Introduction
The term osteonecrosis was first described by Ahlback et al. in 1968. The original description involved the spontaneous appearance of what was thought to represent focal osteonecrosis of the femoral condyle. After the hip, the knee has been determined to be the second most common location affected. Osteonecrosis has subsequently been separated into three distinct groups: spontaneous, secondary and postarthroscopic. Spontaneous osteonecrosis is the most common form, whereas postarthroscopic osteonecrosis is infrequently seen. Secondary osteonecrosis is the only type that likely represents an avascular process. Our understanding of the three types of osteonecrosis continues to evolve as newer studies attempt to further expose the causes.
Spontaneous Osteonecrosis
Spontaneous osteonecrosis of the knee (SONK) is the most common type and often affects patients in their sixth decade of life or later. The exact incidence is not entirely known, but studies suggest the incidence to be as high as 3.4% in patients older than 50 years and 9.4% in patients older than 65. However, the prevalence is likely underestimated as a result of patients being undiagnosed and progressing to end-stage osteoarthritis by the time of presentation. The vast majority of cases are unilateral and involve the medial femoral condyle with a male-to-female ratio of 1:3 to 1:5. , Cadaveric vascular studies by Reddy and Frederick have demonstrated a watershed area of the medial femoral condyle that has been thought to contribute to the propensity for SONK to affect the medial femoral condyle.
Aetiology and Pathogenesis
Historically the disease was thought to be initiated by a vascular insult, resulting in diminished microcirculation and increased intraosseous pressures and ending with ischemia and necrosis of the subchondral bone. However, histopathological studies suggest an insufficiency fracture mechanism. Hatanaka et al. examined a surgically resected medial femoral condyle specimen from a patient with diagnosed SONK. Their analysis revealed a linear fracture line paralleling the subchondral endplate, callus formation with granulation tissue on both sides of the fracture and fracture-related bone debris at the osteochondral side of the fracture. The authors concluded that their findings suggested that SONK was the result of a subchondral fracture rather than primary osteonecrosis. Similarly, Yamamoto and Bullough described insufficiency fractures as the primary occurrences leading to osteonecrosis after their histopathologic evaluation of 14 patients with SONK.
As a result of the subchondral insufficiency fracture theory, a growing level of interest has developed to look at the potential role meniscus tears play in the advancement of this pathological process. The meniscus functions to maintain joint kinematics by dissipating joint reaction forces while protecting the articular cartilage and underlying subchondral bone. A disruption of the posterior medial meniscus root has been shown to result in a loss of hoop tension, increased contact pressures, and subsequent alteration of normal knee biomechanics, mimicking the biomechanics seen in total meniscectomy. Therefore, tears of the meniscus may increase joint reaction forces and lead to subchondral insufficiency fractures and subsequent bone marrow oedema of the knee. Patients with low bone mineral density and vitamin D deficiency are at high risk for this process. A comprehensive review found 21 of 26 articles (80.7%) connecting the meniscus in the development of SONK with the medial meniscus and posterior medial meniscus root tears more commonly involved ( Fig. 25.1 ). Five studies reported posterior medial meniscus root tears in association with the development of SONK. Yamagami et al. noted 62.2% of patients who underwent surgery for SONK had medial meniscus posterior root tears. Similarly, Robertson et al. observed these posterior medial meniscus root tears in 80% of patients with SONK and attributed the condition to femoral condyle overload due to the tear. As a result of this work in elucidating the true aetiology, there is a current trend towards replacing the term SONK with subchondral insufficiency fractures of the knee (SIFK).
Clinical Presentation and Diagnosis
Patients with SONK will typically present with a sudden onset of medial knee pain without a history of trauma. Pain with weightbearing and at night is common. Tenderness over the medial femoral condyle is often present on examination. Routine radiographic studies include a standing anteroposterior (AP), lateral, Merchant and 45-degree flexion posteroanterior (PA) view. The radiographs will be normal in the early course of the disease but may reveal subchondral radiolucencies or flattening of the femoral condyle in the later stages of disease. Magnetic resonance imaging (MRI) is the imaging modality of choice to detect early changes and to determine the extent of disease. The two key findings on MRI include oedema, which indicates stress, and insufficiency fracture with collapse of trabeculae ( Fig. 25.2 ). These findings often present as a subchondral area of low signal intensity on T2 weighted images, focal epiphyseal contour depressions and lines of low signal intensity deep in the affected condyle. The most common classification systems are the Ficat and Arlet, Aglietti and Koshino, which was the first classification for SONK described in 1979. There are four stages in the Koshino staging system, which are based on clinical and radiographic findings. A patient who has knee symptoms but normal radiographs is classified as stage I. Stage II includes flattening of the weightbearing area with subchondral radiolucencies surrounded by osteosclerosis. Stage III reveals extension of the radiolucencies around the affected area and subchondral collapse. Stage IV is the degenerative phase with osteosclerosis and osteophyte formation around the condyles. Prognostic value has been placed on the size of the lesion, with an increased risk of osteoarthritis with increasing size.
