Fig. 7.1
Schematic drawing of the blood supply to the femoral head. (A) Superior retinacular vessels. (B) Inferior retinacular vessels. (C) Lateral epiphyseal artery. (D) Medial epiphyseal artery. (E) Superior metaphyseal artery. (F) Inferior metaphyseal artery. (G) Intramedullary vessels
Histologic and angiographic studies of femoral heads with osteonecrosis have identified consistent involvement of the superior retinacular and lateral epiphyseal vessels. Some attempt at vascular repair can be seen with ingrowth of new vessels from the stumps of occluded vessels and from other vessels in the region. However, this process is usually limited and is often blocked by the presence of necrotic material and collapsed bone (personal observation, MES).
Within hours of the vascular insult, death of marrow elements can be seen. Death of bone also takes place, but cannot be identified histologically until several days later when disappearance of osteocytes from their lacunae is noted (Fig. 7.2). Osteoclasts and phagocytic cells infiltrate the margins of the necrotic region and begin to remove dead tissue. This process is accompanied by the release of lysosomal enzymes. This is followed by the arrival of osteoblasts which attempt to repair the damage by laying down new bone directly upon the surface of dead trabeculae (Fig. 7.3). This composite of living and dead bone results in markedly thickened trabeculae which appear as radiodense or “sclerotic” regions at the margins of the infarct (Fig. 7.4). Adjacent areas from which dead bone has been removed become filled with fibrous tissue and amorphous debris, appearing as radiolucent or “cystic” areas.
Fig. 7.2
Dead bone and marrow elements from the center of the necrotic lesion
Fig. 7.3
Osteoblasts forming new bone directly on old, dead trabeculae
Fig. 7.4
Markedly thickened trabeculae at the margins of the necrotic region are composed of both living and dead bone
(We use the University of Pennsylvania Classification of Osteonecrosis—Table 7.1). Within the first 2–3 weeks after the vascular insult, X rays appear normal but changes can usually be detected on MRI (Stage I) (Fig. 7.5a, b). However, they do not appear on routine radiographs until several weeks to months later (Stage II) (Fig. 7.6). The processes of osteolysis and bone resorption and bone repair continue, during which the affected area steadily loses mechanical strength. Because the superior retinacular and lateral epiphyseal vessels, which supply the antero-superior aspect of the femoral head, are primarily involved, and since this is also the area of maximal weight bearing, collapse of subchondral trabeculae gradually develops in this region. This often takes place before the articular surface itself is affected and may appear as a radiolucent “crescent sign” (Stage III). This stage is not always seen as collapse of the articular surface with the subchondral bone may occur more or less simultaneously. If the necrotic region is small and not close to an area of major weight bearing, the situation may stabilize and the repair process may provide it with sufficient strength so that it does not collapse. It may persist as an area of radiodensity, although occasionally it is resorbed and disappears from radiographs. This corresponds with the clinical observation that very small lesions, especially those located medially, have a good prognosis. However, less than 5 % of lesions meet these criteria [15, 16]. It has also been observed that the prognosis for sclerotic lesions is better than for lesions which appear cystic. This is most likely due to the fact that sufficient new bone has been formed to provide mechanical strength to the region and hence decrease the incidence of collapse [17].
Table 7.1
University of Pennsylvania classification of osteonecrosis
Stage | Criteria | |
---|---|---|
0 | Normal or nondiagnostic radiograph, bone scan, and MRI | |
I | Normal radiograph; abnormal bone scan and/or MRI | |
A: Mild | (<15 % of head affected) | |
B: Moderate | (15–30 %) | |
C: Severe | (>30 %) | |
II | Lucent and sclerotic changes in femoral head | |
A: Mild | (<15 %) | |
B: Moderate | (15–30 %) | |
C: Severe | (>30 %) | |
III | Subchondral collapse (crescent sign) without flattening | |
A: Mild | (<15 % of articular surface) | |
B: Moderate | (15–30 %) | |
C: Severe | (>30 %) | |
IV | Flattening of femoral head | |
A: Mild | (<15 % of surface and <2 mm depression) | |
B: Moderate | (15–30 % of surface or 2–4 mm depression) | |
C: Severe | (>30 % of surface or >4 mm depression) | |
V | Joint narrowing and/or acetabular changes | |
A: Mild | Average of femoral head involvement as determined in Stage IV, and estimated acetabular involvement | |
B: Moderate | ||
C: Severe | ||
VI | Advanced degenerative changes |
Fig. 7.5
Images of a young male with Stage I steroid-induced osteonecrosis of right hip. (a) Plain radiograph appears “normal.” (b) T1 Weighted MRI shows characteristic changes of ON
Fig. 7.6
Sclerosis and lucency within the femoral head are characteristic of Stage II ON
With progressive collapse of subchondral trabeculae, the unsupported articular surface eventually begins to flatten. This represents an irreversible stage in the pathogenesis, Stage IV (Fig. 7.7). The articular cartilage is attached to the subchondral plate and remains viable, since it is nourished by diffusion from the synovial fluid and not by the vascular supply to the femoral head itself. However, the attached bone plate becomes necrotic (Figs. 7.8 and 7.9).
