Paul E. Beaulé
It is estimated that 300,000 to 600,000 people have osteonecrosis in the United States and that osteonecrosis (ON) of the femoral head accounts for approximately 10% of the more than 250,000 total hip replacements performed annually. 1 Patients with osteonecrosis of the femoral head are commonly in their 30s and 40s at the time of onset, and nearly 50% have bilateral disease. Results of total hip arthroplasty in this patient group have been inferior to other diagnostic groups. 2 These factors have led to the development of conservative surgical methods aimed at sparing the femoral head. However, the efficacy and the proper indications for these surgical interventions such as core decompression (with or without grafting), osteotomy, and hemiresurfacing arthroplasty are still debated. The preferred treatment varies according to the severity, extent of the disease, condition of the acetabular articular cartilage, and age of the patient. 2
Etiology and Pathophysiology
Eighty percent of patients with osteonecrosis have one or more risk factors known to predispose patients to developing this disorder. Table 9-1 provides a list of the known risk factors, leaving only about 20% of patients within the idiopathic group. There is a high prevalence of underlying thrombophilia or hypofibrinolysis in patients with osteonecrosis. 3 With the increasing number of etiologic associations, it is important to delineate between these associations and pathogenesis. Mont and others have proposed that the etiology of ON is multifactorial and either there exists an accumulated tissue threshold whereby a number of hits by various etiologic factors eventually meets the threshold for disease or there are certain at-risk patients in whom some other factors trigger the pathologic response. 4
Ficat 5 emphasized that bone necrosis is the end result of severe and prolonged ischemia, which may involve various mechanisms at different sites in extraosseous and/or intraosseous vessels. This blockage of the osseous microcirculation leads to intramedullary stasis, increasing compartmental pressure, metabolic disturbance, anoxia, and eventually death of osteocytes. The osteocytic death is eventually followed by the reparative phase, at which time the femoral head is at greatest risk of collapse when bone resorption exceeds production. 6Because the bone most susceptible to vascular compromise is closest to the joint space, the articular surface of the femoral head can become incongruous after collapse owing to structural failure of the underlying bone. More important, the time lapse between diagnosis and severe joint deterioration of the joint leading to a major surgical procedure is about 3 years in 50% of cases. 7
Diagnosis
Clinical Features
The average age at presentation is less than 40 years old. Males are more likely to be affected than females by a ratio of 8 to 1. 5,6,7,8 The most common symptom, seen in 50% of cases, is sudden pain in the groin area, which may be progressive and associated with radiation to the thigh. The pain is often worse at night. 5 The pain is usually triggered by weight bearing or by moving the affected limb. Fifteen percent of patients will have occurrence of osteonecrosis in other joints. 4
P.68
On physical examination the patient may have a limp and pain with internal hip rotation, but not uncommonly the exam will fail to uncover any striking abnormalities. Thus a high index of suspicion should be maintained. Further investigations to confirm the diagnosis include plain radiographs and magnetic resonance imaging. Sufficient plain radiographs must be ordered to permit evaluation of the contour of the femoral head in more than one plane. In addition to the anteroposterior and frog lateral radiographs, an anteroposterior pelvis radiograph with the hip flexed at 45 degrees will show the anterior contour and a 40-degree caudad view will show the posterior contour.
|
TABLE 9-1 Risk Factors Associated with Osteonecrosis |
|||||||||||||
|
Classification
Femoral head collapse and size and location of the osteonecrotic bone segment are recognized to be the most important prognostic factors. Ficat and Arlet's original classification scheme had four stages and was based on anteroposterior and frog lateral radiographs. Stages II and III represented the distinction between precollapse and postcollapse disease, and this classification system established the premise for staging osteonecrosis and subsequent classification systems. 5,7 Later on a stage 0 was added and qualified as the silent hip as first described by Marcus. 9 Standard radiographs show only the shadow of the mineralized portion of a bone; consequently early bone necrosis has no specific radiographic appearance, and a normal radiograph does not necessarily mean a normal hip. On the MRI, a single-density line on the T1-weighted image demarcates the normal ischemic bone interface, and a double-density line on the T2-weighted image represents the hypervascular granulation tissue. 6 On plain radiographs, the crescent sign originally described by Ficat 5 is characterized by the pathognomic appearance of a sequestrum on the radiograph within which occurs a subchondral fracture, i.e., crescent line.
