surface deterioration and elemental composition of retrieved orthodontic miniscrews

ORIGINAL ARTICLE

Surface deterioration and elemental compositionof retrieved orthodontic miniscrews

Pradnya Patil,a Om Prakash Kharbanda,b Ritu Duggal,c Taposh K. Das,d and Dinesh Kalyanasundarame

New Delhi, India

aPostgCentrNew DbProfeCentrNew DcProfeDentaDelhi,dAddiMicroeAssisTechnAll InAll auPotenAddreand DInstitucom.Subm0889-Copyrhttp:/

S88

Introduction: This study provides insight into surface and elemental analyses of orthodontic retrieved mini-screw implants (MSIs). The sole purpose was to investigate the behavior of MSIs while they are in contactwith bone and soft tissues, fluids, and food in the oral cavity. The information thus gathered may help tounderstand the underlying process of success or failure of MSIs and can be helpful in improving their ma-terial composition and design. Methods: The study was carried out on 28 titanium-alloy MSIs (all from thesame manufacturer) split into 3 groups: 18 MSIs were retrieved after successful orthodontic treatment, 5were failed MSIs, and 5 were as-received MSIs serving as the controls. All MSIs were subjected to energydispersive x-ray microanalysis to investigate the changes in surface elemental composition and to scanningelectron microscopy to analyze their surface topography. Data thus obtained were subjected to suitablestatistical analyses. Results: Scanning electron microscope analysis showed surface manufacturing imper-fections of the as-received MSIs in the form of stripes. Their elemental composition was confirmed to thespecifications of the American Society for Testing of Materials for surgical implants. Retrieved MSIsexhibited generalized surface dullness; variable corrosion; craters in the head, neck, body, and tip regions;and blunting on tips and threads. Energy dispersive x-ray analyses showed deposition of additionalelements: calcium had greater significance in its proportion in the body region by 0.056 weight percent; ironwas seen in greater proportion in the failed retrieved MSIs compared with the successful miniscrews;cerium was seen in greater proportions in the head region by 0.128 weight percent and in the neck regionby 0.147 weight percent than in the body and tip regions of retrieved MSIs. Conclusions: Retrieved MSIsshowed considerable surface and structural alterations such as dullness, corrosion, and blunting of threadsand tips. Their surfaces showed interactions and adsorption of several elements, such as calcium, at thebody region. A high content of iron was found on the failed MSIs, and cerium was seen in the head andneck regions of retrieved MSIs. (Am J Orthod Dentofacial Orthop 2015;147:S88-100)

raduate student, Division of Orthodontics and Dentofacial Deformities,e for Dental Education and Research, All India Institute of Medical Sciences,elhi, India.ssor and head, Division of Orthodontics and Dentofacial Deformities,e for Dental Education and Research, All India Institute of Medical Sciences,elhi, India.ssor, Division of Orthodontics and Dentofacial Deformities, Centre forl Education and Research, All India Institute of Medical Sciences, NewIndia.tional professor, Department of Anatomy; officer in charge, Electronscope Facility, All India Institute of Medical Sciences, New Delhi, India.tant professor, Centre for Biomedical Engineering, Indian Institute ofology, Delhi; assistant professor, Department of Biomedical Engineering,dia Institute of Medical Sciences, New Delhi, India.thors have completed and submitted the ICMJE Form for Disclosure oftial Conflicts of Interest, and none were reported.ss correspondence to: Om Prakash Kharbanda, Division of Orthodonticsentofacial Deformities, Centre for Dental Education and Research, All Indiate of Medical Sciences, New Delhi 110029, India; e-mail, opk15@hotmail.

itted, July 2014; revised and accepted, October 2014.5406/$36.00ight � 2015 by the American Association of Orthodontists./dx.doi.org/10.1016/j.ajodo.2014.10.034

The orthodontic miniscrew, a temporary anchoragedevice, was introduced by Gainsforth and Higley1

in an animal study. In each of 5 dogs, a Vitalliumscrew was placed in the anterior border of the ramus ofthe mandible to apply traction by means of an orthodon-tic elastic connected to a maxillary appliance for skeletalanchorage. However, Kanomi2 in 1997 first describedthe “mini-implant” specifically designed for orthodonticapplications. Over the last 2 decades, titanium mini-screws have gained enormous popularity in orthodonticsand are often regarded as the source of absolute intrao-ral anchorage for clinical purposes.3 The popularity ofthese devices is due to their low cost, small dimensions,ease of insertion and removal, and the possibility ofapplying immediate loading, thereby reducing the totalorthodontic treatment duration.4,5 The introduction ofminiscrew implants (MSIs) in the orthodonticarmamentarium has widened the scope and envelopeof orthodontic treatment to some extent. It is possible

Delta:1_given name

Delta:1_surname

Delta:1_given name

Delta:1_surname

Delta:1_given name

mailto:[email protected]

mailto:[email protected]

http://dx.doi.org/10.1016/j.ajodo.2014.10.034

Patil et al S89

to treat patients with moderate to severe skeletaldiscrepancies and obtain complex tooth movementsthat were not possible previously. Compared with otherforms of compliance-dependent anchorage, MSI-supported anchorage offers a more predictable outcome.

MSIs are manufactured from commercially pure tita-nium and grade V titanium alloy. Titanium alloy isfavored because of its higher strength relative tocommercially pure titanium. Contemporary MSIs aredesigned for ease of insertion and are generally safe touse. However, they have been reported to cause gingivalinjuries and have occasionally been found to undergofracture because of mechanical failure in the oral envi-ronment. Failed MSIs necessitate their removal orreplacement.6 Several factors influence the success oforthodontic MSIs, including careful patient selection,the characteristics of the implantation site, and themacrostructure and microstructure properties of theimplants.7-9 Even though titanium alloys are known tobe exceptionally corrosion resistant because of thestability of the passive titanium oxide layer on thesurface, MSIs have been reported to undergo corrosionafter clinical applications.10 Generally, corrosion isobserved when the titanium oxide film breaks downlocally, and rapid dissolution of the underlying metaloccurs in the form of pits.11 Crevice corrosion occursbetween 2 close surfaces or in constricted places whereoxygen exchange is not available.11 When an implantis milled and placed in bone, the stress on the MSI mightlead to stress corrosion or cracking of the alloy. Thiscracking may propagate in the physiologic or the corro-sive environment.

