Update in diagnosis, treatment, and prevention of osteochondritis dissecans of the capitellum

Abstract

Osteochondritis dissecans of the capitellum is debilitating and is a potentially sports career-ending injury in a young and athletic population. Osteochondritis dissecans typically occurs in patients between the ages of 10 and 24 years, and boys are more commonly affected than girls. Conventional radiographs have low diagnostic accuracy, and magnetic resonance imaging (with or without contrast) or computed tomography may aid in accurate diagnosis. The primary indication for non-operative treatment is the presence of an intact cartilage cap on magnetic resonance imaging, indicating a “stable lesion.” However, if operative treatment is necessary, various surgical procedures are available when operative treatment for an osteochondritis dissecans of the capitellum is considered, including open or arthroscopic removal of loose bodies, with or without microfracturing, fragment fixation, osteochondral autograft transplantation, and osteochondral allograft transplantation. The decision-making process for selecting the appropriate treatment considers factors such as the patient's characteristics, functional limitations, and lesion morphology.

Introduction

Osteochondritis dissecans (OCD) was originally described as an inflammatory pathology affecting cartilage and bone, resulting in localized necrosis and fragmentation.^1,2 However, since the original description by König in 1889, ¹ the presence of an inflammatory component has never been demonstrated. Osteochondral defects predominantly occur in the knee, ankle, and elbow. ³ OCD lesions in the elbow are predominantly located at the capitellum, with rare occurrences in the trochlea.^4,5 OCD of the elbow is a debilitating and potentially sports career-ending injury in a young and athletic population if not recognized early and treated adequately. This article addresses the diagnosis, treatment, and prevention of capitellar OCDs.

Pathogenesis and demographics

OCD of the elbow typically affects adolescent athletes engaged in repetitive overhead or upper extremity weight-bearing activities (e.g. baseball, tennis, volleyball, lacrosse, javelin throwing, weightlifting, and gymnastics). The prevalence of radiologic OCD of the capitellum is 3.4% in adolescent baseball players.^6,7 Patients with symptomatic OCD usually present between the ages of 10 and 24 years, with boys more commonly affected than girls (80%/20%). ⁷ The dominant elbow is more frequently affected than the non-dominant elbow, with bilateral OCDs reported in up to 8% of patients. ⁸

The poor vascularization and abnormal radiocapitellar compressive forces put the capitellum at risk for developing OCDs.^9,10 The blood supply of the capitellum is provided by two end arteries (branches of the radial recurrent and interosseous recurrent arteries) running from posterior to anterior. ¹¹ Blood supply of the capitellum may be disrupted by repetitive microtrauma or a single traumatic event leading to subchondral bone injury. Repetitive compressive forces on the capitellum are generated by either large valgus stress or direct axial compression on the lateral compartment of the elbow. Based on a cadaveric study, Rotman et al. ¹² recently proposed a hypothesis that laxity in the proximal radioulnar joint could cause radial head lag during throwing, resulting in a malarticulation between the radius and capitellum. This malarticulation could lead to abnormal forces on the capitellum and potentially result in the development of OCD. ¹²

Differential diagnosis

It is important to differentiate between capitellar OCD and Panner's disease because of the differences in natural history and treatment strategies between these pathologies. ¹³ However, the term Panner's disease is often misused and conflated with OCD. ¹³ Panner's disease is a self-limiting disease occurring in children aged 4 to 12 years and is characterized by ischemia and necrosis of the capitellar epiphysis, followed by regeneration and recalcification in 1 to 2 years after the initial presentation.^13,14 On the other hand, OCD is typically an isolated lesion and affects patients ranging from 10 to 24 years. To differentiate between the two conditions in patients who are of age to be eligible for both, examining the characteristics of the lesion can be helpful. Panner's disease is characterized by subchondral stress fractures in the capitellum, while OCD presents as osteochondrosis with a single lesion.

Furthermore, medial epicondyle apophysitis (MEA), also known as little league elbow, shows some similarities in the clinical presentation, such as pain and limited range of motion in the elbow joint. The main distinguishing factor is the location of the lesion; MEA involves the medial epicondyle of the humerus, and OCD primarily the capitellum. ¹⁵

In addition, it is crucial to consider that other conditions besides Panner's disease and MEA can lead to similar symptoms. These include rheumatoid or septic arthritis, osteoarthritis, bone cysts, lateral epicondylitis, plicae, radial head fracture, and epicondylar avulsion fracture in older patients.

