ORIGINAL ARTICLE

Influence of common fixed retainers onthe diagnostic quality of cranial magneticresonance images

Miriam Shalish,a Nimrod Dykstein,b Silvina Friedlander-Barenboim,c Eliel Ben-David,d John Moshe Gomori,e

and Stella Chaushuf

Jerusalem, Israel

aClinitics, HbCliniSchoocHeadUnivedCliniJerusaeProfeMedicfAssocHadasAll auPotenAddreHebreIsrael;Subm0889-Copyrhttp:/

604

Introduction: Orthodontists are often asked to remove fixed retainers before magnetic resonance imaging(MRI). This study was undertaken to assess the effects of 2 commonly used fixed retainers on MRI distortionand whether they should be removed. Methods: MRI scans were performed on a dry skull with Twistflex (Den-taurum, Ispringen, Germany) and Ortho Flex Tech (Reliance Orthodontic Products, Itasca, Ill) retainers. Twoneuroradiologists independently ranked the distortions. The influence of the fixed retainers' alloys, their distanceto the area of diagnosis, location, strength of the magnetic field, and the spin-echo sequence were examined.Statistical analysis included kappa and Pearson chi-square tests. Results: Ortho Flex Tech retainers causedno distortion. Twistflex retainers caused distortion in 46% of the tests in areas close to the retainer (tongueand jaws). Maxillary fixed retainers and the combination of maxillary and mandibular fixed retainers furtherincreased the distortion. Greater distortion was observed with 3-T magnetic fields and T1-weighted spin-echosequences. Conclusions: Removal of the Ortho Flex Tech retainer is unnecessary before MRI. Removal ofthe Twistflex should be considered if the MRI scans are performed to diagnose areas close to the fixed retainers,when 3-T magnetic fields and T1-weighted sequences are used, and when both maxillary and mandibular fixedretainers are present. (Am J Orthod Dentofacial Orthop 2015;147:604-9)

Magnetic resonance imaging (MRI) has becomea common diagnostic tool for various medicaland dental conditions (migraine and cluster

headaches, epilepsy and other seizure disorders, multiplesclerosis, head and neck tumors, temporomandibularjoint disorder, and many others).1 The advantages ofMRI over other imaging techniques include excellentsoft-tissue quality and absence of ionizing radiation.

cal lecturer and director, Postgraduate Program, Department of Orthodon-ebrew University-Hadassah School of Dental Medicine, Jerusalem, Israel.cal instructor, Department of Orthodontics, Hebrew University-Hadassahl of Dental Medicine, Jerusalem, Israel., Oral & Maxillofacial Radiology Unit, Oral Medicine Department, Hebrewrsity-Hadassah School of Dental Medicine, Jerusalem, Israel.cal instructor, Department of Radiology, Hadassah Medical Center,lem, Israel.ssor and head, Neuroradiology Unit, Department of Radiology, Hadassahal Center, Jerusalem, Israel.iate professor and chair, Department of Orthodontics, Hebrew University-sah School of Dental Medicine, Jerusalem, Israel.thors have completed and submitted the ICMJE Form for Disclosure oftial Conflicts of Interest, and none were reported.ss correspondence to: Miriam Shalish, Department of Orthodontics,w University-Hadassah School of Dental Medicine, POB 12272, Jerusalem,e-mail, mshalish@hadassah.org.il.itted, July 2014; revised and accepted, November 2014.5406/$36.00ight � 2015 by the American Association of Orthodontists./dx.doi.org/10.1016/j.ajodo.2014.11.022

The common MRI apparatus, often designated by theirmagnetic fields, are of 1.5 and 3 teslas (T), which repre-sent the magnetic intensity of the machine.1 Ferromag-netic materials (several common metals) distort themagnetic field and result in distortion of the magneticresonance image, affecting its diagnostic value.1

Recently, there has been an increase in the numberof orthodontic patients worldwide. According to Proffitet al2 in 1998, over 30% of white youths, 11% ofMexican-Americans, and 8% of African Americansreported receiving orthodontic treatment. At the endof active orthodontic treatment, the results are usuallyretained for many years using fixed retainers made of ametal wire bonded to the lingual side of the anteriorteeth.

