Navicular syndrome (navicular disease) is a degenerative syndrome involving the osseous and soft tissue structures of the podotrochlear apparatus (Wright et al, 1998). The degeneration of the podotrochlear apparatus is thought to be the result of biomechanical overloading from a combination of workload and conformation (Wright and Douglas, 1993). Pain originating from this region is estimated to be responsible for at least one-third of chronic forelimb lameness in Quarter Horses, Thoroughbreds, and Warmbloods (Stashak, 1998; Dyson et al, 2011). Because radiography is commonly the first imaging modality performed following lameness evaluation, accurate radiographic interpretation requires well-positioned radiographs with a thorough understanding of normal variants, artefacts, and the spectrum of radiological lesions that can affect the appearance of the navicular bone. This article reviews methodology for evaluation and interpretation of the most frequently encountered radiographic lesions of the navicular bone, based on a review of the literature and the authors' experiences. Note, although navicular lesions occur in both the fore and hind limbs, for ease of discussion the forelimb terminology will be used.
Normal radiographic appearance
The navicular bone, or distal sesamoid bone, is located on the palmar aspect of the distal interphalangeal joint suspended by the collateral sesamoidean ligament, which attaches to the distal abaxial aspects of the proximal phalanx, and tethered to the distal phalanx by the distal sesamoidean impar ligament. The palmar border (flexor cortex) of the navicular bone is covered with a layer of fibrocartilage which, along with the navicular bursa, helps create a smooth gliding surface for the deep digital flexor tendon. These structures are collectively named the navicular or podotrochlear apparatus. The navicular bone is made of compact cortical bone and internal spongiosa, which are radiographically distinguishable. The conical smooth central defects of the distal aspect of the bone are synovial invaginations that communicate with the distal interphalangeal joint (Poulos and Smith, 1988; Olive and Videau, 2017).
The radiographic projections that best highlight the navicular bone include: lateromedial, dorso-55° proximal-palmarodistal oblique (i.e. upright pedal or D55°Pr-PaDiO), and palmaro-45–55° proximal-palmarodistal oblique (i.e. navicular skyline) (Figure 1). Dorso 60° proximal 45°lateral-palmarodistomedial and dorso 60° proximal 45°medial-palmarodistolateral oblique views are useful for identification of distal border fragments. While an important component of routine foot radiographs, the horizontal beam dorsal-palmar view typically provides the least information regarding the navicular bone.
Precise radiographic acquisition is important because obliquity will result in artefacts that interfere with accurate radiographic evaluation of the navicular bone. For example, poor positioning of the limb and/or insufficient beam angle of the navicular skyline view will result in superimposition of the proximopalmar border of the navicular bone and will artefactually increase the opacity of the spongiosa (Dyson, 2011). During acquisition of the upright pedal view, an insufficient beam angle will not provide adequate separation of the distal border of the navicular bone from the distal interphalangeal joint, limiting evaluation of the distal margin. Additionally, over or under flexion of the fetlock will cause superimposition of the fetlock with the navicular bone (Dyson, 2011). Placing the limb caudal to the contralateral limb, but still weight bearing, will aid radiographic acquisition of the navicular skyline. Whenever possible, removal of shoes and pads will improve the quality of radiographs acquired. The foot should be well cleaned, and debris removed from the sole. Packing the sulci of the frog usually decreases artefacts in this area, but occasionally will accentuate lucent gas lines, and repacking should be considered if there is a clinical concern.
A grading scale for the radiographic appearance of the navicular bone has been previously established, based on the presence and severity of navicular bone sclerosis, conical channels (synovial invaginations), shape, and enthesophytes (Dik and van den Broek, 1995). While a grading scale provides a useful template for consistent evaluation, classification on a grading scale and potential indications of certain grades should be interpreted with caution. In one study, a group of sound horses with mild radiographic changes of the navicular bone (grade 2) showed significant early histopathological lesions including fibrillation and disruption of the deep digital flexor tendon, and thinning of the fibrocartilage and subchondral bone (Komosa et al, 2014). Clinical lameness, pattern of diagnostic analgesia, radiographic lesions, and response to therapeutics should be evaluated together for more individualised prognosis.
