The role of imaging in the diagnosis, staging, and management of the osteochondral lesions of the talus

Abstract

Osteochondral lesions of the talus (OLT) represent an abnormality of the articular cartilage and sub-chondral bone. The abnormality is typically associated with trauma though the exact aetiology remains unknown. Multiple staging systems have been developed to classify the abnormality and management can vary from conservative treatment to different surgical options. Early diagnosis is essential for optimal outcome and all imaging modalities have a role to play in patient management. The aim of this article is to review the pathology, classification, multimodality imaging appearances of OLT, and how the imaging affects patient management.

Introduction

Osteochondral lesions of the talus (OLT) represent an abnormality of the articular cartilage and sub-chondral bone commonly associated with trauma.^1,2 Early diagnosis is essential for optimal outcome, with management ranging from conservative treatment to a variety of surgical options aimed at maintaining pain-free joint function.^1,2 The aim of this article is to review the pathology, classification, and multimodality imaging appearances of OLT, and to describe how imaging informs current management options.

Definition

The term “osteochondral lesion” is used to describe a post-traumatic or developmental condition involving a region of hyaline cartilage and sub-chondral bone at a convex articular surface.^2,3 The talar dome is the third commonest site of OCL after the femoral condyles and capitellum, and accounts for ∼4% of all OCL.^4,5 Various terms have been used to describe talar OCL including osteochondral fracture, trans-chondral fracture, osteonecrosis, and osteochondritis dissecans (OCD).⁶ These terms have been used to describe the underlying proposed traumatic or non-traumatic mechanisms, although they are often used interchangeably, which results in confusion.^7,8 However, the term “osteochondral lesion of the talus” is now considered the most appropriate, with OCD and osteochondral fracture considered as subsets of OLT.^1,2,9–11 It is important to highlight that histological studies have failed to identify an inflammatory basis for this condition, and therefore the use of the word “osteochondritis” is controversial.^6,7

Aetiology

Konig first described loose bodies in the knee arising from the medial femoral condyle as OCD, while Kappis¹² first used the term to describe the lesion at the ankle.¹³ The abnormality occurs predominantly at the medial or lateral margins of the talar dome, although a small minority may occur anteriorly or posteriorly. Medial lesions are commoner, accounting for 53%-89% of cases, while lateral lesions account for 11%-46%.^7,12–15 Lateral lesions typically occur in the mid- or anterior quadrant while medial lesions typically occur in the posterior half.^8,16 The average size of an OLT is 6.9 mm in width (range 1.7-13.1 mm), 9.4 mm in length (range 1.9-9.2 mm), and 5.4 mm in depth (range 1.0-15.5 mm).¹³

The aetiology of OLT can be broadly categorized into acute traumatic and non-traumatic. Berndt and Harty⁸ attributed OLT to trauma, with 20 of 24 patients giving a history of injury. Similarly, 14 of 18 patients in a study by Parisien¹⁷ gave a history of trauma. Hintermann et al^18,19 reported talar cartilage injuries in 55% of patients with chronic ankle instability and 69.4% of patients with acute ankle fractures. Zengerink et al²⁰ showed that 50% of post-traumatic OLT will require some form of surgical intervention.

Lateral lesions occur due to inversion and dorsiflexion stress causing the anterolateral aspect of the talar dome to impact against the fibula, and medial lesions occur with inversion stress on a plantar flexed ankle with lateral rotation of tibia on the talus resulting in posteromedial talar dome impaction against the tibia.^5,8 There is a stronger association of lateral lesions with trauma compared to medial lesions. Canale and Belding²¹ found that all lateral lesions were associated with trauma. However, Raikin et al²² reported that 63% of OLT imaged with MRI were located medially compared to 34% involving the lateral talar dome, with 80% of both occurring in the central portion of the talus in the sagittal plane. Furthermore, medial OLT were shown to be both larger and deeper than their lateral counterparts.²² A report by Orr et al²³ also showed that OLT were predominantly found in the central third of the talus in the sagittal plane, and that medial lesions were larger. However, they reported that lateral lesions (65%) were commoner than medial (35%).²³ Significantly more lateral lesions were symptomatic and underwent operative intervention despite being radiologically smaller, which is likely due to increased contact pressure in the centro-lateral joint compared to the centro-medial joint.

