The Versatility of Distraction Osteogenesis in Craniofacial Surgery

ORIGINAL ARTICLE The Versatility of Distraction Osteogenesis in Craniofacial Surgery

Mario J. Imola, MD, DDS, FRCSC; David D. Hamlar, MD, DDS; Gentry Thatcher, MD; Khalid Chowdhury, MD, FRCSC

Objectives: To review our preliminary results using dis- Conclusions: Preliminary results demonstrated good- traction osteogenesis for the correction of craniofacial de- to-excellent outcome in correcting facial skeletal formities and to determine its role in treating anatomic deformity in 80% of patients. Functional outcomes deformities and functional deficits relative to conven- included resolution or significant improvement of tional craniofacial surgery. upper airway obstruction in 13 of 14 patients and correction of corneal exposure for all 5 patients with pre- Design and Setting: Retrospective clinical review; ter- existing exorbitism. Correction of malocclusion was tiary care center. less reliable. Problems related to the distraction devices, including failure of the advancement mecha- Methods: Twenty-four consecutive patients were treated nism and fixation, were the most prevalent complica- with distraction osteogenesis during a 34-month pe- tions. Distraction osteogenesis represents an exciting riod. Outcomes were compared with preexisting ana- new development in craniofacial surgery with several tomic deformities and functional deficits using records potential benefits, including less invasive surgery, the of clinical assessments, photodocumentation, diagnos- ability for earlier intervention, and the potential for tic imaging, and treatment planning aids. correction of more severe deformities with improved posttreatment stability. The exact role of distraction Main Outcome Measures: Distraction achieved vs osteogenesis relative to conventional techniques planned distraction based on clinical and radiographic requires ongoing assessment. assessment, clinical status of functional deficits before and after treatment, and objective rating of aesthetic improvement. Arch Facial Plast Surg. 2002;4:8-19

LTHOUGH orthopedic sur- nique to lengthen a canine mandible, that geons have used distrac- DO was first applied to surgical osteoto- tion osteogenesis (DO) for mies of the facial skeleton. Interest in cra- several decades to repair niofacial distraction was slow to grow ini- long bone defects, the pro- tially, with sporadic experimental reports cedure is only now gaining acceptance for appearing throughout the ensuing 2 de- A 7,8 correction of craniofacial deformities. Bone cades. However, in the early 1990s, ex- distraction was first introduced by Cod- perimental investigation intensified fol- villa1 nearly 100 years ago and subse- lowing reports that examined lengthening quently popularized during the 1940s by canine mandibles and the use of DO to suc- Ilizarov,2,3 who developed a single-stage cessfully close canine segmental lower jaw From the Craniofacial–Skull 9-13 14-29 Base Center, Denver, Colo procedure to lengthen long bones with- defects. Thereafter, several studies (Drs Imola and Chowdhury); out the use of grafting material. The fea- demonstrated the ability to apply osteo- the Department of sibility of applying Ilizarov’s principles to distraction at several different sites, in- Otolaryngology–Head and different craniofacial deformities was not cluding the mandible, lower maxilla, mid- Neck Surgery, University of considered until several decades after his face, and cranial vault, within a variety of Colorado, Denver (Drs Imola pioneering work in the peripheral skel- animal models. The first clinical results of and Chowdhury); and eton. Strictly speaking, the first reports of craniofacial DO were reported in 1992 by the Department of craniofacial DO were in the early 1960s, at McCarthy et al30 in a small series of pa- Otolaryngology–Head and which time rapid expansion of the palate tients with congenital mandible deformi- Neck Surgery, Division of Facial Plastic and was carried out in growing subjects; how- ties. Since then, several larger series with Reconstructive Surgery, ever, this involved distraction of a natu- longer follow-up periods have ap- 4,5 30-35 University of Minnesota, rally occurring growth interface. There- peared. More recently, the technique Minneapolis (Drs Imola, fore, it was not until 1973, when Snyder has been successfully used for midface and Hamlar, and Thatcher). et al6 first described the Ilizarov tech- upper craniofacial skeletal defects.36-42

(REPRINTED) ARCH FACIAL PLAST SURG/ VOL 4, JAN-MAR 2002 WWW.ARCHFACIAL.COM 8

©2002 American Medical Association. All rights reserved. Downloaded From: https://jamanetwork.com/ on 09/27/2021 PATIENTS AND METHODS Latency time ranged from 4 to 7 days after placement of the devices, and in general a shorter waiting period was used in the younger patients. Perioperative antibiotics A medical record review was conducted to identify all pa- were used in all cases. During the distraction phase, all tients who had undergone DO within the craniofacial skel- the devices were advanced at the rate of 1 mm per day. eton since we began using this technique. A total of 24 pa- Turnings were performed twice daily (0.5 mm each time) tients who were treated during the 34-month period from if possible and once daily (1.0 mm) if the former was not March 1, 1997 to December 31, 1999, were identified and convenient. Once the distraction was completed, the formed the case series for this study. Each patient under- devices were left in place from 4 to 17 weeks with a mean went a comprehensive preoperative assessment, including of 8.3 weeks. The consolidation time was determined by clinical examination, photodocumentation, and appropri- considering the site of distraction, the magnitude of dis- ate diagnostic imaging. Cephalometric and Panorex x-ray films traction, and the presence of any adverse bone healing were used in 19 patients, and in 15 individuals computed to- factors. In addition, weekly radiographs were obtained to mographic (CT) scans with 3-dimensional reconstruction were semiquantitatively assess mineralization of the regenerate. obtained. Dental study casts were fabricated from impres- Following consolidation, all the devices were removed in sions in 18 patients, and in 6 cases acrylic replicas of perti- a second operative procedure and patients were moni- nent facial skeletal segments were constructed from preop- tored periodically in clinic. erative CT scans. The preoperative data were analyzed to define For each patient, the results were reviewed with con- the nature of the deformity and the objectives and goals of sideration of the following variables: planned vs actual dis- the reconstructive procedure being contemplated. The value traction (magnitude and direction) achieved, improvement of DO was weighed relative to other surgical options before in facial form measured using an objective semiquantita- selecting DO as the reconstructive procedure of choice. Skel- tive scale, and functional improvements. The magnitude of etal tracings were used in most cases to plan the necessary distraction was ascertained from radiographic monitoring osteotomies and desired vector of distraction. and clinical measurement using reliable reference surface The distraction procedures included 6 midface and 29 landmarks such as the dentition wherever applicable. Func- mandibular osteotomies with a total of 40 distraction de- tional evaluation included assessment of sinonasal symp- vices inserted. Several designs are available from the vari- toms (obstruction, sinusitis, sleep disturbances), ophthal- ous suppliers, and the devices used in this series are shown mologic symptoms (corneal exposure, conjunctivitis, lacrimal in Figure 3. Distraction devices can be broadly grouped drainage, visual disturbances), and masticatory function (oc- into internal or external types based on the location of the clusion, mandibular movement). The nature of any compli- main body of the distraction mechanism. Different de- cations and their outcome were also noted. Data were then signs are suited to different applications, and careful con- analyzed to formulate relevant observations and conclu- sideration is required to ensure that the appropriate de- sions regarding the clinical utility of craniofacial DO based vice is selected for a given clinical situation. on our experience.