Once the required orthodontic objectives have beenachieved, the MSIs are removed from patient's bone.After removal, the retrieved devices are usually dis-carded. However, economic factors or environmentalconservation might influence clinicians to consider reus-ing MSIs. Not all temporary implant devices can bereused, but metal implants, such as those made fromtitanium alloy, may be more amenable to reuse becausethey can be mechanically and chemically cleaned andresterilized with potentially little or no loss of form orfunction.12

Several studies related to mechanical, chemical, andsurface characteristics of prosthetic dental implants areavailable in the literature.13-15 On the other hand, welocated 4 studies describing the surface and mechanicalnatures of retrieved orthodontic MSIs.8,16-18 Eliadeset al8 characterized the morphologic, structural, andcompositional alterations, and assessed the changes inthe hardness of orthodontic MSIs retrieved after success-ful service. They reported that used titanium-alloy MSIshave morphologic and surface structural alterations.

American Journal of Orthodontics and Dentofacial Orthoped

Their Vickers microhardness testing showed no changein surface hardness of the retrieved specimens comparedwith the controls. Mattos et al16 compared the surfacemorphology and fracture torque resistance ofas-received, sterilized, and retrieved mini-implants toevaluate the fracture risks of reusing orthodontic mini-implants after sterilization. They reported that no defectsor corrosion could be identified in autoclaved andretrieved mini-implants, but worn surfaces and scratchmarks were observed. A statistically significant differencein the fracture torque was observed between theas-received and retrieved groups. Sebbar et al17,18

assessed the surface changes in MSIs retrieved afterusage and compared them with as-received MSIs underan optical microscope. Used MSIs showed signs ofcorrosion mainly at the sites of manufacturing defects.

The literature lacks comprehensive information onused miniscrews regarding their surface and elementalcomposition. Therefore, a study aimed at surface andelemental analyses of successful miniscrews afterretrieval compared with failed and as-received mini-screws was undertaken. Our sole purpose was to investi-gate the behavior of MSIs while in contact with bone andsoft tissues, oral fluids, and food. The information thusgathered may help orthodontists to understand the intri-cacies of success or failure of MSIs related to their designand material composition.

MATERIAL AND METHODS

This study was conducted with 28 MSIs, 8 mm long,1.5 mm in diameter, bracket head type, self-drilling,made from grade V titanium alloy. All MSIs wereprocured from same manufacturer (Absoanchor; Dentos,Daegu, Korea) to prevent any bias in design and materialproperties. Of the 28 MSIs, 18 were retrieved from pa-tients after successful service of 12.89 6 5.33 months.Five MSIs, which had to be retrieved because ofloosening failure during treatment (duration,6.8 6 2.86 months), constituted the failed retrievedgroup. All MSIs were retrieved from buccal interradicularbone between the second premolar and the permanentfirst molar. These retrieved MSIs were used in 8 patients(4 male, 4 female; mean age, 17.75 6 6.08 years) withClass I bimaxillary protrusion malocclusions. TheseMSIs were used for en-masse retraction of anterior teeth(11 in the maxillary arch, 12 in the mandibular arch). Anorthodontic force of 200 g was applied on each MSI. Weused Nitinol closed-coil springs of 9 mm length (GAC In-ternational Inc, Central Islip, NY), which applied aconsistent retraction force. These retrieved miniscrewswere part of the ongoing research in which a standardprotocol of insertion and force application was used.

ics April 2015 � Vol 147 � Issue 4 � Supplement 1

Fig 1. Four zones of investigation on the surface of anMSI.

Table I. Physical properties of maxillary bone, artifi-cial bone model, and MSI

MaterialElastic

modulus (GPa)Density(kg/m3)

Cortical bone in maxillary region 14.7 2000Cancellous bone in maxillary region 1.5 1000Artificial bone model outermost layerrepresenting cortical bone (madewith short fiber-filled epoxy sheets)

16.7 1660

Artificial bone model core layerrepresenting cancellous bone (madewith solid polyurethane foam)

0.759 0640

MSI 96 4620

S90 Patil et al

Informed consent was obtained from all patients beforeMSI placement. Five as-received MSIs served as the con-trol group.

Each MSI was retrieved with a long-hand driver sup-pliedby themanufacturer, consistingof bodyand tip.Afterremoval, each MSI was gently washed and storedcompletely immersed in fresh deionized water in an auto-claved glass vial duly labeled.12,16 Deionized water wasconsidered a suitable medium for storage because it hasbeen reported to be free from any other elements andhas the least effect on the dissolution of elements fromthe cementum of premolars. In addition to its inertness,it also has minimal effects on physical properties such asthe hardness and the elastic modulus of teeth.19

Analysesby energydispersive x-raymicroanalysis (EDX)and scanning electron microscopy (SEM) were done at 4zones of eachMSI: head, neck, body, and tip (Fig 1). Thesezones were considered unique because each zone isexposed to a different environment in the body. Thehead is exposed to oral fluids and food, whereas theneck is in contact with oral mucosa and gingiva, and thebody is in contact with bone. Although the tip is immersedin bone, it may show a difference in behavior comparedwith the body because of its tiny dimensions. Corticaland cancellous bone properties are given in Table I.

Each MSI was mounted on a carbon stub and kept ina dessicator for 24 hours. An EDX detector (Oxford In-struments, Abingdon, Oxfordshire, United Kingdom)was used to investigate their elemental compositionwith an x-ray microanalysis detector. The quantitativeanalysis of the percentage of weight concentration ofthe probed elements was performed with an INCAmicro-analyzer (version 4.06; Oxford Instruments). The carbonquantification was excluded from this study because of

April 2015 � Vol 147 � Issue 4 � Supplement 1 American

the technical limitation of the EDX in an electron micro-scope.20 Elements with lower atomic masses such as car-bon (atomic number 6 and lower) are difficult todistinguish from each other using EDX. The carbonx-rays have low energy and are easily absorbed by thex-ray detector windows. Furthermore, there can be a sig-nificant carbon background signal because of hydrocar-bon contamination. Hydrocarbons from the chambersurfaces, vacuum pumps, and sample surface migrateand react with the electron beam to form a black spotthat is rich in carbon.20

After elemental analysis with the EDX, the retrievedMSIs were subjected to a cleaning cycle of 30 minutesin an ultrasonicator, completely immersed in enzymaticdetergent (Cidezyme; ASP: a Johnson and Johnsoncompany, Irvine, Calif) so that organic debris wouldbe removed and the surface topography of the MSIscould be fully observed under the microscope.16 Aftercleaning, they were again mounted on carbon stubsand kept in a dessicator for 24 hours, and surface im-ages were taken with SEM (Leo 435VP; SEMTech Solu-tions, Cambridge, United Kingdom). Alterations ofsurface changes were looked for: eg, crevice corrosion,corrosion surface damages, dullness, cracks, craters,fractures, and blunting (Table II). After initial scanningof each zone, images of the damaged features or areaswere captured. Multiple images of each zone of theMSI were taken; the number of images varied foreach zone (about 2-4 images per zone of each MSI).Digital images were obtained at various magnifications(20-500 times) in an incremental manner. Experimentalconditions were vacuum at 10�4 mm of mercury; extrahigh tension voltage level, 20 keV; working distance,37 mm; and dead time, below 30%.

