2015_147_6_725_737

download 2015_147_6_725_737

of 13

Transcript of 2015_147_6_725_737

  • 7/23/2019 2015_147_6_725_737

    1/13

    Quantitative evaluation of maxillaryinterradicular bone with cone-beam computed

    tomography for bicortical placement oforthodontic mini-implants

    Lei Yang,a Feifei Li,b Meng Cao,b Hong Chen,c Xi Wang,c Xuepeng Chen,d Le Yang,c Weiran Gao,e

    Joseph F. Petrone,f and Yin Dingg

    Xi'an, Shaanxi, Taiyuan, Shanxi, and Hangzhou, Zhejiang, China, and Pittsburgh, Pa

    Introduction:The purpose of this study was to propose a protocol for safe bicortical placement of mini-implants

    by measuring the interradicular spaces of the maxillary teeth and the bone quality. Methods: Cone-beam

    computed tomography data were obtained from 50 adults. Three-dimensional reconstructions andmeasurements were made with SimplantPro software (Materialise, Leuven, Belgium). For each interradicular

    site, the bone thicknesses and interradicular distances at the planes 1.5, 3, 6, and 9 mm above the

    cementoenamel junction were measured. Standard bone units were dened to evaluate the inuences of bone

    density and the different placement patterns on the stability of the mini-implants. Results: The safe interradicular

    sites in the maxilla for bicortical placement of 1.5-mm-diameter mini-implants were in all planes between therst

    and second premolars, and between the second premolar and the rst molar. The safe palatal sites were

    between the rst and second molars, and the safe labial sites of the 9-mm plane were between the central

    incisors, and between the lateral incisor and the canine. The safe buccal sites of the 6- and 9-mm planes

    were between the rst and second molars, and the safe buccal sites of the 3-, 6-, and 9-mm planes were

    between the canine and the rst premolar. Most bone thicknesses were from 8 to 12 mm. The optimal

    placement angle between the second premolar and the rst molar was 58. Bicortical placement could have

    more standard bone units than unicortical placement in the maxilla. Conclusions:Bicortical placement would

    be more stable in the maxilla. For the site between the molars, special care should be taken at a plane higher

    than 6 mm to prevent maxillary sinus penetration. The most favorable interradicular area in the maxilla was be-

    tween the second premolar and the rst molar. (Am J Orthod Dentofacial Orthop 2015;147:725-37)

    Appropriate anchorage in orthodontic treatmentis important. Mini-implants have gained consid-

    erable popularity because of their low cost,effectiveness, easy clinical management, and stability.Among the factors related to mini-implant stability,alveolar bone thickness, bone density, placement a

    ngle, and location appear to be critical for successful

    placement. Adequate bone quantity at the placementsite can affect the success of the mini-implants.1 This

    has prompted further research for the ideal sites andthe greatest stability for mini-implants.2-7

    Bone density appears to be a key determinant for thestability of mini-implants in sites with inadequate

    cortical bone thickness because primary retention of

    a

    Postgraduate student, Department of Orthodontics, School of Stomatology,State Key Laboratory of Military Stomatology, Fourth Military Medical

    University, Xi'an, Shaanxi, China; 264th Hospital of Chinese People's Liberation

    Army, Taiyuan, Shanxi, China.bAssociate professor, Department of Orthodontics, School of Stomatology, State

    Key Laboratory of Military Stomatology, Fourth Military Medical University,

    Xi'an, Shaanxi, China.cPostgraduate student, Department of Orthodontics, School of Stomatology,

    StateKey Laboratoryof Military Stomatology, Fourth Military Medical University,

    Xi'an, Shaanxi, China.dAssociate professor, Department of Orthodontics, Hospital of Stomatology,

    Zhejiang University, Hangzhou, Zhejiang, China.eAttending doctor, 264th Hospital of Chinese People's Liberation Army, Taiyuan,

    Shanxi, China.fChair, Department of Orthodontics and Dentofacial Orthopedics, School of

    Dental Medicine, University of Pittsburgh, Pittsburgh, Pa.

    g

    Professor and head, Department of Orthodontics, School of Stomatology, StateKey Laboratory of Military Stomatology, Fourth Military Medical University,

    Xi'an, Shaanxi, China.

    All authors have completed and submitted the ICMJE Form for Disclosure of Po-

    tential Conicts of Interest, and none were reported.

    Supported by a grant from the National Natural Science Foundation of China

    (31200706; http://www.nsfc.gov.cn/publish/portal1).

    Address correspondence to: Yin Ding, State Key Laboratory of Military Stomatol-

    ogy, Department of Orthodontics, School of Stomatology, Fourth Military Med-

    ical University, No. 17, Changle West Road, Xi'an, Shaanxi 710032, China;

    e-mail,[email protected].

    Submitted, July 2014; revised and accepted, February 2015.

    0889-5406/$36.00

    Copyright 2015 by the American Association of Orthodontists.

    http://dx.doi.org/10.1016/j.ajodo.2015.02.018

    725

    ORIGINAL ARTICLE

    http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://www.nsfc.gov.cn/publish/portal1mailto:[email protected]://dx.doi.org/10.1016/j.ajodo.2015.02.018http://dx.doi.org/10.1016/j.ajodo.2015.02.018mailto:[email protected]://www.nsfc.gov.cn/publish/portal1http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-
  • 7/23/2019 2015_147_6_725_737

    2/13

    mini-implants during the early stages of placement is

    achieved by mechanical means rather than through os-seointegration.8 The distribution of mechanical stress

    occurs primarily where bone contacts the implant.Bone density inuences the amount of bone in contactwith the implant surface, and thus the stress can alsobe reduced by increasing the functional area over which

    the force is applied by increasing either the length or thediameter of the implant. The results of previous studieshave suggested that bone of higher density might ensurea better biomechanical environment for mini-implants.7,9Moreover, longer screw-type mini-implantscould be a better choice in a jaw with low bone density.In the comparatively weak cortical bone area, stress is

    known to be distributed to both cancellous and corticalbone, whereas where the cortex is thick and dense, stressis centered on the cortical bone.10 When this is consid-ered with the study of Hedia,11 showing that stress can

    be concentrated at the cortical bone with weak or no

    cancellous bone, the cancellous bone in the maxilla

    might have a greater inuence on success than that inthe mandible. Unicortical anchorage occurs when the

    mini-implant penetrates only 1 cortical plate, whereaswith bicortical anchorage, the mini-implant is longenough to penetrate 2 cortical plates. The in-vitro exper-imental ndings of Brettin et al12 showed that bicortical

    mini-implants provide superior anchorage resistance,reduced overall cortical bone stress, and superior stabil-ity compared with unicortical mini-implants.

    Clinically, computed tomography (CT) is currently theonly diagnostic imaging technique that allows for arough determination of the structure and density of

    bone in the jaws.13,14 It is also an excellent tool for

    assessing the relative distributions of cortical andcancellous bone in an anatomic structure.15,16 Inrecent years, cone-beam CT (CBCT), which offers clear3-dimensional (3D) images with small voxel size, has

    been widely used in orthodontics and implant dentistry

    Fig 1. Measurement of the narrowest interradicular distance in different axial images: A, reorientation

    of the axial images to the occlusal plane; B,measured sites shown in the axial images;C, coronal im-

    age shows the measurement reference planeCEJ planeandthe axial measurement planes (1.5, 3,

    6, and 9 mm); D, narrowest interradicular distances. U, Maxillary; R, right; L, left; interradicular dis-

    tances: 11, central incisors;12, central and lateral incisors; 23, lateral incisor and canine; 34, canineand rst premolar; 45, premolars; 56, second premolar and rst molar; 67, rst and second molars;B, buccal;P, palatal.

    726 Yang et al

    June 2015 Vol 147 Issue 6 American Journal of Orthodontics and Dentofacial Orthopedics

  • 7/23/2019 2015_147_6_725_737

    3/13

    for accurate surgical guidance of mini-implant place-ment17-19 and is a reliable tool to objectively deter-mine site-specic bone density.20

    The objectives of this study were to determine the

    interradicular spaces between the maxillary teeth using3D CBCT data, and to determine the optimal sites, direc-

    tions, and angles for the placement of bicortical mini-implants in orthodontic treatment. Additionally, wesought to evaluate bone density and quality at commonorthodontic implant sites with quantitative measure-

    ments of the simulated placement of mini-implantsin the maxillary interradicular bone, and to propose aprotocol for safe bicortical mini-implant placement.

    MATERIAL AND METHODS

    Our sample consisted of the CBCT data from 50

    adults (22 men, 28 women; ages, 18-39 years; averageage, 25.7 years). All patients met the following criteria:no periodontitis or posterior arch discrepancy; posteriorteeth not rotated or malformed; and no history of ortho-dontic treatment before the collection of the CBCT

    images. These subjects gave written informed consentto publish the resulting data and case details.