Fig. 7.7
Marked collapse and flattening of the femoral head without radiographic evidence of acetabular abnormality represents Stage IV ON
Fig. 7.8
Photomicrograph of a section of articular cartilage attached to its subchondral bony plate from a Stage IV hip. The cartilage remains viable whereas the bone is dead, as indicated by the empty osteocyte lacunae
Fig. 7.9
Low power photomicrograph of a section through the femoral head shows a large lesion with elements of necrosis and attempted repair
Radiographs of the hip continue to show a normal appearing acetabulum for quite some time after femoral head collapse. This can be misleading as histological changes in the articular cartilage are already taking place. In a study of 41 hips with ON which underwent total hip replacement despite a radiographic diagnosis of a “normal acetabulum,” 40 hips showed gross changes in the acetabular cartilage, and all 41 showed histologic degeneration [18]. It is important to keep this in mind when considering a hemi-arthroplasty involving only the femoral head with the assumption that the acetabulum is “normal.”
Progressive degenerative changes take place in the acetabulum secondary to the abnormal mechanical stresses imposed by the collapsed femoral head. Initially they involve only the articular cartilage as indicated by radiographic narrowing of the joint line. Later the underlying bone becomes affected and radiolucent and sclerotic regions appear in the roof of the acetabulum, often accompanied by marginal osteophyte formation. This represents Stage V radiographically. In a small number of cases this process continues until the joint is almost completely obliterated, which represents Stage VI [19].
Classification and Staging
The pathophysiologic sequence of events outlined usually follows a relatively predictable course. As a result, it is possible to describe the status of the osteonecrotic hip by means of a system of classification and staging.
The first classification system for ON was described in the early 1960s by Arlet and Ficat [20] and included three specific stages. A fourth stage was added in the 1970s and this is the version most widely used today, although in 1985 six stages were described [21, 22]. MRI was not originally included as it was not available at the time, and there was no attempt to indicate the size of the infarct nor the extent of joint involvement. Other classifications followed including those described by Marcus et al. [23], Sugioka [24], and the Japanese Investigation Committee for Avascular Necrosis [15].
The University of Pennsylvania Classification was developed in the early 1980s and identified seven clearly defined radiographic stages. It was the first to employ MRI as a specific modality for determining the stage, and was the first to include direct measurement of lesion size and the extent of joint involvement [19, 25, 26] (Table 7.1). In 1991 this classification was endorsed by the Association Research Circulation Osseous (ARCO), although modifications were made in 1992 and 1993 [27–29]. In 1992 it was also endorsed by the Committee on the Hip of the American Academy of Orthopedic Surgeons.
Recognizing the importance of the size of the infarct, a number of methods for measuring lesion size have been described during the past several years. However, most have relied on simple angular measurements made on plain radiographs or MRI, which are approximations rather than accurate measurements. In addition, these measurements have been used primarily to supplement non-quantitative classifications rather than as an integral part of one system [30–32].
MRI is currently the best modality for early diagnosis of ON, before changes appear on plain radiographs [33, 34]. This is important as the best results are obtained by early treatment, which in turn requires early diagnosis. The prognosis and treatment of hips with ON is also directly related to the size of the necrotic lesion and the extent of joint involvement. Accordingly, the clinical importance of using a comprehensive classification that indicates the extent of necrosis in addition to the stage is well recognized [19, 35–37]. This helps establish a prognosis, follow improvement or progression, compare different treatment options, and determine the best method of management for patients with different stages of ON. The uniform use of such a classification will help clarify the current confusion regarding both the natural history and the treatment of ON, and improve our management of patients with this perplexing disorder. A recent review of the literature shows a steady trend in this direction [36, 38].
At the present time, there are ongoing efforts to reach a consensus regarding the uniform use of a single effective classification. With advances in imaging techniques, it is now considerably easier than it was initially to measure accurately the size of the necrotic segment and the extent of joint involvement.