|
Figure 9-1 A 45-year-old female presents with an 8-month history of bilateral hip pain. A: Plain radiographs show bilateral collapse of femoral heads: stage IVC. Inset is the frog lateral of the right hip. B: Magnetic resonance imaging of the left hip shows the extent of necrotic lesion. |
Although not included in the Ficat classification, the size of the necrotic segment has been shown to influence the outcome of head-sparing procedures and was first quantified by Kerboull et al. 10 His sum included angles of the necrotic segment measured on the anteroposterior (AP) and lateral radiographs, and found a sum of angles greater than 200 degrees represented a worse prognosis. 2 (Fig. 9-1). The more recently developed University of Pennsylvania system incorporates size of lesion and magnetic resonance imaging findings (Table 9-2) and therefore represents a valuable classification system that aids in management. 11
Osteonecrosis of the femoral head is associated with secondary involvement of the acetabular cartilage owing to mechanical damage caused by collapse of the femoral head, as demonstrated by both Steinberg 12 and Beaulé. 13 Thus acetabular cartilage must be carefully evaluated at the time of the operative procedure, and when there is severe damage, total hip replacement may be the preferred method of treatment, especially when patients are older than 40 years of age. 2
Nishii et al. 14 reviewed 54 hips to determine the risk of collapse and progression in hips with osteonecrosis. At a mean followup of 6 years, only 31% of hips went on to collapse when the necrotic area was less than two-thirds of the weightbearing area compared to 68% with the larger areas.
P.69
In addition, the percentage of hips that went on to arthritic changes once collapsed was 92% with necrotic areas greater than two-thirds compared to 11% with the smaller area. All hips with >2 mm of initial collapse continued to collapse. Impending femoral head collapse is seen on radiographs as a crescent sign, which represents a fracture of the bone and probably the cartilage as well, which will become further depressed depending on the size of the lesion and joint loading forces.
|
TABLE 9-2 University of Pennsylvania Staging System for Osteonecrosis of the Femoral Head11 |
|||||||
|
Operative Treatment Options
The age of the patient at presentation should be considered carefully; the younger the patient, the more strongly a joint- or bone-preserving procedure should be favored because it is probable that he or she may require another operative procedure. Because with osteonecrosis the acetabular cartilage of the hip is secondarily mechanically damaged after the femoral head has collapsed and this phenomenon is progressive, the duration of symptoms (pain) provides some indication of the quality of the acetabular cartilage. The extent of acetabular damage will ultimately affect the outcome of joint and bone preserving procedures, i.e., osteotomy, free vascularized fibular graft, hemiresurfacing arthroplasty, and the trapdoor procedure. With respect to the cause and risk factors of osteonecrosis, the continued use of high-dose steroids, alcohol, and/or systemic lupus erythematosus is usually associated with a poorer outcome with certain head-sparing procedures. 15 The judicious selection of the best treatment option is dependent on careful analysis of patient demographics and cause, the extent of head involvement, and articular cartilage damage of the acetabulum. The goal of treatment is to optimize outcome while simultaneously minimizing morbidity and maintaining treatment options for potential subsequent secondary procedures.
Core Decompression with or Without Grafting
The role of core decompression is for precollapse stages of osteonecrosis. The best outcomes are for lesions with <15% involvement with a 78% survivorship at an average of 63 months 16 and in those hips with sclerosis with a 95% survivorship at 10 years. 17 The value of bone grafting in addition to core decompression is uncertain. Aaron and associates 18 reported on 28 hips treated with demineralized bone matrix compared with another group without grafting, and there was no significant difference in success rate between these two groups for stage III with survivorships of 83% and 72%, respectively. More recently, Hernigou and Beaujean 19 presented their midterm results of autologous bone marrow grafting after core decompression using a 3-mm-diameter trephine. Hips with a lesion <25% in stages I and II had the best outcome with only 9 out of 145 hips requiring a total hip replacement versus 25 out of 44 hips for the rest of the stages.
Although we continue to use core decompression alone in some patients with Steinberg stage III, it has no role in the treatment of stage IV and V disease because it cannot restore articular sphericity or remove the collapsed segment from the weight-bearing area, and further cartilage degeneration is inevitable. Smith et al. 20 reported a 0% survivorship following core decompression at average follow-up of 3 years for stage IV with all patients requiring another operative procedure. Mont et al. 21 reviewed the radiographic predictors of outcome for 68 Ficat stage III (Steinberg stage IV) hips following core decompression with mean follow-up of 12 years. Only 29% had a satisfactory outcome. If core decompression is combined with grafting in patients with Ficat stage II/Steinberg stage III lesions, satisfactory results have been reported for both nonvascularized and vascularized grafts. 22,23
Free Vascularized Fibular Graft
Free vascularized fibular graft techniques have been used to support the subchondral surface and enhance revascularization in combination with core decompression and osteoinductive cancellous bone. These grafting procedures have a role in the precollapse stages, but have inferior results in the postcollapse stages because of articular cartilage involvement and the inability to restore femoral head sphericity. For stages IV and V, several authors have reported survivorships <75% at 5 years with surviving patients having Harris hip scores <80. 20,22,24The best results have been with earlier-stage lesions with a survivorship of 89% at 50 months for stage III compared with 65% for core decompression alone. However, one must carefully consider the potential disadvantages of this surgery such as donor site morbidity at 24%25 and a 2.5% incidence of subtrochanteric fracture. 26
Open Grafting of the Femoral Head
Another grafting technique, the trapdoor procedure, first described by Merle D'Aubigné, 27 has been used. In this procedure, an arthrotomy is performed to dislocate the hip anteriorly, the necrotic segment of the head is curetted out, and iliac crest bone graft is packed inside. This is
P.70
done through a cartilage window of the femoral head. Mont et al. 28 reported on a series of 24 hips with Ficat stage III (Steinberg stage IVA) disease and 6 hips with early stage IV (Steinberg IVB) disease treated with this procedure. At a mean duration of follow-up of 56 months (range 30 to 60 months), 73% had good to excellent results. Of the eight hips with poor results, five had undergone subsequent operative procedures. All of the poor results had combined necrotic angles greater than 200 degrees. The results are encouraging, but the procedure has limited indications because of the difficulty in restoring the sphericity of the femoral head. Further study with longer duration of follow-up will be required to assess the utility of this procedure.