Statistical analysis

The statistical analysis of the data was performed us-ing SPSS statistical software (version 20; IBM, Armonk,NY). The Kruskal-Wallis and the Wilcoxon rank sum tests

Journal of Orthodontics and Dentofacial Orthopedics

Table II. Surface changes on MSIs and comparisons between 3 groups and 4 zones: values depict the number of siteson which the surface changes were present

Zone/Group

Corrosion surfacedamage

Pvalue

DullnessP

value

CracksP

value

CratersP

value

FractureP

value

BluntingP

valueS T S T S T S T S T S THeadC 0 15 0 5 0 5 0 5 0 5 NAFR 7 15 0.49 5 5 0* 0 5 0.4 1 5 0.79 0 5 1SR 31 54 18 18 5 18 4 18 2 18

NeckC 0 15 0 5 0 5 0 5 0 5 NAFR 7 15 0.72 5 5 0* 0 5 NA 3 5 0.14 0 5 NASR 28 54 18 18 0 18 8 18 0 18

BodyC 0 15 0 5 0 5 0 5 0 5 0 5FR 12 15 0.009* 5 5 0* 0 5 1 3 5 0* 0 5 0.41 5 5 0*SR 54 54 18 18 3 18 18 18 4 18 18 18

TipC 0 15 0 5 0 5 0 5 0 5 0 5FR 7 15 0.001* 5 5 0* 1 5 1 1 5 0.35 0 5 1 5 5 0*SR 47 54 18 18 3 18 7 18 2 18 18 18

S, Number of sites in each zone where the parameter was present; T, total number of sites examined in each zone; NA, not applicable; C, controlMSIs; FR, failed MSIs; SR, successful MSIs.*P\0.05 was statistically significant; Fischer exact test applied.

Patil et al S91

were applied to compare the EDX data on elementalcomposition, and the chi-square test was used for thecomparison of the SEM results. We had 3 groups (con-trol, and successful and failed MSIs) and 4 zones(head, neck, body, and tip). To understand the varia-tions, 2 types of analysis were performed: comparisonsamong the groups and zones, and multiple comparisonsbetween any 2 groups at a time. The Kruskal-Wallismethod was used to find the P values among the 3groups; P values less than 0.05 were considered signifi-cant. In the case of multiple comparisons within the 3groups—analyzing any 2 groups at a time—post hocanalysis of variance was done with the Wilcoxon ranksum method and the Bonferroni correction; P valuesless than 0.016 were considered significant. TheKruskal-Wallis and Wilcoxon rank sum test results areshown in Tables III and IV.

To understand the effects of frictional resistanceencountered during miniscrew insertion and removal inbone, an experiment was simulated using artificialbone model (Sawbones; Pacific Research Laboratories,Vashon Island, Wash). The artificial bone block has shortfiber-filled epoxy sheets (1 mm) and solid rigid polyure-thane foam (40 per cubic foot) representing corticaland cancellous bones, respectively (Table I).21 Five as-received MSIs with similar specifications were used.Each MSI was inserted into the artificial bone at an angleof 45� with a screwdriver in an artificial bonemodel blockof 23 2 in mounted on a fixer. TheMSI was immediately


removed by unscrewing it with the same screwdriver. Thetechnique of insertion and removal simulated themethod used in the clinical environment. Each MSI waspacked in an airtight bag and observed by SEM accordingto the protocol described above. The artificial boneexperiment was done after our study as an additionalsubset experiment to corroborate our findings: ie, toanalyze whether bone friction alone could have causedthe kind of surface damage seen in the retrieved MSIs.This subset was not used for data analysis with theabove-mentioned 3 groups for evaluation.

RESULTS

The as-received MSIs (control) showed a relativelysmooth appearance to the naked eye but had surfacemilling and polishing defects (manufacturing defects)in the form of stripes and scratches (Figs 2, A, and 3,A and C). Dullness was present in retrieved MSIs (suc-cessful and failed) at all the sites. In the retrieved MSIsfrom the artificial bone model, minimal blunting at thetip and dullness was seen compared with the as-received group; however, no other findings such ascorrosion or surface damage were evident.

Crevice corrosion, which occurs in the crevices be-tween the implant and the tissue, was prominent onthe retrieved MSIs. Other forms of corrosion such aspitting and fretting would be difficult to distinguishwithout elaborate experiments. Hence, for this study,


Table III. EDX findings on surface elements (weight percent) of control (C), failed retrieved (FR), and successfulretrieved (SR) MSIs: single comparison (Kruskal-Wallis test)

Element/Group Head Neck Body Tip

P value forcomparisonwithin groups

P value forcomparisonwithin zones

TitaniumC 89.75 6 0.07 89.75 6 0.07 89.75 6 0.07 89.75 6 0.07 0.03* 0.67FR 73.264 6 14.084 75.202 6 13.848 82.018 6 9.728 73.295 6 16.854SR 84.395 6 5.489 74.222 6 14.408 90.322 6 1.641 70.871 6 14.16

AluminumC 5.97 6 0.01 5.97 6 0.01 5.97 6 0.01 5.97 6 0.01 0.001* 0.656FR 3.762 6 1.536 2.755 6 2.055 3.659 6 2.926 2.92 6 1.568SR 4.304 6 2.739 2.769 6 2.341 0.805 6 0.300 2.534 6 1.421

VanadiumC 4.07 6 0.07 4.07 6 0.07 4.07 6 0.07 4.07 6 0.07 0.001* 0.639FR 3.276 6 2.248 1.69 6 1.477 1.786 6 1.581 1.502 6 1.349SR 0.811 6 1.404 0.865 6 1.499 2.071 6 3.558 1.249 6 1.093

OxygenC 0.10 6 0.01 0.10 6 0.01 0.10 6 0.01 0.10 6 0.01 0.009* 0.384FR 11.354 6 18.002 19.011 6 9.385 10.186 6 8.183 19.985 6 12.829SR 8.795 6 4.008 21.302 6 11.253 5.813 6 5.105 23.742 6 11.344