Owing to the increased number of MRI referrals,3

orthodontists are frequently requested to remove metalorthodontic appliances before the MRI procedures.Removal and reinstallation of an orthodontic applianceis laborious and bears a risk of relapse of teeth to unde-sired positions. As a result, orthodontists are interestedin understanding the effects of orthodontic applianceson MRI imaging to minimize the need for removal.

The presence of orthodontic appliances has 3 impor-tant implications for MRI diagnosis. First, magnetic field

Fig 1. Mandibular Ortho Flex Tech fixed retainerembedded into a polyvinylsiloxane key.

Fig 2. The skull with the maxillary and mandibular polyvi-nylsiloxane keys.

Shalish et al 605

interactions may pose a risk of detachment to thepatient. Loose orthodontic devices pose a significantdanger for the patient.4 Second, metals in the appliancesmay cause heating.5 Third, they may produce image dis-tortions that could affect the diagnostic quality ofthe MRI.6

Only a few studies have been conducted to evaluatethe influence of different fixed orthodontic applianceson the diagnostic quality of magnetic resonance im-ages.4-9 One study showed that stainless steel bracketscause significant distortion, rendering several cranialregions nondiagnostic, whereas plastic, ceramic, andtitanium brackets cause minimal interference.6

The effect of different types of fixed retainershas scarcely been addressed so far.7,8 When bondedfirmly, fixed retainers pose no risk to the patient in themagnetic field of a clinical MRI scanner.9 However,image quality may be significantly affected.8

The purpose of this study was to assess the in-vitroeffects of the 2 most common types of fixed retainerson MRI distortion and answer the dilemma of whetherthey should be removed before MRI.

MATERIAL AND METHODS

A dry skull of a female was used for this study. Algi-nate impressions of the maxillary and mandibular dentalarches were taken, and plaster models were poured. The2 types of fixed retainers were fabricated directly on thelingual aspects of the maxillary and mandibular incisorsand canines of the plaster models. The fixed retainer wirewas attached to the lingual surface of each tooth usingdental floss. Polyvinylsiloxane key (Elite HD 1 type0 putty consistency; Zhermack, Badia Polesine, Italy)was prepared for each jaw to hold the retainer in place.The retainers were embedded into this material (Fig 1).

For each retainer key, the skull was immersed in acontainer including a diluted 2% solution of gadoliniumcontrast agent (Dotarem; Guerbet, Villepinte, France)

American Journal of Orthodontics and Dentofacial Orthoped

and placed in the scanner. The MRI sequences were per-formed for that particular set; then a different tray wasinserted, and subsequent magnetic resonance scanswere completed (Fig 2). The diluted gadolinium solutioncreated a background signal to the bone that simulatedthat of soft tissues. It did not affect the metal distortionartifacts but allowed them to be seen using MRIsequences identical to those in clinical imaging.10,11

To determine the effect of the polyvinylsiloxane keyon the MRI scans, 2 scans were carried out on the keywithout the retainers, which served as the controls.

Two experienced neuroradiologists (J.M.G. andE.B.-D.) independently evaluated the images to deter-mine the amount of distortion. All images were pre-sented at once on the neuroradiologists' computerscreens. The images were randomly labeled. The neuro-radiologists were asked to rank the images according tothe distortions in these regions using a modified receiveroperating characteristic method of distortion classifica-tion (Table I).

The following 5 variables were examined.

1. The type of alloy of the fixed retainer. Twocommonly used fixed retainers were chosen forthis study: Twistflex, 0.18 in (Dentaflex, triplestrand twisted; Dentaurum, Ispringen, Germany),an alloy of stainless steel (chromium, 17%-20%;nickel, 8%-12%; carbon, 0.08%-0.15%; the rest is

ics May 2015 � Vol 147 � Issue 5

Table I. Distortion classification

Score Image appearanceDiagnostic ornondiagnostic

1 No distortion or artifact Diagnostic2 Minimal distortion or artifact Diagnostic3 Moderate distortion or artifact Moderately diagnostic4 Severe distortion Nondiagnostic