The underlying principle of DO, as described by Iliz- that underlie DO have been well documented in long bones arov,is“themechanicalinductionofnewbonebetweenbony and more recently in the craniofacial skeleton.2,3,10,43-48 Dur- surfaces that are gradually distracted.”2,3 The process of DO ing the distraction phase, bone formation occurs in re- begins with careful preoperative assessment and planning, sponse to the tension-stress forces exerted on the regen- which are critical to success. During the initial surgery, os- erate, and healing proceeds primarily by a reparative teotomies are performed and the distraction device is in- membranous ossification process (Figure 2).10,43-45 The serted. A waiting period (latency phase) is allowed to elapse middle of the regenerate consists of a fibrous central zone so that osseous healing is initiated at the bony gap, perios- where osteoid is deposited with collagen fibers oriented teal integrity is restored, and callus formation begins.43 The parallel to the direction of distraction. Ossification oc- bone segments at either end of the gap are then progressively curs as a primary mineralization front advances from ei- distractedoverseveraldays(distractionphase)duringwhich ther end of the fibrous central zone, resulting in a bridge osteogenesis is induced, producing a so-called regenerate of immature bone across the distraction gap. Bone remod- of immature bone laid down between the cut bone ends. eling begins during the consolidation phase and contin- With time, the bone remodels into a more mature state (con- ues throughout a 1- to 2-year period, eventually trans- solidation phase), and the surrounding soft tissues accom- forming the regenerate into a mature osseous structure modate to their new positions and lengths.43 similar in size and shape to the adjacent preexisting bone. Distraction can be divided into either elongation DO Although the volume and architecture of the new bone are or transport DO, depending on whether a foreshortened comparable to the adjacent bones, animal studies have bone is being increased in length (elongation) or a seg- shown that mineral content and radiodensity are less. The mental defect is being repaired within a bone that is oth- tensile strength of the regenerated segment is approxi- erwise of normal length (transport). In addition, DO has mately 75% of the native bone.13 In addition to bony been categorized into monofocal, bifocal, and trifocal types, changes, there are effects on the adjacent soft tissues that depending on the number of foci at which osteogenesis occur in response to osseous distraction.46 Muscle and soft occurs (Figure 1). Monofocal elongation DO currently tissue mass increases via a process referred to as distrac- represents most of the clinical applications in the cranio- tion histogenesis. Clinically, this offers a distinct advan- facial skeleton. The histologic and physiologic principles tage since several craniofacial anomalies have soft tissue

(REPRINTED) ARCH FACIAL PLAST SURG/ VOL 4, JAN-MAR 2002 WWW.ARCHFACIAL.COM 9

4 321

Figure 2. Schematic representation of the stages of bone formation during distraction osteogenesis: (1) zone of fibrous tissue, (2) zone of bone mineralization, (3) zone of bone remodeling, and (4) zone of mature bone. Adapted from Karp et al.10(p7)

Figure 1. Three types of distraction osteogenesis have been described: monofocal, bifocal, and trifocal. A, Monofocal distraction is used to lengthen successful osteogenesis, and therefore many advocated that abnormally shortened bones and involves separation of 2 bone segments a corticotomy-only bone cut be performed through a mini- across a single osteotomy. B, Bifocal distraction is used to repair a segmental mal periosteal opening.2,3 More recently, however, inves- defect and requires creation of a transport disk, which is then distracted across the defect until it docks with the opposing bony segment. C, Trifocal tigators have demonstrated that the periosteum alone can distraction is similar to bifocal distraction attempts to halve the distraction time provide sufficient osteogenic capacity to form a viable re- by transporting 2 disks from opposite ends of a defect to dock in the middle. generate in the well-vascularized membranous bone of the Arrows indicate distraction vectors; large arrow heads, distraction regenerate; craniofacial skeleton.9,11,18,21,22,26,27,43 Therefore, although and small arrow heads, docking site. Reprinted from Costantino PD, Shybut G, Friedman CD, et al.11(p543) some clinicians continue to advocate corticotomy, most reports31,35,43 of craniofacial DO describe the use of a complete osteotomy, taking care to preserve as much of the hypoplasia in addition to deficient bony structures. Neu- surrounding periosteum as possible. Prior radiation therapy rovascular elements contained within distracted bony seg- to the distraction site has been shown to not adversely in- ments are also stimulated to elongate. Experimental stud- fluence the results of distraction in the canine model, and ies in dogs have demonstrated a recanalized mandibular 47 when using DO to repair segmental defects, the status of canal that contains both neural and vascular elements. the surrounding soft tissues will likely be the key factor The functional level of the regenerated neurovascular struc- that influences outcome.48 tures, however, was less than normal. Latency, rate, and rhythm of distraction are vari- Several factors influence the physiologic process of ables that influence the quality of the regenerate. Of these DO, and these can be separated into 2 basic groups: bone factors, the effect of latency is the most controver- healing factors and distraction factors as outlined here: sial.37,39,43 Most craniofacial surgeons have empirically ap- Bone-Healing Factors plied the conclusions from long bone studies and recom- Local Systemic Distraction Factors mend waiting periods of 4 to 7 days following osteotomy Osteoprogenitor supply Age Rate of distraction and before initiating the distraction process. In younger Blood supply Metabolic disorders Frequency of distraction children, the high rate of bone metabolism favors a shorter Infection Vitamin D deficiency Latency period waiting period. Some clinicians, however, use a zero la- Soft tissue scarring Connective tissue disease Rigidity of fixation tency period and begin distracting right at the time of ap- Bone stock Steroid therapy Adequate consolidation pliance insertion. They claim no adverse effects on out- Prior radiation therapy Calcium deficiency period come while substantially shortening the treatment Length of regenerate period.37,39 Waiting too long before distraction (beyond 10 Factors that affect bone healing can be local or systemic to 14 days) substantially increases the risk of premature in nature. Viability of osteocytes and osteoblasts is essen- bone union. In contrast to latency, the rate and fre- tial to provide an adequate source of osteogenic activity quency (rhythm) of distraction are believed to be impor- at the distraction site. Hence, careful surgical technique tant factors.43 If widening of the osteotomy site occurs too should be used to minimize thermal or mechanical in- rapidly (Ͼ2 mm per day), then a fibrous nonunion will jury to the periosteum and endosteum, which are the main result, whereas if the rate is too slow (Ͻ0.5 mm per day), sources of osteoblast precursors. Similarly, an adequate premature bony union prevents lengthening to the de- blood supply to the distraction site is critical to osteogen- sired dimension. These findings in long bones have been esis. Arterial insufficiency may lead to ischemic fibrogen- empirically applied to the craniofacial skeleton, and most esis within the regenerate, yielding a loose, irregular col- studies have described a rate of 1.0 mm per day. Accord- lagen network instead of the desirable dense, regular ing to Ilizarov’s work in long bones, the ideal rhythm of collagen pattern. Venous outflow obstruction has been as- DO is a continuous steady-state separation of the bone frag- sociated with cystic degeneration of the regenerate. The ments.2,3,44,45 However, this is impractical from a clinical clinician, therefore, needs to ensure that the soft tissues standpoint, and therefore, most reports30-33,35,38 recom- that surround the site of the proposed distraction are well mend distraction frequencies of 1 or 2 times daily. The vascularized. Early studies in long bones concluded that length of the consolidation phase has been recom- both an intact periosteum and endosteum were critical to mended as ranging from 6 to 12 weeks in long bones, de-

