Osteonecroses of the Skeletal System

6.1 Anatomy, Etiology, and Pathogenesis

Anatomy. Bone varies significantly in its composition, depending on age and location. It has an organic matrix, comprised of primarily collagen (osteoid), into which inorganic mineral components (especially calcium hydroxyapatite) are incorporated. The mineralized matrix is subject to constant resorption and formation by specific cells (osteoblasts and osteoclasts). This matrix harbors both red (hematopoietic) and yellow (fatty) marrow. These are the components that, together with the mineralized bony matrix, constitute the organ “bone.” Survival of the bone is not possible, however, without arterial inflow and venous drainage.

For purposes of clarification. Osteonecrosis takes place at a cellular level and only histopathology can identify the cellular alterations. Fat cells and hematopoietic marrow are always involved where ischemia is present. Cellular death in bone is a nonspecific, commonly subclinical, process that occurs more often when bone cells are subjected to abnormal stress. Therefore, histopathology will reveal small areas of necrosis along with severe osteoarthritis, stress or insufficiency fracture, an acute fracture, tumor, or infection. This chapter addresses clinically relevant forms of osteonecrosis which can be readily demonstrated by imaging studies. Imaging reflects macroscopic anatomy and reveals the effects of cell death on the bone (or parts of it). Common terms such as “osteonecrosis,”“bone infarction,”“avascular necrosis,” and “aseptic necrosis” are poorly defined and are applied inconsistently. These terms do not provide information about prognosis or etiology.

Note

In everyday language, the term “osteonecrosis” generally refers to bone necrosis located in the epiphysis or apophysis or involving the entire bone ( Fig. W6.1). If the osteonecrosis is located in the metaphysis or diaphysis, then this is referred to as “bone infarct.” A task for the future is to devise a classification system for bone necrosis that provides prognostic information independent of the skeletal location.

Etiology. In many cases the etiology and pathogenesis are obvious, such as interruption of blood supply secondary to a dislocation. If an etiology cannot be clearly defined, then it is better to refer to risk factors that may result in bone necrosis. Table 6.1 presents important risk factors. Additionally, some genetic factors that may predispose to bone necrosis have been identified.

Osteonecrosis without a clear etiology is referred to as “primary,”“idiopathic,” or “spontaneous” osteonecrosis.

Pathology. A number of theories have been advanced to describe the processes involved in the pathogenesis of osteonecrosis. It is generally accepted that all of them ultimately end in a reduced or interrupted supply of oxygenated blood to the bone. Ischemia prevents the normal repair processes of microfractures; results in the death of osteocytes, fat cells, and cells of hematopoiesis; and culminates in the loss of normal bone architecture.

Revascularization of an area of bone necrosis starts from the periphery. Osteoclasts are activated to absorb dead trabeculae. Fibrovascular tissue is formed to enclose the dead bone. This tissue is partially converted to bone. If the zone of necrosis involves the metaphysis or diaphysis, it will have no biomechanical impact. The same holds true for small epiphyseal lesions. If a larger necrotic zone and/or a lesion located within a weight-bearing area of a joint is placed under significant stress, then disruption of bone architecture will lead to functional failure with subsequent subchondral fracture and collapse of the joint surface.

Table 6.1 Risk factors for necrosis of the skeletal system

Pathogenesis	Risk factors
External factors	• Trauma • Surgery • Decompression sickness (caisson disease)
Iatrogenic	• Corticosteroids • Bisphosphonates • Radiotherapy
Nutritional	• Alcohol
Hematologic/oncologic	• Previous renal transplant (even without steroids) • Hemoglobinopathies (sickle cell anemia, thalassemia) • Leukemia
Metabolic	• Gaucher’s disease • Pancreatitis with fat embolism
Rheumatological	• Systemic lupus erythematosus • Necrotizing arteritis
Infections	• Osteomyelitis

Note

Beware of simplification. A “multiple-hit hypothesis” is now favored for the common cortisone-induced osteonecrosis. Cortisone results in an imbalance between osteoblasts and osteoclasts related to fatty degeneration and damage to the cell membrane. Steroids damage endothelial cells of blood vessels within bone. Thrombophilia and hypofibrinolysis are augmented by steroids; the latter are also responsible for alterations in fat metabolism resulting in increased adipogenesis, fat cell hypertrophy, and fat emboli.

6.2 Bone Infarction

Pathology. A bone infarction involving fatty marrow presents as a circumscribed lesion, while an infarct involving red marrow tends to be a poorly marginated lesion within the hematopoietic marrow. Ultimately, the necrotic area slowly becomes surrounded by a reparative margin. Bone infarcts may become smaller over time (this is common; see Fig. 6.5), or may even be completely absorbed.