    The images were taken with a CBCT apparatus (New-Tom VGi; QR Srl, Verona, Italy) at 110 kV, 0.07 mAs, slice

    thickness of 0.3 mm, and pixel size of 0.3 mm. The CBCTdata were saved as DICOM les. Three-dimensional

    reconstruction procedures and measurements weremade with SimplantPro software (Materialise, Leuven,

    Belgium).To measure the interradicular distances, the axial im-

    ages were reoriented to the occlusal plane (Fig 1,A), andthen sequential axial plane images of 1.5, 3, 6, and 9 mmfrom the cementoenamel junction (CEJ) apical and par-allel to the occlusal plane were constructed (Fig 1,BandC). The CEJ plane, used as the reference, was dened as aplane through the midpoints of the CEJ of 2 adjacentteeth and parallel to the occlusal plane. The narrowest

    interradicular distances between neighboring rootswere measured in each axial plane (Fig 1, D). Becauseof the multiroot nature of the molars, the buccal andpalatal root distances in the molar areas were measuredindividually.

    Fig 2. Measurement of the distances between the sinus oor and the CEJ, the buccal and palatal

    cortical bone thickness, and the alveolar-process bone thickness of each interradicular area: A, slice

    view along the panoramic curve; B, axial view shows the panoramic curve and the measurement of

    the thickness;C, 3D view with cross-section clipping.

    Yang et al 727

    American Journal of Orthodontics and Dentofacial Orthopedics June 2015 Vol 147 Issue 6

  • 7/23/2019 2015_147_6_725_737

    4/13

    To measure the distance between the sinus oor andthe CEJ of each interradicular area, a panoramic curve onthe axial image was drawn, and then the shortest dis-tance was measured on the slice images (Fig 2, A). On

    the sequential axial plane images, the alveolar processbone thickness of each interradicular area was measured

    (Fig 2,B).At each site, the optimal bicortical placement angle

    was measured in relation to the sagittal plane. In thesite mesial to the rst molar, the optimal placement of

    the mini-implant was along a line bisecting the angleof the adjacent roots. In the buccal site between the rstand second molars, the optimal placement of the mini-implant was along the angle bisecting the second mo-lar's mesiobuccal root and that of the rst molar.

    When the maxillary molar was rotated, unicortical place-ment was selected in the palatal site between the rst

    and second molars (U67PinFig 3,B), with the optimalplacement of the mini-implant along the angle bisectingthe rst and second molars' palatal roots. Mini-implants

    with a diameter of 1.5 mm were always used in clinicalpractice and thus were used in the simulation for each

    interradicular site along the optimal angles at thedifferent axial planes (Fig 3, A and D). The angles be-tween the mini-implants and the sagittal plane areshown in the implant properties (Fig 3, C), and the

    mean angles were recorded.Misch and Kircos21 classied bone density into 5

    types based on Hounseld units (HU): D1, morethan 1250 HU; D2, 1250 to 850 HU; D3, 850 to350 HU; D4, 350 to 150 HU; and D5, less than 150

    HU. In our effort to determine bone quality, we dened

    standard bone units using 1-mm-thick bone slicesfrom the CBCT images as follows: standard boneunits5

    mean bone density150200 3bone thickness. Thus, mean

    bone density higher than 350 HU would be rated as1 standard bone unit, making a 1-mm slice of

    bone at a mean density of 350 HU the basis for astandard bone unit of 1. Then a 1-mm slice of D1

    bone with a mean density of 1650 HU may have 7.5standard bone units, a 1-mm slice of D2 bone witha mean density of 1050 HU may have 4.5 standard

    bone units, a 1-mm slice of D3 bone with a meandensity of 600 HU may have 2.25 standard bone units,

    Fig 3. Optimal bicortical placementangle of the mini-implant simulated in differentaxialplanes: A, coronal

    view; B, axial image shows the denition of the optimal angle of each interradicular site; C, implant

    properties show the optimal angle of each mini-implant; D, 3D view of the simulated mini-implant place-

    ment. Maxillary interradicular spaces:67, First and second molars; 56, second premolar and rst molar.

    728 Yang et al

    June 2015 Vol 147 Issue 6 American Journal of Orthodontics and Dentofacial Orthopedics

  • 7/23/2019 2015_147_6_725_737

    5/13

    and a 1-mm slice of D4 bone with a mean density of

    250 HU may have only 0.5 standard bone unit.To evaluate the inuence of bone density and the

    different placement patterns on the stability of mini-implants, the prole lines of 4 placement patterns inmaxillary interradicular bone were used to visualize theintensity of the varying densities along the dened lines

    (Fig 4). Placement patterns 1 and 3 were bicortical place-ments, and placement patterns 2 and 4 were unicortical

    placements. In the prole line list, the Hounseld unitsalong the created prole line are presented. The colorsof the different bone density types are indicated on thegraphs and facilitate the interpretation of the Hounseld

    units. The prole picture was imported into Photoshopsoftware (Adobe, San Jose, Calif) to calculate the totalarea between the prole line and the bottom line ofthe D4 bone (150 HU). The area of 1-mm D4 bonemay be interpreted as 1 standard bone unit (Fig 5, Aand C). Then the total area was divided by the area of1-mm D4 bone to identify the total bone units of the

    prole (Fig 5, B and D). The bone density graph givesthe mean bone density along the prole line (Fig 6).

    In Figure 3, the placement of a 1.5-mm-diametermini-implant along the optimal angle was simulatedfor each interradicular site. The bone width and mean

    bone density along with the implant were measured,

    and the standard bone units of each site were calculated.

    Statistical analysis

    To determine the intraresearcher and interresearcherperformances, the method was applied to 10 randomlyselected subjects; 1 of the 3 researchers (L.Y., F.L., and

    M.C.) repeated the measurements a week later. Themethod had an intraresearcher difference of 0.17 mm,

    with a measurement error of 0.15 mm. The interre-searcher difference was 0.21 mm, with a measurement

    error of 0.17 mm. Both were well within the 0.3-mmrange that is the voxel size of the CBCT data and consid-

    ered acceptable.The data are presented as means and standard devi-

    ations, frequencies, and percentages, as appropriate.SPSS software (version 17.0; SPSS, Chicago, Ill) wasused for all statistical analyses in this study. P\0.05

    was considered to be statistically signicant.

    RESULTS

    The means and standard deviations of the interra-dicular spaces at the different planes are showed in

    Table I and Figure 7. The safe bone thickness around

    Fig 4. Prole lines of 4 placement patterns (PL) in the maxillary interradicular bone.

    Yang et al 729

    American Journal of Orthodontics and Dentofacial Orthopedics June 2015 Vol 147 Issue 6

  • 7/23/2019 2015_147_6_725_737

    6/13

    the mini-implants was dened as 0.5 mm or more, sothe safe maxillary sites beyond 2.5 mm in width werelocated between the rst and second premolars,

    between the second premolar and the rst molar, and

    between the rst and second molars in all planes; be-tween the central incisors, and between the lateral

    incisor and the canine in the plane of 9 mm; buccally,between the rst and second molars in the planes of 6and 9 mm; and between the canine and the rstpremolar in the planes of 3, 6, and 9 mm. The palatal

    aspects between the second premolar and the rstmolar, and between the rst and second molars hadgreater interradicular spaces than did the buccal as-pects. The palatal aspect between the second premolarand the rst molar had the greatest interradicularspaces of any plane. As for the effect of the verticallevel, there were signicantly greater interradicular

    spaces in the apical regions than in the cervical regions,especially for the posterior teeth. In contrast, the buccalspaces between the second premolar and the rstmolar, and between the rst and second molars showedsignicant increases at the 1.5-mm plane compared

    with the 3-mm plane because of the constriction ofthe teeth at the CEJ.

    The alveolar process bone thickness measurementsshowed that overall bone thicknesses were greater in

    the posterior areas. Most bone thicknesses were between8 and 12 mm, except between the rst and second mo-

    lars (15.7 mm) (Table II,Fig 8).Measurements of the mean heights and the ranges

    between the sinus oor and the CEJ are shown inTable III and Figure 9, with all mean heights greater

    than 6 mm. The ratios of heights less than 6 mmwere respectively 4%, 12%, and 14% at sites betweenthe rst and second premolars, between the secondpremolar and the rst molar, and between the rstand second molars; and the ratios of heights lessthan 9 mm were 8%, 64%, and 72%, respectively, atsites between the rst and second premolars, between

    the second premolar and the rst molar, and betweenthe rst and second molars.