Clinical presentation

Early recognition of capitellar OCD is essential to prevent progression. Patients with capitellar OCD initially present with activity-related tenderness and pain. Some will have swelling and stiffness caused by reactive synovitis of the elbow. Mechanical symptoms of locking or catching caused by intra-articular loose bodies may present in later stages. Some patients present with a decreased range of motion, and a fixed flexion deformity of 15° to 20° is not uncommon. Pain in the lateral compartment with radiocapitellar compression while rotating the forearm may be suggestive of capitellar OCD. However, physical examination is not very specific for capitellar OCD. A high index of suspicion for capitellar OCD in a young athletic population is essential to ensure a timely and accurate diagnosis.

Radiology and classification

Plain anteroposterior radiographs with the elbow in 45° of flexion provide the best visualization of a possible capitellar OCD lesion with conventional imaging.^16,17 Radiographic signs of an OCD are flattening or lucency of the capitellum, a focal defect of the bony articular surface (Figure 1), and loose bodies. ¹⁸ Given the moderate inter-rater agreement, radiographs should be used with other imaging modalities to diagnose OCD. However, it can be used for follow-up to estimate the size of the OCD. ¹⁹

Figure 1.

Conventional bilateral plain radiograph: different signs of osteochondritis dissecans (OCD) with flattening of the capitellum (left) and lucency (right) (courtesy of MoRe Foundation).

Ultrasound (US) of the elbow is of limited value in diagnosing capitellar OCD because the capitellum is partially obscured by the radial head, but reactive synovitis may sometimes be seen. ² However, recently, Shinohara et al. ²⁰ evaluated the use of an artificial intelligence algorithm to diagnose OCD in young baseball players using ultrasonography. The study included 40 elbows with confirmed OCD and 100 US images per elbow. The results showed that the algorithm can accurately (0.87) identify OCD in US images, focusing on irregularities or discontinuities in the subchondral bone surface. ²⁰

Computed tomography (CT) is as reliable as magnetic resonance imaging (MRI) in the latter stages of the disease (Figure 2(a): CT and Figure 2(b): three-dimensional (3D) CT) and is more sensitive (Figure 3) to detect OCD fragmentation and secondary changes, such as loose bodies, radial head abnormalities, and osteophytes.^21–23 However, especially in this young patient group, radiation exposure is a drawback.

Figure 2.

(a) CT and (b) 3D reconstruction of an advanced-stage capitellar OCD lesion (courtesy of MoRe Foundation).

CT: computed tomography; 3D: three-dimensional; OCD: osteochondritis dissecans.

Figure 3.

CT is particularly helpful in advanced stages of OCD with an improved visualization of loose bodies, when compared to MRI (courtesy of MoRe Foundation).

CT: computed tomography; OCD: osteochondritis dissecans; MRI: magnetic resonance imaging.

Both MRI (Figure 4) and magnetic resonance arthrography (MRA) help identify bony edema, found in the early stages of OCD.^22,24 However, the main reason for advanced imaging of the elbow is to determine OCD lesion stability, as this predominantly dictates treatment. MRI is particularly valuable in determining OCD lesion stability and viability. ²⁵ MRA (Figure 5) is particularly helpful in detecting loose bodies, but it does not improve the sensitivity for detecting cartilage lesions. ²⁶

Figure 4.

MRI of an OCD lesion with bone edema at the capitellum but intact cartilage. This image highlights the challenge of accurately classifying OCD. Although the cartilage appears intact in the image, the presence of a bone edema of this size indicates instability (courtesy of MoRe Foundation).

MRI: magnetic resonance imaging; OCD: osteochondritis dissecans.

Figure 5.

MRA of an unstable OCD lesion with decreased signal intensity in the subchondral bone and a complete fissure in the cartilage (courtesy of MoRe Foundation).

MRA: magnetic resonance arthrography; OCD: osteochondritis dissecans.