Proximal Femoral Osteotomy
The clinical results of proximal femoral osteotomy for osteonecrosis have been variable and sometimes disappointing because of the difficulty rotating the necrotic segment out from the weight-bearing zone of the hip, especially when the lesion is large. Langlais and Forestier 29 reported their results with either the Sugioka (anterior) or Kempf (posterior) rotational osteotomy in 20 patients with Ficat stages II and III (Steinberg III and IV). The Kempf osteotomy was used for the atypical case where the necrotic zone extended >30 to 40 degrees posteriorly. Patients older than 45 years and patients with either steroid- or alcohol-induced osteonecrosis were excluded from their study because of predictably poor results. They recommended the Sugioka procedure only for Steinberg stage III and the Kempf procedure for stages III and IV where the depth of the necrosis was not more than one third of the head diameter and located posteriorly.
Others advocate an intertrochanteric osteotomy because it avoids the need for a greater trochanteric osteotomy and does not risk compromising the blood supply to the femoral head. Mont et al. 30 reported on this technique in 31 hips with Ficat stage III (Steinberg IV) disease at a mean follow-up of 11.5 years. Good to excellent results were obtained in 74% of the hips. Poor results were associated with age older than 45 years, a combined necrotic angle >200 degrees, and Dinulescu et al. 31 reported similar findings with survivorship of 70% at 5 years and 45% at 10 years in 50 stage II and stage III hips. One of the main disadvantages to intertrochanteric osteotomy is the apparent negative effect on the complication rate and subsequent total hip arthroplasty, both of which are probably related to the femoral deformity created by the osteotomy. 30,32,33 Nevertheless, for surgeons with experience with the procedure, a proximal femoral osteotomy is a reasonable head-sparing operation in patients younger than 45 years of age, with Kerboull angle <200 degrees, and no continuing steroid use.
Hemiresurfacing Arthroplasty
Hemiresurfacing hip arthroplasty with cement fixation was first preformed in the early 1980s as a custom device in the young, active population to preserve femoral bone stock and permit conversion to a total hip replacement with minimal morbidity. 34,35,36,37,38 (Fig. 9-2). The results of this procedure have varied in different reports. Beaulé et al. 13 reviewed a series of 37 hips with a mean follow-up of 6.5 years (range 2 to 18 years), in which the acetabular cartilage involvement was photographed at the time of the operative procedure and subsequently graded. It was found that the longer duration of preoperative symptoms the more severe the acetabular damage, and that hips that had been converted to total hip replacement had a longer duration of symptoms prior to their hemiresurfacing arthroplasty than the ones that were still functioning (17 versus 12 months, respectively). The survivorship in this series was 79% at 5 years, 62% at 10 years, and 45% at 15 years. Other centers using different hemiresurfacing designs have reported comparable survivorship (i.e., 80% at 5 years and 60% at 10 years) and pain relief for osteonecrosis of the hip with Ficat stage III and early IV disease (Steinberg IVA and IVB). However, as shown by Mont et al., 40 the pain relief is not as predictable as that following a total hip replacement: In their patients a hemiresurfacing arthroplasty had a mean Harris hip score of 88 compared with 93 for patients with a total hip resurfacing (THR). Proper component sizing and acetabular cartilage quality are probably the two most important factors affecting outcome. A remaining challenge is durability of the articular cartilage against the hemiresurfacing component.
|
Figure 9-2 Radiograph of patient shown in fig. 9-1 after bilateral hemiresurfacing arthroplasty. |
Conclusion and Recommendations
The treatment of patients with osteonecrosis remains controversial; however, it is generally believed that the size of the lesion, as well as the stage of the disease at the time of treatment, determine the outcome of head-sparing procedures. Core decompression and or bone grafts have a role in the precollapse stages of osteonecrosis. The size of the necrotic segment plays a critical role in the outcome of proximal femoral osteotomies. Thus, unless a relatively small lesion (Kerboull angle <200 degrees) is present, hemiresurfacing arthroplasty provides a better clinical outcome compared with other head-sparing procedures with the caveat that the predictability of the pain relief is not that of a total hip replacement. Because the acetabular cartilage damage increases over time once collapse has occurred,
P.71
early surgery is recommended for any type of head-sparing procedure. In the presence of notable acetabular involvement or advanced Steinberg V disease, total hip replacement offers the most predictable outcome.
Reference