Management
Despite the increasing interest in osteonecrosis and the advances in understanding its etiology and pathophysiology, we still do not have an entirely satisfactory treatment. This is of particular concern because it affects most often younger adults, involves major weight bearing joints, and is usually progressive without appropriate treatment.
Prevention
A number of risk factors have been identified and these should be eliminated to the extent possible. These include alcohol ingestion, smoking, exposure to hyperbaric conditions, and corticosteroid administration. The postoperative management of organ transplantations has changed over the years, modifying the role of steroids, and accordingly the incidence of ON has diminished. When guidelines for divers and others working under hyperbaric conditions are followed, the incidence of ON decreases. During the past few years a number of genetic abnormalities have been identified which predispose patients to ON. In this group at risk, particular efforts should be exerted to minimize exposure to factors which could lead to the development of ON. Patients with hyperlipidemias might benefit from measures to control circulating lipid levels. In patients with certain coagulation abnormalities, some authorities have suggested long-term anticoagulation [4]. However, this approach has not been generally accepted since there is little evidence that this treatment is effective in preventing ON and the dangers of routine anticoagulation most likely outweigh the theoretical benefits.
Non-operative Management
A primary goal is to diagnose ON as early as possible, before collapse of the femoral head begins. This enables us to initiate measures designed to retard or prevent progression. A number of non-operative measures have been described. Patients are often placed on limited or non-weight bearing when the hip or lower extremity is affected. Although this may help to decrease pain, there is no evidence that it will retard progression and prevent eventual joint collapse. Various physical modalities have been advocated, including ultrasound and different types of electrical stimulation. At present they are used infrequently, and further evaluation and development may be indicated [39, 40]. There was also earlier enthusiasm about the role of hyperbaric oxygen, however there is little evidence that it is effective and it is rarely used [41]. Bisphosphonates have been given to slow the progress of bone resorption and thereby delay or prevent collapse. This approach is theoretically attractive and a limited number of studies have shown early promise. However, other investigators have failed to demonstrate a positive effect in patients followed over 2 years [42] (ref). Other agents, such as vasodilators and fulleral, a powerful antioxidant, have been suggested but their effectiveness has not yet been established.
Treatment Before Femoral Head Collapse
When osteonecrosis is diagnosed before femoral head collapse has taken place, a number of surgical procedures have been employed to delay or halt progression and promote healing. Technically, they vary considerably from one another, but essentially all are based upon physiologic principles, which address one or more aspects of the pathology involved. The results and complications reported have varied widely from one series to another. This section gives only a very brief overview of some of these procedures, and the reader is urged to look elsewhere if more information is required. Some surgeons have been reluctant to treat asymptomatic lesions, especially when complicated techniques are being considered. However, prior to trabecular collapse there is little correlation between the degree of pain per se and the outcome, and the majority of asymptomatic lesions do eventually become painful. Therefore, treatment designed to preserve the femoral head should not be withheld or delayed solely because the osteonecrotic lesion is asymptomatic or minimally painful [2, 43–45].
Core Decompression
One of the earliest and most often used methods of treating ON of the femoral head is “core decompression” [46, 47]. During the 1960s Arlet and Ficat, as part of their study of ON, removed diagnostic cores of bone from the femoral head and neck [20]. Patients noted prompt relief of pain following this procedure, which was felt to be due to relieving the high intraosseous pressure found to be present. This procedure became known as “core decompression” and was widely used to treat early cases of ON. Subsequently it has undergone several modifications including the use of several small perforations into the lesion rather than a single large core track. It has also been supplemented with electrical stimulation [40, 48] and by the addition of bone grafts and various agents to stimulate vascular ingrowth and bone formation, such as VEGF, bone morphogenetic protein (BMP), demineralized bone matrix (DBM), and mesenchymal stem cells, which will be discussed later.
The results reported following conventional core decompression have varied widely, but a review of the literature found a very low incidence of complications and a satisfactory result in 65–70 % of patients treated early [49–51]. Core decompression is now the most widely used joint preserving procedure in the United States. It can act through several mechanisms including decreasing elevated intraosseous pressure, removing areas of necrotic bone, stimulating the ingrowth of new vessels, and possibly as a channel for the introduction of materials that can stimulate vascular and bone growth. It is a relatively simple procedure with a very low rate of complications, when performed properly. Results with smaller lesions are better than with larger lesions, and it has been suggested that lesions which occupy less than 15 % of the femoral head, especially if located medially in a region of minimal weight bearing, may heal spontaneously and do not necessarily require treatment. In the cases that fail core decompression, later conversion to hip arthroplasty is not compromised.