Imaging features associated with a worse prognosis are a displaced fragment, physeal closure of the capitellum or lateral epicondyle, and irregular contours of the articular surface or a high signal interface on T2-weighted MRI. Although none of these predictors have 100% sensitivity, they can guide decision-making. ²⁷ Various classification systems have been described (Table 1). ²⁸ Each classification system provides valuable information in a specific setting and can be used in conjunction with other clinical findings, such as the size of the affected area and the degree of joint instability, to determine the appropriate treatment for each patient. Most classifications serve as an aid to a specific type of imaging to classify the lesion preoperatively: Minami et al. ²⁹ described a classification based on anteroposterior radiographs. Itsubo et al. ³⁰ introduced a T2-weighted MRI staging system. The International Cartilage Repair Society has proposed an arthroscopic classification system for OCD lesions. ³¹ Recently, Sayani et al. ³² described a four-stage grading system that is based on unified existing validated classification systems and can be used to classify the severity of the condition and to guide treatment decisions. Wang et al. ³³ have introduced an MRI-based scoring system designed for postoperative evaluation of the healing after single-plug osteochondral autograft transplantation (OAT). No single classification is universally accepted or used to guide treatment decisions or predict prognosis in OCD.

Table 1.

Classifications for grading osteochondritis dissecans (OCD) lesions.

Preoperative grading classifications
Minami classification 29
Grade I: Lesions demonstrated a shadow in the middle of the capitellum
Grade II: Lesions had a clear zone between the lesion and the adjacent subchondral bone
Grade III: Displaced or detached fragment

Prevention

Capitellar OCD has a prevalence of 3.4% in (adolescent) baseball players. Several preventive measures have been developed to avoid injury in this vulnerable group. Recommendations primarily focus on the prevention of ulnar collateral ligament lesions,^34–36 but the same valgus extension stresses that cause ulnar collateral ligament injuries in the overhead athlete are also responsible for repetitive compressive forces on the lateral elbow compartment, potentially liable for the development of capitellar OCD.

Several risk factors for capitellar OCD have been identified and can be defined as either non-modifiable or modifiable. Non-modifiable factors include age and height. Modifiable factors include body mass index, coaching, altered throwing mechanics, frequency, and volume of throwing activities, and individual physical characteristics, including decreased flexibility, fatigue, weakness of upper and lower limbs, rotator cuff imbalance, and altered core stability.^34,35,37 It needs no discussion that the focus should be on modifiable factors to avoid future OCD of the capitellum.

Non-operative treatment

Despite the abundance of literature on the treatment of capitellar OCD, an evidence-based treatment algorithm has not yet been developed. Treatment options for capitellar OCD vary from a total cessation of any aggravating activities to immediate surgery. The primary indication for non-operative treatment is the presence of an intact cartilage cap because this is indicative of a stable lesion. These patients generally do well with conservative measures. ³⁸ The result of non-operative treatment of patients with disruption of the articular cartilage cap, radial head enlargement, or advanced skeletal age of the throwing side relative to the non-throwing side is less favorable.^39–41

Non-operative treatment consists of relative rest or sports restriction. Loading of the elbow is not permitted. In addition, non-steroidal anti-inflammatory drugs and/or a short course of immobilization may be considered.^41,42 Protection and treatment of OCD lesions with an off-loading hinged elbow brace has been described, but its use is not widespread. ⁴³ Non-operative management with restriction of activities is continued for a minimum of 12 weeks until symptoms resolve. We prefer to repeat the MRI at 3-month intervals until healing occurs. When the MRI demonstrates signs of healing of the lesion (decreased edema on T2 sequences), progressive loading of the elbow may be initiated.

An important aspect of conservative treatment of capitellar OCD in these aspiring athletes is the counseling of the patient and their parents, trainers, and coaches. It should be emphasized that adherence to these guidelines is necessary to decrease the chance of progression of the OCD lesion, after which surgical intervention may be inevitable. Unfortunately, despite perfect compliance with the guidelines, it cannot be guaranteed that the OCD lesion will not progress to a higher stage.