Table II. Effect of the type of fixed retainer alloy:distortion scores of Twistflex (TF) vs Ortho Flex Tech(OF) retainers

Wires

Distortion

TotalNone Minimal Moderate SevereTFCount 54 20 7 19 100% 54.0 20.0 7.0 19.0 100.0

OFCount 80 0 0 0 80% 100.0 0.0 0.0 0.0 100.0

ControlCount 20 0 0 0 20% 100.0 0.0 0.0 0.0 100.0

606 Shalish et al

mainly iron); and Ortho Flex Tech (Reliance Ortho-dontic Products, Itasca, Ill), an alloy composed ofgold, 58.33%-58.70%; silver, 1.5%-6.0%; copper,28.0%-31.0%; nickel, 0.3%-6.5%; and zinc,5.0%-6.3%.

2. The distance between the region of interest and theretainer. Ten regions of the head were assessed:tongue, mandible, anterior maxilla, posteriormaxilla, orbits, nasopharynx, pituitary gland, frontallobe, temporal lobes, and brain stem.

3. The location of the fixed retainer. Six trays werefabricated: maxillary and mandibular Twistflex,maxillary and mandibular Ortho Flex Tech, andmaxillary and mandibular control polyvinylsiloxanekeys without retainers.

4. The strength of the magnetic field. The scans wereperformed in the MRI unit, Hadassah Medical Cen-ter, Jerusalem, Israel, on both 1.5-T (Avanto) and3-T (Trio) systems (both, Siemens Medical Solu-tions, Erlangen, Germany) using standard multi-element head coils.

5. The spin-echo sequence. Imaging sequencesincluded axial fast-spin echo T2-weighted images(TR/TE 3500/90 ms), axial and sagittal conven-tional spin-echo T1-weighted images (TR/TE500/14 ms), with a slice thickness of 5 mm anda 20% gap.

Statistical analysis

Statistical analysis included descriptive statistics,kappa tests to evaluate interexaminer calibrations, andPearson chi-square tests. The Pearson chi-square testwas used to compare the distortions caused by thedifferent retainers and different MRI settings. The signif-icance level for the statistical testing was set at P\0.05.The analysis was carried out using SPSS software(version 21.0; IBM, Armonk, NY).

RESULTS

Statistically significant differences were foundbetween the distortion scores assigned to the Twistflexand Ortho Flex Tech retainers (P\0.001). No distortionwas observed with the Ortho Flex Tech retainers, similarto the control, whereas the Twistflex retainers caused

May 2015 � Vol 147 � Issue 5 American

distortion in 46% of the tests. In 20% of the cases, thedistortion was minimal; in 7%, it was moderate; and in19%, it was severe (Table II).

The Twistflex retainers mainly affected anatomic re-gions in their vicinity, such as tongue, mandible, andmaxilla. Other anatomic regions, not near the Twistflexretainer, were unaffected (Table III). The differences be-tween the distortion scores in the different anatomic re-gions were highly statistically significant (P\0.001).

Significant differences related to the location of theTwistflex fixed retainer were found. Maxillary retainerscaused more severe distortion than did mandibular re-tainers, and both maxillary and mandibular retainersfurther increased the severity of the distortion(P 5 0.05) (Table IV).

The intensity of the magnetic field (3 T vs 1.5 T) had amarked impact on the distortion scores. Greater distor-tion was observed for the stronger magnetic field (3 T)(P 5 0.005) (Table V).

The spin-echo sequence also significantly affected thedegree of distortion, with T1-weighted images causingmore severe distortion than T2-weighted images(P5 0.03) (Table V).

The polyvinylsyloxane tray (control) caused nodistortion of the magnetic resonance images.

The evaluations of the 2 neuroradiologists werecompared using kappa tests. Satisfactory calibrationbetween the 2 examiners was obtained (kappa values,0.6-0.8).

