INTRODUCTION

Orthognathic surgery is an important treatment method for patients with significant skeletal dysplasias. For patients who choose orthognathic surgery to correct their skeletal problems involving the mandible, the most common approach to treatment today is the bilateral sagittal split osteotomy (BSSO) and rigid internal fixation (RIF).1

In the past, the majority of orthognathic surgical procedures used wire fixation to stabilize segments at the time of surgery. This was followed with approximately six weeks of intermaxillary fixation while the bony segments healed. Although this produced satisfactory results, the flexibility between segments with wire fixation allowed for some movement of the segments that contributed to increased relapse. Stability has been enhanced with the introduction of RIF.2

The effect of RIF on the temporomandibular joint (TMJ) is an area of concern. A common concern is that rigid fixation may torque the condyles relative to the glenoid fossa, unlike wire fixation that appears less rigid. Alteration of TMJ morphology may adversely affect function leading to temporomandibular disorders (TMD). Furthermore, secondary osseous remodeling in the TMJ may lead to decreased stability of the surgical result.

The purpose of this review was to evaluate the effect of BSSO mandibular advancement surgery with RIF on TMJ morphology. The specific research questions were:

1. What are the effects on TMJ osseous morphology?

2. What are the effects on disk position?

3. What are the effects on condylar position in all planes of space?

METHODS

The following criteria were used to consider studies for this review:

1. Adult patients (male or female) treated with BSSO mandibular advancement surgery and RIF (screws or plates).

2. Utilization of transcranial radiographs, submental vertex (SMV) radiographs, tomographic radiography, computed tomography (CT) scan, or magnetic resonance imaging (MRI) to view the TMJ.

3. Controlled study (minimum internal control with pre- and postintervention imaging).

4. Case reports were excluded.

To identify the relevant articles for this systematic review, a search was conducted for Medline (1966 to October 2001). PubMED was also searched with the keywords and subjects used for Medline in an attempt to locate more recent publications that may have been omitted with Medline. None were found. The researcher reviewed the abstracts of related articles and hand searched the reference list of articles that met the inclusion criteria. The search strategy used for this study is shown in Table 1.

TABLE 1.  Search Strategy

View Fullscreen

Reading the abstracts eliminated many articles but others required a full read of the methods section. Two of the authors read all eligible reports, and agreement of inclusion was reached by discussion.

RESULTS

Eight articles that met the inclusion criteria were identified. The study by Spitzer et al3 was eliminated because of unclear methodology even though it met all the inclusion criteria. The study by Harris et al4 used the same sample and same imaging as Alder et al5 and was eliminated. The remaining six articles are summarized in Table 2 and were published between 1989 and 1999. All the six articles had different primary authors but there was some overlap in the group with second and third authors.

TABLE 2.  Studies Included for Review

View Fullscreen

The type of diagnostic imaging used varied significantly. Only SMV radiographs were used in more than one of the studies. Two different forms of CT were used (3D CT and linear CT).

There were no random clinical trials and no multicenter studies in the sample. One study took records at three points in time but the remaining studies had two sets of records. The timing of data collection varied widely between studies. Only one study took records more than 6–12 months after surgery.

The statistics used were primarily t-tests to compare groups and some multiple regression analyses to look at contributing factors such as the amount of advancement. Some studies had limited statistical analysis and displayed descriptive statistics only.

This study was intended to answer three questions. Each question addressed a possible or perceived problem that could occur after orthognathic surgery to advance the mandible. Data relevant to each question will be presented separately.

DISCUSSION

Randomized clinical trials are not feasible because of the nature of intervention. Internal controls comparing pretreatment (T1) and posttreatment (T2) diagnostic images are appropriate where examiner blinding is ensured and method error is clearly reported. This approach would be strengthened by introducing an additional imaging sequence (T0) taken a period of time before surgery that would be equivalent to the time interval T1 to T2. Error introduced by repeated imaging including patient positioning, tomographic and MRI slice depth, MRI slice orientation, and remodeling unrelated to surgery could be analyzed. Blinding is important in protecting against investigator bias.1314 To assess validity of reported results, method error must be reported. If method error exceeds the mean difference between comparison groups, no conclusion can be made.

