Lateral impingements of the temporomandibular joint: a classification system and MRI imaging characteristics


Finite element analysis of dynamic temporomandibular joint (TMJ) loading reveals a predominance of localization of loading laterally towards the collateral ligament regions and disc/capsule attachments to the mandibular condyle. A previous publication (Kirk, Kirk. OMS Clin North Am 2006; 18 :345–68) introduced biomechanical principles for surgeons to consider in the diagnostic phase of management as well as initial surgical procedure selection. The concept of impingements and their impact with development of derangement is presented in this paper with an expanded collection of imaging characteristics. Diagnostic coronal imaging using a dual photon imaging technique is presented. This technique is superior to traditional T1 and T2 weighted imaging sequences when sagittal imaging is employed. Coronal imaging using this technique adds a new dimension to preoperative imaging. Impingement presence and the discernment of early lateral disc/capsule rupture from the condyle of the mandible is superior with the dual photon technique. Images and a classification of degrees of impingement are presented. The biomechanical importance of diagnosis of impingement is discussed.

Macro-orthopaedic pathologies of the temporomandibular joint (TMJ) such as osteophytes, cartilage erosion, and early disc perforation have consistently and repeatedly been described as being localized in the lateral most compartments of the joint by observers in different eras. For more than a decade, a resurgence of interest in TMJ function among biomechanical engineers has resulted in better understanding of orthopaedic biomechanical properties and functional load distributions within the disc/capsule and traversing fossa/eminence and condylar surfaces. In the past it was difficult to discern actual reactive loads in the joint. Now, independent biomechanical research groups worldwide have contributed much to the understanding of loading characteristics of the TMJ in normal and pathologic circumstances by using computer finite element analyses. Such analyses have produced the concept of force fields that are present within this system during orthopaedic function.

In 2006, a mathematic model based on mechanical engineering and subsequently on orthopaedic loading principles in the TMJ during static (occlusion) and dynamic (translational and eccentric) loading was introduced to the surgical community. The concept of ‘micro-biomechanical’ distributed force field principles was introduced to provide a better understanding of pre-surgical conditions which might be used in initial surgical procedure selection and to suggest aetiologies of genesis of orthopaedic derangement development and impairment. The eventual impact of laterally directed kinetic loading force fields and joint surface and compartment loading led to the concept of potential initiating factors that explain predominance of lateral joint tissue and structural damage in advancing derangement due to destruction by shear and torque. Converging tension force fields distributed to small areas along the osseous rim of the glenoid fossa, to the lateral tubercle of the eminence and the lateral pole of the condyle were suggested to be responsible for the evolution of bone pathologies of the fossa and condyle which interfered with normal joint function. The use of coronal magnetic resonance imaging (MRI) to identify impingement processes of varying degrees was introduced. The development and pathogenesis of non-traumatic derangement of the TMJ is now suggested to mimic development of impingement syndrome of the shoulder. The eventual damage to the rotator cuff of the shoulder from impingement processes can be applied as a model for the development of impingement of the TMJ due to the predominance of tension forces delivered to the lateral most aspects of these joints. Joint space narrowing or frank development of osteophyte processes would follow due to bone adaptation and growth when subject to pathologic tension forces by way of Wolfe’s law.

This mathematic model and postulates have been independently supported in degrees by independent researchers using computer finite element analysis. Concepts that physical loads within the TMJ complex are greatest in the lower joint compartment support previous discussions of surgical observations and demonstration that tissue damage in advancing derangement accelerates and is far advanced beyond that of the superior disc surface in advancing Wilkes derangement Types II–IV with eventual pathologic disc/capsule rupture and resultant stages III–V. Initiation of TMJ derangements is now seen as initial detachment of the disc/capsule complex from attachments to the lateral aspect of the condyle and condylar pole. Progressive and further detachment secondary to the impact of unstable and destructive shear and torque give way to gross orthopaedic joint instability progressing to various mechanical dysfunctions, tissue damage, and eventual impairment.

Koolstra’s finite analysis studies have suggested load thresholds of 80 N during mouth opening and joint translation. Such load estimates exceed actual load detachment thresholds calculated by Ben Amour and associates in cadaver studies. This work suggested shearing loads of 29–58 N were capable of creating disc/capsule detachment from the lateral pole and surrounding regions of attachment to the lateral condyle. These detachment thresholds were also noted to be age and gender independent. TMJ developmental impingement processes are suggested to be a cause whereby these detachment force thresholds are exceeded during eccentric and translational jaw movements and the advancement of impairing derangement due to the impact of destructive shear.

The purpose of this paper is to expand the concept of importance of coronal MRI of the TMJ to complement established diagnostic parameters and staging of sagittal imaging evaluation of TMJ derangements. A classification system of impingement processes is further expanded and documented from previous reports to support the concept of their influence in the development and progression of derangement.

Materials and methods

Sagittal and coronal images of 49 consecutive patients with TMJ orthopaedic functional impairment and advancing pain, referred for diagnostic evaluation and potential surgical management were imaged with coronal plane imaging and sagittal imaging. Their ages ranged from 13 to 45 years with a mean of 26 years. All patients were examined with sagittal and coronal MRI imaging with and without an oral orthotic. The all acrylic oral orthotic was custom made to fit the mandibular teeth with posturing of the mandible to a near edge to edge dental incisor position with approximately 3–5 mm of occlusal separation. This position allowed assessment of early condyle translation along the articular eminence and out of the glenoid fossa. This condyle position change was confirmed with sagittal plane tomography prior to MRI.