CalciumC 0.00 6 0.00 0.00 6 0.00 0.00 6 0.00 0.00 6 0.00 0.031* 0.529FR 0.215 6 0.187 0.17 6 0.03 0.165 6 0.181 0.189 6 0.17SR 0.162 6 0.169 0.15 6 0.026 0.221 6 0.229 0.139 6 0.144

IronC 0.00 6 0.00 0.00 6 0.00 0.00 6 0.00 0.00 6 0.00 0.045* 0.218FR 0.504 6 0.161 0.87 6 0.116 0.387 6 0.374 0.266 6 0.076SR 0.305 6 0.329 0.59 6 0.102 0.160 6 0.204 0.114 6 0.129

CeriumC 0.00 6 0.00 0.00 6 0.00 0.00 6 0.00 0.00 6 0.00 0.136 0.049*FR 0.126 6 0.149 0.184 6 0.318 0.00 6 0.00 0.00 6 0.00SR 0.131 6 0.228 0.111 6 0.097 0.00 6 0.00 0.00 6 0.00

NitrogenC 0.00 6 0.00 0.00 6 0.00 0.00 6 0.00 0.00 6 0.00 0.293 0.412FR 2.58 6 3.62 3.04 6 4.58 1.54 6 1.45 0.00 6 0.00SR 1.97 6 3.90 1.70 6 2.62 5.86 6 11.61 0.02 6 0.06

*P\0.05 was statistically significant.

S92 Patil et al

we called all such changes “corrosion surface damage.”Corrosion surface damage and craters were greater inthe retrieved MSIs (greater in the successful retrievedgroup than in the failed retrieved group); there werenone in the control group. Corrosion surface damagewas greater at the body and tip regions (Figs 2, B, and3, B). Craters were most common at the body regionand least common at the head region. No significantcracks and fractures were evident in the control groupbut were seen on 4 retrieved MSIs (Fig 2, C and D).Blunting was present in all retrieved MSIs at the tipsand threads and was absent in the controls (Fig 3, Cand D) (Table II). In the immediately retrieved MSIsfrom the artificial bone model, no other findings suchas corrosion or other surface damage were evident; how-ever, minimal blunting at the tip and dullness were seencompared with the as-received group.


The surface elemental composition of as-receivedMSIs was found to be within the ASTM standard(F136-13). Retrieved MSIs showed adsorption of oxy-gen, nitrogen, calcium, phosphorus, iron, fluorine,sodium, chlorine, magnesium, cerium, and potassiumin varying amounts (Table III; Fig 4, A and B). Statisti-cally significant findings were as follows. Calcium wasin greater proportion in the successful retrieved groupthan in the failed retrieved group in the body regionby 0.056 weight percent (P 5 0.0149) (Fig 5, A)(Tables III and IV). Iron was in greater proportion at all4 zones in the failed MSIs than in the successful MSIs(P 5 0.045) (Tables III and IV) by 0.199 weightpercent in the head region, by 0.28 weight percent inthe neck region, by 0.227 weight percent in the bodyregion, and by 0.152 weight percent in the tip region(Fig 5, B). Cerium was in greater proportion in the


Table IV. EDX findings on surface elements (weightpercent) of control (C), failed retrieved (FR), and suc-cessful retrieved (SR) MSIs: multiple comparisonswithin groups (Wilcoxon rank sum test): numbersdepict P values of comparisons

Zone/ElementC (n 5 5)

vs FR (n 5 5)FR (n 5 5)

vs SR (n 5 18)SR (n 5 18)vs C (n 5 5)

HeadTi 0.0086* 0.4561 0.0008*Al 0.0086* 0.5023 0.0008*V 0.0086* 0.4123 0.0008*O 0.0086* 0.6018 0.0008*Ca 0.0411 0.0568 0.0568Fe 0.0647 0.0121* 0.0324Ce 0.0539 0.3653 0.1094N 0.0521 0.7815 0.0527

NeckTi 0.0086* 0.4123 0.0008*Al 0.0086* 0.4123 0.0008*V 0.0273 0.4123 0.0022*O 0.0086* 0.136 0.0008*Ca 0.0973 0.9178 0.2594Fe 0.0860 0.0098* 0.0457Ce 0.0539 0.3923 0.0396N 0.0451 0.3615 0.0578

BodyTi 0.0086* 0.2051 0.0017*Al 0.115 0.8815 0.0017*V 0.115 0.8815 0.0112*O 0.0086* 0.1547 0.1339Ca 0.0489 0.0179 0.0079*Fe 0.0539 0.0111* 0.1478Ce NA NA NAN 0.0524 0.0203 0.0186

TipTi 0.0459 0.8231 0.0017*Al 0.0459 0.8231 0.0017*V 0.115 0.8815 0.0029*O 0.115 0.1563 0.009*Ca 0.0317 0.0187 0.0014*Fe 0.0539 0.0132* 0.446Ce NA NA NAN 0.1547 0.8812 0.8812

NA, Not applicable: indicates that both sets of data points had only anumeric value of zero; hence, no comparison could be made.Ti, Titanium; Al, aluminum; V, vanadium; O, oxygen; Ca, calcium;Fe, iron; Ce, cerium; N, nitrogen.*P\0.016 was statistically significant.

Patil et al S93

head region by 0.128 weight percent and the neck regionby 0.147 weight percent than in the body and tip regionsof the retrieved MSIs (P5 0.049), and there was none inthe control group (Tables III and IV; Fig 5, C). Nosignificant difference was found in the MSIs retrievedfrom the maxilla and the mandible in the SEM andEDX investigations.

In this study, no prior sample size calculation couldbe done at the planning stage because of the nature of


the experiment and the scarce information available.At the end of the study, the power calculation wascomputed separately for surface changes (SEM) andEDX data based on the results. For the surface-change(SEM) data, the surface alterations of all 4 zones ofeach MSI (head, neck, body, and tip) were summed asthe number of alterations per MSI. Subsequently, meansand standard deviations of surface changes were derivedand computed for each group separately: control, failedretrieved, and successful retrieved. The power values ofthe study (at the 95% confidence level and 5% alpha)were control vs failed retrieved, 100%; control vs suc-cessful retrieved, 100%; and successful retrieved vsfailed retrieved, 91%.

For the EDX data, we used calcium and cerium asrepresentative denominators because these exhibitedmajor variations. Quantitative values of calcium for thebody and tip regions and of cerium for the head andneck regions were used. The computed power valuesto detect mean differences of calcium in the tip regionbetween control vs failed retrieved, control vs successfulretrieved, and failed retrieved vs successful retrievedwere 14%, 82%, and 61%, respectively. For the samegroups in the body region, the power values were58%, 60%, and 60%, respectively. The power for ceriumwas weak. In the control vs failed retrieved, control vssuccessful retrieved, and failed retrieved vs successfulretrieved, the power values were 38%, 30%, and 18%for the neck region and 26%, 23%, and 10% for thehead region, respectively.