Treatment
The treatment options for patients with SONK include operative and nonoperative options and depend on the stage of the disease, associated structural abnormalities and symptoms. The size of the lesion, which is taken by measuring the greatest width in the AP radiograph and the greatest length in the lateral radiograph, can also help guide management. Small lesions (<3.5 cm 2 ) have a high potential to respond well to nonoperative management and heal. On the other hand, larger lesions (>5 cm 2 ) have a higher likelihood of failure with nonoperative treatment and tend to progress to advanced osteoarthritis. Another method to assess size involves calculating the width of the lesion in the AP radiograph as a percentage of the entire affected condyle width. With this method Lotke et al. determined that small lesions involving 32% of the condyle responded well to conservative treatment, whereas large lesions involving more than 50% of the condyle progressed rapidly to collapse and required arthroplasty.
Nonsurgical management
The initial management for most small precollapse lesions consists of rest, activity modification, antiinflammatory medications, physical therapy, protected weightbearing and bracing. Unloader braces function to redistribute joint reaction forces and create a more favourable biomechanical environment for healing of the subchondral bone to occur. Pharmacological agents such as vitamin D and bisphosphonates should be considered to maximise the healing potential of the subchondral region. Vitamin D deficiencies need to be restored to normal levels. Bisphosphonates are thought to decrease the resorption of bone while allowing healing to occur and reducing the chances of subchondral collapse. The goals of nonoperative treatment are to control symptoms and reduce stress on the subchondral bone to maximize healing. Nonsurgical management has been associated with favourable results when initiated during the early precollapse stages of small lesions with greater than 80% of patients in most studies experiencing success without progression of the disease process. , , However, Nakayama et al did find that the presence of varus deformity was significantly associated with a poor prognosis resulting in progressive deformity and prolonged disability.
Other areas of nonoperative management being studied are hyperbaric oxygen chamber therapy (HBOT) and pulsed electromagnetic field (PEMF) therapy. Increased levels of reactive oxygen species occur with HBOT and trigger a set of responses that lead to increased vascularisation and the modulation of impaired proinflammatory cytokine productions. A retrospective study of 37 SONK patients treated with HBOT documented improved Oxford Knee scores in 86% of patients after 30 sessions and in 100% of patients after 50 sessions. In addition, MRI 1 year after HBOT completion demonstrated resolution of oedema at the femoral condyle in all but 1 patient. In another study, 28 patients with symptomatic Kashino stage 1 SONK were treated with local electromagnetic field therapy and followed for 24 months. The patients were treated 6 hours daily for 90 days. At final follow-up, visual analogue scale (VAS), Tegner and EuroQol-5D (EQ-5D) scores and Knee Society Score (KSS) results all significantly improved compared with baseline. Furthermore, MRI evaluation at 6 months was favourable, with a significant reduction of mean total Whole-Organ Magnetic Resonance Imaging Score (WORMS) and mean femoral bone marrow lesion area. Only 4 of the 28 patients (14.3%) required total knee arthroplasty within 24 months of ending PEMF therapy because of persistent pain and symptoms.
Surgical management
Surgical intervention should be considered for patients who have failed 12 weeks of nonoperative treatment and for patients with large lesions that are at high risk for advanced collapse. Joint-preserving procedures include arthroscopy, core decompression, bone grafting, subchondroplasty and osteochondral allograft/autograft transplantation. Unicompartmental and total knee arthroplasty are reserved for failed joint-preserving procedures and advanced collapse disease.
Arthroscopy allows for management of associated intraarticular abnormalities and has been described to facilitate decompression of the subchondral lesion. Loose chondral flaps can be debrided and meniscal pathological conditions may be addressed. Meniscus root repair restores meniscus function and unloads the medial compartment, allowing for subchondral healing. Akgun et al. performed arthroscopic debridement and microfracture on 26 patients with SONK and followed them for a mean of 27 months. All patients failed nonoperative management before surgical intervention, and the average lesion size was 1.62 cm 2 . The authors reported a statistically significant improvement in Lysholm score at mean follow-up, and 71% of patients were able to return to strenuous sports with no or minimal limitation. Of note, 14 of the 26 patients (53.8%) were found to have meniscus tears, and partial meniscectomy was performed in those cases.
Core decompression has been described to help facilitate healing of the subchondral bone and prevent additional surgical treatment in precollapse disease. This technique typically involves extraarticular drilling under fluoroscopic guidance. It is meant to decrease intraosseous pressure and enhance microcirculation to the affected area of subchondral pathological conditions. Forst et al. achieved clinical improvement in 15 of 16 patients with early-stage SONK treated with core decompression. They reported improved mean KSSs at a mean follow-up of 35 months and normalisation of bone marrow signalling on MRI.