Different causes for OLT have been proposed, particularly for medial lesions where a history of a preceding acute traumatic event is not always present. Athanasiou et al²⁴ studied the articular cartilage at the ankle in human cadavers and found that the tibial cartilage was stiffer compared to the talar cartilage in the posterior aspect of the joint. They hypothesized that this mechanical disparity led to repetitive microtrauma, cartilage damage, and avascular necrosis.²⁴ The finding of bilateral OLT and the increased prevalence of OLT in siblings has led to a proposal for a genetic or familial predisposition, although this is considered to be rare.^25,26

The term OCD has been used to describe an acquired idiopathic lesion of sub-chondral bone that can produce delamination and sequestration with or without articular cartilage involvement, and instability.⁷ It is most frequent in the second decade of life with an incidence of 0.09% and a prevalence of 0.002/100 000 person per year in the talus.^27–29

Clinical features

OLT can be asymptomatic. Symptomatic patients typically present with chronic ankle pain following inversion injury, with intermittent pain on weight-bearing or running. Symptom severity increases with detachment, which may lead to joint swelling, a feeling of instability and rarely locking.^5,7 Physical examination may demonstrate tenderness behind the medial malleolus with the ankle dorsi-flexed for medial lesions, and at the anterolateral aspect of the ankle joint in maximal plantar-flexion for lateral lesions.⁵ The age at presentation is variable, ranging from young children to the elderly with the average age being 20-40 years.^8,14,29

Staging

The first OLT staging system was published by Berndt and Harty⁸ (Table 1). While not specified that the system was based on radiographic findings, it was used in the description of their cases and the subsequently published literature has assumed it to be based on radiography. They also had a separate Stage 3a for fractures of the lateral margin of the talus, with a diagram showing marked displacement of the fragment but still present adjacent to the native bone.

Table 1.

Staging systems for OLT.

Modality	Name of the staging system, year of publication, modality, and stages
Radiography	Brendt and Harty (1959) Stage I—Sub-chondral impaction fracture Stage II—Partially detached osteochondral fragment Stage III—Completely detached osteochondral fragment with no displacement Stage IV—Displaced osteochondral fragment
Arthroscopy	Pritsch et al (1986) Stage I—Intact, firm and shiny cartilage Stage II—Intact soft cartilage Stage III—Frayed cartilage
MRI	Anderson et al (1989) Stage I—Sub-chondral trabecular fracture with normal radiographs, bone marrow oedema on MRI, and positive bone scan Stage IIA—Articular cartilage injury with sub-chondral cyst formation and bone marrow oedema on MRI Stage IIB—Fluid SI incompletely separating the osteochondral fragment from the underlying bone Stage III—Fluid SI completely separating the osteochondral fragment from the underlying bone with no displacement Stage IV—Displaced osteochondral fragment
MRI	Nelson et al (1990) Stage I—Intact cartilage with signal changes in sub-chondral bone Stage II—High SI breach of articular cartilage Stage III—A high SI rim extending incompletely between the osteochondral fragment and underlying bone Stage IV—Mixed or low SI loose body within the centre of the lesion or displaced within the joint
CT	Ferkel and Sgaglione (1990) Stage I—Intact roof/cartilage with cystic lesion underneath Stage IIA—Cystic lesion with communication to the surface Stage IIB—Open surface lesion with overlying fragment Stage III—Non-displaced fragment with lucency underneath Stage IV—Displaced osteochondral fragment
MRI & arthroscopy	Dipaola et al (1991) Stage I—Injury limited to articular cartilage; softening and irregularity of cartilage on arthroscopy; low SI in cartilage on intermediate-weighted or spin-echo MR images Stage II—Articular cartilage breached; non-displaced fragment on arthroscopy; damage to cartilage on MRI, with low SI rim deep to the fragment indicating fibrous union Stage III—Articular cartilage breached on arthroscopy with a partially displaced fragment attached by overlying cartilage; MRI shows hyperintense SI on T2 between the fragment and underlying bone Stage IV—Displaced osteochondral fragment seen on arthroscopy and MRI
CT	Loomer et al (1993) Stage I—Compression fracture Stage II—Partially detached fragment Stage III—Completely detached osteochondral fragment without displacement Stage IV—Displaced osteochondral fragment Stage V—Radiolucent fibrous defect in sub-chondral bone
Arthroscopy	Ferkel and Chen (1995) Stage A—Smooth and intact but soft and ballotable cartilage-stable Stage B—Rough surface of cartilage-stable Stage C—Fibrillation/fissures in cartilage-stable Stage D—Cartilage flap present or bone exposed-unstable Stage E—Loose undisplaced fragment-unstable Stage F—Displaced fragment-unstable
MRI	Hepple et al (1999) Stage I—Articular cartilage damage with low SI on MR Stage IIA—cartilage injury with underlying fracture and bone marrow oedema Stage IIB—stage IIA without bone marrow oedema Stage III—Completely detached osteochondral fragment; no displacement Stage IV—Displaced osteochondral fragment Stage V—Sub-chondral cyst formation
MRI & arthroscopy	University of Pittsburgh (1999) Stage I—Sub-chondral trabecular compression with bone marrow oedema on MRI Stage II—Sub-chondral cysts developed from stage I lesions; not seen acutely Stage III—Partially or completely detached osteochondral fragment without displacement Stage IV—Displaced osteochondral fragment At each stage, cartilage is assessed on arthroscopy and classified as viable (A) or non-viable/breached (B)
MRI	Mintz et al (2003) Modification of the Ferkel and Chen criteria with application to MRI Grade 0—Normal Grade 1—hyperintense but morphologically intact cartilage Grade 2—Cartilage fibrillation/fissuring not extending to bone Grade 3—Cartilage flap/exposed bone Grade 4—Loose undisplaced fragment Grade 5—Displaced fragment