    The angulations for optimal bicortical placement atthe different sites are shown inTable IV. The mean an-gulations for the most commonly used sites between

    Fig 5. Calculation of standard bone units of different proles. The standard bone unit (1 u) was dened

    as the area of 1 mm of D4 bone as the threshold in the prole picture. A and C, The standard bone unit

    has differentareas in different prole lines; B and D, total bone units along theprole were calculated by

    the area of 1 mm of D4 bone as the threshold in Photoshop.

    730 Yang et al

    June 2015 Vol 147 Issue 6 American Journal of Orthodontics and Dentofacial Orthopedics

  • 7/23/2019 2015_147_6_725_737

    7/13

    the second premolar and the rst molar, between therst and second molars, and palatally between the rstand second molars were 58, 68, and 59, respectively.

    The standard bone units of the different placement

    patterns are given inTable V. This table shows that allplacement patterns had similar mean bone density

    values but different standard bone units, with bicorticalplacement patterns involving more standard bone unitsthan unicortical placement patterns in the maxilla.

    DISCUSSION

    We undertook this study to identify a reasonable and

    feasible protocol for the placement of mini-implants inthe posterior regions of the maxilla and to offer a remedyfor monocortical mini-implant anchorage failure. Somepatients have thinner buccal cortical bones and lower

    bone density than would normally be expected, espe-cially in the posterior maxilla. Mini-implant placementin these patients is a heightened challenge in relationto implant stability. When monocortical mini-implantsfail during treatment, clinicians can now take advantage

    of bicortical mini-implant anchorage to fulll theirtreatment goals.

    CBCT technology has been used to provide 3Dimages, enabling more detailed 3D visualization

    and quantication of mini-implant status in themaxilla.6,22,23 Although the mini-implant placement

    sites in the maxilla have been studied extensively, asystematic evaluation of maxillary interradicular bonehas not been done.1-6 Clinically, it is important forclinicians to be familiar with the anatomy of mini-

    implant placement sites. In this study, we analyzed themaxillary interradicular bone for bicortical mini-implant placement and the risks of sinus penetration.

    Most mini-implants have a thread diameter from1.2to 2.0 mm and a length from 4.0 to 12.0 mm,24-26

    although some are also available at lengths of 14 to 17mm, or even 21 mm.27,28 Decreased thread diameters

    facilitate placement into sites with small rootproximities and reduce the risk of root contact.

    However, a major concern regarding the threaddiameter of mini-implants is the increased riskof frac-ture noted with diameters less than 1.2 mm.29,30 In

    Fig 6. Calculation of standard bone units of the 4 placement patterns in maxillary interradicular bone.

    The prole lines of the 4 placement patterns in the maxillary interradicular bone are shown. The mini-

    implants with the 1.5-mm dimension and correlated length were simulated to implant along the prole

    lines. Standard bone units5mean bone density150

    200 3bone thickness:

    Yang et al 731

    American Journal of Orthodontics and Dentofacial Orthopedics June 2015 Vol 147 Issue 6

  • 7/23/2019 2015_147_6_725_737

    8/13

    circumstances involving narrow interradicular spaces,we chose 1.5-mm-diameter implants as the standardmini-implants used in our clinical practice.

    In the anterior maxilla, most interradicular distanceswere not sufcient to accommodate a mini-implant.Possible placement sites in the maxillary anterior regionwould be between the central incisors, between the

    lateral incisor and the canine, and between the canineand the rst premolar. Nevertheless, the mini-implantshould be placed farther apically to prevent root dam-age. If intrusion of the anterior teeth is indicated in

    well-aligned deepbite patients, the equiapical or

    subapical portion between the maxillary central incisorswould be adequatefor a mini-implant with a diameterless than 1.5 mm.31

    Inammation of the mini-implant site can be attrib-uted to the mini-implant's position above the mucogin-gival limit and close to the vestibular oor. Miniscrewsplaced in the alveolar mucosa have a greater likelihood

    to trigger inammation, making placement in the ante-rior maxilla a challenge. In such conditions, the clinicianshould allow the mucosa to cover the miniscrew implant,

    with onlya wire or an attachment passing through themucosa.24

    Table I. Mesiodistal interradicular space measurements at planes of 1.5, 3, 6, and 9 mm (mean 6 SD in millimeters)

    Interradicular space 1.5 mm 3 mm 6 mm 9 mm

    Central incisors 1.86 6 0.35 1.77 6 0.50 2.24 6 0.41 3.18 6 0.86

    Central and lateral incisors 1.42 6 0.34 1.49 6 0.43 1.93 6 0.64 2.4 6 0.51

    Lateral incisor and canine 1.77 6 0.34 2.09 6 0.58 2.46 6 0.49 3.18 6 0.38Canine and rst premolar 2.14 6 0.61 2.52 6 0.57 2.54 6 0.51 2.69 6 1.05

    First and second premolars 2.63 6 0.31 2.8 6 0.46 3.03 6 0.37 3.2 6 0.42

    Second premolar and rst molar, buccally 2.82 6 0.49 2.62 6 0.54 3.11 6 0.71 4.01 6 0.68

    Second premolar and rst molar, palatally 4.2 6 0.84 4.9 6 0.81 5.67 6 0.68 6.84 6 0.96

    First and second molars , bucally 2.39 6 0.79 2.29 6 0.74 2.62 6 0.88 3.49 6 1.29

    First and second molars , palatally 3.09 6 0.96 3.36 6 0.92 4.07 6 1.19 4.39 6 1.14

    Fig 7. Means for the mesiodistal interradicular space measurements at 1.5, 3, 6, and 9 mm planes.

    Safe placement areas for 1.5-mm-diameter mini-implants starting from 2.5-mm width (dashed line

    across the figure) can be located between the rst and second premolars (U45), between the second

    premolar and the rst molar (U56), and between the rst and second molars (U67P) palatally at all

    heights; between the central incisors (U11) and between the central and lateral incisors (U12) at a

    height of 9 mm; between the rst and second molars (U67-B) buccally at heights of 6 and 9 mm;

    and between the canine and the second premolar (U34) at heights of 3, 6, and 9 mm. The palatal sides

    of the rst and second molars have greater interradicular spaces than do the buccal sides.

    732 Yang et al

    June 2015 Vol 147 Issue 6 American Journal of Orthodontics and Dentofacial Orthopedics

  • 7/23/2019 2015_147_6_725_737

    9/13

    Mini-implants should be inserted in keratinizedgingiva when possible,25 and the frenulum and muscletissues should be avoided.32,33 According to Kim,34 theaverage width of the buccal attached gingiva is be-tween 3.5 and 5.3 mm. From these studies, the best

    initial point for mini-implant placement should benear the mucogingival line in the attached gingiva. Inour study, we found that almost all the posterior sitesin all planes from the rst premolar to the secondmolar were feasible for safe placement of 1.5-mm

    mini-implants. Therefore, we chose the 3- to 6-mmplane as the standard plane for the routine insertionsite in the posterior alveolar bone.

    Several studies reported that perforation of themaxillary sinus membrane during dental implant

    placement is not a signicant cause of postsurgicalclinical complications.35,36 However, in cases ofsinus involvement by orthodontic mini-implants, ithas been recommended to monitor patients for po-tential development of sinusitis and mucoceles.

    Table II. Alveolar process bone thickness measurements (millimeters)

    Interradicular space 1.5 mm 3 mm 6 mm 9 mm Mean 6 SD

    Central incisors 8.24 8.32 8.53 9.22 8.58 6 0.45

    Central and lateral incisors 8.78 8.85 8.82 9.32 8.94 6 0.25

    Lateral incisor and canine 7.67 9.14 10.63 9.83 9.32 6 1.26Canine and rst premolar 7.78 9.03 10.21 11.34 9.59 6 1.53

    First and second premolars 9.58 10.26 10.56 11.75 10.54 6 0.91

    Second premolar and rst molar 11.28 11.47 12.49 13.41 12.16 6 0.99

    First and second molars 14.93 15.12 15.84 16.93 15.71 6 0.91

    First and second molars , palatally 7.45 8.65 10.23 11.65 9.49 6 1.07

    Fig 8. Alveolar process bone thickness measurements. Most bone thicknesses were between 8 and

    12 mm except for between the rst and second molars (U67).

    Table III. Mean heights and ranges between the CEJ and the sinus oor (mean 6 SD in millimeters)

    Interradicular space Mean 6 SD Range Median \6-mm ratio* \9-mm ratio*

    Canine and rst premolar 16.02 6 3.30 9.68-24.21 15.86 0 0

    First and second premolars 14.72 6 4.51 5.59-23.32 14.40 4% 8%

    Second premolar and rst molar 9.54 6 3.80 2.87-19.30 7.90 12% 64%

    First and second molars 8.42 6 2.40 3.83-19.57 8.22 14% 72%

    *The occurrence rate of penetration of the maxillary sinus oor for bicortical placement at different sites of the 6-mm and 9-mm planes.