Takahara et al.^17,44 reported a 50% success rate of conservative treatment involving activity restriction after a follow-up of 12 years. These poor results led to the conclusion that capitellar OCD has a limited tendency to heal with conservative treatment, especially in patients with unstable lesions. A later study by Takahara et al. ⁴⁵ stated that activity restriction might generally be insufficient for treating capitellar OCD and that cast immobilization for around 4 weeks followed by a splint for approximately 7 weeks appeared to be more promising for stable OCD cases, with a success rate of 92% in terms of complete healing and a mean return to sport within 6 months. Sakata et al. ⁴⁶ showed a significantly greater return to sports for early stage capitellar OCD treated with early motion therapy compared to advanced-stage capitellar OCD. Bradley and Petrie ⁴⁷ formulated three predictors for full recovery with a complete return to sports: (a) an open capitellar growth plate; (b) localized flattening or radiolucency of the subchondral bone; and (c) full range of motion. Mihara et al. ⁴¹ also found that the spontaneous healing rate of OCD is higher in patients with open capitellar growth plates. Conversely, it has been shown that unstable OCD lesions in patients with closed growth plates have very limited healing potential.^41,47,48

Operative treatment

Various surgical procedures are available, including open^17,42,48 or arthroscopic removal of loose bodies, with or without microfracturing of the OCD lesion,^49–56 OCD fragment fixation,^{42,53,57–60} and OAT.^61–67 Different surgical options are discussed in more detail below.

Fragment removal

Depending on the size of the lesion, long-term results of open fragment removal were discouraging, with only 40% to 49% returning to sports.^17,42,48 The introduction of arthroscopic techniques has significantly improved these results.^42,54,68

Arthroscopic surgery is typically performed in the lateral decubitus position, although other positions are possible. A 30° or 70° arthroscope can be used to visualize the OCD lesion. Our arthroscopic approach involves beginning with an assessment of the anterior compartment, as this is not a universally adopted practice (Figure 6). The anterior capitellum is inspected and is usually normal. Often there is some mild synovitis, and a synovectomy is performed through a lateral portal. Any loose bodies are removed. Next, a posterolateral portal is made, and the posterior compartment is assessed. If present, loose bodies are usually found in the posterolateral gutter or olecranon fossa. Assessment of medial elbow laxity in patients with OCD is important because it may indicate the presence of valgus extension overload syndrome (VEOS), which is a condition that can cause OCD in children and adolescents.^69,70 By identifying medial elbow laxity, clinicians can anticipate the possibility of other future problems associated with VEOS, such as posterior impingement and ulnar collateral ligament complaints. This factor is examined by performing a valgus stress test with the elbow in 70° flexion. A central posterior working portal is made under direct view, ensuring the ulnar nerve is protected. Loose bodies are removed, and other pathologies, such as posteromedial osteophytes or synovitis, are addressed. To visualize the posterior and distal aspects of the capitellum, the scope is directed into the radiohumeral gutter, and a needle is used to confirm the correct position of the soft spot portal. The lateral soft spot portal may be used as a working portal but can alternatively be used as a viewing portal. A smaller 2.7 mm arthroscope may be used to assess the lesion from this portal. Once adequate visualization of the OCD lesion has been obtained, it is graded. Unstable lesions are treated by removing loose fragments and establishing stable cartilage borders. It is often sufficient to shave onto bleeding bone (Figure 7), but microfracturing and subchondral drilling may also be performed (Figure 8). From the available literature, it is not possible to recommend between abrasion with a shaver, retrograde or antegrade drilling, or microfracturing of subchondral bone. Even though OCD is one of the most common indications for arthroscopy in pediatric patients, there is limited evidence for the use of elbow arthroscopy for OCD. Some studies with level IV evidence showed promising results that warrant further investigation.^71–73 However, it should be noted that patient selection plays a crucial role in determining the success of arthroscopic intervention. ⁷⁴

Figure 6.

Intraoperative arthroscopic view of the anterior compartment of the elbow, showing a loose body from osteochondritis dissecans (OCD) (courtesy of MoRe Foundation).

Figure 7.

Arthroscopic posterior view from the radial gutter. The osteochondritis dissecans (OCD) lesion is debrided, using a shaver. Note some cartilage abrasion on the radial head (courtesy of MoRe Foundation).

Figure 8.

An icepick is introduced from the soft spot portal. Microfracturing is used if the subchondral bone is sclerotic, and insufficient marrow stimulation can be obtained from shaving only (courtesy of MoRe Foundation).