DISCUSSION

Retrieval analysis is an emerging field in biomedicalmaterials because of the critical information it provideson the performance of the material in the environmentin which it is intended to function. The developmentof international standards for the retrieval analysis ofbiomaterials strongly indicates the significance of thismethod in studying the performance of materials.22,23

Our results indicate substantial changes in the surfaceprofiles of MSIs and the presence of impuritiesprecipitated on the surfaces of retrieved MSIs, whichare attributed to contact of the implant with oraltissues, biologic fluids, blood and exudates, saliva, andfood.8

Retrieved MSIs showed loss of gloss and surface fin-ish, resulting in a dull surface at all 4 examined zones.8 Itwas assumed that insertion and removal of miniscrewcauses the surface to wear out to some extent. Bluntingof threads and tip (less sharp thread and tip comparedwith as-received screws)16 was also linked to wear duringthe process of insertion and removal (Fig 3, D), and so


Fig 2. A, Threaded body region of as-received MSI at 500-times magnification (arrow) (micromarker,200 mm);B,wear type of corrosion defect on thread of retrieved MSI at 500-times magnification (arrow)(micromarker, 200 mm); C, crack on thread of retrieved MSI at 500-times magnification (arrow) (micro-marker, 200 mm); D, fracture on a thread of retrieved MSI at 500-times magnification (arrow) (micro-marker, 200 mm).

S94 Patil et al

the occasional fracture of the thin thread at its outer-most border or tip was attributed to poor mechanicalstrength in these regions (Fig 2, C and D). A subset of5 miniscrews used in a simulated experiment on artificialbone showed only minor blunting of the tips and dull-ness of the surfaces of the screws. Therefore, it appearsthat in addition to physical bone contact during inser-tion and removal, other factors may have also contrib-uted to the major surface changes on the MSIs, andthese will be elaborated in the ensuing paragraphs.

If the implant pierces the mucosa and drills in bone, itcauses traumatic inflammation, including reduction ofthe pH in the early exudative phases, activation of cellsincluding polymorphonuclear granulocytes and macro-phages, and release of proteins, enzymes, and oxidizingagents that might contribute to the surface alterations.8

A primary requisite for any metal used in the mouth isthat it must be biocompatible and should not corrodewhen in contact with the tissues. Factors such as temper-ature, quantity and quality of saliva, plaque, pH, pro-teins, the physical and chemical properties of foodsand liquids, and oral health conditions can influencecorrosion.9 The composition of saliva varies considerably


between persons and at different times. Saliva, a physi-ologic fluid in oral cavity, also has variations in pH atdifferent times of the day and among different persons(pH 5 6.2-7.4).24 Moreover, oral hygiene has a strongeffect on the corrosiveness of the oral environment.

Titanium alloys used to manufacture MSIs are lessresistant to corrosion because the alloys represent dis-continuities in the protective oxide film.11 The surfacemilling and polishing defects during the manufacturingprocess are seen in the form of stripes and scratches.These tiny defects can be a starting point for electro-chemical attack when miniscrews are inserted in thebody.17 Although titanium alloys are considered highlycorrosion resistant because of the stable passive titaniumoxide layer on the surface, they are not inert to corrosiveattack. When the stable surface oxide layer breaks downor is removed and cannot reform on parts of the surface,titanium can be corrosive, as are many other basemetals.25 Titanium has a high coefficient of friction.Corrosion takes place when the passive oxide film isworn off from contact with bone. Wear particles mayform because of corrosion (Fig 2, B).11 This type ofcorrosion is responsible for most of the metal released


Fig 3. A, Junction between the body and thread regions of as-received MSI at 70-times magnification(arrow) (micromarker, 1 mm); B, crevice corrosion at the junction between the body and thread regionsof MSI at 70-times magnification (arrow) (micromarker, 1 mm); C, tip of as-received MSI at 500-timesmagnification (arrow shows the tip region; circle shows the smooth surface) (micromarker, 200 mm); D,blunting of the tip of retrieved MSI at 500-times magnification (arrow shows the blunt tip region; circleshows the deposition) (micromarker, 200 mm).

Patil et al S95

into tissues.26-28 In contrast, pure titanium is highlyresistant to corrosion in any in-vivo environment likelyto be encountered; however, its low strength limits itsuse as an implant.

The mechanical load on orthodontic MSIs is nonvi-brating and nondynamic. Also, because the bone issofter than the implant, the damaging effects on theimplant surfaces from mechanical loading are assumedto be minimal. Therefore, the surface damage on theoxide layer during insertion can be significant in contrib-uting to corrosion.29 Crevice corrosion (Fig 3, B) occursbetween 2 close surfaces or in constricted places, trap-ping a stagnant layer of solution where oxygenexchange is not available. This is because pH in the crev-ices can be as low as 2 (highly acidic) for a 7 pH (neutral)solution.30 The reduction in pH further causes initiationand propagation of the crevice corrosion phenomenon.When the acidity of the milieu increases with time, thepassive layer of the alloy dissolves, and it acceleratesthe local corrosion process. Natural convection no


longer allows the trapped solution to mix with thebulk solution outside, so that diffusion is the onlyform of mass transport by which dissolved oxygen canenter the occluded region. In such crevices, the supplyof dissolved oxygen in the trapped solution can bedepleted. As a result, the anodic corrosion reaction oc-curs in the crevices, and the supporting cathodic reduc-tion reaction occurs on the much larger surfaces outsidethe crevices.30 Crevice-like geometric patterns wereobserved on threads, tips, heads, and junctions of thehead-neck and neck-thread regions. This phenomenonexplains why more crevice corrosion occurred in theseareas of the retrieved miniscrews.

A crater is a cavity or hole in any surface.31 A crack is aline on the surface of something along which it has splitwithout breaking apart.32 A fracture is the cracking orbreaking of a hard object or material.13 In our study,corrosion surface damage and craters were found to besignificantly present in successful retrieved MSIs becausethe extended periods of retention (12.896 5.33 months)


Fig 4. A, Representative graph of EDX plot of as-received MSI (elemental composition confirmed toASTM standards); B, representative graph of EDX plot of retrieved MSI (the additional significant ele-ments are oxygen, calcium, iron, and cerium). Ti, Titanium; V, vanadium; O, oxygen; Al, aluminum; N,nitrogen;Mg, magnesium;Ce, cerium;Na, sodium; Fe, iron; P, phosphorus;Ca, calcium; keV, kiloelec-tronvolt.