DISCUSSION

Fixed lingual retainers, bonded to mandibular ante-rior teeth, were introduced in the 1970s12 and recog-nized as an important part of orthodontic treatment toprevent relapse or secondary crowding of the mandib-ular incisors.13 The main advantages of mandibular fixedretainers compared with removable retainers are that

Journal of Orthodontics and Dentofacial Orthopedics

Table IV. Effect of location of fixed retainers: distor-tion scores of Twistflex in the jaws

Jaw

Distortion

TotalNone Minimal Moderate SevereBoth jawsCount 28 10 6 16 60% in jaw 46.7 16.7 10.0 26.7 100.0

MaxillaCount 13 3 1 3 20% in jaw 65.0 15.0 5.0 15.0 100.0

MandibleCount 13 7 0 0 20% in jaw 65.0 35.0 0.0 0.0 100.0

Table III. Effect of the distance from the Twistflexretainer: distortion scores in different anatomicregions

Distortion

None Minimal Moderate SevereTongueCount 26 6 3 5% in anatomic site 65.0 15.0 7.5 12.5

MandibleCount 27 11 0 2% in anatomic site 67.5 27.5 0.0 5.0

Maxilla, anteriorCount 23 1 4 12% in anatomic site 57.5 2.5 10.0 30.0

Maxilla, posteriorCount 38 2 0 0% in anatomic site 95.0 5.0 0.0 0

Orbit and other areasCount 40 0 0 0% in anatomic site 100.0 0.0 0.0 0

Table V. Effect of magnetic field strength and spin-echo sequence: distortion scores of Twistflex at 1.5 Tvs 3 T, and T1-weighted vs T2-weighted sequence

MRI

Distortion

None Minimal Moderate Severe1.5 TCount 37 15 2 6% in MRI 61.7 25.0 3.3 10.0

3 TCount 17 5 5 13% in MRI 42.5 12.5 12.5 32.5

T1Count 28 15 2 15% in T 46.7 25.0 3.3 25.0

T2Count 26 5 5 4% in T 65.0 12.5 12.5 10.0

T, Teslas; T1, T1-weighted; T2, T2-weighted.

Shalish et al 607

they are invisible and compliance free.14 The disadvan-tages include the demanding technique,15 the potentialfor distortion of the wire, and increased plaque andcalculus accumulation compared with removableretainers.16 A survey showed that nearly a third of thepractitioners in the United States use fixed retainers inthe mandibular arch.17 Another more recent studyshowed that 40% of the members of the American Asso-ciation of Orthodontists use fixed retainers and do notinstruct patients to have them removed at a specifictime.18 In Norway, a fixed retainer bonded to all anteriorteeth in the mandible was most common (66.4%).Although retention lasted 2 to 3 years (34.7%) in themaxilla, or 3 to 5 years (23.8%) in the mandible,41.5% of the orthodontists left the retainers in placefor more than 5 years.19

American Journal of Orthodontics and Dentofacial Orthoped

The use of MRI has recently increased,20 as well as3-T MRI for pediatric patients.21 A recent study, basedon electronic records of members of 6 large integratedhealth systems from different regions of the UnitedStates, reported a nearly quadruple increase in MRIbetween 1996 and 2010.3

The influence of orthodontic appliances on MRI hasalready been studied in several articles. Elison et al6

showed that stainless steel brackets cause significantdistortions. Another study showed that orthodonticappliances have a significant influence on imaging ofthe frontal and temporal lobes of the brain and shouldbe removed for accurate MRI diagnosis of epilepsy.22

In contrast, and despite the fact that there is a defin-itive trend to use fixed retention for long-term stabil-ity,17 there is limited information regarding its effectson MRI scans.7,8 The constant rise in the frequencyof fixed retainers use, the trend to extend the periodsof fixed retention, and the increase in the frequency ofMRI referrals increase the chances of finding a fixedretainer in a patient referred for MRI. Since themagnetic cross-section of the fixed retainer is small rela-tive to common orthodontic appliances, it should beexpected to cause a smaller perturbation of the field.Therefore, a careful study of the effects could minimizeconsiderably the need for removal of fixed retainersbefore MRI.

The most commonly used fixed retainers are Twist-flex and Ortho Flex Tech. There are additional lesscommonly used fixed retainers based on stainless steel,nickel-titanium alloy, or fiberglass.