Subchondroplasty (SBC; Zimmer Knee Creations; West Chester, PA, USA) is a proprietary term to describe the technique of surgically injecting calcium phosphate into the trabeculae of subchondral cancellous bone. It is meant to reinforce the subchondral region and stabilise insufficiency fractures. A literature review reported on 164 patients from eight studies with bone marrow lesions who went on to surgical subchondral calcium phosphate injection. All studies demonstrated significant functional improvements despite 25% of patients in one of the studies with some type of persistent pain complaint. Furthermore, the largest series included in the review acknowledged a 70% reduction in conversion to total knee arthroplasty (TKA) at 2 years. Despite optimism regarding the technique, there is a lack of long-term outcome data.
Osteochondral autograft transfer (OAT) is a technique in which one or multiple osteochondral cylinders are harvested from a nonweightbearing area of the knee and implanted within the SONK lesion. This procedure is often limited to small lesions and can result in donor site morbidity. Duany et al. reported success in nine patients treated with this technique at a mean follow-up of 42 months.
Fresh osteochondral allograft (OCA) transplantation has demonstrated excellent efficacy in treating large SONK lesions of the knee. This procedure is a viable option for cases of early or late collapse without associated diffuse osteoarthritis. The disrupted chondral surface and abnormal subchondral bone are replaced with an appropriately sized and matched OCA. A case series of seven patients treated with OCA for large SONK lesions (mean 4.6 cm 2 ) of the medial femoral condyle reported no failures and improved subjective outcome scores at an average follow-up of 7.1 years. In addition, all patients were extremely satisfied with the results of the procedure.
High tibial osteotomy (HTO) is a common surgical procedure for the treatment of degenerative conditions of the medial compartment and can be used to treat SONK. It facilitates unloading of the lesion and can be combined with concomitant drilling or bone grafting. Takeuchi et al. followed 30 patients with SONK of the medial femoral condyle who underwent HTO combined with curetting and drilling of the lesion to a mean follow-up of 40 months. At final follow-up, American Knee Society Score and Function Score increased from 51 to 93 and 58 to 93, respectively. Osteotomy hardware was removed in 24 patients, and arthroscopic findings at that time revealed complete fibrocartilage fill of the previous SONK lesions.
Unicompartmental knee arthroplasty (UKA) may be an appropriate option for patients with advanced collapsed disease isolated to a single condyle or compartment. It can also be used as a salvage option after failed joint preservation surgery. Chalmers et al. reported a 5- and 10-year survivorship of 93% in 41 patients treated with UKA for primary osteonecrosis of the knee. Similarly, Heyse et al. revealed a survival rate of 93.1% at 10 years and 90.6% at 15 years, with 97.3% of patients reporting feeling satisfied (21.6%) or very satisfied (75.7%). TKA is the treatment of choice for advanced disease affecting more than one compartment.
Secondary Osteonecrosis
Secondary osteonecrosis of the knee affects patients younger than 45 years old and is often bilateral. As opposed to SONK, secondary osteonecrosis may involve both femoral condyles and the proximal tibia. More than 80% of patients will have bilateral disease or other joint involvement.
Aetiology and Pathogenesis
The cause of secondary osteonecrosis is multifactorial, and the pathogenesis remains unclear. The two most common risk factors, which contribute to a suspected 90% of cases, are alcohol abuse and corticosteroid use. Despite the known association between these risk factors and the disease process, the exact mechanism is not entirely understood. Based on a femoral head autopsy analysis it is hypothesised that corticosteroids cause proliferation of adipocytes, leading to increased intraosseous pressure and subsequent necrosis. It is also postulated that alcohol abuse causes fatty emboli from the liver to occlude vessels within subchondral bone. Other conditions linked to secondary osteonecrosis include various haematological and endocrine conditions such as sickle cell disease, systemic lupus erythematosus, Gaucher disease and myeloproliferative disorders. Sickle cell is thought to cause a direct occlusion to small perfusing blood vessels of the subchondral bone. Gaucher disease and myeloproliferative diseases are suspected to displace marrow and subsequently increase intraosseous pressures. All these pathways are thought to lead to ischemia and ultimate necrosis.
Clinical Presentation and Diagnosis
Patients with secondary osteonecrosis usually present with a gradual onset of pain. The pain may be bilateral, be localised to multiple areas and involve other joints. A thorough history is paramount in identifying potential risk factors. The physical examination is commonly nonspecific but may demonstrate tenderness over the femoral condyles or the proximal tibia. As with SONK, routine radiographic studies include a standing AP, lateral, Merchant and 45-degree flexion PA view. Radiographs can demonstrate late disease process findings such as subchondral collapse but will usually be normal during early stages. The two most common radiographic classification systems are the Kashino system and the modified Ficat and Arlet staging system, which has been adapted from the hip staging system. According to the modified Ficat and Arlet system, radiographs appear normal in stage I. Joint space is maintained but sclerosis of the condyles is observed with stage II. The characteristic finding in stage III is the ‘crescent sign’ representing subchondral fracture, and radiographs may exhibit early joint space narrowing. Stage IV is defined by joint space narrowing and subchondral collapse. Nonspecific bone marrow oedema is commonly seen on MRI, along with disease-specific serpentine lesions that are well delineated ( Fig. 25.3 ).