No clear correlation has been shown between the size of the lesion or the age of the patient and the outcome of arthroscopic surgery.^52,56 In the short and medium terms, arthroscopy has greatly improved the results of open debridement with respect to return to sports. Rahusen et al. ⁵⁴ showed a return to sports in 80% of patients following arthroscopic debridement and marrow stimulation by shaving to subchondral bone. Others also reported an 80% return to sports, albeit in more heterogeneous groups of patients.^49,51,52 Some authors have reported less favorable results. Ruch et al. ⁵⁵ only reported a 25% return to sports rate, although mechanical symptoms resolved in 11 out of 12 patients. Schoch and Wolf ⁵⁶ reported that only 4 out of 10 patients returned to sports but noted that it was unsure if patients stopped their sports activities due to elbow problems or due to other reasons.

Fragment fixation

In larger lesions, open fixation of unstable OCD fragments has yielded good results with varying methods of fixation: compression screws, bioabsorbable implants, cannulated screws, sutures, wires, nails, and pins.^{42,53,57–60} Takahara et al. ⁴² reported significantly better results with open refixation than fragment removal alone if the osteochondral defect was >50% of the capitellar articular width. However, Hennrikus et al. ⁵³ reported on 26 patients operated on over a 12-year period and found that patients aged older than 15.3 years or patients with a sagittal plane lesion width of more than 13 mm were found to be at higher risk for non-union. This suggests that refixation could be challenging for fracture fragments larger than 13 mm.

Excellent results have been reported with arthroscopic fragment fixation. Using a hybrid arthroscopic and mini open suture fixation, Koehler et al. ⁵⁸ showed healing of the fragment on MRI in all four patients at the 3-month follow-up. The authors did not comment on the size of the lesion. Uchida et al. ⁶⁰ reported on arthroscopic fixation of OCD lesions with bioabsorbable pins in 18 patients. Excellent results were found in 17 of 18 patients. All but one patient returned to sports, with five performing at a higher level. One patient changed from baseball to soccer, one stopped with sports altogether due to unrelated causes, and one needed revision arthroscopic loose body removal after 1 year. Takeba et al. ⁵⁹ used a similar technique of fixation with bioabsorbable pins in 13 patients. During the same procedure, four patients underwent additional removal of loose bodies, and in one patient, microfracturing was performed. All patients were able to return to play.

Osteochondral autograft transplantation

Patients with lesions >13 mm and involving the lateral margin of the capitellum may not benefit from refixation, due to the high chance of non-union.^53,66 In these patients, OAT may be indicated. Osteochondral plugs are most commonly harvested from the lateral femoral condyle of the knee, but costochondral grafts are also an option.^61,64,67 A potential disadvantage of an OAT procedure is donor-site morbidity, which can be up to 8% using knee grafts and 2% using rib grafts. ⁷⁵ At the moment, both harvesting and transplanting the osteochondral plug require an arthrotomy. However, a cadaveric study showed that an all-arthroscopic procedure is feasible. ⁶²

Based on their experience, Lyons et al. ⁶⁵ have developed an algorithm for the treatment of OCD lesions. Their algorithm indicates an OAT procedure when the fragment is not amendable to fixation and the lesion is greater than 1 cm. ⁶⁵ Despite the higher-grade lesions treated with OAT, good to excellent results are generally found in over 80%, and most patients are able to return to their previous level of sports.^63,66 In extreme cases, such as gross deformation of the capitellum or radius head, even OAT is no longer possible.

Osteochondral allograft transplantation

To avoid OAT-related issues as donor-site morbidity and mismatch of curvature and cartilage thickness, a fresh distal humerus allograft transplantation can be done. Although there is limited literature on the use of this technique for treating capitellar OCD, a study conducted by Mirzayan and Lim ⁷⁶ has demonstrated promising rates of return to sports (100%) and favorable functional outcomes. A study by Samuelsen et al. ⁷⁷ describes a surgical technique where the patient is positioned supine. An arthrotomy is made in the interval between the anconeus and extensor carpi ulnaris. After the arthrotomy, the OCD lesions are identified and measured. Based on these measurements, a suitable deep osteochondral plug containing the OCD is removed from the recipient. It is important that the donor plug harvest is of the same size and location to make sure it matches the recipient capitellum. Before final placement, the graft is irrigated and bathed in platelet-rich plasma. Possible disadvantages of osteochondral allograft transplantation are high costs and a risk of disease transmission. ⁶⁴

Postoperative rehabilitation

Postoperative rehabilitation depends on the type of surgery performed, but the objectives remain the same: reducing pain and swelling and restoring range of motion. A specialized physical therapist may supervise the postoperative rehabilitation.