Our study showed that in most cases, the observeddistortion was minimal for both types of retainers; for

ics May 2015 � Vol 147 � Issue 5

608 Shalish et al

most MRI screenings, there is actually no indication forremoval of the fixed retainer. Distortion was observedonly with Twistflex retainers. The physics underlyingthese results appears to be rather simple. Ortho FlexTech retainers contain only minor amounts (0.3%-6.5%) of the ferromagnetic metal nickel, and all theother metals in the alloy are nonferromagnetic. Theferromagnetic cross-section is therefore small and so isthe perturbation it can cause to the flow of the magneticfield. In contrast, the Twistflex retainer is a stainless steelalloy with a high percentage of ferromagnetic metals,such as iron. Our results support clinical findings onMRI scans of 3 patients wearing Orthoflex retainers7

and in-vitro findings on MRI scans of stainless steel vsnickel-titanium retainers.8However, in this last study,the authors used a simplified in-vitro model, in whichstainless steel and nickel-titanium wires were embeddedin a cylinder placed in the MRI field.8 To the best of ourknowledge, our study is the first in-vitro study thatmimics the actual position of retainers in a patient'smouth and compares the widely used Ortho Flex Techand Twistflex retainers in a controlled manner.

Even though a distortion was observed with Twist-flex, it was rather limited, affecting only the closestanatomic regions, such as tongue, mandible, or maxilla.As already mentioned, the reason for this rather smalleffect was most probably the small mechanical cross-section of the fixed retainer. Our results support previousstudies on the effects of stainless steel brackets, whichalso showed that the distortions diminish with the dis-tance from the appliance.6

Importantly, the amount and location of the materialinfluenced the scores of distortion, meaning that lon-ger maxillary retainers, or maxillary and mandibularretainers together, pose higher risks than a mandibularretainer alone. Fortunately, maxillary fixed retainersare less commonly used than mandibular ones and areleft in situ for shorter periods after treatment.19

Thedistortionwasdirectly proportional to the strengthof themagneticfield—ie, a greater distortionwas observedfor the stronger intensity field (3 T)—and was also influ-enced by the spin-echo sequence, with T1-weightedcausing much more distortion than T2-weighted.

Based on our results, we can therefore conclude thatfixed retainers in which the metal alloy contains greateramounts of ferromagnetic metals, such as Twistflex,may distort MRI images of the maxilla, mandible,or tongue, and their removal should be limited toMRI scans performed to analyze pathology in theseanatomic regions. However, the use of fixed retainersmade of nonferromagnetic metals, or small percentagesthereof, such as Ortho Flex Tech, avoids removal inall cases.

May 2015 � Vol 147 � Issue 5 American

CONCLUSIONS

1. According to our findings, removal of Ortho FlexTech fixed retainers is not necessary.

2. Removal of Twistflex fixed retainers is not requiredbefore most MRI scans. Removal should be consid-ered only if the MRI scan is aimed at diagnosing pa-thologies in the jaws or tongue, especiallywhenusingthe3-T, T1-weighted, spin-echo sequence, andwhenboth maxillary and mandibular retainers are present.

3. According to our findings, if the region of interest ofthe MRI is not near the fixed retainer, there is noneed to remove it.

REFERENCES

1. Yousem DM, Grossman RI. Neuroradiology: the requisites. 3rd ed.Philadelphia: Mosby/Elsevier; 2010. p. 4-15.

2. Proffit WR, Fields HW Jr, Moray LJ. Prevalence of malocclusionand orthodontic treatment need in the United States: estimatesfrom the NHANES III survey. Int J Adult Orthodon OrthognathSurg 1998;13:97-106.

3. Smith-Bindman R, Miglioretti DL, Johnson E, Lee C, Feigelson HS,Flynn M, et al. Use of diagnostic imaging studies and associatedradiation exposure for patients enrolled in large integrated healthcare systems, 1996-2010. JAMA 2012;307:2400-9.

4. Kemper J, Priest AN, Schulze D, Kahl-Nieke B, Adam G, Klocke A.Orthodontic springs and auxiliary appliances: assessment of mag-netic field interactions associated with 1.5 T and 3 T magneticresonance systems. Eur Radiol 2007;17:533-40.