Note

From a prognostic viewpoint, a bone infarction located in the metaphysis or diaphysis may be regarded as a “benign” form of osteonecrosis. Because of their location, with little cancellous bone and strong cortex, infarcts here are irrelevant for the structural integrity of the bone and are often clinically occult. Similar areas of necrosis in the epiphysis of tubular bones, in flat bones such as the ilium, and in irregularly formed bones such as the sacrum resemble metaphyseal infarctions but are referred to as osteonecrosis merely because of their location.

Clinical presentation. As a general rule, infarcts are often incidental findings. In the majority of cases these lesions are asymptomatic; however, they may be associated with chronic or acute pain, with the latter often occurring with acute infarctions related to a hemoglobinopathy (especially sickle cell anemia).

Radiography/CT. The early stage of bone infarction/osteonecrosis is radiographically undetectable, with subsequent poorly marginated rarefaction of the trabeculae ( Fig. 6.1). Reparative tissue develops at the edge of the infarct and slowly mineralizes to become evident as a peripheral sclerotic margin surrounding an area of central lucency ( Fig. 6.2). Extensive, intralesional calcifications are recognizable in the later stages. In very rare cases, metaphyseal and diaphyseal infarcts result in periosteal reaction and a widening of the bone.

NUC MED. In the initial phase a “cold spot” (i.e., decreased uptake) will be present in the area of necrosis, and eventually a “cold in hot spot” ( Fig. 6.3a) will be seen owing to the increased peripheral uptake related to the vascularized reparative tissue along its margin.

MRI. An infarction within yellow, fatty marrow, will demonstrate fat-equivalent signal in its center on T1W sequences. The area is typically bordered by a low–signal intensity margin, although the appearance may vary depending on the age of the infarct ( Figs. 6.3c, 6.4b, and 6.5). A hyperintense line (granulation tissue; Figs. 6.3b and 6.4a) is often present around the zones of necrosis on fat-suppressed water-sensitive sequences (see Fig. 6.3c). On T2W sequences without fat suppression (not routinely used) a “double-line sign” may be seen. Areas of cystic degeneration (fluid signal intensity on T2W images) and amorphous calcifications (hypointense on all sequences) may be present within the necrosis.

An infarction in areas with predominantly hematopoietic marrow or with pathologic bone marrow infiltration displays an area of low signal intensity on T1W images (provided it is visible at all against the already dark marrow) and increased signal intensity on fat-saturated PDW or T2W sequences ( Fig. 6.6). The enhancement pattern of an infarct after contrast administration reflects its pathophysiology: If the diagnosis is made early, there is little or no contrast enhancement within the center of the infarction. Later, strong marginal enhancement of the entire border zone will be seen. With advancing age of the infarction, the nonenhancing region becomes progressively smaller.

Fig. 6.1 Bone infarction in the tibial plateau. Fine sclerotic margin (arrows).

DD. A confident diagnosis of a bone infarct within yellow marrow is established by the detection of fat within the lesion on MRI. Other lesions containing fat include:

• Bone lesions with the potential for spontaneous remission (fibrous cortical defect, a brown tumor in renal osteodystrophy).

• Intraosseous lipomas.

The differential diagnosis of an infarction within red marrow is particularly difficult. Osteomyelitis, stress fractures, and necrotic tumors must be differentiated with the aid of the clinical history and presentation, laboratory findings, and follow-up imaging studies.

Enchondroma. If intralesional fat is not identified within an infarct due to a large amount of reparative fibrous tissue (rare), then it is not always distinguishable on T1W sequences from a chondroid tumor. Differentiation is also difficult on T2W images due to the juxtaposition of bright (cysts, cartilage) and dark signal intensity (calcifications). The typical lobular pattern of an enchondroma is often helpful in diagnosis ( Fig. 6.7). After contrast administration, an enchondroma demonstrates a number of “septations,” reflecting its lobular structure. Enhancement in bone infarction is more marginal or—when within the lesion—patchy.

Fig. 6.2 Bone infarction of the distal femur.

Fig. 6.3 Bone infarction of the tibial plateau. (a) Classic “cold in hot spot” appearance on the bone scan. (b) Serpentine, high–signal intensity margin surrounds the central necrosis on this fluid-sensitive, fat-saturated image. (c) The central necrosis is even brighter than the surrounding bone marrow due to cell death of the fat cells.

Fig. 6.4 Typical bone infarctions in femur and tibia. (a) Multilobular signal-intense lines, similar to an enchondroma, predominate on the fat-saturated PDW image. (b) Markedly irregular margins surround the extensive central necrosis.

Fig. 6.5 Radiological course of a bone infarction in the distal femur. (a) Initial appearance. (b) After one year, reduction in size and increasing marginal sclerosis is evident.