Radiographic evaluation should include the following regions:
- Spongiosa:
- Sclerosis of the spongiosa (decreased corticomedullary definition)
- Increased lucent regions (cyst-like changes, dilated synovial invaginations)
- Dorsal cortex:
- Distal interphalangeal osteophytes
- Proximal border:
- Enthesopathy of the collateral sesamoidean ligament
- Dilated vascular channels
- Distal border:
- Synovial invagination size
- Cyst-like lesions
- Impar ligament enthesopathy
- Distal border fragments
- Palmar border (flexor cortex):
- Full or partial thickness lysis
- Cortical thickening
Normal anatomical variants of the navicular bone
Normal anatomical variants occur in many locations in the horse, and knowledge of these variants will help prevent misdiagnosis or negative advice in pre-purchase exams (Becht et al, 2001; Hinkle et al, 2019). The shape of the navicular bone, specifically the proximal border, is considered heritable in Dutch and Hanoverian Warmblood horses (Dik and van den Broek, 1995; Dik et al, 2001). Horses with a convex or horizontal proximal border are reportedly at lower risk of navicular syndrome than those with an undulating or concave margin (Dik and van den Broek, 1995). The navicular bone shapes considered to be at the highest risk for development of navicular syndrome (the undulating or concave articular margins) were not present in the older horse population, indicating that these shapes likely resulted in either early retirement and/or shorter lifespan. Horses with navicular bones considered to be a lower risk (a convex or horizontal border) were more prevalent in the older population, but still developed navicular syndrome, indicating that accumulated chronic biomechanical overload is harmful to all navicular bones regardless of shape (Dik et al, 2001).
Elongation of the navicular bone flexor cortex is a frequent finding identified on the lateral view in sound horses, occurring in over half of sampled sound horses at the distal margin and in a third of sound horses at the proximal margin (Kaser-Hotz and Ueltschi, 1992) (Figure 2a). Elongation of the navicular bone flexor cortex is not considered a valuable criterion to support the diagnosis of navicular syndrome because of its high incidence in sound horses (Kaser-Hotz and Ueltschi, 1992). The articulation between the distal phalanx and the navicular bone was more frequently parallel, although a convergent joint space was also found in sound horses (Kaser-Hotz and Ueltschi, 1992) (Figure 2b).
The morphology of the sagittal ridge (i.e. central eminence) of the flexor cortex of the navicular bone is best outlined on the navicular skyline radiograph. It can be variably shaped, including the typically smooth rounded shape, flattened, pointed, or raised with a central plateau (Kaser-Hotz and Ueltschi, 1992; Becht et al, 2001). The variable presence of an internal crescent shape lucency in the sagittal ridge is frequently identified on the navicular skyline radiograph and represents non-compact bone between a reinforcement line of compacted cancellous bone and the flexor cortex (Berry et al, 1992) (Figure 3). Horses with a more prominent central lucency on the skyline image correspondingly tend to have a more prominent concave curvature or smoothly marginated dimple of the flexor cortex on the lateral image (Dyson, 2011). Unlike a true flexor cortical erosion, the crescent-shaped lucency on the navicular skyline radiograph is well defined, of consistent shape and bordered by smooth compact bone. Additionally, the appearance of the normal lucent region is generally symmetrical between limbs on the same horse.
Spongiosa: sclerosis and loss of corticomedullary definition
The spongiosa is frequently called the medulla or medullary cavity of the navicular bone, but this is a misnomer since sesamoid bones lack a true medullary cavity. The opacity of the normal spongiosa is always less than that of the flexor cortex in the normal horse, with a sharp distinct margin between the flexor cortex and spongiosa.
Sclerosis of the spongiosa and loss of the distinction between the flexor cortex and spongiosa are frequently referred to as ‘loss of corticomedullary definition’, which is generally accepted terminology despite the lack of a true medullary cavity. Sclerosis of the spongiosa is associated with increased forces exerted by the deep digital flexor tendon on the flexor cortex (Pool et al, 1989). Sclerosis is an early and consistent finding in navicular syndrome, although it has reportedly been found in 16% of sound horses (Kaser-Hotz and Ueltschi, 1992). Most frequently, sclerosis is identified as a focal or diffuse region of increased opacity, often extending from the endosteal surface of the flexor cortex (Figure 4). A diffuse haze across the entire spongiosa or more linear increased opacity bisecting the long axis of the bone should flag the interpreter to consider superimposition as the source of this appearance.
If the navicular skyline image is not obtained at the appropriate angle, which is horse and hoof dependent, then the borders of the proximal or distal margin become superimposed with the spongiosa. To prevent this error, the examiner must ensure that the joint space between the navicular bone and the middle phalanx should be sharp and clear, as shown in Figure 1c. If this interface is ill-defined on the skyline view, there is a greater likelihood of artefactual increased opacity of the navicular bone. Focal round regions of sclerosis on the abaxial aspects of the spongiosa on the navicular skyline image may represent superimposition of a distal border fragment.