5. G€org€ul€u S, Ayyildiz S, Kamburoglu K, G€okce S, Ozen T. Effect oforthodontic brackets and different wires on radiofrequency heat-ing and magnetic field interactions during 3-T MRI. Dentomaxil-lofac Radiol 2014;43:20130356.

6. Elison JM, Leggitt VL, Thomson M, Oyoyo U, Wycliffe ND. Influ-ence of common orthodontic appliances on the diagnostic qualityof cranial magnetic resonance images. Am J Orthod DentofacialOrthop 2008;134:563-72.

7. Aizenbud D, Hazan-Molina H, Einy S, Goldsher D. Craniofacialmagnetic resonance imaging with a gold solder-filled chain-likewire fixed orthodontic retainer. J Craniofac Surg 2012;23:e654-7.

8. Tymofiyeva O, Vaegler S, Rottner K, Boldt J, Hopfgartner AJ,Proff PC, et al. Influence of dental materials on dental MRI. Den-tomaxillofac Radiol 2013;42:20120271.

9. Klocke A, Kahl-Nieke B, Adam G, Kemper J. Magnetic forces onorthodontic wires in high field magnetic resonance imaging(MRI) at 3 tesla. J Orofac Orthop 2006;67:424-9.

10. Yoshioka H, Schlechtweg PM, Kose K. Magnetic resonance imag-ing. In: Weissman B, editor. Imaging of arthritis and metabolicbone disease. Philadelphia: Mosby/Elsevier; 2009. p. 34-48.

11. Hornak JP. The basics of MRI. 1996. Available at: http://www.cis.rit.edu/htbooks/mri/. Accessed October 27, 2014.

12. Knelrim RW. Invisible mandibular cuspid to cuspid retainer. AngleOrthod 1973;43:218-20.

13. Little RM, Riedel RA,�Artun J. An evaluation of changes in mandib-ular anterior alignment from 10 to 20 years postretention. Am JOrthod Dentofacial Orthop 1988;93:423-8.

14. Zachrisson BU. Clinical experience with direct-bonded orthodonticretainers. Am J Orthod 1977;71:440-8.

15. Zachrisson BU. The bonded lingual retainer and multiple spacingof anterior teeth. J Clin Orthod 1983;17:838-44.

Journal of Orthodontics and Dentofacial Orthopedics

Shalish et al 609

16. Pandis N, Vlahopoulos K, Madianos P, Eliades T. Long-term peri-odontal status of patients with mandibular lingual fixed retention.Eur J Orthod 2007;29:471-6.

17. Keim RG, Gottlieb EL, Nelson AH, Vogels DS 3rd. 2002 JCO studyof orthodontic diagnosis and treatment procedures. Part 1. Resultsand trends. J Clin Orthod 2002;36:553-68.

18. Valiathan M, Hughes E. Results of a survey-based study to identifycommon retention practices in the United States. Am J OrthodDentofacial Orthop 2010;137:170-7.

19. Vandevska-Radunovic V, Espeland L, Stenvik A. Retention:type, duration and need for common guidelines. A survey

Nostalgia Advertisement from a

American Journal of Orthodontics and Dentofacial Orthoped

of Norwegian orthodontists. Orthodontics (Chic.) 2013;14:e110-7.

20. Dang CM, Zaghiyan K, Karlan SR, Phillips EH. Increased use of MRIfor breast cancer surveillance and staging is not associated withincreased rate of mastectomy. Am Surg 2009;75:937-40.

21. Dagia C, Ditchfield M. 3T MRI in paediatrics: challenges and clin-ical applications. Eur J Radiol 2008;68:3009-19.

22. Costa AL, Appenzeller S, Yasuda CL, Pereira FR, Zanardi VA,Cendes F. Artifacts in brain magnetic resonance imaging due tometallic dental objects. Med Oral Patol Oral Cir Bucal 2009;1:E278-82.

1935 issue of the Journal

ics May 2015 � Vol 147 � Issue 5