Postsurgical TMJ remodeling and degenerative changes have been implicated in postsurgical relapse.1516 Arnett2 discussed potential sources of relapse in BSSO mandibular advancement. He suggested that skeletal relapse could occur by osteotomy slippage, condylar sag, and condylar compression with morphologic change. Condylar sag relates to lack of seating of the condyle into the glenoid fossa at the time of surgery. Because the condyle “reseats” into its preoperative position within the glenoid fossa, the alveolar base of the mandible would rotate down and back. Condylar compression with secondary remodeling would be a source of relapse, 9–18 months after surgery. Compression could occur in medial, lateral, or posterior directions. Arnett et al1718 have also described idiopathic condylar resorption as a source of relapse evidenced by progressive mandibular retrusion. Some individuals may adapt to small joint changes whereas others may not. Whether or not adaptation occurs may be related to pretreatment TMJ disk status, the amount of change produced by surgery, the nature of the original malocclusion, or host remodeling capacity.1819

Unfortunately, there is inadequate published information to establish the effect of BSSO with RIF on the osseous articular tissues within the TMJ. The Gaggl et al7 article described degenerative changes but did not supply adequate information on methodology to be conclusive. Furthermore, there was no report of blinding or method error. The Schultes et al6 article did not report method error or blinding.

The relative risk of developing TMJ internal derangement or inducing a progression of internal derangement with BSSO with RIF is an important issue in treatment planning and in obtaining informed consent. TMJ internal derangement is a common abnormality in the general population. Tallents et al20 reported that 66% of their symptomatic sample and 33% of their asymptomatic volunteer sample had disk displacement. The prevalence of internal derangement in asymptomatic patients seeking orthodontic treatment may be even higher.21 Clinical examination of disk status does not provide accurate determination of disk status and a definitive diagnosis requires MRI.22 The etiology of disk displacement remains obscure, and the natural progression of internal derangement has not been well defined.2324 It is generally accepted that disk displacement with reduction precedes disk displacement without reduction. The time interval for progression is not well defined and risk factors in progression are also uncertain. The role of internal derangement in degenerative joint disease is uncertain25 but disk abnormality may reduce the joint capacity to adapt to changes induced by condylar displacement associated with orthognathic surgery.

The only publication that evaluated disk position7 was not conclusive. Methodology was not adequately described, there was no report of examiner blinding, and method error was not reported.

Condyle position changes within the glenoid fossa should be analyzed in three dimensions. This could be achieved using SMV radiographs with axially corrected sagittal and coronal tomography. Williamson et al26 reported measurement error associated with several SMV analyses. Their findings highlight the relevance of individual variability and importance of reporting error. Change in condylar angulation is best reported in relation to stable landmarks rather than in relation to the other condyle. The Hackney et al12 article did not report blinding or method error. The Stroster et al9 article did report method error but there was no report of blinding and no statistical analysis of T1 to T2 change in condylar angulation with BSSO and RIF. The Rotskoff et al8 article used axially corrected sagittal tomography but they did not report examiner blinding or method error.

The Alder et al5 study had sound well-described methodology. Although method error was not provided it is fair to conclude that all the condyles had some 3D displacement. The magnitude and direction of displacement had large individual variability. The clinical significance of these displacements is unknown, and no long-term studies to assess postsurgical remodeling have been published.

CT scan data provides a useful tool because displacement in three dimensions can be analyzed from a single imaging sequence. Acceptable accuracy of two-dimensionally reformatted CT images of the TMJ has been reported in other studies.2728 One cause for concern is the high dosage of radiation associated with CT scanning.29 The value of the data set obtained for research vs the risk of the radiation dosage should be considered. This is particularly important in studies with repeated exposure. Accessibility and cost of CT scans may also be a consideration.

This systematic review has demonstrated the need for additional research to determine the effect of BSSO with RIF on TMJ morphology. Future research should include MRI assessment of disk position and disk morphology, radiographic assessment of osseous morphology, and 3D assessment of condyle position. To establish internal control and to evaluate method error, imaging should be assessed 6–8 weeks before surgery, immediately before surgery, and 6–8 weeks after surgery. To determine clinical relevance of changes induced by surgery long-term follow up imaging should also be acquired. The interrelationship between condyle position, disk status, and bony morphology should be explored using combined data obtained for a single sample. It may also prove useful to correlate cephalometric data with TMJ findings. Sample size can be estimated from the data of the existing studies.

CONCLUSIONS

The following conclusions regarding the effect of BSSO with RIF on TMJ morphology can be made.

1. The nature of condylar and glenoid fossa remodeling has not been established.

2. Changes in disk position and morphology have not been established.

3. Changes in condyle position present with large individual variability.

There is a need for research to determine the effect of BSSO with RIF on TMJ morphology. Future research requires effective use of controls, examiner blinding, and reporting of method error. The interrelationship between condyle position change, disk status, and osseous morphology need to be explored. It is important to establish the relative risk ratio of developing TMD, causing exacerbation, and/or progression of TMD. The long-term consequences of altered TMJ morphology and the relationship TMJ morphology to treatment stability should also be established.