Abbreviations: OLT = osteochondral lesions of the talus; SI = signal intensity.

Advances in imaging techniques have led to better visualization of the pathology and identification of features not evident on radiography, such as cartilage defects and bone marrow oedema. Similarly, progress in arthroscopy has led to better assessment of OLT and correlation with imaging findings. Multiple staging systems have been published based on imaging alone or in combination with arthroscopy, in order to guide management. These have been summarized in Table 1.^{15,16,30–36}

While the staging systems allow for uniform description of OLT, there is little evidence of their prognostic value and they rarely guide treatment.^1,5 They do not take lesion size into account, which is an important criterion in determining and planning appropriate intervention. Some staging systems consider the presence of sub-chondral cysts as a higher-grade lesion, while others either do not consider them or have included them in earlier stages.^{15,16,30,31,33} Shearer et al³⁷ published the results of 34 patients with grade V lesions according to the classification by Loomer et al,¹⁶ treated with conservative management. They found good-to-excellent results in 54% of patients with osteoarthritis presenting as a rare sequalae. However, these would be classified as Stage IIA according to the classification by Anderson et al³⁰ on MRI, and the Ferkel-Sgaglione³³ classification on CT. Similarly, Dipaola et al³¹ recommended conservative management for Stage I lesions and surgical intervention for higher-grade lesions, while others have shown that although higher-grade lesions may be associated with longer symptom duration, this does not necessarily correlate with poorer clinical outcome.¹³

Imaging of OLT

Imaging has a significant role to play in the management of patients with OLT, including lesion identification, which may be occult on initial post-injury radiography (Figure 1), and delineating the abnormality in detail on cross-sectional imaging to help plan appropriate intervention.

Figure 1.

A 28-year-old male presenting with persistent pain following ankle injury. (A) anteroposteriorAP radiograph demonstrates soft tissue swelling (arrow) below the lateral malleolus, but no bone lesion. (B) Sagittal T1W TSE and (C) coronal STIR MR images demonstrate a small area of bone bruising (arrows) in the lateral margin of the talar dome and a swollen, oedematous deltoid ligament (arrowhead—C). The features are consistent with a Stage 1 lateral osteochondral lesions of the talus (OLT).