S96 Patil et al

in the jaws allowed longer contact with oral tissues andbiologic fluids. Surface alterations were seen morefrequently in the body region because of the increasedsurface contact area of the threads. Cracks and fractureswere seen mostly in the thread and tip regions. This canbe attributed to the decreased thickness of material atthese regions.

Orthodontic devices in the mouth are subjected toelectrochemical corrosion, which leads to gradual surfacebiodegradation of the surface material by a process ofoxidation, and they trigger the release of potentially toxicor allergenic substances.10 Corrosion can severely limitthe fatigue life and ultimate strength of the material,leading to mechanical failure of the implants.33,34

Corrosion can lead to roughening of the surface,


weakening of the implant, liberation of elements fromthe metal or alloy, and toxic reactions. The liberation ofalloy elements and corrosion products can producediscoloration of adjacent soft tissues and allergicreactions such as oral edema, perioral stomatitis,gingivitis, and extraoral manifestation such aseczematous rashes in susceptible patients. Corrosionproducts have been implicated in causing local pain orswelling near implants in the absence of infection, andthey may be carcinogenic.33 Kasemo35 demonstratedthe dissolution of corrosion products into the bioliquidand adjacent tissues. The pathomechanism of theimpaired wound healing is modulated by specific metalions,36 and corrosion products released during corrosioninfluence the function of the participating cell types—eg,


Fig 5. Graphs of the relative weight percentages of elements on EDX: A, calcium was in greater pro-portion in the successful retrieved group than in the failed retrieved group in the body region by 0.056weight percent;B, iron on the surface of the failed MSIs compared with the successful MSIs had higheradsorption rates at all 4 zones (head region, 0.199 weight percent; neck region, 0.28 weight percent;body region, 0.227 weight percent; tip region, 0.152 weight percent); C, cerium was in greater propor-tion on the head and neck regions of retrieved MSIs (head region, 0.128 weight percent; neck region,0.147 weight percent). C, Control; FR, failed retrieved; SR, successful retrieved. *P\0.05 was statis-tically significant.

Patil et al S97

endothelial cells. Olmedo et al34 reported that macro-phages in peri-implant soft tissues induced by acorrosion process play an important role in implant fail-ure. The corrosion products that are released are phago-cytosed by macrophages, stimulating the release ofinflammatory mediators such as cytokines toward thebone surface, contributing to its resorption by osteoclastactivation, and the metallic particles that result fromcorrosion may directly inhibit osteoblast function, lead-ing to local osteolysis and loss of clinical stability of theMSI.34

The frictional forces during insertion and removal ofMSIs alone do not contribute to major surfacealterations in the absence of interactions with bodyfluids and tissues. Corrosion surface damage seen on ti-tanium alloy MSIs is the result of complex interactions


and a time-consuming phenomenon. Therefore, corro-sion effects were not seen in the immediately retrievedMSIs in a simulated experiment on artificial bone.

For the EDX analysis, the as-received MSIs conformedto the ASTM standards for surgical implants. Titanium,aluminum, and vanadiumare the parent elementsof gradeV titanium alloy, the presence of which was maskedbecause of other elements deposited on the surface ofretrieved MSIs: oxygen, nitrogen, calcium, phosphorus,iron, fluorine, sodium, chlorine, magnesium, cerium, andpotassium (Fig 4). Since EDX provides data in weight per-centages of elements, the 3 parent elements were maskedin the retrieved MSIs because of adsorption of otherelements over the surface from the local biologic milieu.

Oxygen was present relatively in greater proportionsin the retrieved MSIs than in the control MSIs


S98 Patil et al

(Table III); this may have its source from contact withtissues or exposure to the external environment whilethe MSIs were transported from the autoclaved packetto the bone until insertion and removal.12 It is knownthat the titanium oxide passive layer is formed on thesurface because of contact with oxygen.8 Calcium phos-phate precipitates on the titanium oxide layer, changingthe outer oxide layer to complex titanium and calciumphosphates. The calcium is derived mainly from the con-tact of the implant surface with blood involving mainlyadsorption of proteinaceous integuments, which arecalcified later with the precipitation of calcium andphosphorus.37 Also, bone particles, seen adhering tothe miniscrew implant, were derived through intimatecontact with alveolar bone.8,38 Calcium was seen ingreater proportions in the successful retrieved groupthan in failed retrieved group in the body region by0.056 weight percent. Its presence in the successfulretrieved MSIs was enhanced by the extended periodof retention in alveolar bone of 12.89 6 5.33 months(Fig 5, A).8

X-ray photoelectron spectroscopy characterizationstudy on titanium has shown the presence of titanium21, titanium 31, and titanium 41 on its surface.39 Inother words, the surface oxide film on titanium is notstoichiometric and still has scope for further oxidation.Calcium phosphate was formed when titanium wasimmersed in Hanks solution. Calcium and phosphatecannot exist stably alone on titanium and eventuallyformed calcium phosphate on it. The surface oxidefilm on titanium is thus not completely oxidized and isrelatively reactive.39

The greater proportion of iron in the failed MSIscompared with the successful MSIs by 0.214 weightpercent (Fig 5, B) was attributed to contact with blood.8

Since peri-implantitis accounts for about 30% ofminiscrew failures, we hypothesized that iron comesfrom the increased blood flow caused by inflamma-tion.40 There were soft tissues covering or in contactwith the head and neck of MSIs, but the body and tip re-gions were in contact with bone, with no soft-tissuecontact. Also, there was more iron in the failed MSIsthan in the successful ones at the body and tip regions;this suggests that apart from peri-implantitis, bone-related changes also contribute toMSI failures. The pres-ence of nitrogen on the surface implies proteins thatwere adsorbed from body fluids.8 More nitrogen wasfound in the body region of the retrieved MSIs becauseof the greater surface contact area.