Cyst-like lesions in the spongiosa separate from synovial invaginations or flexor cortex erosions have been identified in some navicular bones but have not been investigated with histology. In one study, the presence of cyst-like lesions was positively associated with the severity of distal sesamoid impar ligament injury based on magnetic resonance imaging (MRI) and histopathology (Dyson et al, 2010). The majority of round lucent defects found in the spongiosa of the navicular bone represent flexor cortical erosions superimposed with the spongiosa on the D55°Pr-PaDiO oblique or abnormal dilation of the synovial invaginations rather than true cystic lesions.
Proximal border: enthesopathy of the collateral sesamoidean ligament and vascular channels
Osseous proliferation of the lateral and medial proximal margins of the navicular bone is best identified on the upright pedal view and represents enthesophyte formation at the attachment of the collateral sesamoidean ligament of the navicular bone (Dyson, 1988; Pool et al, 1989). This change has been reported to be found in 5% of sound horses as well as horses affected by navicular syndrome, so the clinical significance remains unclear (Kaser-Hotz and Ueltschi, 1992) (Figure 5). In the authors' opinion, this finding may occur with other significant navicular changes, but on its own it is rarely a primary clinically significant lesion.
In the course of navicular syndrome, progressive enlargement of the arterial supply of the proximal border can be detected radiographically as a change in shape on the proximal border on the lateral view, with increased concavity to the proximal cortex. In more advanced cases lucent regions may be viewed on the proximal aspect of the navicular bone on the upright pedal view. The lucent vascular channels in the proximal border of the navicular bone are superimposed with the distal border synovial invaginations on the navicular skyline image, limiting the value of the skyline view for assessing this change. These changes indicate a shift in arterial supply to the navicular bone: decreased distal arterial supply to an increased proximal arterial supply to the point that distal arteries may become absent in horses with navicular syndrome (Rijkenhuizen et al, 1989).
Distal border: synovial invaginations and distal border fragments
A few conical lucent regions are normally present on the distal border of the navicular bone highlighted by the upright pedal view and found in the centre of the spongiosa on the navicular skyline image. These lucent regions represent synovial invaginations from the distal interphalangeal joint and do not communicate with the navicular bursa as confirmed by computed tomography comparison of bursography vs arthrography and gross dissection (Poulos and Smith, 1988; Olive and Videau, 2017). Previous reports of up to seven synovial invaginations have been reported in normal sound horses with an average of four or five typically present (Colles, 1983; Dyson, 1988; Poulos and Smith, 1988; Park, 1989). Hindfeet typically have fewer synovial invaginations than forefeet (Butler et al, 2017). Normal invaginations should have a greater height than width (Colles, 1983).
Overall, radiographs underestimate the number and depth of synovial invaginations of the navicular bone, with computed tomography identifying an average of two invaginations more than radiography (Claerhoudt et al, 2012). Computed tomography of 50 Warmblood horses identified an average total number of synovial invaginations of 5.9±1.56 (Claerhoudt et al, 2012). In the authors' experience, the number and size of synovial invaginations are also breed dependent. Warmblood horses typically have larger sized synovial invaginations than small breed horses (Figure 6).
Given the variation in the normal shape and number, there can be a grey area between normal anatomical variants and pathological changes including size, shape and number. In those cases, comparison between limbs is of value as asymmetry is more suggestive of abnormal changes. One report found abnormal invaginations in 11% of sound horses, which indicates that abnormal invaginations identified in isolation may not be conclusive for a diagnosis of navicular syndrome (Kaser-Hotz and Ueltschi, 1992). In the authors' experience, it is not uncommon to find horses wiThenlarged synovial invaginations, without other navicular changes, that are unaffected clinically. In more advanced cases, enlarged synovial invaginations may be associated with localised osteonecrosis extending into the cortical and spongiosa (Figure 7) (Blunden et al, 2006). The way in which synovial invaginations enlarge can be of importance; enlargement of the synovial invaginations in a dorsalpalmar dimension can result in pressure resorption of the palmar endosteum of the navicular bone and, in more advanced cases, result in flexor cortical lysis.
Since synovial invaginations are extensions of the distal interphalangeal joint, it has been postulated that this abnormality is an indicator of joint disease rather than navicular syndrome (Olive and Videau, 2017). However, this hypothesis warrants further investigation. It has been the authors' observation on radiography and MRI that abnormal trabecular bone quality of the navicular bone may allow synovial invaginations from the distal interphalangeal joint to enlarge in the absence of identifiable distal interphalangeal joint disease.