Radiography

Radiographs are the initial imaging investigation for OLT.³⁸ They are easily available, are cost effective and the first line of imaging bearing in mind that OLT are frequently associated with trauma or that patients may present later with suspected osteoarthritis.³⁹ In most cases, standard anteroposterior and lateral radiographs are considered sufficient, although others have advocated an additional mortise view.^1,27,40 OLT can appear as contour irregularity, radiolucency at the articular surface (Figure 2), or as a well-circumscribed osteochondral fragment demarcated from the adjacent bone by a radiolucent line (Figures 3 and 4).^38,41 Radiographs can also delineate the morphology of the lesions, with lateral OLTs typically demonstrating a shallow wafer-shaped appearances and medial OLTs usually appearing deeper and cup-shaped (Figure 4A).²¹ However, radiography still suffers from low sensitivity, particularly related to early stage lesions limited to the articular cartilage.³⁹ In a study by Verhagen et al,⁴² radiography missed 41% of OLT, the sensitivity being only 59% but with a specificity of 91%. In a study by Hepple et al¹⁵ of 18 patients with OLT, only 13 were visible on radiography. Nelson et al⁴³ showed a poor correlation between radiographic findings and low-grade lesions, with all lesions misclassified on radiography. Even when radiographs identify the abnormality, they demonstrate poor correlation with arthroscopic findings and are not accurate in terms of identifying lesion extent, any sequestration or sub-chondral cyst formation.^34,44

Figure 2.

A 52-year-old male presenting with post-traumatic ankle pain. (A) AP radiograph demonstrates an area of subtle radiolucency (arrow) in the medial margin of the talar dome. (B) Sagittal STIR MR image shows a corresponding area of bone marrow oedema (arrow).

Figure 3.

A 24-year-old male presenting with medial ankle pain. (A) AP radiograph demonstrates an irregular sclerotic lesion (arrow) in the medial talar dome margin with surrounding lysis. (B) Sagittal T1W TSE and (C) coronal STIR MR images show a medial osteochondral lesions of the talus (OLT) (arrows). The lesion appears hypointense suggesting that it is necrotic. There is thinning of the overlying articular cartilage and associated bone marrow oedema (arrowhead—C).

Figure 4.

A 15-year-old female investigated for right ankle pain. (A) AP radiograph demonstrates a well-defined oval bone fragment in the medial talar dome with surrounding lysis. (B) Sagittal T1W TSE, (C) sagittal STIR, and (D) axial SPAIR MR images show the fragment (arrows) in situ with surrounding bone marrow oedema (arrowheads—B-D). There is no fluid signal intensity (SI) between the fragment and the surrounding bone suggesting that the lesion may be stable. (E) Sagittal and (F) coronal CT arthrogram multiplanar reformats showing the fragment in situ (arrows) with no contrast medium extending between the fragment and host bone, confirming a stable lesion.

MRI and MR arthrography

In patients with clinically suspected OLT and normal radiographs, MRI is the imaging technique of choice since it provides excellent visualization of the articular cartilage, sub-chondral bone and adjacent soft tissues (Figures 1-4).^45,46 MRI has a sensitivity of up to 96%, specificity of 96%-100%, positive predictive value (PPV) between 89% and 100% and negative predictive value (NPV) between 88% and 99%.^36,42 There is close correlation of MRI findings and grading with arthroscopic findings, along with good interobserver agreement.³⁶ MRI also demonstrates a sensitivity of 97% and specificity of 100% for identifying unstable lesions and is superior to CT for identifying early stage lesions limited to the articular cartilage.^30,47,48 The perceived disadvantage of MRI is due to the presence of bone marrow oedema in the acute stage (Figure 5), which may lead to overestimation of lesion size or obscure lesion margins, information which may be required for surgical planning.^42,49 Imaging the abnormality in all 3-planes with small interslice gap allows for accurate measurement and staging.^4,50–53

Figure 5.

A 32-year-old male with persistent post-traumatic ankle pain. (A) Coronal PDW FSE and (B) sagittal STIR MR images of the left ankle demonstrate a transverse sub-chondral fracture line (arrows) and extensive bone marrow oedema (arrowheads—B) in the medial talar dome.

MR arthrography is superior to conventional MRI.⁵⁴ Injection of intra-articular contrast can lead to more accurate staging of lesions and allow differentiation between Stages II and III lesions according to the Berndt and Harty classification by demonstrating fluid around the OLT.^55,56 The accuracy of MR arthrography for detecting OLT is up to 88%,⁵⁶ but the procedure carries a small risk of infection, as well as potential allergic reaction to contrast medium.