    Yang et al 733

    American Journal of Orthodontics and Dentofacial Orthopedics June 2015 Vol 147 Issue 6

    http://-/?-http://-/?-
  • 7/23/2019 2015_147_6_725_737

    10/13

    Maxil lary sinus penetration should be considered apotential risk factor in mini-implant retention.1 As

    we measured in this study, only 12% of the alveolarheights at the sites between the second premolarand the rst molar were less than 6 mm, and only14% of the alveolar heights at the sites between therst and second molars were less than 6 mm. Whenconsidering the interradicular distances, the sites

    with the greatest margins of safety in the maxillafor the 1.5-mm mini-implants were between the rstand second premolars in all planes; between the

    second premolar and the rst molar, and palatally,between the rst and second molars in the 1.5-, 3-,and 6-mm planes; and between the rst and secondmolars in the 6-mm plane.

    For lingual orthodontics, a good site for the place-ment of mini-implants is the palatal alveolar bone

    between the rst and second molars.37 As we discovered,this site has enough space for placement of a mini-implant. Additionally, the palatal mucosa is rm, thick,and resistant to inammation and may have apositiveinuence on the success of mini-implants.38-40

    Fig 9. Measurements of the mean heights and ranges between the sinus

    oor and the CEJ. All meanheights were higher than 6 mm. Maxillary interradicular distances: U34, Canine andrst premolar; U45,

    premolars; U56, second premolar and rst molar;U67, rst and second molars.

    Table IV. Optimal bicortical placement pattern at different sites ()

    Interradicular space Mean 6 SD Range Median

    Central and lateral incisors 17.39 6 4.52 7.06-31.15 17.05

    Lateral incisor and canine 27.16 6 5.46 13.20-41.2 28.26

    Canine and rst premolar 47.18 6 7.24 26.86-63.17 48.80

    First and second premolars 54.37 6 6.40 38.32-69.36 53.93

    Second premolar and rst molar 58.00 6 4.82 45.57-71.34 57.87

    First and second molars 68.19 6 4.43 57.97-83.40 67.52

    First and second molars , palatally 58.76 6 5.17 48.16-74.13 59.45

    Table V. Standard bone units and mean bone density values of 4 placement patterns between the maxillary rstmolar and the second premolar as shown inFigure 4

    Pattern

    Prole linestandard

    bone units (u)

    Simulated implantstandard

    bone units (u)Mean bone

    density (HU)Whole

    length (mm)

    Buccalcortical bonelength (mm)

    Cancellousbone

    length (mm)

    Palatalcortical bonelength (mm)

    1 38.27 38.09 841 11.03 2.15 7.06 1.82

    2 16.07 16.43 721 5.76 1.87 3.89 0

    3 33.49 34.84 787 10.94 1.41 7.54 1.96

    4 32.95 33.23 786 11.08 2.45 8.63 0

    734 Yang et al

    June 2015 Vol 147 Issue 6 American Journal of Orthodontics and Dentofacial Orthopedics

  • 7/23/2019 2015_147_6_725_737

    11/13

    The placement angle between the long axes of mini-

    implants and the sagittal plane of the maxilla is impor-tant for bicortical placement. In this study, we foundthat the sites between the second premolar and therst

    molar, and between the

    rst and second molars palatallyhad enough space for safe placement of mini-implants ifthe placement angles were about 58 to the sagittal

    plane. The depth for bicortical placement at the secondpremolar and therst molar site was about 12 mm, andthe depth for unicortical placement at the palatal site ofthe rst and second molars was about 8 mm.

    Poggio et al23 examined maxillary alveolar widths be-tween the rst and second premolars and found anaverage thickness of 9.9 mm. This value was slightly

    smaller than the average width of 10.54 mm in thisstudy. As we measured, most of the maxillary alveolar

    widths were between 8 and 12 mm, except betweenthe rst and second molars.

    In contrast to dental implants, a mechanical lockrather than bone integration is required to sustain amini-implant throughout orthodontic treatment. Theamount of cortical bone in contact with the mini-implant threads plays an important role in mechanicallocking.41,42 Furthermore, this mechanical lock can be

    either unicortical or bicortical.Bone quality is an important factor affecting the suc-

    cess of dental implants. Bone density is strongly relatedto bone strength, and the compressive strength ofboneis proportional to the square of its density.43,44

    Hounseld units can be used to identify thequantitative properties of tissues. Misch and Kircos21

    classied the bones into 5 types according to density:D1 is found in the anterior mandible, buccal shelf, andmidpalatal region; D2 is found in the anterior maxilla,the midpalatal region, and the posterior mandible; D3

    is found in the posterior maxilla and the mandible;and D4 is found in the tuberosity region.21,45 (D5 bone

    is immature bone.) Bones in regions D1 to D3 areadequate for orthodontic mini-implant placement.

    Mini-implants placed in D1 and D2 bones exhibit lessstress at the screw-bone interface and may provide

    greater stationary anchorage during loading.46

    Factorsaffecting the success of mini-implants are most likelymultifactorial. However, research suggests that thedensity of bone is important for the success of mini-implants.

    Since less dense bone is found in the posteriormaxilla, it has a smaller area of contact with the body

    of the implant. Consequently, a greater implant surfacearea is required to obtain a similar amount of bone-implant contact in soft bone than in denser bone.

    Bone density is directly related to its strength. Misch47

    observed a 10-fold difference in bone strength from

    D1 to D4 bones. D2 bone exhibited 47% to 68% greater

    ultimate compressive strength compared with D3 bone.Because a positive correlation of the preoperative bonedensity quantitatively assessed by CT with torque inten-

    sity during implant placement has been found, bonedensity measured by CT can be used to estimate thelikelihood of primary implant stability.

    Recently, Park et al10 evaluated bone density at or-thodontic implant sites. They measured the density atpoints on the cortical layer and in cancellous bone.

    Because the thickness of the cortical layer varies by jawand area, the data from that study cannot representactual bone densities for mini-implant placement. It is

    believed that measurements of bone density with simu-

    lated placement of orthodontic mini-implants wouldprovide practical information for the placement ofmini-implants. In this study, we used the area of 1 mmof D4 bone as the threshold in the prole picture as a

    standard bone unit to evaluate the effective bone quan-tity. This unit compounded 2 factors relating to bonequantity: bone density and mini-implant length. Thismeasurement was calculated to evaluate the mean

    bone density around the simulated mini-implant. Theresults of these 2 methods had no signicant difference

    statistically.There are many placement patterns for mini-

    implants. Many investigators studied mini-implantangulations in relationto thelong axesof the teeth inunicortical placements.1,3,48,49 Melsen48 recommended

    the placement of mini-implants at an oblique angle inthe maxilla. Kyung et al49 proposed inserting mini-

    implants at a 30 to 40 angle tothe long axes of theteeth in the maxilla. Carano et al3 also suggested anangle of 30 to 45 in the maxilla, but they advisedinserting the mini-implant more perpendicularly near

    the maxillary sinus to prevent any damage to the sinus.We evaluated the standard bone units of different

    mini-implant placement patterns in the maxilla andfound that bicortical placement patterns could havemore standard bone units around the mini-implantsthan any unicortical placement pattern in the maxilla,

    and that mini-implant parallel to the occlusal planehad almost no chance to penetrate the sinus oor,

    was much longer, and would be closer to the center ofresistance of the anterior teeth than the inclined mini-implant (Fig 10). Thisresult is consistent with the nd-ings of Brettin et al.12

    This study could provide valuable information when

    selecting sites and choosing placement methods formini-implants. In a clinical circumstance involving nar-row interradicular spacing, an extended maxillary sinus,or severe alveolar bone loss, an orthodontic miniplateinstead of a mini-implant should be used for skeletal

    Yang et al 735

    American Journal of Orthodontics and Dentofacial Orthopedics June 2015 Vol 147 Issue 6

  • 7/23/2019 2015_147_6_725_737

    12/13

    anchorage.6,22 We are aware of the limitations of

    accurate placement because of individual differencesand the use of average calculated data. However, wehave suggested general guidelines for simple and safeplacement of mini-implants in the maxilla. In clinicalcases involving complex anatomy such as an extendedsinus or alveolar bone loss, the clinician should consider

    using a precise surgical guide with CBCT.

