Ischemic Bone Lesions

Etiology

Ischemic bone lesions occur in several conditions, although they are frequently idiopathic. Several etiologic factors are associated with osteonecrosis: fracture and joint dislocation, systemic corticosteroid use and Cushing disease, alcohol abuse, sickle cell disease and other hemoglobinopathies, vasculitis, trauma, renal transplantation and osteodystrophy, radiation therapy, pancreatitis, gout, Gaucher disease, connective tissue diseases (e.g., systemic lupus erythematosus), caisson disease, and cytotoxic agents (e.g., vinblastine, vincristine, cisplatin, cyclophosphamide, methotrexate, bleomycin, or 5-fluorouracil). Several diseases including acquired immunodeficiency syndrome (AIDS) and severe acute respiratory syndrome (SARS) seem to be associated with an increased prevalence of ischemic bone lesions, although lesions could be related either to the diseases themselves or to their treatments. A Japanese survey of femoral head osteonecrosis estimated that 34.7% were due to corticosteroid use, 21.8% to alcohol abuse, and 37.1% to idiopathic mechanisms.

Prevalence and Epidemiology

The prevalence of ischemic bone lesions is unknown, mainly because many lesions are clinically silent or because they are part of more complex lesions such as displaced cortical fractures. The rate of symptomatic epiphyseal osteonecrosis of the hip is 2 to 4.5 cases per patient-year, with approximately 15,000 new cases reported each year in the United States. Femoral head osteonecrosis accounts for more than 10% of total hip replacement surgeries performed in the United States. A Japanese survey estimated that 2500 to 3300 cases of epiphyseal osteonecrosis of the hip occur each year.

There is no racial predilection, except for osteonecrosis associated with sickle cell disease and hemoglobin S and SC disease, which predominantly occurs in African and Mediterranean populations.

Age at onset of epiphyseal osteonecrosis, male-to-female ratio, and bilaterality depend on the underlying causes. Idiopathic femoral head osteonecrosis most often develops in male subjects age between 35 and 55 years, with a mean age of approximately 40 years. It is more common in men, with an overall male-to-female ratio ranging from 4 to 8 : 1. On average, osteonecrosis presents almost 10 years later in women than in men. In idiopathic osteonecrosis, the disorder is bilateral in 40% to 80% of cases. Presentation of symptoms may be asynchronous because progression to fracture may occur at a different pace.

In posttraumatic osteonecrosis, epiphyseal fracture that reveals the underlying clinically occult osteonecrotic lesion develops generally several months to 2 years after a displaced fracture or joint dislocation.

Spontaneous osteonecrosis generally occurs in elderly patients. Spontaneous osteonecrosis of the knee typically occurs in the medial femoral condyles of elderly women with medial meniscal root tear or in younger patients after large medial meniscectomy. Spontaneous osteonecrosis of a vertebral body typically develops at the thoracolumbar junction of elderly patients under steroid therapy, after radiation therapy, or after a spontaneous insufficiency fracture.

Clinical Presentation

Clinical presentation ranges from fortuitous discovery at MRI in totally asymptomatic patients to deep excruciating bone pain in patients with sickle cell crisis. Nonfractured systemic osteonecrosis is usually clinically occult. Symptoms generally develop asynchronously in different joints, mainly in the lower limbs in association with the development of fractures of the epiphyseal surfaces. Joint pain associated with the spontaneous fracture is usually worse at weight bearing.

In posttraumatic osteonecrosis, symptoms related to osteonecrosis are also generally absent. They may develop spontaneously several months after the fracture or joint dislocation, frequently in association with the development of a fracture secondary to the osteonecrotic lesion.

In spontaneous osteonecrosis, pain always seems to be present from the beginning, as there is no predisposing underlying lesion. Pain is usually stress modulated.

Pathophysiology

The pathophysiology of osteonecrosis remains poorly understood, and three different conditions are recognized ( eTable 72-1 ).