(REPRINTED) ARCH FACIAL PLAST SURG/ VOL 4, JAN-MAR 2002 WWW.ARCHFACIAL.COM 10

1 in

C D

Titanium Head Cap Screw E

Carbon Rod, 120 mm

Hexagonal Nuts for Central Shaft

Upper Part, Aluminum Distraction Screws Holder for Horizontal Cross-Bar Splint Traction Hooks With Eyelet

Figure 3. Distraction appliances can be divided into internal and external types. Internal appliances have the main body of the distraction mechanism submerged below the skin or mucosa and are fixated to bone using conventional plate and screw metal hardware. A connecting rod passes from the distraction mechanism and exits through the skin cover or mucosa to allow activation. Internal appliances used in this study include the Stryber-Leibinger (Kalamazoo, Mich) intraoral device (A), KLS Martin (Jacksonville, Fla) modular (B), and Stryber-Leibinger modular internal distractor (C). External devices are positioned outside the body and are secured to bony segments using transcutaneous pin fixation. External appliances used in this study include the Leibinger multiguide (D) and the KLS rigid external distractor (E).

pending on the length of the distraction segment. In the role and future place in the management of the difficult chal- craniofacial skeleton, most authors advocate 4 to 8 lenges that craniofacial patients typically present. weeks, with the general rule that the consolidation period should be at least twice the duration of the distrac- RESULTS tion phase.30-33,35,38 Distraction in load-bearing bones, such as the mandible, is an indication for a longer con- Congenital deformities comprised most of the cases (22), solidation time. Finally, appliance rigidity during distrac- whereas 2 patients had acquired defects. Table 1 gives tion and consolidation is a critical element to ensure that the various underlying diagnoses present in the overall bending or shearing forces do not result in microfrac- group. The youngest group consisted of 8 patients who tures of the immature columns of new bone within the re- presented with glossoptosis-micrognathia and upper air- generate, which lead to focal hemorrhage and cartilage way obstruction as the main indication for distraction. interposition.43 These patients had a mean age of 32 months (range, 6 We have been using DO since early 1997 for several days to 5 years) and consisted of 6 individuals with iso- indications within the craniofacial skeleton, including the lated glossoptosis-micrognathia (3 with Pierre Robin se- correction of form and function. This study was under- quence), 1 patient with Nager syndrome, and 1 patient taken to review our experience with DO and evaluate its with maxillomandibular syngnathia. Craniofacial dys-

(REPRINTED) ARCH FACIAL PLAST SURG/ VOL 4, JAN-MAR 2002 WWW.ARCHFACIAL.COM 11

©2002 American Medical Association. All rights reserved. Downloaded From: https://jamanetwork.com/ on 09/27/2021 Table 1. Number, Age, and Diagnoses Table 2. Distraction Procedures Performed of Patients Included in the Study Group and Lengths of Distraction