Distal border fragments are variably sized and shaped osseous bodies present at the distal lateral or medial angles of the navicular bone. Distal border fragments are most easily identified on the upright pedal or oblique upright pedal radiograph (Figure 8). These fragments have been described as avulsion fracture of the navicular bone, fracture of an enthesophyte at the origin of the distal sesamoidean impar ligament, a separate centre of ossification, synovial osteoma, or dystrophic mineralisation in the distal sesamoidean impar ligament (Poulos et al, 1989; Wright, 1993). Larger fragments can also be identified on the lateromedial view and as a rounded region of increased opacity superimposed with the spongiosa on the navicular skyline. One study found distal border fragments in 13.6% of horses and 9.8% of feet, which were identified equally in a single foot or both front feet, were more commonly uniaxial and could be medial or lateral (Biggi and Dyson, 2011).
Radiography has low sensitivity (37.8–42.8%) but high specificity (99.4–100%) for the detection of distal border fragments (Schramme et al, 2005; Biggi and Dyson, 2010). Fragments with more severe signal changes in the distal border on MRI and medium- and large-sized fragments of the navicular bone were most likely to be identified radiographically (Biggi and Dyson, 2010). Fragments observed radiographically were likely to be associated with other pathological abnormalities within the navicular bone, but up to 45% of fragments with severe signal change in the distal border on MRI and 43% of the large-sized fragments were not detected radiographically (Biggi and Dyson, 2010).
Image quality has a substantial impact on the likelihood of detecting fragments. A poorly prepared hoof with solar debris, inadequate radiographic technique, insufficient beam angle of the upright pedal view, and lack of oblique upright pedal views are all factors that decrease fragment identification. If the beam angle is too shallow (flat) in the upright pedal view, the distal border of the navicular bone will superimpose with the distal interphalangeal joint, making it more challenging to detect small fragments. Increasing the angle to be more upright will lift the distal border of the navicular bone to be superimposed to the mid aspect of the middle phalanx, making the distal border more conspicuous.
Previous studies have described distal border fragments in both lame and non-lame horses (Poulos et al, 1989; Kaser-Hotz and Ueltschi, 1992; Wright, 1993; Schramme et al, 2005; Dyson, 2011), but these occur more frequently in lame horses (Wright et al, 1998; Blunden et al, 2006). Other research found no increased probability of being lame if a fragment was present, although there was a slight increased probability of lameness if both medial and lateral fragments were present (Yorke et al, 2014). In the authors' experience, smooth, small ovoid fragments of otherwise normal navicular bones are often found incidentally—both radiographically and with MRI.
Examining the image for radiolucent defects on the distolateral or distomedial aspects of the bone can help improve identification of distal border fragmentations and their associated fragment beds. While these lesions can correspond with fragmentation, false positives have been recognised following MRI evaluation (Biggi and Dyson, 2010). False-positive radiolucencies at the angles of the navicular bone are typically the result of artefact from superimposition of solar gas or incidental contour variation in radiography.
Dorsal border: osteophytes
The dorsal cortex of the navicular bone articulates with the distal interphalangeal joint. The most common abnormality that affects the dorsoproximal aspect of the navicular bone is osteophytosis. Osteophytes at this site most likely represent osteoarthritis of the distal interphalangeal joint rather being than a sign of navicular syndrome (Wintzer and Dämmrich, 1971; Hertsch, 1977).
Palmar border (flexor cortex): erosive lesions
Flexor cortical erosions have been identified in fewer than 1% of sound horses, making them a strong indicator of clinically significant navicular syndrome (Kaser-Hotz and Ueltschi, 1992). This finding is considered highly relevant by the authors because of the likelihood of a career-limiting prognosis. Flexor cortical defects are most frequently accompanied by other radiographic signs of navicular syndrome, such as sclerosis of the spongiosa, thickening of the flexor cortex, and loss of a distinct interface between cortex and spongiosa (Kaser-Hotz and Ueltschi, 1992) (Figure 9). However, it is possible to have small flexor cortex erosions identified on MRI that do not have any identifiable bone change on the radiographs; this is most likely in the early phase of disease. Increased availability of MRI and owners' willingness to have their horses undergo advanced imaging has revealed these smaller changes which are not evident radiographically, but could be the source of lameness.