The time to recovery and return to sports after arthroscopy is typically faster than after open surgery. Active-assisted motion exercises are started within 1 or 2 days after debridement. The range of motion is unrestricted and may progress as pain tolerates. Seo and Yoon ⁷⁸ stated that additional exercises targeting the rotator cuff and elbow flexors significantly improved muscle strength and range of motion compared with basic rehabilitation exercises following surgery for capitellar OCD in middle and high school baseball players. For patients treated with more extensive techniques such as refixation or OAT, the range of motion may be restricted for the first weeks to protect the reconstructed articular surface. Some surgeons advise bracing the elbow in the initial postoperative period. Resistance exercises are initiated 8 weeks after arthroscopic treatment and 12 weeks after open treatment. For overhead athletes, a throwing program is started 5 months after surgery once the athlete has no pain and a normal range of motion. Return to full participation in sports is usually advised at 6 months.

Prognosis

Although substantial evidence is lacking, non-operatively treated stable OCD lesions are considered to follow a more benign course. Patients with capitellar OCD may develop (early) degenerative changes in the elbow joint.^17,48 The relation between the size of the cartilage defects and the development of osteoarthritis in the long term has not been completely elucidated. ⁷⁹ Bauer et al. ⁴⁸ concluded that one-third of 31 patients with capitellar OCD had radiographic degenerative changes, and 42% complained of pain and/or reduced range of motion at a mean follow-up of 23 years. Larger lesions and older age at the time of onset seem to be associated with worse clinical outcomes and an increase in radiographic degenerative changes in the long term.^17,44,48,79 Currently, there is no evidence that surgical treatment of capitellar OCD improves long-term outcomes, or that repair strategies decrease elbow joint degeneration. ⁷⁹

Future treatment options and future research

Future research and guidelines should primarily be aimed at the prevention of OCD. Firstly, risk factors should be identified and logged (patient, elbow, and activity characteristics). An apparent genetic factor has not been found, but a predisposition may exist and could be used to identify the individual athlete at risk.

Basic research should focus on exposing the correct pathogenesis of OCD. For example, research on the intraosseous blood supply of the capitellum with micro-CT could give some additional insight into pathomechanism. If the pathomechanism is elucidated, specific growth factors or stem cells influencing OCD disease progression could be identified, which could pave the way for optimal treatment strategies.

Long-term data on the development of degenerative changes should be obtained for both conservatively and surgically treated patients. This data could serve as a baseline for comparing older and more recently developed techniques.

Summary

Patients with OCD typically present between the ages of 10 and 24 years. Although the pathogenesis has not been conclusively described, abnormal compressive forces on the elbow that compromise capitellar microvascularization are thought to be the underlying cause. Young athletes participating in overhead sports are particularly at risk for developing capitellar OCD.

Patients with early stage capitellar OCD often present with lateral-sided, activity-related elbow pain, whereas locking or catching may be present in later stages. Radiographs are unreliable. MRI is very helpful in diagnosing and staging the lesion. Additional CT images are useful in the presence of bony changes, such as loose bodies.

Stable lesions are best treated with relative rest and sports restrictions, with repeated MRIs at three-monthly intervals. Conservative treatment is usually unsuccessful in unstable lesions or patients with closed growth plates.

Operative treatment is indicated in unstable lesions or lesions accompanied by loose bodies. Arthroscopic debridement, with or without marrow stimulation, yields good results in the majority of patients. The treatment outcome is influenced by lesion size and location. Arthroscopic fragment fixation can be performed if the fragment is large enough, and low-evidence studies suggest that the results are excellent in selected patients. OAT may be indicated in lateral lesions or lesions >1 cm. Transplantation of osteochondral allografts could avoid problems associated with autografts, such as donor-site morbidity, but is associated with higher costs and a risk of disease transmission.

Unfortunately, little research has been performed focusing on the prevention of OCD. Good guidelines have been issued on the prevention of valgus overload syndrome. Similar advice could be used for throwing athletes to prevent OCD. More specific guidelines are needed for non-throwing athletes, such as gymnasts.