Fig. 6.6 Bone infarction of red marrow in sickle cell anemia. (a) Extensive increased signal intensity in the metaphysis. (b) The necrosis becomes visible after contrast administration due to the lack of enhancement of the central components. (c) It is not possible to differentiate the red infarction on the unenhanced T1W image due to the diffuse bone marrow infiltration in sickle cell anemia. (d) “Normal” contralateral side without infarction.

Fig. 6.7 Enchondroma versus bone infarction. (a) The center of the lesion appears hypointense on the T1W image (cartilaginous matrix). (b) Typical lobular high signal intensity architecture of an enchondroma.

6.3 Osteonecrosis

Pathology. The term “osteonecrosis” is used when an area of necrosis occurs within an epiphysis. Alterations involving an entire bone are also covered by the term “osteonecrosis” (e.g., necrosis of the lunate). From a pathophysiological aspect, there is no difference from a bone infarction. If the necrosis takes up large areas of the epiphysis or is situated in the weight-bearing part of the bone, the subchondral bone plate may collapse in that area. The entire subchondral bone then collapses into the necrotic zone, together with the overlying cartilage.

Location. Common sites of osteonecrosis related to the risk factors and pathogenesis described in Chapter 6.1 include the femoral head, humeral head ( Figs. 6.8, W6.2 and W6.3), scaphoid, lunate, femoral condyles ( Figs. 6.9, W6.4 and W6.5), and talus. Other less commonly involved sites include the proximal tibia, patella, navicular bone of the foot, and vertebrae.

Note

Osteonecrosis of the talus related to the risk factors in Table 6.1 is not uncommon. Trauma (especially talar neck fractures, cf. Chapter 2.15.4), and corticosteroids are the most common causes. Chronic osteochondral lesions of the talar dome (cf. Chapter 2.15.3) are not true osteonecroses, but are posttraumatic injuries sometimes associated with small, necrotic fragments. Osteochondritis dissecans of the talar dome is considered an osteochondrosis and not osteonecrosis (cf. Chapter 7.2.5).

6.3.1 Osteonecrosis of the Femoral Head

Osteonecrosis of the femoral head is the most common form of epiphyseal osteonecrosis; however, the same features may be seen in other locations, such as the femoral condyles and humeral head, and will therefore not be repeated for those sites.

Clinical presentation. The clinical spectrum ranges from a total lack of symptoms to severe pain and an inability to walk. The vast majority of symptomatic patients have a poor prognosis, with subsequent loss of function of the hip joint. There is also a risk of progression in asymptomatic cases discovered by MRI; however, these lesions may remain constant over a long period of time or may sometimes heal spontaneously ( Fig. W6.5).

Pathology. Osteonecrosis of the femoral head is found more commonly in men than in women, usually between the ages of 35 and 55 years. It is commonly bilateral. See Chapter. 6.1 regarding the etiology and pathophysiology of osteonecrosis of the femoral head. Progression results in collapse of the femoral head and subsequent secondary osteoarthritis of the hip joint.

Prognosis. Prognosis depends on the underlying risk factors (e.g., steroid therapy) and the degree of mechanical stress. As a rule, osteonecrosis with joint surface collapse does not regenerate over time nor is it influenced in its progression by surgical measures. With an intact joint surface, the risk of a poor outcome is related to the size of the necrotic area, so that MRI findings are helpful for predicting prognosis. The literature does provide some rules for determining the size of the affected area of the femoral head (expressed as a percentage), picturing the femoral head as an idealized hemisphere:

• If only 15 to 25% of the joint surface of the femoral head is involved, then a stable lesion without tendency to collapse may be expected.

• If over 25% of the joint surface is involved, then the development of a collapse is probable; surgery should be considered.

• If the area of necrosis is situated in the medial third of the stress distribution zone (on the coronal image), then prognosis is favorable. The “best” lesion, therefore, is small and located medially in the femoral head.

Classification systems. The presence of a fracture involving the joint surface is an important feature in all current classification systems because this typically portends collapse of the femoral head. Fractures of the joint surface correspond in almost all classification systems to Stage III. Ficat’s classification ( Table 6.2) is based on radiographic findings and functional evaluation of bone (by intraosseous phlebography and measurement of bone marrow pressure). The ARCO classification system (Association Internationale de Recherche sur la Circulation Osseuse) introduced MRI into the classification and took the size and site of the necrotic zone into account ( Table. 6.3). The Steinberg classification is also a modification of the Ficat system, with stages ranging from 0 to VI. As with the ARCO system, the extent of involvement is also taken into consideration.

Treatment. Currently, surgical treatment of the early stages primarily involves core decompression of the femoral head, with or without the insertion of a bone graft. If the femoral head has collapsed, then total hip replacement is basically the only therapeutic option. If the collapsed area is small, then a displacement osteotomy may be attempted.