Retrieved MSIs exhibited cerium in greater propor-tions in the head and neck regions by 0.137 weightpercent than in the body and tip regions (Fig 5, C).Cerium is a component of a few mouthwashes for its


antimicrobial properties. Cerium is also present insome foods, such as tubers grown in cerium-rich soilsand water.41,42 We attributed mouthwashes and foodsas the sources of the cerium on the heads and necks ofthe retrieved MSIs in our study. Cerium chloride and itscombination with fluoride can significantly reducevarious mineral losses and the progression of the depthof enamel lesions.41 It is also said to have a bacteriostaticeffect.43 A strong binding of cerium nanoparticles to theouter membrane of Staphylococcus aureus causes theinhibition of active transport, dehydrogenase and peri-plasmic enzyme activity, and thus eventually the inhibi-tion of RNA, DNA, and protein synthesis leading to celllysis.43 A cerium coating could lead to reduction inperi-implantitis and therefore a reduction in the failurerate of MSIs. This subject needs further exploration.Minute amounts of other elements were seen: nitrogen,fluorine, sodium, chlorine, magnesium, and potassium.Most of these elements are present in drinking water,foods, mouthwashes, toothpastes, and beverages.

A study of retrieved MSIs by Eliades et al8 showed nodifference in the Vickers microhardness testing betweenas-received and retrieved MSIs, implying that no strain-hardening phenomena occurred despite the self-drillingprocess. The SEM and EDX analyses did not provide sub-stantial information on the greater inflammation associ-ated with loosened MSIs. One possibility of the looseningof MSIs could be due to the biofilm formed on metallicimplants as reported by Arciola et al44 and Schaeret al45; biofilm formation can be significantly reducedby hydrophobic polycationic coatings. We hypothesizedthat the loosening ofMSIs could be due to biofilm forma-tion on metallic implants leading to microbial attack onsurrounding bone and subsequent inflammation. Theconfirmation of biofilm formation would need anelaborative study, as we envisage in a future study.

Thus, the outermost atomic layers of MSIs are criticalregions associated with biochemical surface interactionsof the implant-tissue interface. This should have atremendous influence on a high degree of standardiza-tion and surface control in the production of MSIs.35

This knowledge will be helpful in exploring possibleresearch approaches for determining the biologicproperties of implant materials.

CONCLUSIONS

The results of this retrieval analysis of MSIs havesuggested the following.

1. Retrieved MSIs exhibit morphologic surface changesin the form of dullness, blunting of threads and tips,corrosion, craters, and occasional tearing of thinthreads.


Patil et al S99

2. The surface elemental composition of retrievedMSIs differs from that of as-received MSIs, withadditional elements; the most conspicuous wascalcium in the body region (P\0.05).

3. The surfaces of failed MSIs compared with success-ful MSIs have higher adsorption of iron at all 4 zones(mean, 0.214 weight percent), the source of whichcan be attributed to inflammation around theMSIs in the bone and gingiva.

4. Although the head and neck regions of retrievedMSIs show cerium in greater proportions (mean,0.137 weight percent), these findings should beconfirmed in a larger sample with adequate poweralong with other elements such as iron.

ACKNOWLEDGMENTS

We thank the ElectronMicroscope Facility of All IndiaInstitute of Medical Sciences, New Delhi; and Dr R. M.Pandey, Professor & Head, Biostatistics; Guresh Kumar;and Ashish DU for their contributions to the statisticalanalyses.

REFERENCES

1. Gainsforth BL, Higley LB. A study of orthodontic anchorage possi-bilities in basal bone. Am J Orthod Oral Surg 1945;31:406-17.

2. Kanomi R. Mini-implant for orthodontic anchorage. J Clin Orthod1997;31:763-7.

3. Tsui WK, Chua HDP, Cheung LK. Bone anchor systems for ortho-dontic application: a systematic review. Int J Oral Maxillofac Surg2012;41:1427-38.

4. Papadopoulos MA, Tarawneh F. The use of miniscrew implants fortemporary skeletal anchorage in orthodontics: a comprehensivereview. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 2007;103:e6-15.

5. Wiechmann D, Meyer U, B€uchter A. Success rate of mini- andmicro-implants used for orthodontic anchorage: a prospectiveclinical study. Clin Oral Implants Res 2007;18:263-7.

6. Papageorgiou SN, Zogakis IP, Papadopoulos MA. Failure rates andassociated risk factors of orthodontic miniscrew implants: a meta-analysis. Am J Orthod Dentofacial Orthop 2012;142:577-95.e7.

7. Sbordone L, Traini T, Caputi S, Scarano A, Bortolaia C, Piattelli A.Scanning electron microscopy fractography analysis of fracturedhollow implants. J Oral Implantol 2010;36:105-11.

8. Eliades T, Zinelis S, Papadopoulos MA, Eliades G. Characterizationof retrieved orthodontic miniscrew implants. Am J Orthod Dento-facial Orthop 2009;135:10.e1-7.

9. Locci P, Lilli C, Marinucci L, Calvitti M, Belcastro S, Bellochio S,et al. In vitro cytotoxic effects of orthodontic appliances. J BiomedMater Res 2000;53:560-7.

10. Mikulewicz M, Chojnacka K. Release of metal ions from orthodon-tic appliances by in vitro studies: a systematic literature review.Biol Trace Elem Res 2011;139:241-56.

11. Hin TS. Engineering materials for biomedical applications. 1st ed.Singapore: World Scientific Publishing Company; 2004. p. 3-24. 29.

12. Mouhyi J, Sennerby L, Pireaux JJ, Dourov N, Nammour S, VanReck J. An XPS and SEM evaluation of six chemical and physical


techniques for cleaning of contaminated titanium implants. ClinOral Implants Res 1998;9:185-94.

13. Yokoyama K, Ichikawa T, Murakami H, Miyamoto Y, Asaoka K.Fracture mechanisms of retrieved titanium screw thread in dentalimplant. Biomaterials 2002;23:2459-65.

14. Sch€upbach P, Glauser R, Rocci A, Martignoni M, Sennerby L,Lundgren A, et al. The human bone-oxidized titanium implantinterface: a light microscopic, scanning electron microscopic,back-scatter scanning electron microscopic, and energy-dispersive x-ray study of clinically retrieved dental implants. ClinImplant Dent Relat Res 2005;7(Suppl 1):S36-43.

15. Al Jabbari YS, Fournelle RA, Zinelis S, Iacopino AM. Biotribologicalbehavior of two retrieved implant abutment screws after long-termuse in vivo. Int J Oral Maxillofac Implants 2012;27:1474-80.

16. Mattos CT, de Oliveira Ruellas AC, Elias CN. Is it possible to re-usemini-implants for orthodontic anchorage? Results of an in vitrostudy. Mater Res 2010;13:521-5.

17. SebbarM, Bourzgui F, Aazzab B, Elquars F. Anchorageminiscrews:a surface characterization study using optical microscopy. IntOrthod 2011;9:325-38.

18. Sebbar M, Bourzgui F, Lazrak L, Aazzab B, El Quars F. Microscopiccomparison of the miniscrew’s surface used in orthodontics:before and after use. Rev Stomatol Chir Maxillofac 2012;113:365-9.