    CONCLUSIONS

    Bicortical placement may include more standard

    bone units around the mini-implants than anyinclined unicortical placement in the maxilla, and

    bicortical placement could be more stable in themaxilla. The most favorable interradicular site in themaxilla for bicortical placement was between the sec-

    ond premolar and the rst molar with an implant of12 mm in length at 58 relative to the sagittal plane.

    Bicortical placement of mini-implants in the anteriorregion is not always viable because of the narrow in-terradicular spaces. For sites between the molars,

    special care should be taken to consider the oor of

    the maxillary sinus when the placement of the mini-implant starts at a plane higher than 6 mm from theCEJ.

    REFERENCES

    1. Deguchi T, NasuM, Murakami K, Yabuuchi T, Kamioka H, Takano-

    Yamamoto T. Quantitative evaluation of cortical bone thickness

    with computed tomographic scanning for orthodontic implants.

    Am J Orthod Dentofacial Orthop 2006;129:e7-12.

    2. Park HS. An anatomical study using CT image for the implantationof micro-implants. Korean J Orthod 2002;32:435-41.

    3. Carano A, Velo S, Incorvati C, Poggio P. Clinical applications of the

    mini-screw-anchorage-system (M.A.S.) in the maxillary alveolar

    bone. Prog Orthod 2004;5:212-35.

    4. Kim HJ, Yun HS, Park HD, Kim DH, Park YC. Soft-tissue and

    cortical-bone thickness at orthodontic implant sites. Am J Orthod

    Dentofacial Orthop 2006;130:177-82.

    5. Chung KR, Kim SH, Kook YA. The C-orthodontic micro-implant. J

    Clin Orthod 2004;38:478-86.

    6. Kim SH, Yoon HG, Choi YS, Hwang EH, Kook YA, Nelson G. Eval-

    uation of interdental space of the maxillary posterior area for or-

    thodontic mini-implants with cone-beam computed

    tomography. Am J Orthod Dentofacial Orthop 2009;135:635-41.

    Fig 10. The relationship between the centers of resistance of the anterior teeth and 2 mini-implants

    with different placement patterns between the second premolar and the rst molar. Mini-implant num-

    ber 1 parallel to the occlusal plane had almost no chance to penetrate the sinus oor, was much longer,

    and would be closer to the center of resistance of the anterior teeth than the inclined mini-implant

    number 2.

    736 Yang et al

    June 2015 Vol 147 Issue 6 American Journal of Orthodontics and Dentofacial Orthopedics

    http://refhub.elsevier.com/S0889-5406(15)00140-7/sref1http://refhub.elsevier.com/S0889-5406(15)00140-7/sref1http://refhub.elsevier.com/S0889-5406(15)00140-7/sref1http://refhub.elsevier.com/S0889-5406(15)00140-7/sref1http://refhub.elsevier.com/S0889-5406(15)00140-7/sref2http://refhub.elsevier.com/S0889-5406(15)00140-7/sref2http://refhub.elsevier.com/S0889-5406(15)00140-7/sref3http://refhub.elsevier.com/S0889-5406(15)00140-7/sref3http://refhub.elsevier.com/S0889-5406(15)00140-7/sref3http://refhub.elsevier.com/S0889-5406(15)00140-7/sref3http://refhub.elsevier.com/S0889-5406(15)00140-7/sref4http://refhub.elsevier.com/S0889-5406(15)00140-7/sref4http://refhub.elsevier.com/S0889-5406(15)00140-7/sref4http://refhub.elsevier.com/S0889-5406(15)00140-7/sref5http://refhub.elsevier.com/S0889-5406(15)00140-7/sref5http://refhub.elsevier.com/S0889-5406(15)00140-7/sref6http://refhub.elsevier.com/S0889-5406(15)00140-7/sref6http://refhub.elsevier.com/S0889-5406(15)00140-7/sref6http://refhub.elsevier.com/S0889-5406(15)00140-7/sref6http://refhub.elsevier.com/S0889-5406(15)00140-7/sref6http://refhub.elsevier.com/S0889-5406(15)00140-7/sref6http://refhub.elsevier.com/S0889-5406(15)00140-7/sref6http://refhub.elsevier.com/S0889-5406(15)00140-7/sref6http://refhub.elsevier.com/S0889-5406(15)00140-7/sref5http://refhub.elsevier.com/S0889-5406(15)00140-7/sref5http://refhub.elsevier.com/S0889-5406(15)00140-7/sref4http://refhub.elsevier.com/S0889-5406(15)00140-7/sref4http://refhub.elsevier.com/S0889-5406(15)00140-7/sref4http://refhub.elsevier.com/S0889-5406(15)00140-7/sref3http://refhub.elsevier.com/S0889-5406(15)00140-7/sref3http://refhub.elsevier.com/S0889-5406(15)00140-7/sref3http://refhub.elsevier.com/S0889-5406(15)00140-7/sref2http://refhub.elsevier.com/S0889-5406(15)00140-7/sref2http://refhub.elsevier.com/S0889-5406(15)00140-7/sref1http://refhub.elsevier.com/S0889-5406(15)00140-7/sref1http://refhub.elsevier.com/S0889-5406(15)00140-7/sref1http://refhub.elsevier.com/S0889-5406(15)00140-7/sref1
  • 7/23/2019 2015_147_6_725_737

    13/13

    7. Chung KR,Nelson G, KimSH, Kook YA.Severe bidentoalveolarpro-

    trusion treated with orthodontic microimplant-dependent en-

    masseretraction.Am J OrthodDentofacialOrthop 2007;32:105-15.

    8. Jafn RA, Berman CL. The excessive loss of Branemark xtures in

    type IV bone: a 5-year analysis. J Periodontol 1991;62:2-4.

    9. Tada S, Stegaroiu R, Kitamura E, Miyakawa O, Kusakari H. In

    u-ence of implant design and bone quality on stress/strain distribu-

    tion in bone around implants: a 3-dimensional nite element

    analysis. Int J Oral Maxillofac Implants 2003;18:357-68.

    10. Park HS, Lee YJ, Jeong SH, Kwon TG. Density of the alveolar and

    basal bones of the ma xilla and the mandible. Am J Orthod Dento-

    facial Orthop 2008;133:30-7.

    11. Hedia HS.Stress andstraindistribution behavior in thebone dueto

    the effect of cancellous bone, dental implant material and the

    bone height. Biomed Mater Eng 2002;12:111-9.

    12. Brettin BT, Grosland NM, Qian F, Southard KA, Stuntz TD,

    Morgan TA, et al. Bicortical vs monocortical orthodontic skeletal

    anchorage. Am J Orthod Dentofacial Orthop 2008;134:625-35.

    13. Watzek G, Ulm C. Compromised alveolar bone quality in edentu-

    lous jaws. In:Zarb G, LekholmU, AlbrektssonT, TenenbaumH, ed-

    itors. Aging, osteoporosis, and dental implants. Chicago:Quintessence; 2002. p. 67-84.

    14. Quirynen M, Mraiwa N, van Steenberghe D, Jacobs R. Morphology

    and dimensions of the mandibular jaw bone in the interforaminal

    region in patients requiring implants in the distal areas. Clin Oral

    Implants Res 2003;14:280-5.

    15. Norton MR, Gamble C. Bone classication: an objective scale of

    bone density using the computerized tomography scan. Clin Oral

    Implants Res 2001;12:79-84.

    16. Shahlaie M, Gantes B, Schulz E, Riggs M, Crigger M. Bone density

    assessments of dental implant sites: 1. Quantitative computed to-

    mography. Int J Oral Maxillofac Implants 2003;18:224-31.

    17. Huang J, Buman A, Mah J. Three-dimensional radiographic anal-

    ysis in orthodontics. J Clin Orthod 2005;39:421-8.

    18. Wagner JD, Baack B, Brown GA, Kelly J. Rapid 3-dimensional pro-

    totyping for surgical repair of maxillofacial fractures: a technical

    note. J Oral Maxillofac Surg 2004;62:898-901.

    19. Hamada Y, Kondoh T, Noguchi K, Iino M, Isono H, Ishii H, et al.

    Application of limited cone beam computed tomography to clin-

    ical assessment of alveolar bone grafting: a preliminary report.

    Cleft Palate Craniofac J 2005;42:128-36.

    20. Gonzalez-Garca R, Monje F. The reliability of cone-beam

    computed tomography to assess bone density at dental implant

    recipient sites: a histomorphometric analysis by micro-CT. Clin

    Oral Implants Res 2013;24:871-9.

    21. Misch CE, Kircos LT. Diagnostic imaging and techniques. In:

    Misch CE, editor. Contemporary implant dentistry. 2nd ed. St

    Louis: Mosby; 1999. p. 73-87.