In systemic osteonecrosis (e.g., corticosteroid-induced osteonecrosis), impaired perfusion with subsequent necrosis of bone and marrow can be caused by several mechanisms, including thrombotic or embolic occlusion of blood vessel (e.g., fat embolism, sickle cell crisis, caisson disease), injury to vessel wall (e.g., vasculitis, connective tissue diseases such as systemic lupus erythematosus, radiation, infection), and increased pressure on the vessel wall (e.g., extravasated blood in marrow, inflammation caused by lipid accumulation in osteocytes, intraosseous hypertension from proliferating Gaucher cells in Gaucher disease). The term systemic osteonecrosis is chosen to stress the fact that a systemic marrow disease is at the origin of bone marrow necrosis ( eFig. 72-1 ). The marrow condition can be either primary (e.g., anemias) or secondary (e.g., corticosteroid-induced osteonecrosis). It can involve any bone segment and any marrow type (red or yellow marrow).

In posttraumatic epiphyseal osteonecrosis, anatomic disruption of the blood supply after bone fracture or joint dislocation can cause bone osteonecrosis, given the terminal blood supply of the epiphysis ( eFig. 72-2 ). In posttraumatic cortical osteonecrosis, displaced fragment of cortical bone may become necrotic after separation from normal periosteal vasculature. Direct trauma to the periosteal vessels with extracorporeal shock wave lithotripsy, minimally displaced distal tibial fractures in children, and skin loss can also cause cortical bone osteonecrosis.

The pathogenesis of s pontaneous osteonecrosis is still controversial. Nowadays, the most widely accepted hypothesis is that osteonecrosis results from fractures of a subchondral bone plate and/or the trabecular bone, which do not heal because of persisting joint movements ( eFig. 72-3 ). This subchondral bone plate fracture may lead to pseudoarthrosis and subsequent necrosis. Therefore, disruption of the microvasculature is limited to the immediate subchondral area of the epiphysis. An alternative hypothesis is that bone marrow edema is the initial lesion and causes secondary bone marrow ischemia. Most likely, both mechanical disruption of the microvasculature and increased marrow pressure lead to ischemia and subsequent osteonecrosis.

Whatever the mechanism, fracture is a key event in the physiopathogenesis and natural history of osteonecrosis. In systemic osteonecrosis and in posttraumatic osteonecrosis, fracture is generally a relatively late event because it is preceded by bone marrow osteonecrosis. In spontaneous osteonecrosis, fracture seems to be an early event because it could favor osteonecrosis.

Histopathology

At the microscopic level, cell necrosis may take several patterns, but coagulation necrosis is the most common pattern of necrosis in bones. It results in complete absence of osteocytes within the bone trabeculae, loss of adipocyte nuclei with lipid cyst formation, and death of hematopoietic cells ( eFig. 72-4 ).

An infarct is a localized area of necrosis in tissue resulting from reduction of either its arterial supply or venous drainage. The term bone infarct is generally used to describe a metaphyseal or diaphyseal ischemic lesion, but not an epiphyseal lesion, for cultural rather than medical reasons ( eFig. 72-5 ). The histopathologic changes that occur in infarcts merely depend on the type of marrow vasculature. In yellow marrow, as in epiphyses and metaphyses, arterial occlusion produces a bloodless or white infarct because of the poor vasculature of yellow marrow. Marrow remains fatty and demonstrates normal signal intensity on MR images ( eFig. 72-6 ). The reactive interface of fibrovascular tissue that progressively appears and surrounds the ischemic lesion is the hallmark of yellow marrow infarct (see eFig. 72-6 ). In red marrow, as in the spine, hemorrhagic or red infarcts may develop because of the rich vascular network of red marrow. Marrow becomes edematous, and MR images may demonstrate the marrow edema pattern.