19. Darendeliler MA, Kharbanda OP, Chan EK, Srivicharnkul P, Rex T,Swain MV, et al. Root resorption and its association with alter-ations in physical properties, mineral contents and resorptioncraters in human premolars following application of light andheavy controlled orthodontic forces. Orthod Craniofac Res 2004;7:79-97.

20. Rolland P, Carlino V, Vane R. Improved carbon analysis with evac-tron plasma cleaning. Microsc Microanal 2004;10:964-5.

21. Reynders RM, Ronchi L, Ladu L, Etten-Jamaludin FV, Bipat S.Insertion torque and orthodontic mini-implants: a systematicreview of the artificial bone literature. Proc Inst Mech Eng H2013;227:1181-202.

22. International Organization for Standardization. Implants for sur-gery. Retrieval and analysis of surgical implants—part 1: retrievaland handling:ISO 12891-1; 2011.

23. American Society for Testing Materials International. Specificationfor wrought titanium-6 aluminum-4 vanadium ELI (extra lowinterstitial) alloy for surgical implant applications (UNS R56401):ASTM F136-13; 2013.

24. Washington N, Washington C, Wilson CG. Physiological pharma-ceutics barriers to drug absorption. 2nd ed. London, UnitedKingdom: Taylor & Francis; 2001. p. 42.

25. Abey S, Mathew MT, Lee DJ, Knoernschild KL, Wimmer MA,Sukotjo C. Electrochemical behavior of titanium in artificial saliva:influence of pH. J Oral Implantol 2014;40:3-10.

26. Bianco PD, Ducheyne P, Cuckler JM. Titanium serum and urinelevels in rabbits with a titanium implant in the absence of wear.Biomaterials 1996;17:1937-42.

27. Billi F, Onofre E, Ebramzadeh E, Escudero ML, Garc�ıa-Alonso MC.Characterization of modified Ti6Al4V alloy after fretting-corrosiontests using near-field microscopy. Surface Coat Technol 2012;122:134-44.

28. Ebramzadeh E, Billi F, Sangiorgio SN, Mattes S, Schmoelz W,Dorr L. Simulation of fretting wear at orthopaedic implant inter-faces. J Biomech Eng 2005;127:357-63.

29. Rajeswari S, Suresh Kumar Danadurai K, Sridhar TM,Narasimhan SV. Surface characterization and pitting behavior ofhigh-Cr-Ni-Mo alloys in simulated white water environment.Corrosion 2001;57:465-75.


http://refhub.elsevier.com/S0889-5406(14)01146-9/sref1



































































































S100 Patil et al

30. Sato N. 1989 Whitney Award Lecture: toward a more fundamentalunderstanding of corrosion processes. Corrosion 1989;45:354-68.

31. Cochis A, Fini M, Carrassi A, Migliario M, Visai L, Rimondini L. Ef-fect of air polishing with glycine powder on titanium abutmentsurfaces. Clin Oral Implants Res 2013;24:904-9.

32. Gonz�alez-Garc�ıa R, Monje A, Fern�andez-Calder�on MC, Hierro-Oliva M, Gonz�alez-Mart�ın ML, Monje F. Three-dimensional andchemical changes on the surface of a 3-year clinically retrievedoxidized titanium dental implant. J Mech Behav Biomed Mater2014;34:273-82.

33. Wetterhahn KE, Demple B, Kulesz-Martin M, Copeland ES.Workshop report from the Division of Research Grants, NationalInstitutes of Health. Metal carcinogenesis—a chemical pathologystudy section workshop. Cancer Res 1992;52:4058-63.

34. Olmedo D, Fern�andez MM, Guglielmotti MB, Cabrini RL. Macro-phages related to dental implant failure. Implant Dent 2003;12:75-80.

35. Kasemo B. Biological surface science. Surface Sci 2002;500:656-77.

36. Kirkpatrick CJ, Barth S, Gerdes T, Krump-Konvalinkova V, Peters K.Pathomechanisms of impaired wound healing by metalliccorrosion products. Mund Kiefer Gesichtschir 2002;6:183-90.

37. Eliades T, Papadopulos JS, Eliades G, Silikas N, Watts DC.Multi-technique characterization of retrieved bone cementfrom revised total hip arthroplasties. J Mater Sci Mater Med2003;14:967-72.

38. Mangano C, Piattelli A, Raspanti M, Mangano F, Cassoni A, Iezzi G,et al. Scanning electron microscopy (SEM) and x-ray dispersive


spectrometry evaluation of direct laser metal sintering surfaceand human bone interface: a case series. Lasers Med Sci 2011;26:133-8.

39. Hanawa T, Tsutsumi Y. Calcium phosphate formation on titaniumand zirconium and its application to medical devices. BioceramDev Appl 2010;1:1-4.

40. Miyawaki S, Koyama I, Inoue M, Mishima K, Sugahara T, Takano-Yamamoto T. Factors associated with the stability of titaniumscrews placed in the posterior region for orthodontic anchorage.Am J Orthod Dentofacial Orthop 2003;124:373-8.

41. Wegehaupt FJ, Buchalla W, Sener B, Attin T, Schmidlin PR. Ceriumchloride reduces enamel lesion initiation and progression in vitro.Caries Res 2014;48:45-50.

42. Environmental Protection Agency. Toxicological review ofcerium oxide and cerium compounds (CAS No. 1306-38-3):Avail-able at: http://www.epa.gov/iris/toxreviews/1018tr.pdf; 2009.Accessed October 26, 2013.

43. Negahdary M, Mohseni G, Fazilati M, Parsania S, Rahimi G,Rad S, et al. The antibacterial effect of cerium oxide nanopar-ticles on Staphylococcus aureus bacteria. Ann Biol Res 2012;3:3671-8.

44. Arciola CR, Campoccia D, Speziale P, Montanaro L, Costerton JW.Biofilm formation in Staphylococcus implant infections. A reviewof molecular mechanisms and implications for biofilm-resistantmaterials. Biomaterials 2012;33:5967-82.

45. Schaer TP, Stewart S, Hsu BB, Klibanov AM. Hydrophobic polyca-tionic coatings that inhibit biofilms and support bone healingduring infection. Biomaterials 2012;33:1245-54.


























http://refhub.elsevier.com/S0889-5406(14)01146-9/sref35d

http://refhub.elsevier.com/S0889-5406(14)01146-9/sref35d























http://www.epa.gov/iris/toxreviews/1018tr.pdf












surface deterioration and elemental composition of retrieved orthodontic miniscrews

Documents