    22. Chung KR, Kim YS, Lee YJ. The miniplate with tube for skeletal

    anchorage. J Clin Orthod 2002;36:407-12.23. Poggio PM,Incorvati C, Velo S, Carano A. Safe zones: a guide for

    miniscrew positioning in the maxillary and mandibular arch. Angle

    Orthod 2006;76:191-7.

    24. Melsen B, Verna C. Miniscrew implants: the Aarhus anchorage sys-

    tem. Semin Orthod 2005;11:24-31.

    25. Herman R, Cope JB. Miniscrew implants: IMTEC mini ortho im-

    plants. Semin Orthod 2005;11:32-9.

    26. Maino BG, Mura P, Bednar J. Miniscrew implants: the spider screw

    anchorage system. Semin Orthod 2005;11:40-6.

    27. GrayJB, Smith R. Transitionalimplants fororthodontic anchorage.

    J Clin Orthod 2000;34:659-66.

    28. Lin JC, Liou EJ. A new bone screw for orthodontic anchorage. J

    Clin Orthod 2003;37:676-81.

    29. Cattaneo PM, Dalstra M, Melsen B. Analysis of stress and strain

    around orthodontically loaded implants: an animal study. Int J

    Oral Maxillofac Implants 2007;22:213-25.

    30. Lim JW, Kim WS, Kim IK, Son CY, Byun HI. Three dimensional

    nite element method for stress distribution on the length and

    diameter of orthodontic miniscrew and cortical bone thickness.Korean J Orthod 2003;33:11-20.

    31. Kanomi R. Mini-implant for orthodontic anchorage. J Clin Orthod

    1997;31:763-7.

    32. Park Y, Lee SY, Kim DH, Jee SH. Intrusion of posterior teeth using

    miniscrew implants. Am J Orthod Dentofacial Orthop 2003;123:

    690-4.

    33. Paik C, Woo Y, Boyd R. Treatment of an adult patient with vertical

    maxillary excess using miniscrew xation. J Clin Orthod 2003;37:

    423-8.

    34. Kim JS. Study of attached gingival width between healthy gingival

    and early gingivitis patients [thesis]. Seoul, Korea: Yonsei Univer-

    sity; 1997.

    35. Ardekian L, Oved-Peleg E, Mactei EE, Peled M. The clinical

    signicance of sinus membrane perforation during augmenta-

    tion of the maxillary sinus. J Oral Maxillofac Surg 2006;64:277-82.

    36. Kravitz ND, Kusnoto B. Risks and complications of orthodontic

    mini-implants. Am J Orthod Dentofacial Orthop 2007;131

    (4 Suppl):S43-51.

    37. Lee JS, Park HS, Kyung HM. Micro-implant anchorage for lingual

    treatment of a skeletal Class II malocclusion. J Clin Orthod 2001;

    35:643-7.

    38. Kim JH, Joo JY, Park YW, Cha BK, Kim SM. Study of maxillary

    cortical bone thickness for skeletal anchorage system. J Korean

    Oral Maxillofac Surg 2002;28:249-55.

    39. Stipetic J, Hrala Z, Celebic A. Thickness of masticatory mucosa in

    the human hard palate and tuberosity dependent on gender and

    body mass index. Coll Antropol 2005;29:243-7.

    40. Park HS,Jeong SH,Kwon OW.Factors affectingthe clinical success

    of screw implants used as orthodontic anchorage. Am J Orthod

    Dentofacial Orthop 2006;130:18-25.

    41. Ono A, Motoyoshi M, Shimizu N. Cortical bone thickness in the

    buccal posterior region for orthodontic mini-implants. Int J Oral

    Maxillofac Surg 2008;37:334-40.

    42. Monnerat C, Restle L, Mucha JN. Tomographic mapping of

    mandibular interradicular spaces for placement of orthodontic

    mini-implants. Am J Orthod Dentofacial Orthop 2009;135:

    428-9.

    43. Carter DR, Hayes WC. Bone compressive strength: the inuence of

    density and strain rate. Science 1976;194:1174-6.

    44. Rice JC, Cowin SC, Bowman JA. On the dependence of the elastic-

    ity and strength of cancellous bone on apparent density. J Bio-

    mech 1988;21:155-68.

    45. Misch CE. Density of bone: effect on treatment planning, surgicalapproach, and healing. In: Misch CE, editor. Contemporary

    implant dentistry. St Louis: Mosby; 1993. p. 469-85.

    46. Kravitz ND, Kusnoto B, Tsay TP, Hohit WF. The use of temporary

    anchorage devices for molar intrusion. J Am Dent Assoc 2007;138:

    56-64.

    47. Misch CE. Density of bone: effect on treatment plans, surgical

    approach, and healing, and progressive loading. Int J Oral Implan-

    tol 1990;6:23-31.

    48. Melsen B. Mini-implants: where are we? J Clin Orthod 2005;39:

    539-47.

    49. Kyung HM, Park HS,Bae SM, SungJH, Kim IB. Development of or-

    thodontic micro-implants for intraoral anchorage. J Clin Orthod

    2003;37:321-8.