CT and CT arthrography

CT also clearly demonstrates the location and extent of an OLT (Figure 6). It can accurately define sub-chondral cysts and has been advocated as an essential part of pre-operative planning.^44,49,57 The main drawback of CT is reduced sensitivity compared with MRI. While CT has a specificity of 99%, with PPV of 96%, and NPV of 94%, its sensitivity is reported as 81%.⁴² Reduced sensitivity is due to low-grade lesions limited to the articular cartilage, which are not identifiable on CT. However, injection of intra-articular contrast increases the accuracy of the technique (Figure 4). CT arthrography (CTA) increases the accuracy of detection of OLT to between 90% and 92% and is considered a superior technique compared to standard MRI.^56,58 CTA can also detect tissue growth at the injury site following microfracture, although this assessment is not related to clinical outcome.⁵⁹ However, the procedure carries a small risk of infection, allergic reaction, and radiation dose.

Figure 6.

A 23-year-old male with medial left ankle pain. (A) Sagittal and (B) coronal CT MPR of the left ankle showing a shallow elongated medial osteochondral lesions of the talus (OLT) (arrows). The fragmented sclerotic appearance of the lesion would suggest that it is necrotic. (C) Sagittal T1W TSE and (D) STIR MR images demonstrate linear fluid signal intensity (SI) at the junction of the lesion and host bone suggesting that the fragment is unstable.

Nuclear medicine

Bone scintigraphy demonstrates increased uptake at the OLT, increased activity being seen in all 3-phases although stable lesions do not show uptake in the blood-flow phase.^27,48 Uptake in blood-pool phase is 100% sensitive and 83% specific for identifying an unstable OLT, while increased uptake in the late-phase also significantly corresponds to an unstable lesion.^48,60 Despite excellent sensitivity, the technique is no longer utilized in the management of OLT because of a lack of specificity, radiation exposure, and the availability of MRI.³⁸

On SPECT-CT, the functional component using bone-specific radioactive tracers demonstrates increased uptake while the CT component will demonstrate the anatomy (Figure 7). A study by Leumann et al⁶¹ showed a change in management in 48%-52% of patients with OLT based on information provided by SPECT-CT (Figure 8). Meftah et al⁶² demonstrated that patients with activity on SPECT-CT more commonly underwent intervention compared to those with no increased uptake who were treated conservatively. In patients with multiple lesions, SPECT-CT can also identify the symptomatic lesion due to its activity and help in surgical planning by demonstrating the extent of the lesion. Hence, while it is not considered the primary imaging modality, there is some evidence that SPECT-CT can act as an adjunct in surgical planning and has therefore been recommended as part of a comprehensive assessment of the abnormality.⁶³

Figure 7.

A 32-year-old male imaged for left ankle pain. (A) Coronal SPAIR MR image demonstrates a medial osteochondral lesions of the talus (OLT) manifest by a sub-chondral cyst (arrow) with associated bone marrow oedema (arrowhead). (B) Coronal SPECT-CT MPR demonstrates the cyst with a sclerotic margin (arrow) and increased uptake in the surrounding bone.

Figure 8.

A 50-year-old female imaged for poorly defined left ankle/foot pain. (A) Coronal SPAIR MR image shows a stage 1 medial osteochondral lesions of the talus (OLT) (arrow). (B) Sagittal STIR MR image shows a sub-chondral cyst in the intermediate cuneiform (arrow) with mild associated reactive marrow oedema. (C and D) Sagittal SPECT-CT MPRs show activity related to the cuneiform cyst (arrows), but no activity related to the OLT (arrowhead—D).