No. of Mean Age No. of Distractions Diagnosis Patients (Range) Mean Length Achieved of Planned Congenital anomalies No. of Distraction, 0- to 2-mm Ͼ2-mm Glossoptosis-micrognathia 6 27 mo (6 d-5 y) Procedure Patients mm (Range) Deficit Deficit Crouzon, Apert, or Pfeiffer syndrome 5 5 y (4-7 y) Hemifacial microsomia 6 7.6 y (5-11 y) Le Fort III 5 23.7 (22-26) 5 . . . Other Le Fort I 1 22 . . . 1 . . . Bilateral lengthening 10 20.5 (12-30) 7 3 Treacher Collins syndrome 1 12 y Unilateral lengthening 6 19.2 (16-26) 4 2 Hallermann-Streiff syndrome 1 31 y Segmental repair 2 22 (2-24) 1 1 Maxillomandibular sygnathia 1 4.5 y Nager 1 4.0 y Cleft lip or palate 1 14 y Acquired deformities Segmental mandible defect (oncologic 152y Table 3. Cosmetic Results in 20 Patients resection) Bilateral condylectomies 156y Deformity (temporomandibular joint ankylosis) Unilateral Bilateral Midface Mandibular Mandibular Result Deficiency Deficiency Deficiency Total ostosis patients (3 with Crouzon syndrome, 1 with Ap- Excellent 4 2 3 9 ert syndrome, and 1 with Pfeiffer syndrome) who re- Good 2 3 2 7 quired midfacial advancement had a mean age of 5 years, Fair 0 1 3 4 whereas 6 patients with oroauriculovertebral spectrum Total 6 6 8 20 (hemifacial microsomia) underwent distraction at a mean age of 7.6 years. Follow-up in the overall group ranged from 6 to 36 months, with a mean of 14 months follow- primarily for airway obstruction included 2 infants with ing the end of the consolidation phase. glossoptosis-micrognathia who required urgent endotra- The cases were separated into midface and man- cheal intubation after birth for whom consultations were dible procedures (Table 2), and the actual length of dis- requested for tracheostomy to manage the airway. Both of traction achieved was compared with the preoperative these patients underwent external bilateral mandibular body desired length. In 18 patients, the devices were turned distraction and eventually were successfully extubated, the planned number of revolutions, and in 14 of these thereby avoiding tracheostomy. The remaining 6 patients the desired length of distraction was fully achieved, had preexisting tracheotomies: 4 of these patients under- whereas in the other 4 the distracted length was within went decannulation, 1 was pending tube removal follow- 2 mm of the desired length based on radiologic following unrelated surgery, and 1 has failed attempts at decan- up. In 6 other patients, the devices could not be ad- nulation. All 5 patients who underwent Le Fort III vanced the desired number of revolutions because of me- advancement demonstrated some degree of preexisting oro- chanical problems with the distraction devices (4), pharyngeal or nasopharyngeal obstruction and disturbed compliance problems (1), and soft tissue breakdown with sleep patterns. Formal sleep studies were available for 3 of device exposure at the osteotomy site (1). these patients before and after advancement, demonstrat- Cosmetic results were rated objectively by the treating resolution of moderate sleep apnea based on the res- ing surgeon using an arbitrary scale of excellent, good, fair, piratory index. One patient with Hallerman-Streiff syn- or unsatisfactory. The results were tallied for 20 patients drome noted subjective clinical improvement of associated for whom aesthetic considerations were applicable, and the airway obstruction. Exorbitism and varying degrees of prob- results were good to excellent in 80% of the group lems with corneal exposure were present in all 5 patients (Table 3). Midfacial advancements achieved the best cos- with craniofacial dysostosis. Midface advancement was able metic outcome (Figure 4). Most hemifacial microsomia to sufficiently improve corneal protection and decrease the patients were rated as good; however, within this group it conjunctival inflammation in these patients. was noted that lower face asymmetry was difficult to fully Malocclusion and related masticatory disturbances correct. Cosmetic results were less predictable in 8 pa- were the most prevalent functional problem in our group tients, with symmetric mandibular deficiency ranging from of patients, presenting in 16 of 24 patients overall. Unfor- fair to excellent (Figure 5 and Figure 6). tunately, they were also the most refractory to correction. Functional deformities fell into 3 basic groups: air- In the 6 patients with maxillary deficiency, anteroposte- way obstruction, exorbitism (corneal exposure), and mal- rior malrelationships ranged from moderate to severe class occlusion. Airway obstruction was identified in 14 pa- III skeletal disharmony. These patients underwent midfa- tients as follows: 8 individuals with obstruction as the cial advancement and 3 were corrected to class 1, whereas main indication for distraction (6 with glossoptosis- the remaining 3, although substantially improved, were left micrognathia, 1 with Nager syndrome, 1 with maxillo- with some degree of class III relationship (Figure 4). Hemi- mandibular syngnathia) and 6 patients in whom obstruc- facial microsomia patients (6) presented with mild to mod- tion was an associated finding (5 with midfacial deficiency, erate degrees of ipsilateral class II molar relationships, mid- 1 with Hallermann-Streiff syndrome). The group treated line shift, and occlusal cant. Although the anteroposterior

(REPRINTED) ARCH FACIAL PLAST SURG/ VOL 4, JAN-MAR 2002 WWW.ARCHFACIAL.COM 12

DO Vector S-N 62 mm 20 12 Po-Or 52 mm 23 Ba-S-N 124° SN-MxP 4° N-ANS 37 mm ANS-Me 63 mm SNA 75° ANB –4°

D E F

Figure 4. A 5-year-old boy with Crouzon syndrome who had undergone prior bifrontal-orbital advancement and cranial vault remodeling during infancy. A, Preoperative lateral cephalogram demonstrating severe horizontal and vertical midfacial deficiency. Functional deficits included exorbitism with chronic conjunctival exposure, oronasal airway obstruction with obstructive sleep apnea, and recurrent rhinosinusitis. B, Corresponding cephalometric tracing illustrating proposed vector of distraction. We planned for 20 mm of anterior displacement of orbitale (Or) and 12 mm of inferior midface repositioning. Addition of the 2 component vectors yielded the resultant distraction vector with a magnitude of 23 mm and direction as illustrated (green triangle). S indicates sella; N, nasion; Po, porion; Ba, basion; SN, sella-nasion plane; MxP, maxillary plane; ANS, anterior nasal spine; and Me, menton. C, Preoperative articulated dental models showing class III molar relationship and severe negative overjet. D, A Le Fort III osteotomy was performed and bilateral Stryber-Leibinger modular internal distractors were secured along the planned line of distraction. E, Photograph of the patient during the distraction phase with activation rods exiting the temporal region. F, Lateral cephalogram following distraction and consolidation phases. Comparison to preoperative cephalogram reveals marked improvement in midfacial proportions. G and H, (next page) Preoperative frontal and lateral photographs. I and J, Postoperative frontal and lateral views 16 months following distraction demonstrating excellent cosmetic outcome with correction of exorbitism and a normalized facial profile. K, Before and L, after base views showing improved midfacial projection. M, Pretreatment and N, posttreatment intraoral photographs illustrating occlusal correction into a class I relationship and interval exfoliation of the maxillary anterior deciduous incisors.