    Yang et al 737

    http://refhub.elsevier.com/S0889-5406(15)00140-7/sref7http://refhub.elsevier.com/S0889-5406(15)00140-7/sref7http://refhub.elsevier.com/S0889-5406(15)00140-7/sref7http://refhub.elsevier.com/S0889-5406(15)00140-7/sref8http://refhub.elsevier.com/S0889-5406(15)00140-7/sref8http://refhub.elsevier.com/S0889-5406(15)00140-7/sref8http://refhub.elsevier.com/S0889-5406(15)00140-7/sref8http://refhub.elsevier.com/S0889-5406(15)00140-7/sref8http://refhub.elsevier.com/S0889-5406(15)00140-7/sref8http://refhub.elsevier.com/S0889-5406(15)00140-7/sref9http://refhub.elsevier.com/S0889-5406(15)00140-7/sref9http://refhub.elsevier.com/S0889-5406(15)00140-7/sref9http://refhub.elsevier.com/S0889-5406(15)00140-7/sref9http://refhub.elsevier.com/S0889-5406(15)00140-7/sref9http://refhub.elsevier.com/S0889-5406(15)00140-7/sref9http://refhub.elsevier.com/S0889-5406(15)00140-7/sref9http://refhub.elsevier.com/S0889-5406(15)00140-7/sref9http://refhub.elsevier.com/S0889-5406(15)00140-7/sref10http://refhub.elsevier.com/S0889-5406(15)00140-7/sref10http://refhub.elsevier.com/S0889-5406(15)00140-7/sref10http://refhub.elsevier.com/S0889-5406(15)00140-7/sref11http://refhub.elsevier.com/S0889-5406(15)00140-7/sref11http://refhub.elsevier.com/S0889-5406(15)00140-7/sref11http://refhub.elsevier.com/S0889-5406(15)00140-7/sref12http://refhub.elsevier.com/S0889-5406(15)00140-7/sref12http://refhub.elsevier.com/S0889-5406(15)00140-7/sref12http://refhub.elsevier.com/S0889-5406(15)00140-7/sref13http://refhub.elsevier.com/S0889-5406(15)00140-7/sref13http://refhub.elsevier.com/S0889-5406(15)00140-7/sref13http://refhub.elsevier.com/S0889-5406(15)00140-7/sref13http://refhub.elsevier.com/S0889-5406(15)00140-7/sref14http://refhub.elsevier.com/S0889-5406(15)00140-7/sref14http://refhub.elsevier.com/S0889-5406(15)00140-7/sref14http://refhub.elsevier.com/S0889-5406(15)00140-7/sref14http://refhub.elsevier.com/S0889-5406(15)00140-7/sref15http://refhub.elsevier.com/S0889-5406(15)00140-7/sref15http://refhub.elsevier.com/S0889-5406(15)00140-7/sref15http://refhub.elsevier.com/S0889-5406(15)00140-7/sref15http://refhub.elsevier.com/S0889-5406(15)00140-7/sref15http://refhub.elsevier.com/S0889-5406(15)00140-7/sref16http://refhub.elsevier.com/S0889-5406(15)00140-7/sref16http://refhub.elsevier.com/S0889-5406(15)00140-7/sref16http://refhub.elsevier.com/S0889-5406(15)00140-7/sref17http://refhub.elsevier.com/S0889-5406(15)00140-7/sref17http://refhub.elsevier.com/S0889-5406(15)00140-7/sref18http://refhub.elsevier.com/S0889-5406(15)00140-7/sref18http://refhub.elsevier.com/S0889-5406(15)00140-7/sref18http://refhub.elsevier.com/S0889-5406(15)00140-7/sref19http://refhub.elsevier.com/S0889-5406(15)00140-7/sref19http://refhub.elsevier.com/S0889-5406(15)00140-7/sref19http://refhub.elsevier.com/S0889-5406(15)00140-7/sref19http://refhub.elsevier.com/S0889-5406(15)00140-7/sref20http://refhub.elsevier.com/S0889-5406(15)00140-7/sref20http://refhub.elsevier.com/S0889-5406(15)00140-7/sref20http://refhub.elsevier.com/S0889-5406(15)00140-7/sref20http://refhub.elsevier.com/S0889-5406(15)00140-7/sref20http://refhub.elsevier.com/S0889-5406(15)00140-7/sref20http://refhub.elsevier.com/S0889-5406(15)00140-7/sref20http://refhub.elsevier.com/S0889-5406(15)00140-7/sref20http://refhub.elsevier.com/S0889-5406(15)00140-7/sref21http://refhub.elsevier.com/S0889-5406(15)00140-7/sref21http://refhub.elsevier.com/S0889-5406(15)00140-7/sref21http://refhub.elsevier.com/S0889-5406(15)00140-7/sref22http://refhub.elsevier.com/S0889-5406(15)00140-7/sref22http://refhub.elsevier.com/S0889-5406(15)00140-7/sref23http://refhub.elsevier.com/S0889-5406(15)00140-7/sref23http://refhub.elsevier.com/S0889-5406(15)00140-7/sref23http://refhub.elsevier.com/S0889-5406(15)00140-7/sref23http://refhub.elsevier.com/S0889-5406(15)00140-7/sref23http://refhub.elsevier.com/S0889-5406(15)00140-7/sref23http://refhub.elsevier.com/S0889-5406(15)00140-7/sref23http://refhub.elsevier.com/S0889-5406(15)00140-7/sref23http://refhub.elsevier.com/S0889-5406(15)00140-7/sref24http://refhub.elsevier.com/S0889-5406(15)00140-7/sref24http://refhub.elsevier.com/S0889-5406(15)00140-7/sref25http://refhub.elsevier.com/S0889-5406(15)00140-7/sref25http://refhub.elsevier.com/S0889-5406(15)00140-7/sref26http://refhub.elsevier.com/S0889-5406(15)00140-7/sref26http://refhub.elsevier.com/S0889-5406(15)00140-7/sref27http://refhub.elsevier.com/S0889-5406(15)00140-7/sref27http://refhub.elsevier.com/S0889-5406(15)00140-7/sref28http://refhub.elsevier.com/S0889-5406(15)00140-7/sref28http://refhub.elsevier.com/S0889-5406(15)00140-7/sref29http://refhub.elsevier.com/S0889-5406(15)00140-7/sref29http://refhub.elsevier.com/S0889-5406(15)00140-7/sref29http://refhub.elsevier.com/S0889-5406(15)00140-7/sref30http://refhub.elsevier.com/S0889-5406(15)00140-7/sref30http://refhub.elsevier.com/S0889-5406(15)00140-7/sref30http://refhub.elsevier.com/S0889-5406(15)00140-7/sref30http://refhub.elsevier.com/S0889-5406(15)00140-7/sref30http://refhub.elsevier.com/S0889-5406(15)00140-7/sref31http://refhub.elsevier.com/S0889-5406(15)00140-7/sref31http://refhub.elsevier.com/S0889-5406(15)00140-7/sref32http://refhub.elsevier.com/S0889-5406(15)00140-7/sref32http://refhub.elsevier.com/S0889-5406(15)00140-7/sref32http://refhub.elsevier.com/S0889-5406(15)00140-7/sref33http://refhub.elsevier.com/S0889-5406(15)00140-7/sref33http://refhub.elsevier.com/S0889-5406(15)00140-7/sref33http://refhub.elsevier.com/S0889-5406(15)00140-7/sref33http://refhub.elsevier.com/S0889-5406(15)00140-7/sref33http://refhub.elsevier.com/S0889-5406(15)00140-7/sref34http://refhub.elsevier.com/S0889-5406(15)00140-7/sref34http://refhub.elsevier.com/S0889-5406(15)00140-7/sref34http://refhub.elsevier.com/S0889-5406(15)00140-7/sref35http://refhub.elsevier.com/S0889-5406(15)00140-7/sref35http://refhub.elsevier.com/S0889-5406(15)00140-7/sref35http://refhub.elsevier.com/S0889-5406(15)00140-7/sref35http://refhub.elsevier.com/S0889-5406(15)00140-7/sref35http://refhub.elsevier.com/S0889-5406(15)00140-7/sref35http://refhub.elsevier.com/S0889-5406(15)00140-7/sref36http://refhub.elsevier.com/S0889-5406(15)00140-7/sref36http://refhub.elsevier.com/S0889-5406(15)00140-7/sref36http://refhub.elsevier.com/S0889-5406(15)00140-7/sref37http://refhub.elsevier.com/S0889-5406(15)00140-7/sref37http://refhub.elsevier.com/S0889-5406(15)00140-7/sref37http://refhub.elsevier.com/S0889-5406(15)00140-7/sref38http://refhub.elsevier.com/S0889-5406(15)00140-7/sref38http://refhub.elsevier.com/S0889-5406(15)00140-7/sref38http://refhub.elsevier.com/S0889-5406(15)00140-7/sref38http://refhub.elsevier.com/S0889-5406(15)00140-7/sref39http://refhub.elsevier.com/S0889-5406(15)00140-7/sref39http://refhub.elsevier.com/S0889-5406(15)00140-7/sref39http://refhub.elsevier.com/S0889-5406(15)00140-7/sref40http://refhub.elsevier.com/S0889-5406(15)00140-7/sref40http://refhub.elsevier.com/S0889-5406(15)00140-7/sref40http://refhub.elsevier.com/S0889-5406(15)00140-7/sref41http://refhub.elsevier.com/S0889-5406(15)00140-7/sref41http://refhub.elsevier.com/S0889-5406(15)00140-7/sref41http://refhub.elsevier.com/S0889-5406(15)00140-7/sref42http://refhub.elsevier.com/S0889-5406(15)00140-7/sref42http://refhub.elsevier.com/S0889-5406(15)00140-7/sref42http://refhub.elsevier.com/S0889-5406(15)00140-7/sref42http://refhub.elsevier.com/S0889-5406(15)00140-7/sref43http://refhub.elsevier.com/S0889-5406(15)00140-7/sref43http://refhub.elsevier.com/S0889-5406(15)00140-7/sref43http://refhub.elsevier.com/S0889-5406(15)00140-7/sref43http://refhub.elsevier.com/S0889-5406(15)00140-7/sref44http://refhub.elsevier.com/S0889-5406(15)00140-7/sref44http://refhub.elsevier.com/S0889-5406(15)00140-7/sref44http://refhub.elsevier.com/S0889-5406(15)00140-7/sref45http://refhub.elsevier.com/S0889-5406(15)00140-7/sref45http://refhub.elsevier.com/S0889-5406(15)00140-7/sref45http://refhub.elsevier.com/S0889-5406(15)00140-7/sref46http://refhub.elsevier.com/S0889-5406(15)00140-7/sref46http://refhub.elsevier.com/S0889-5406(15)00140-7/sref46http://refhub.elsevier.com/S0889-5406(15)00140-7/sref47http://refhub.elsevier.com/S0889-5406(15)00140-7/sref47http://refhub.elsevier.com/S0889-5406(15)00140-7/sref47http://refhub.elsevier.com/S0889-5406(15)00140-7/sref48http://refhub.elsevier.com/S0889-5406(15)00140-7/sref48http://refhub.elsevier.com/S0889-5406(15)00140-7/sref49http://refhub.elsevier.com/S0889-5406(15)00140-7/sref49http://refhub.elsevier.com/S0889-5406(15)00140-7/sref49http://refhub.elsevier.com/S0889-5406(15)00140-7/sref49http://refhub.elsevier.com/S0889-5406(15)00140-7/sref49http://refhub.elsevier.com/S0889-5406(15)00140-7/sref49http://refhub.elsevier.com/S0889-5406(15)00140-7/sref48http://refhub.elsevier.com/S0889-5406(15)00140-7/sref48http://refhub.elsevier.com/S0889-5406(15)00140-7/sref47http://refhub.elsevier.com/S0889-5406(15)00140-7/sref47http://refhub.elsevier.com/S0889-5406(15)00140-7/sref47http://refhub.elsevier.com/S0889-5406(15)00140-7/sref46http://refhub.elsevier.com/S0889-5406(15)00140-7/sref46http://refhub.elsevier.com/S0889-5406(15)00140-7/sref46http://refhub.elsevier.com/S0889-5406(15)00140-7/sref45http://refhub.elsevier.com/S0889-5406(15)00140-7/sref45http://refhub.elsevier.com/S0889-5406(15)00140-7/sref45http://refhub.elsevier.com/S0889-5406(15)00140-7/sref44http://refhub.elsevier.com/S0889-5406(15)00140-7/sref44http://refhub.elsevier.com/S0889-5406(15)00140-7/sref44http://refhub.elsevier.com/S0889-5406(15)00140-7/sref43http://refhub.elsevier.com/S0889-5406(15)00140-7/sref43http://refhub.elsevier.com/S0889-5406(15)00140-7/sref42http://refhub.elsevier.com/S0889-5406(15)00140-7/sref42http://refhub.elsevier.com/S0889-5406(15)00140-7/sref42http://refhub.elsevier.com/S0889-5406(15)00140-7/sref42http://refhub.elsevier.com/S0889-5406(15)00140-7/sref41http://refhub.elsevier.com/S0889-5406(15)00140-7/sref41http://refhub.elsevier.com/S0889-5406(15)00140-7/sref41http://refhub.elsevier.com/S0889-5406(15)00140-7/sref40http://refhub.elsevier.com/S0889-5406(15)00140-7/sref40http://refhub.elsevier.com/S0889-5406(15)00140-7/sref40http://refhub.elsevier.com/S0889-5406(15)00140-7/sref39http://refhub.elsevier.com/S0889-5406(15)00140-7/sref39http://refhub.elsevier.com/S0889-5406(15)00140-7/sref39http://refhub.elsevier.com/S0889-5406(15)00140-7/sref38http://refhub.elsevier.com/S0889-5406(15)00140-7/sref38http://refhub.elsevier.com/S0889-5406(15)00140-7/sref38http://refhub.elsevier.com/S0889-5406(15)00140-7/sref37http://refhub.elsevier.com/S0889-5406(15)00140-7/sref37http://refhub.elsevier.com/S0889-5406(15)00140-7/sref37http://refhub.elsevier.com/S0889-5406(15)00140-7/sref36http://refhub.elsevier.com/S0889-5406(15)00140-7/sref36http://refhub.elsevier.com/S0889-5406(15)00140-7/sref36http://refhub.elsevier.com/S0889-5406(15)00140-7/sref35http://refhub.elsevier.com/S0889-5406(15)00140-7/sref35http://refhub.elsevier.com/S0889-5406(15)00140-7/sref35http://refhub.elsevier.com/S0889-5406(15)00140-7/sref35http://refhub.elsevier.com/S0889-5406(15)00140-7/sref34http://refhub.elsevier.com/S0889-5406(15)00140-7/sref34http://refhub.elsevier.com/S0889-5406(15)00140-7/sref34http://refhub.elsevier.com/S0889-5406(15)00140-7/sref33http://refhub.elsevier.com/S0889-5406(15)00140-7/sref33http://refhub.elsevier.com/S0889-5406(15)00140-7/sref33http://refhub.elsevier.com/S0889-5406(15)00140-7/sref32http://refhub.elsevier.com/S0889-5406(15)00140-7/sref32http://refhub.elsevier.com/S0889-5406(15)00140-7/sref32http://refhub.elsevier.com/S0889-5406(15)00140-7/sref31http://refhub.elsevier.com/S0889-5406(15)00140-7/sref31http://refhub.elsevier.com/S0889-5406(15)00140-7/sref30http://refhub.elsevier.com/S0889-5406(15)00140-7/sref30http://refhub.elsevier.com/S0889-5406(15)00140-7/sref30http://refhub.elsevier.com/S0889-5406(15)00140-7/sref30http://refhub.elsevier.com/S0889-5406(15)00140-7/sref29http://refhub.elsevier.com/S0889-5406(15)00140-7/sref29http://refhub.elsevier.com/S0889-5406(15)00140-7/sref29http://refhub.elsevier.com/S0889-5406(15)00140-7/sref28http://refhub.elsevier.com/S0889-5406(15)00140-7/sref28http://refhub.elsevier.com/S0889-5406(15)00140-7/sref27http://refhub.elsevier.com/S0889-5406(15)00140-7/sref27http://refhub.elsevier.com/S0889-5406(15)00140-7/sref26http://refhub.elsevier.com/S0889-5406(15)00140-7/sref26http://refhub.elsevier.com/S0889-5406(15)00140-7/sref25http://refhub.elsevier.com/S0889-5406(15)00140-7/sref25http://refhub.elsevier.com/S0889-5406(15)00140-7/sref24http://refhub.elsevier.com/S0889-5406(15)00140-7/sref24http://refhub.elsevier.com/S0889-5406(15)00140-7/sref23http://refhub.elsevier.com/S0889-5406(15)00140-7/sref23http://refhub.elsevier.com/S0889-5406(15)00140-7/sref23http://refhub.elsevier.com/S0889-5406(15)00140-7/sref22http://refhub.elsevier.com/S0889-5406(15)00140-7/sref22http://refhub.elsevier.com/S0889-5406(15)00140-7/sref21http://refhub.elsevier.com/S0889-5406(15)00140-7/sref21http://refhub.elsevier.com/S0889-5406(15)00140-7/sref21http://refhub.elsevier.com/S0889-5406(15)00140-7/sref20http://refhub.elsevier.com/S0889-5406(15)00140-7/sref20http://refhub.elsevier.com/S0889-5406(15)00140-7/sref20http://refhub.elsevier.com/S0889-5406(15)00140-7/sref20http://refhub.elsevier.com/S0889-5406(15)00140-7/sref20http://refhub.elsevier.com/S0889-5406(15)00140-7/sref20http://refhub.elsevier.com/S0889-5406(15)00140-7/sref19http://refhub.elsevier.com/S0889-5406(15)00140-7/sref19http://refhub.elsevier.com/S0889-5406(15)00140-7/sref19http://refhub.elsevier.com/S0889-5406(15)00140-7/sref19http://refhub.elsevier.com/S0889-5406(15)00140-7/sref18http://refhub.elsevier.com/S0889-5406(15)00140-7/sref18http://refhub.elsevier.com/S0889-5406(15)00140-7/sref18http://refhub.elsevier.com/S0889-5406(15)00140-7/sref17http://refhub.elsevier.com/S0889-5406(15)00140-7/sref17http://refhub.elsevier.com/S0889-5406(15)00140-7/sref16http://refhub.elsevier.com/S0889-5406(15)00140-7/sref16http://refhub.elsevier.com/S0889-5406(15)00140-7/sref16http://refhub.elsevier.com/S0889-5406(15)00140-7/sref15http://refhub.elsevier.com/S0889-5406(15)00140-7/sref15http://refhub.elsevier.com/S0889-5406(15)00140-7/sref15http://refhub.elsevier.com/S0889-5406(15)00140-7/sref14http://refhub.elsevier.com/S0889-5406(15)00140-7/sref14http://refhub.elsevier.com/S0889-5406(15)00140-7/sref14http://refhub.elsevier.com/S0889-5406(15)00140-7/sref14http://refhub.elsevier.com/S0889-5406(15)00140-7/sref13http://refhub.elsevier.com/S0889-5406(15)00140-7/sref13http://refhub.elsevier.com/S0889-5406(15)00140-7/sref13http://refhub.elsevier.com/S0889-5406(15)00140-7/sref13http://refhub.elsevier.com/S0889-5406(15)00140-7/sref12http://refhub.elsevier.com/S0889-5406(15)00140-7/sref12http://refhub.elsevier.com/S0889-5406(15)00140-7/sref12http://refhub.elsevier.com/S0889-5406(15)00140-7/sref11http://refhub.elsevier.com/S0889-5406(15)00140-7/sref11http://refhub.elsevier.com/S0889-5406(15)00140-7/sref11http://refhub.elsevier.com/S0889-5406(15)00140-7/sref10http://refhub.elsevier.com/S0889-5406(15)00140-7/sref10http://refhub.elsevier.com/S0889-5406(15)00140-7/sref10http://refhub.elsevier.com/S0889-5406(15)00140-7/sref9http://refhub.elsevier.com/S0889-5406(15)00140-7/sref9http://refhub.elsevier.com/S0889-5406(15)00140-7/sref9http://refhub.elsevier.com/S0889-5406(15)00140-7/sref9http://refhub.elsevier.com/S0889-5406(15)00140-7/sref8http://refhub.elsevier.com/S0889-5406(15)00140-7/sref8http://refhub.elsevier.com/S0889-5406(15)00140-7/sref7http://refhub.elsevier.com/S0889-5406(15)00140-7/sref7http://refhub.elsevier.com/S0889-5406(15)00140-7/sref7