Spontaneous (nontraumatic) fracture of an epiphysis is a key event in epiphyseal osteonecrotic lesions ( eFig. 72-7 ). The fracture may involve the plate of cortical bone that is covered by hyaline cartilage on the articular side (the so-called subchondral bone plate) and that covers the necrotic marrow on the other side. The subchondral bone plate fracture frequently starts at the margin of the underlying necrotic lesion. The fracture may also involve the cancellous bone within the necrotic epiphyseal segment. It frequently runs parallel underneath the subchondral bone plate itself (see eFig. 72-7 ), although it may run more deeply in large osteonecrotic lesions. In the vast majority of cases, this fracture never heals, because it extends within necrotic bone that cannot repair the fracture. Usually, the fracture does not extend into the adjacent normal bone. This fracture leads to a radioclinical condition characterized by pain and functional disability of the joint at clinical examination and irreversible deformity of the epiphyseal contour on radiographs ( eFig. 72-8 ). This condition subsequently leads to early osteoarthritis due to joint incongruency.

Diagnosis of Osteonecrosis

The diagnosis of ischemic bone lesions relies heavily on medical imaging. The alert clinician may suspect the diagnosis when facing a patient with acute and spontaneous articular pain and risk factors for ischemic lesions. Clinical examination has little diagnostic value except for accurate localization of the involved area (knee pain due to femoral head osteonecrosis).

Blood tests generally do not contribute to the diagnosis of the ischemic lesions, but it is of importance in the recognition of an eventual underlying disease (e.g., anemia, hyperuricemia, alcohol abuse).

MRI is the imaging modality of choice for the diagnosis of osteonecrosis. T1-weighted spin-echo images should first be obtained because of their sensitivity to the presence of any marrow changes. Normal T1-weighted images may be sufficient to rule out ischemic lesions in yellow marrow in the vast majority of cases. T2-weighted spin-echo or fat-saturated intermediate-weighted images are also necessary to stage the disease because they help to assess the overlying articular cartilage and the subchondral bone plate in epiphyseal osteonecrosis. Contrast-enhanced MR images generally do not contribute significantly to the assessment of ischemic bone lesions in adult patients, except in posttraumatic ischemic lesions, in which the interface may be lacking at the early stage, or when complications such as infection or tumor are suspected. MRI protocols should include both sides of the body, at least in patients with unilateral hip symptoms and suspicion of osteonecrosis, because of the high frequency of bilateral involvement in systemic epiphyseal osteonecrosis (see eFig. 72-4 ). Technical characteristics of the MR unit such as magnetic field strength do not influence its accuracy in the detection of ischemic bone lesions. Whole-body MR imaging may reveal the true extent of the systemic disease.

Rarely, ischemic bone lesions develop in red marrow areas. They may remain occult on T1-weighted SE images because of the frequently abnormal background marrow. Fluid-sensitive MR sequences or enhanced T1-weighted sequences become mandatory for their detection ( eFig. 72-9 ). Actually, these lesions frequently demonstrate a marrow edema pattern.

Radiographs are relatively insensitive for the detection of the early stages of ischemic bone lesions. At a more advanced stage, osteonecrotic lesion become visible because of the progressive sclerosis of their margins ( eFig. 72-10 ). Multidetector computed tomography (MDCT) has not been widely used for the assessment of ischemic bone lesions. A preliminary study suggested that MDCT could be valuable in the detection of subchondral bone plate fractures. MDCT could also contribute to the assessment of posttraumatic osteonecrosis treated with metallic hardware.

Bone scintigraphy using technetium diphosphonate has been used for decades in the assessment of articular pain, mainly because of the relatively high negative predictive value in the setting of symptomatic epiphyseal osteonecrosis. Its use has progressively decreased because of the availability of MRI and its well-known limitations due to its poor specificity and poor spatial resolution. There are limited literature data on the value of positron emission tomography (PET) in ischemic bone lesions. Ischemic bone lesions could remain occult on PET images except for slight marker accumulation in the joint capsule.

Imaging Features/Manifestations of the Disease