Role of imaging in management planning

The role of imaging is to diagnose the lesion and to inform an appropriate treatment pathway. While important factors for clinical decision making include skeletal maturity, lesion size, and evidence of instability, other characteristics of the lesion, such as integrity of the overlying cartilage and sub-chondral changes, are also important to help decide the appropriate type of surgical intervention.³⁸

There is a higher healing potential for OLT in children, with studies showing significant partial or complete union and good clinical outcome.^28,64 In adults, OLT rarely heals. Cross-sectional imaging, such as MRI or CT, is important in symptomatic patients who have failed conservative therapy or in those with displaced fragments, to assess lesion size. Choi et al demonstrated that defect size was the single most important prognostic factor in OLT, and that factors, such as age, symptom duration, history of trauma, or lesion location, had no association with clinical failure.⁶⁵ The choice of intervention is usually dictated by OLT size, with lesions measuring <10 mm in diameter, <100 mm² in cross-sectional area and <5 mm in depth considered ideal for reparative procedures, and larger lesions being suitable for replacement procedures. However, others have advocated that lesions up to 15 mm in diameter or measuring up to 150 mm² can also be treated with reparative techniques.^65–70 A length of 15 mm or cross-sectional area of 150 mm² are considered a critical cut-off point, with poor clinical outcome associated with larger lesions undergoing debridement and microfracture.^65,71 Autologous osteochondral transplantation can be performed for lesions <2 cm², while matrix-induced autologous chondrocyte implantation (MACI) can be performed for defects measuring >2 cm².⁷² Similarly, a bone marrow derived cell transplant technique is used for lesions >1.5 cm² in area and <5 mm deep.⁷³

OLT stability/instability can be assessed on imaging. On radiographs, lesions measuring 0.8 cm² are likely unstable and stable lesions typically measure <0.2 cm², while 56% of OLT measuring 0.2-0.8 cm² are unstable. Another sign of instability on radiography is the presence of sclerotic margin measuring ≥3 mm, with 42% of lesions with sclerotic margins measuring 0-3 mm demonstrating instability. The presence of an ossified centre on radiography (Figures 3 and 4) is not significantly associated with instability.⁴⁸

On MRI, T2-weighted (T2W) sequences are important in identifying the 4 signs of instability as described by Smet et al.⁴⁷ The commonest sign of instability is a high-signal intensity (SI) line at the interface between the OLT and native bone, seen in 72% of cases (Figure 6). This high signal most likely represents fibrovascular granulation tissue.⁴⁵ However, this finding is not reliable in differentiating stable from unstable lesions in children.⁷⁴ Other signs of instability on T2W images include a > 5mm rounded high SI region deep to the lesion (Figure 7), this representing a cyst, which may either be due to intrusion of joint fluid into the cancellous bone or cancellous bone resorption, a focal defect in the articular cartilage and sub-chondral bone region measuring 5 mm or more in width representing partial or complete displacement of the lesion into the joint, and a high SI line traversing the articular cartilage and sub-chondral bone into the lesion.^47,75 These 3 signs of instability are seen in 22%-31% of patients with an unstable OLT, but ∼ 50% of patients demonstrate only 1 of the 4 signs of instability on MRI.⁴⁷ MRI has proven excellent for detecting an unstable OLT, with sensitivity ranging between 92% and 97% and specificity of 90%.^47,48 A displaced OLT is inherently unstable. Conversely, stable lesions do not demonstrate discontinuity of the hyaline cartilage.⁴⁸

Continuity of cartilage and the presence of sub-chondral cystic changes are also important imaging features for management planning. Large cystic OLT can be treated with osteochondral autograft transfer system, combined cancellous allografts inserted into the sub-chondral cysts, whereas cystic lesions with intact cartilage can be treated with drilling.⁷⁶ Studies have shown that the cartilage from the OLT is viable and can be used as a chondrocyte source for procedures, such as ACI or MACI.^77,78 However, despite extensive literature on OLT, there are no studies assessing the viability of the lesions on MRI imaging. This may be since most lesions are either treated conservatively, while others undergo either debridement or some form of transplantation. Taranow et al³² in their classification stated that cartilage condition can be suggested by MRI, although final assessment was done on arthroscopy. They defined viable and intact cartilage on arthroscopy as Grade A, with breached and non-viable cartilage as Grade B (Table 1). Studies have shown that these lesions, even when chronic, can serve as a harvest site of chondrocytes for further procedures, such as ACI or MACI, and therefore likely remain viable long term.^77,78

Conclusions

The term OLT describes an abnormality of the cartilage and sub-chondral bone, which is predominantly associated with trauma. Treatment varies from conservative management to surgery, with multiple techniques available. It is important to be familiar with the pre-operative appearances to identify lesion characteristics which inform treatment decisions.

Funding

None declared.

Conflicts of interest

None declared.