18.SeiberlingKA,DutraJC,GrantT,BajramovicS.Roleof intrathy-roidal calcifications detected on ultrasound as a marker of malignancy.Laryngoscope.2004;114:1753---7.
19.CappelliC,CastellanoM,Pirola I,Cumetti D,Agosti B, Gan-dossiE,etal.Thepredictivevalueofultrasoundfindingsinthe managementofthyroidnodules.QJM.2007;100:29---35. 20.AzarN,Lance C,Nakamoto D,MichaelC, Wasman J.
Ultra-sonographicthyroidfindings suspiciousfor malignancy. Diagn Cytopathol.2013;41:1107---14.
21.FratesMC,BensonCB,DoubiletPM,CibasES,MarquseeE.Can colorDopplersonographyaidinthepredictionofmalignancyof thyroidnodules?JUltrasoundMed.2003;22:127---31.
22.Horvath E,Majlis S, RossiR, Franco C,Niedmann JP, Castro A, etal.Anultrasonogramreportingsystemfor thyroid nod-ules stratifying cancer risk for clinical management. J Clin EndocrinolMetab.2009;94:1748---51.
23.KwakJY,HanKH,YoonJH,MoonHJ,SonEJ, ParkSH,etal. ThyroidimagingreportinganddatasystemforUSfeaturesof nodules:a stepinestablishingbetterstratificationofcancer risk.Radiology.2011;260:892---9.
24.GharibH,PapiniE,PaschkeR,DuickDS,ValcaviR,HegedüsL, etal.AmericanAssociationofClinicalEndocrinologists, Associ-azioneMediciEndocrinologi,andEuropeanThyroidAssociation medical guidelinesforclinicalpracticefor thediagnosis and managementofthyroidnodules.EndocrPract.2010;16:1---43. 25.Ozel A, Erturk SM,Ercan A, YılmazB, Basak T, CantisaniV,