Flexor cortical thickness can vary with breed and individual horse. The mean width of the flexor cortex as measured on the navicular skyline radiograph was 3.6 mm in a large study of sound Warmblood horses and reported as 2 mm in Thoroughbred horses (O'Brien et al, 1975; Kaser-Hotz and Ueltschi, 1992). Flexor cortical widths greater than 4.3 mm were found in horses considered to have navicular syndrome (Ueltschi, 1983). Thickening of the flexor cortex is usually identified more prominently on the proximal aspect of the bone and often corresponds wiThendosteal sclerosis.
Central round lucent regions of the navicular bone identified on the upright pedal view can represent cyst-like lesions of the spongiosa or flexor cortex erosions. This must be correlated with the appearance of the flexor cortex on the lateral and navicular skyline images to differentiate the disease processes. Acquiring additional navicular skyline images of varying degrees has been proven to be helpful in the identification of flexor cortical erosions (Johnson et al, 2018). A more shallow angle for the navicular skyline will better highlight the distal aspect of the navicular bone where lysis is known to occur (Figure 10). This can also be helpful to better differentiate true lesions from artefact created by gas within the sulci of the frog.
Navicular bone fractures vs bipartite or tripartite
Previous hypotheses for the underlying cause of bipartite or tripartite navicular bones include formation of multiple ossification centres and/or a disturbance in blood supply leading to incomplete ossification of the cartilaginous bone model or traumatic fracture during the neonatal or juvenile period (van der Zaag et al, 2016). Distinguishing the partite navicular bone from traumatic fractures depends on degree and duration of lameness and radiographic features. Horses with bipartite or tripartite navicular bones usually present with chronic, persistent lameness with a gradual onset (van der Zaag et al, 2016).
On radiographic evaluation, these navicular bones have a uniform distribution of partition sites at approximately one-third of the length of the navicular bone without substantial peri-articular remodeling (van der Zaag et al, 2016). The partitions usually have rounded bone edges, and can have lysis or cyst-like lesions develop at the partitions' interface depending on the stage of degeneration (Figure 11). Conversely, fractures in the acute phase will have sharp margins whereas chronic fractures may have rounded margins but are unlikely to occur at such a distinct interval as a partite navicular bone (Figure 12). Comparison to the contralateral limb can be helpful to distinguish the processes as well because fractures are rarely bilateral. Both disease processes are often associated with progressive degenerative disease, although some horses with fractures will make a full recovery.
Distal phalanx angle and navicular bone relationship
Not only should the radiographic appearance of the navicular bone be examined, but the distal phalanx angle should also be assessed radiographically. The relationship between the distal phalanx angle and the biomechanical forces placed on the navicular bone are important considerations for making recommendations to owners and farriers for shoeing. A low palmar angle of the distal phalanx will result in biomechanical overload on the navicular bone by the deep digital flexor tendon (Figure 13) (Wilson et al, 2001; Eliashar et al, 2004; Weaver et al, 2009). Supporting the palmar angle with therapeutic shoeing, such as heel wedges, can decrease the force of the deep digital flexor tendon on the navicular bone by 4% for every 1° the palmar angle is increased (Eliashar et al, 2004).
Conclusions
High quality radiographs are paramount to the accurate assessment of the navicular bone. Degenerative changes of the navicular bone will typically begin with sclerosis of the spongiosa, with multiple radiographic lesions appearing during progression. Enthesopathy of the proximal border at the attachments of the collateral sesamoidean ligament and distal border fragments may not indicate navicular syndrome. Changes such as enlarged synovial invaginations are of variable clinical significance and should be interpreted with caution. Flexor cortical erosion is a relevant finding that can indicate career-limiting, progressive disease. Absence of radiographic abnormalities does not exclude navicular bone pathology; MRI can help identify radiographic occult lesions as well as concurrent soft tissue injury and can, therefore, be used to augment the radiographic exam.
KEY POINTS
- Patient positioning, radiographic beam angle and technique are key elements in producing high quality radiographs.
- The spectrum of radiographic navicular bone changes includes normal variants and clinically incidental findings to career-ending lesions. Accurate interpretation of these findings is key to reaching a diagnosis.
- A methodical approach to radiographical interpretation will increase accuracy and decrease over- or under-interpretation.
- Decreased corticomedullary distinction of the navicular bone is an important indicator of pathological change.
- Flexor cortical erosions are the most significant radiographic lesion of the navicular bone.
- Additional radiographic beam angles, such as a more shallow angle on the navicular skyline projection, have been found to aid in detection of flexor cortical erosions.