relationships underwent favorable changes, 5 of these pa- 4 patients, the desired direction or magnitude of distrac- tients developed posterior open bites on the distracted side, tion was not attained at the completion of the distraction which did not fully close with postoperative guided erup- phase.In2otherpatients,patientnoncompliancewasclearly tion of the posterior maxillary teeth. Furthermore, shifts identified, also resulting in significantly less advancement in the mandibular midline to the contralateral side were than planned for. Late complications were primarily related noted in 4 patients. Distraction of the shortened hemiman- to adverse soft tissue effects, including linear scarring at the dible to the desired length was complicated by deviation transcutaneous pin sites with external devices (4), tempo- of the mandibular midline to the opposite side and devel- ral wasting (2), transient trismus (2), and infection (2). The opment of crossbites in 4 of the 6 patients. Of the various patient with the oncologic segmental mandibular defect de- preexisting occlusal deformities, open bite and transverse veloped wound breakdown, infection, and hardware expo- crossbites were the most refractory to correction. In 6 pa- sure, leading to abandonment of the distraction procedure. tients with anterior open bites, 1 experienced closure and Notably, there were no cases of fibrous nonunion across the 1 was improved, whereas the remaining 4 were no better distraction sites, and at the time of internal device removal, at the end of treatment. Posterior crossbite, present in 8 healthy bone regenerate was verified (Figure 7). patients before DO, was corrected in only 3. A variety of complications manifested during the dis- COMMENT traction phase of the process. Problems with the distraction device occurred in 4 patients. In 2 cases, the turning mecha- Our results demonstrate that DO is a useful technique and nism failed and revision surgery with device replacement can be a powerful tool for the correction of various struc- was undertaken (Figure 6). In the other 2 patients, the fixa- tural deformities and functional deficits throughout the cra- tion became unstable during the distraction phase and sur- niofacial skeleton. The early outcomes described herein sup- gical reinsertion was also performed. Notably, in 3 of these port ongoing evaluation to assess the long-term efficacy of

(REPRINTED) ARCH FACIAL PLAST SURG/ VOL 4, JAN-MAR 2002 WWW.ARCHFACIAL.COM 13

I J

K L M N

Figure 4.

(REPRINTED) ARCH FACIAL PLAST SURG/ VOL 4, JAN-MAR 2002 WWW.ARCHFACIAL.COM 14

C D

E F

Figure 5. A 5-year-old girl with primary mandibular-maxillary syngnathia and associated glossoptosis-micrognathia requiring a tracheostomy since infancy. A, Preoperative acrylic model used for treatment planning demonstrates a markedly micrognathic and retrognathic mandible. B, Intraoperative photograph showing bilateral posterior body osteotomies and insertion of internal Stryber-Leibinger (Kalamazoo, Mich) intraoral devices. Panorex x-ray films: C, before and D, after 21 mm of horizontal advancement. Profile views: E, before and F, 12 months after treatment showing good improvement in lower face profile.

(REPRINTED) ARCH FACIAL PLAST SURG/ VOL 4, JAN-MAR 2002 WWW.ARCHFACIAL.COM 15

C D

Figure 6. A 31-year-old woman with Hallerman-Streiff syndrome had residual micrognathia and severe malocclusion following a previous attempt at correction with bilateral sagittal split osteotomies and conventional mandibular advancement. A, Lateral 3-dimensional computed tomographic scan demonstrating mandibular hypoplasia and marked displacement of the condyles anteriorly out of the hypoplastic glenoid fossae. Treatment consisted of posterior body osteotomies and distraction using internal distraction appliances. Planned bilateral advancement was 30 mm. The distraction vector was horizontal in orientation. Midway during the distraction phase, one of the devices was noted to have failed with lack of advancement when activated. B, Despite replacement of the device, only 24 mm of distraction was achieved because of premature consolidation at the osteotomy sites. Profile views: C, before and D, after show a fair cosmetic outcome 18 months following surgery.

craniofacial DO and its role relative to conventional cra- vancements were achieved within 2 mm based on clinical niofacial techniques. The extent of bony advancement that or radiographic follow-up. Factors identified as contribut- we achieved has been retained throughout the follow-up ing to a shortfall in the distraction length included device period in all patients, which would suggest an increased failure and patient noncompliance. resistance to relapse, a well-documented problem with con- Careful selection and planning are critical to a success- ventional craniofacial advancements. Furthermore, in pa- ful outcome.49-53 Initially, all of the necessary preoperative tients with severe skeletal deficiencies or soft tissue scar- data are acquired and analyzed to precisely define the ex- ring, the advancements planned and achieved using DO isting deformities. The various subunits of the craniofacial were well beyond what could have been hoped for using skeleton that need mobilization and repositioning can then conventional surgery. In most of our cases, the desired ad- be identified to determine the nature of the desired oste-

(REPRINTED) ARCH FACIAL PLAST SURG/ VOL 4, JAN-MAR 2002 WWW.ARCHFACIAL.COM 16

©2002 American Medical Association. All rights reserved. Downloaded From: https://jamanetwork.com/ on 09/27/2021 otomies and the vector (direction and magnitude) of distraction (Figure 8). Preoperative considerations for mandibular distraction include the preexisting shape of the lower jaw (ramus height, body length, gonial angle), mandibular size (mandibular length), position of the lower jaw relative to the maxilla and cranial base, presence of asymmetry, and mandibular plane angles. Midface considerations during planning include the degree of exorbitism, the presence of anteroposterior deficiency within the midfacial bony structures, midface position relative to the cranial base and themandible,verticalmidfacedeficiencyorelongation,maxillary plane, and occlusal plane angles. Whenever the desired movements will affect occlusion, a complete occlusal analysis needs to be undertaken, including anteroposterior occlusal relationships, transverse relationships, the presence of open bite deformities, occlusal plane angles, and incisor inclinations. The planning phase requires coopera- tion between the surgeon and orthodontist if the skeletal movements will affect the dentition. Once the osteotomy design and vector of displacement have been determined, selection of the appropriate mode of distraction and the specific appliances to be used requires careful evaluation.54-58 Internalappliancesaremoresuitedforuncomplicatedmove- ments where displacement in only one direction is required. Once internal appliances are in situ, modification of the distraction vector is not possible. External devices have the ad- vantage of allowing multidirectional distraction in several Figure 7. Intraoperative view at the time of removal of internal distraction devices following the consolidation phase in a young patient who underwent different planes of movement. They provide a greater de- Le Fort III distraction. Note the quality of the maturing bone regenerate in the gree of freedom in selecting the desired vector of distrac- distraction gap (instrument tip). tion (both angular and linear movements are possible) and

A Go-B 78 mm B S-N 77 mm Ar-Go 52 mm Po-Or 73 mm Ar-B 108 mm Ba-S-N 130° SN-MdP 33° SN-MxP 8.5° Ar-Go-Me 122° N-ANS 58 mm SE N N-Me 127 mm S SE-PNS 51 mm SNA 82 mm Ar ANB 2° Po Or Ba ANS MxP PNS A V O