All the patients in this group that were imaged failed with continuous orthotic wear for at least a month; it was prescribed for pain and functional management. Long term continuous subjective symptoms of well localized joint pain, joint locking and other dysfunction (>1–2 years) were reported by all patients in this group.

Repeatable patterns of degrees of significant lateral joint space narrowing, loss of condyle and fossa curvature congruencies, and condyle and fossa osteophyte characteristics from previous reports were analysed and 3 simple categories of classification were created and are introduced. Images of dysfunctional joints are compared to normal controls obtained from asymptomatic orthopedically stable and non-painful joints present in the studied group. These normal images were defined as joints with absence of Wilkes derangement in sagittal images and congruent structures in medio-lateral coronal imaging.

Changes in MRI protocols were made at the author’s institution when coronal scans were added to sagittal images for three-dimensional examination. Currently, a four channel 1.5 T magnet (General Electric, Milwaukee, WI, USA) and a dedicated dual TMJ coil are used. A matrix of 288 × 192, a bandwidth of 14 Hz and a small field of view (range 10–12 cm) along with multiple numbers of excitations (2–4 NEX) are required in order to improve signal-to-noise ratio. Standard T1 and T2 images are obtained for sagittal imaging with variations noted below.

Improved coronal images are obtained when proton density (PD) weighted sequences with a short time/echo (TE) pulse sequence is used. The images in this report employed PD weights with a time to repetition (TR) of 2150 and a TE of 42 ms for the PD sequences. Fast spin echo (FSE) sequences are utilized to reduce scanning time given the longer TR inherent with PD imaging. To improve diagnostic quality and to decrease image blurring, an echo to train length (ETL) of 6 is used.

A standard T2-weighted image sequence is also performed in the sagittal plane in the closed mouth position and with the orthotic to evaluate for abnormal marrow oedema, joint effusions or synovitis. T1 sagittal sequences are performed with gradient echo recalled (GRE) images. A stepwise mouth opening technique with employment of varying sizes of oral bite blocks is used to produce a dynamic motion study. A 2D GRE sequence with 4 mm slice thickness and 1 mm skip, TR 367 ms and TE 20 ms with flip angle of 20 is used for this part of the examination.

One difference between coronal and sagittal imaging is that in coronal imaging, fewer images can be obtained. Technicians therefore are educated to program the examination to provide a central condyle image in the coronal plane.


In this observational study, there were 38 normal joints, 11 examples of Type III, 29 examples of Type II and 20 of Type I. There was no correlation between age and advancing degree of impingement, but all patients with Type III impingement related subjective joint pain and dysfunction for more than 5 years.

Normal coronal MRI scans reveal concentric and congruent relationships of the medio-lateral arcs of curvature of the fossa, inferior and superior disc/capsule surfaces and the head of the condyle. Joint space anatomy does not show significant compromise laterally and is generally equal to that of more central and medial regions ( Fig. 1 A–D ).

Fig. 1
(A) Normal coronal image in centric dental occluded position. Note relative concentricity of fossa and condyle osseous surfaces and symmetry of medial–lateral joint space. Condyle medial–lateral width is within m-l arc of fossa. (B) Same joint imaged with orthotic. Adaptation of disc/capsule to condyle and integrity of lateral disc/capsule attachments are confirmed in image. (C) Asymptomatic and normal joint with corresponding normal coronal image in centric dental occlusion. (D) Same joint with orthotic in place. Note degree of decompression as well as much weaker lateral disc/capsule attachment compared to (A) and (B). In all examples note medial lateral concentricity of condyle and its width contained within the medial–lateral arch of curvature of the glenoid fossa.

Type I impingement is characterized by a lack of congruency of the medio-lateral arc of curvature of the glenoid fossa relative to the same curvatures of the condyle and the articular disc capsule. Type I is the simplest form, there is significant narrowing of the lateral joint space when viewed in the coronal dimension. Similarly, there are varying degrees of tissue reaction or visualization of detachment of the disc/capsule to the lateral pole of the condyle of the mandible.

Compared to Type I impingement, Type II impingement extends more deeply towards the apex of the arc of curvature of the TMJ glenoid fossa. Evidence of appositional bone is more visible, particularly as imaging advances towards the eminence of the fossa, and in particular the articular tubercle. Clinically, this impingement is associated with tearing or detachment of the articular disc/capsule in its lateral region of attachment to the lateral condyle. Often, chronic subluxation and progressing spontaneous dislocation will be associated with Type II impingement, particularly boney hyperplasia at the tubercle. In both Types I and II impingement, sagittal images can be relatively unremarkable.

Type III impingements are osteophytic changes of the condylar head along with fossa changes similar to those seen in Types I and II. Stress fractures are often a finding as are Wilkes III–V sagittal derangement patterns. Condylar head width is often greater than the medio-lateral width of the glenoid fossa.

Only gold members can continue reading. Log In or Register to continue

Jan 24, 2018 | Posted by in Oral and Maxillofacial Surgery | Comments Off on Lateral impingements of the temporomandibular joint: a classification system and MRI imaging characteristics
Premium Wordpress Themes by UFO Themes