H MP B B MP Go

Me Me

Figure 8. Careful preoperative planning to determine the vector of distraction is a critical element for success. A, Distraction of the lower jaw can be horizontal, vertical, or oblique, depending on the nature of the preexisting deformity. Assessment should consider the height of the posterior ramus (Ar-Go), the length of the mandibular body (Go-B), and the overall length of the mandible (Ar-B). A horizontal vector (H) is used to primarily correct deficient length of the mandibular body. A vertical vector (V) is selected to increase an abnormally short posterior ramus. Oblique vectors (O) are useful to increase both ramal height and body length. Note that cases with an abnormal gonial angle (Ar-Go-Me) are best managed with an external device capable of angular distraction. Go indicates gonion;B,“B” point; Ar, articulare; SN, sella-nasion plane; MP, mandibular plane; and Me, menton. B, Planning for midface distraction should consider the linear and angular measurements that describe the size and relationship of the orbitomaxillary complex to the cranial base. Key measurements are illustrated and the accompanying numeric values are approximate norms for adults. Each patient should be assessed relative to age-specific norms, keeping in mind, however, that inherent growth deficiencies will persist despite distraction osteogenesis. It is therefore useful to overcorrect close to adult values when treating patients who have completed most of their midfacial growth. S indicates sella; N, nasion; Po, porion; Or, orbitale; Ba, basion; MxP, maxillary plane; ANS, anterior nasal spine; Me, menton; SE, sphenoethmoid intersect; PNS, posterior nasal spine; and A, “A” point.

(REPRINTED) ARCH FACIAL PLAST SURG/ VOL 4, JAN-MAR 2002 WWW.ARCHFACIAL.COM 17

©2002 American Medical Association. All rights reserved. Downloaded From: https://jamanetwork.com/ on 09/27/2021 the ability to change the vector of incremental advancements The timing of intervention has always been a contested during the distraction phase as necessary. This is advanta- issue in craniofacial surgery. During the past several years, geous in controlling occlusal relationships. Furthermore, improved expertise and biotechnology have allowed bet- external appliances are typically retained via transcutane- ter results during manipulation of immature bone using con- ous threaded pins, which facilitates use in very young pa- ventional techniques. Nonetheless, considerable drawbacks tients whose facial skeleton size may preclude using the plate exist when attempting to reposition and internally fixate andscrewfixationnecessarytosecureinternaldevices.Lastly, bone segments in young patients. Uncontrolled and exces- external devices are easily removed following consolida- sive fragmentation during manipulation of fragile bone seg- tion in contrast to internal devices, which often require a ments,inadequatebonestockinwhichtosecuremetalplates major surgical procedure to expose and retrieve. Disadvan- and screws, and injury to developing tooth buds are all well- tages with external devices are unavoidable scars as a re- recognized risks in this setting. Distraction osteogenesis rep- sult of the transcutaneous fixation pins. These produce at resents a major advancement in addressing this problem times a severe linear scar that grows as the pins are splayed owing to the lack of intraoperative repositioning and the during the distraction process. Also, external devices can minimal fixation requirements. be bulky, and having the appliances on for 8 to 12 weeks The less invasive nature of DO permits advancements can be a drawback for some patients. in earlier age groups than conventional surgery whenever In our series, the best aesthetic results were ob- such intervention would be appropriate. In our group, the tained in the midfacial advancement cases and to a lesser distraction procedures were carried out well ahead of the degree in patients with hemifacial microsomia. Cases with age groups considered typical when using conventional bilateral micrognathia did not achieve as favorable a cos- surgical repositioning. The results yielded substantial im- metic outcome. This may have been contributed to by provements in form and function at an earlier developmen- the younger age group of the micrognathia patients. Be- tal period. Another argument in favor of DO that was borne cause of the lack of a proven track record using distrac- out in our study is the ability to achieve very large degrees tion techniques in the very immature facial skeleton, we of advancement. When attempting such movements with tended to be conservative in setting our distraction length conventional craniofacial repositioning, the restrictive soft objectives with these patients. Furthermore, in these pa- tissueenvelopeoftenphysicallyprecludesextremeadvance- tients, the primary objective of distraction was to im- ments. It seems that the gradual movement in DO is able prove or eliminate upper airway obstruction. to overcome the soft tissue pull through a stepwise adap- Functional results were encouraging, particularly in tive process. The long-term outcome of earlier interven- resolving or significantly improving airway obstruction. Dis- tion with DO, particularly as it relates to stability, remains traction osteogenesis holds great potential for addressing tobeseenbecausemoststudiesstilldonothavelongenough the structural problems that contribute to acute respiratory follow-up. Early reports39,58 demonstrate excellent stabil- obstruction in patients with glossoptosis-micrognathia or ity of distraction procedures relative to conventional cra- other disorders with small jaws. Our results indicate that niofacial repositioning, particularly in cases with large ad- early intervention in infants and young children with DO vancements or restrictive soft tissue envelopes. However, can prevent the need for tracheotomy or allow early decan- despite the potential for improved results, the distraction nulationinchildrenwithexistingtracheotomy.Becausemost procedure will not reverse or correct the underlying glossoptosis-micrognathia patients improve spontaneously syndrome-specific abnormalities, and growth will even- over time, the cases should be carefully selected. We have tually lag at the affected areas as the patient ages after treat- reserved mandibular distraction for those with severe mi- ment.34 Thus, intervention at an earlier age with DO may crognathia and obstruction in which case conservative sur- or may not preclude additional surgery. In severe defor- gical techniques have a lower success rate. These patients mities, however, the patient will have several years of con- include those who would otherwise go on to tracheotomy siderably improved form and function, and if revision pro- or those with an existing tracheotomy tube who are not dem- cedures are necessary at the end of skeletal maturity, they onstrating “catch-up” growth. Less severe cases can still be will be technically easier to perform. adequately managed with more conservative techniques, such as lip-tongue adhesion. Impressive improvements in CONCLUSIONS the nasopharyngeal airway were also seen in the patients undergoing midfacial advancement. Furthermore, these pa- In summary, DO has great potential for several congen- tients experienced resolution of conjunctival exposure as- ital and acquired osseous defects that can be encoun- sociated with exorbitism. Occlusal correction proved less tered within the craniofacial skeleton. Numerous advan- amenable to correction using DO. Anteroposterior discrep- tages have been cited among the growing number of ancies had the best improvement; this is not surprising con- advocates for craniofacial DO: (1) The surgery is less in- sidering that the planned vectors of distraction used to cor- vasive and associated with a shorter hospital stay, less rect the skeletal base disharmony coincided with the nec- tissue dissection and bone manipulation, and decreased essary movements needed to improve the dental bases in blood loss. (2) There is no need for bone grafting to main- most patients. Transverse deficiencies, crossbites, and open tain repositioned segments. (3) There is no risk of growth bites, on the other hand, were not of primary concern when restrictions secondary to plate fixation. (4) In severe skel- planning the distraction strategy. Notably, most of the pa- etal deficiencies, there is the potential for substantially tients were left with some occlusal discrepancies that are larger osseous movements and greater stability postop- within the realm of orthodontic correction, although a few eratively, especially in cases with a scarred soft tissue bed. will eventually require finishing orthognathic procedures. (5) There is the potential for improved soft tissue aug-

(REPRINTED) ARCH FACIAL PLAST SURG/ VOL 4, JAN-MAR 2002 WWW.ARCHFACIAL.COM 18

©2002 American Medical Association. All rights reserved. Downloaded From: https://jamanetwork.com/ on 09/27/2021 mentation associated with distraction histogenesis. 25. Persing JA, Morgan EP, Cronin AJ, Wolcott WP. Skull base expansion: craniofacial effects. Plast Reconstr Surg. 1991;87:1028-1033. (6) Finally, surgical intervention is possible in younger 26. Remmler D, McCoy FJ, O’Neil D, et al. Osseous expansion of the cranial vault by age groups. Experience with the technique is still lim- craniostasis. Plast Reconstr Surg. 1992;89:787-794. ited, and more investigation is necessary to refine tech- 27. Tschakaloff A, Losker HW, Mooney MD, Siegel MI, Losker A, Swan J. Internal calvarial bone distraction in rabbits with experimental coronal suture immobili- nique and improve the design and reliability of distrac- zation. J Craniofac Surg. 1994;5:318-323. tion devices. Investigations are under way to determine 28. Barone CM, Ferder M, Jimenez DF, et al. Distraction of the frontal bone outside the cranial plane: a rabbit model. J Craniofac Surg. 1993;4:177-183. the feasibility of combining endoscopic techniques to per- 29. McCormick SU, McCarthy JG, Grayson BH, Staffenberg D, McCormick SA. Effect form the initial osteotomies with insertion of external dis- of mandibular distraction on the temporomandibular joint, part 2: clinical study. traction appliances. As experience with distraction in- J Craniofac Surg. 1995;6:364-369. 30. McCarthy JG, Schrieber J, Karp N, Thorne CH, Grayson BH. Lengthening of the creases, its exact role in correcting craniofacial deformities human mandible by gradual distraction. Plast Reconstr Surg. 1992;89:1-12. will eventually become apparent. Currently, it seems that 31. Molina F, Monasterio FO. Mandibular elongation and remodeling by distraction: DO is most useful for more severe anomalies in which a farewell to major osteotomies. Plast Reconstr Surg. 1995;96:825-840. 32. Diner PA, Kollar EM, Martinez H, Vazquez MP. Intraoral distraction for mandibu- there is a need for earlier intervention, the risk of re- lar lengthening: a technical innovation. J Craniomaxillofac Surg. 1996;24:92-95. lapse with conventional techniques is high, and conven- 33. Cohen SR, Tutrick RE, Burnstein F. Distraction osteogenesis of the human craniofacial skeleton: initial experience with a new distraction system. J Craniofac tional osteotomies are not feasible. Surg. 1995;6:368-374. 34. Kusnoto B, Figueroa AA, Polley JW. A longitudinal three-dimensional evaluation Accepted for publication May 7, 2001. of the growth pattern in hemifacial microsomia treated by mandibular distraction Corresponding author and reprints: Mario J. Imola, MD, osteogenesis: a preliminary report. J Craniomaxillofac Surg. 1999;10:480-486. 35. McCarthy JG. Distraction of the mandible and craniofacial skeleton. J Cranio- DDS, FRCSC, Craniofacial–Skull Base Center, 1601 E 19th maxillofac Surg. 1996;24:193-199. Ave, Denver, CO 80218 (e-mail: [email protected]). 36. Polley JW, Figueroa AA, Charbel A, et al. Monoblock craniomaxillofacial distraction osteogenesis in a patient with severe craniosynostosis: a preliminary report. J Craniofac Surg. 1995;6:421-424. REFERENCES 37. Chin M, Toth BA. Le Fort III advancement with gradual distraction using internal devices. Plast Reconstr Surg. 1997;100:819-832. 1. Codvilla A. On the means of lengthening in the lower limbs, the muscles and tissues 38. Polley JW, Figueroa AA. Management of severe maxillary deficiency in child- which are shortened through deformity. Am J Orthop Surg. 1905;2:353-369. hood and adolescence through distraction osteogenesis with an external, ad- 2. Ilizarov GA. Basic principles of transosseous compression and distraction os- justable, rigid distraction device. J Craniofac Surg. 1997;8:181-185. teosynthesis. Orthop Traumatol Protez. 1971;32:7-9. 39. Toth BA, Kim JW, Chin M, Cedars M. Distraction osteogenesis and its applica- 3. Ilizarov GA. The principles of the Ilizarov method. Bull Hosp Jt Dis Orthop Inst. tion to the midface and bony orbit in craniosynostosis syndromes. J Craniofac 1988;48:1-12. Surg. 1998;9:100-113. 4. Haas AJ. Rapid palatal expansion of the maxillary dental apron and nasal cavity 40. Amaral CMR, Domizio GD, Tiziani V, et al. Gradual bone distraction in cranio- by opening the midpalatal suture. Angle Orthop. 1961;31:73-79. synostosis. Scand J Plast Reconstr Surg Hand Surg. 1997;31:25-37. 5. Cleall JF, Bayne DI, Posen JM, Subtelny JD. Expansion of the midpalatal suture 41. Sugawara Y, Hirabayashi S, Sakurai A, Harii K. Gradual cranial vault expansion in the monkey. Angle Orthop. 1965;35:23-26. for the treatment of craniofacial synostosis: a preliminary report. Ann Plast Surg. 6. Snyder CC, Levine GA, Swanson HM, Browne EZ. Mandibular lengthening by 1998;40:554-563. gradual distraction: preliminary report. Plast Reconstr Surg. 1973;51:506-508. 42. Kobayashi S, Honda T, Saitoh A, Kashiwa K. Unilateral coronal synostosis treated 7. Michieli S, Miotti B. Lengthening of the mandibular body by gradual surgical orth- by internal forehead distraction. J Craniofac Surg. 1999;10:467-472. odontic distraction. J Oral Surg. 1977;35:187-192. 43. Aronson J. Experimental and clinical experience with distraction osteogenesis. 8. Kutsevlink VI, Sukachev VA. Distraction of the mandible in an experiment. Sto- Cleft Palate Craniofac J. 1994;31:473-482. matologia. 1984;63:13-15. 44. Ilizarov GA. The tension-stress effect on the genesis and growth of tissues, part 9. Karp NS, Thorne CH, McCarthy JG, Sissons HA. Bone lengthening in the cranio- II: the influence of stability of fixation and soft-tissue preservation. Clin Orthop. facial skeleton. Ann Plast Surg. 1990;24:231-237. 1989;238:249-281. 10. Karp NS, McCarthy JG, Schrieber JS, et al. Membranous bone lengthening: a 45. Ilizarov GA. The tension-stress effect on the genesis and growth of tissues, part serial histologic study. Ann Plast Surg. 1992;29:2-7. II: the influence of the rate and frequency of distraction. Clin Orthop. 1989;239: 11. Costantino PD, Shybut G, Friedman CD, et al. Segmental mandibular regenera- 263-285. tion by distraction osteogenesis. Arch Otolaryngol Head Neck Surg. 1990;116: 46. Fisher E, Staffenberg DA, McCarthy JG, Miller DC, Zeng J. Histopathologic and 535-545. biochemical changes in the muscles affected by distraction osteogenesis of the 12. Costantino PD, Friedman CD. Distraction osteogenesis: applications for man- mandible. Plast Reconstr Surg. 1997;99:366-371. dibular regrowth. Otolaryngol Clin North Am. 1991;24:1433-1443. 47. Block MS, Daire J, Stover J, Mathews M. Changes in the inferior alveolar nerve 13. Costantino PD, Friedman CD, Shindo ML, et al. Experimental mandibular re- following mandibular lengthening in the dog using distraction osteogenesis. growth by distraction osteogenesis. Arch Otolaryngol Head Neck Surg. 1993; J Oral Maxillofac Surg. 1993;51:652-660. 110:511-516. 48. Gantous A, Phillips JH, Catton P, Holmberg D. Distraction osteogenesis in the 14. Karaharju-Suvanto T, Karaharju EO, Ronta R. Mandibular distraction: an experi- irradiated canine mandible. Plast Reconstr Surg. 1994;93:164-170. mental study in sheep. J Craniomaxillofac Surg. 1990;18:280-284. 49. Losken HW, Patterson GT, Lazarou SA, Whitney T. Planning mandibular distrac- 15. Karahaju-Suvanto T, Peltonen J, Kahri A, Karaharju EO. Distraction osteogen- tion: preliminary report. Cleft Palate Craniofac J. 1995;32:71-76. esis of the mandible. J Oral Maxillofac Surg. 1992;21:118-122. 50. Losken HW, Patterson GT, Tate D, Coit DW. Geometric evaluation of mandibular 16. Califano L, Cortese A, Zupi, Tajana G. Mandibular lengthening by external distrac- distraction. J Craniofac Surg. 1995;6:395-400. tion: an experimental study in rabbits. J Oral Maxillofac Surg. 1994;52:1179-1184. 51. Grayson BH, McCormick S, Santiago PE, McCarthy JG. Vector of device placement 17. Annino OJ, Goguen LA, Karmody CS. Distraction osteogenesis for reconstruc- and trajectory of mandibular distraction. J Craniofac Surg. 1997;8:473-482. tion of mandibular symphyseal defects. Arch Otolaryngol Head Neck Surg. 1994; 52. Tharanon W, Sinn DP. Mandibular distraction osteogenesis with multidirec- 120:911-916. tional extraoral distraction device in hemifacial microsomia patients. J Cranio- 18. Komuro Y, Takato T, Harii K, Yonemara Y. The histologic analysis of distraction fac Surg. 1999;10:202-213. osteogenesis of the mandible in rabbits. Plast Reconstr Surg. 1994;94:152-156. 53. Ortiz-Monasterio F, Molina F, Andrade L, et al. Simultaneous mandibular and max- 19. McCormick SU, McCarthy JG, Grayson BH, Staffenberg D, McCormick SA. Effect illary distraction in hemifacial microsomia: avoiding occlusal disasters. Plast Re- of mandibular distraction on the temporomandibular joint, part 1: canine study. constr Surg. 1997;100:852-861. J Craniofac Surg. 1995;6:358-363. 54. McCarthy JG, Staffenberg DA, Wood RJ, Cutting CB, Grayson BH, Thorne CH. 20. Rachmiel A, Potparic P, Jackson IT, et al. Midface advancement by gradual dis- Introduction of an intraoral bone-lengthening device. Plast Reconstr Surg. 1995; traction. Br J Plast Surg. 1993;46:201-207. 96:978-981. 21. Rachmiel A, Jackson IT, Potparic Z, Laufer D. Midface advancement in sheep by 55. McCarthy JG, Williams JK, Grayson BH. Controlled multiplanar distraction of the gradual distraction. J Oral Maxillofac Surg. 1995;53:525-529. mandible: device development and clinical application. J Craniofac Surg. 1998; 22. Block MS, Brister GD. Use of distraction osteogenesis for maxillary advance- 9:322-329. ment: preliminary results. J Oral Maxillofac Surg. 1994;52:282-287. 56. Diner PA, Kollar EM, Martinez H, Vazquez MP. Submerged intraoral device for 23. Glat PM, Staffenberg DA, Karp NS, Holliday RA, Steiner GS, McCarthy JG. Mul- mandibular lengthening. J Craniomaxillofac Surg. 1997;25:116-123. tidimensional directional osteogenesis: the canine zygoma. Plast Reconstr Surg. 57. Cohen SR. Craniofacial distraction with a modular internal distraction system: 1994;94:753-759. evolution of design and surgical techniques. Plast Reconstr Surg. 1999;103: 24. Staffenburg DA, Wood RJ, McCarthy JG, Grayson BH, Glasburg SD. Midface dis- 1592-1607. traction advancement in the canine without osteotomies. Ann Plast Surg. 1995; 58. Polley JW, Figueroa AA. Rigid external distraction: its application in cleft max- 6:368. illary deformities. Plast Reconstr Surg. 1998;102:1360-1372.

(REPRINTED) ARCH FACIAL PLAST SURG/ VOL 4, JAN-MAR 2002 WWW.ARCHFACIAL.COM 19