JBJS | Atraumatic Osteonecrosis of the Knee*

The Journal of Bone & Joint Surgery, Volume 82, Issue 9

Articles | September 01, 2000

Atraumatic Osteonecrosis of the Knee*

Michael A. Mont, M.D.; Keith M. Baumgarten, M.D.; Aiman RifaI, D.O.; David A. Bluemke, M.D.Ph.D.; Lynne C. Jones, Ph.D.; David S. Hungerford, M.D.

View Disclosures and Other Information

Investigation performed at the Division of Arthritis Surgery, Department of Orthopaedic Surgery, The Johns Hopkins University School of Medicine, Baltimore, Maryland
*No benefits in any form have been received or will be received from a commercial party related directly or indirectly to the subject of this article. No funds were received in support of this study.
Division of Arthritis Surgery, Department of Orthopaedic Surgery, The Johns Hopkins University School of Medicine, The Good Samaritan Hospital, Professional Office Building, G-1, 5601 Loch Raven Boulevard, Baltimore, Maryland 21239. E-mail address for M. A. Mont: rhondamont@aol.com.

The Journal of Bone & Joint Surgery. 2000; 82:1279-1279

5 Recommendations (Recommend) | 3 Comments | Saved by 3 Users Save Case

text A A A

Abstract

Background: The purposes of this study were to define the clinical, demographic, and radiographic patterns of atraumatic osteonecrosis of the distal part of the femur and the proximal part of the tibia at presentation and to report the outcome of treatment of this condition.

Methods: Two hundred and forty-eight knees in 136 patients who were younger than the age of fifty-five years were treated at our institution between July 1, 1974, and September 15, 1998, for atraumatic osteonecrosis of the distal part of the femur or the proximal part of the tibia, or both. Demographic and radiographic features were characterized. The results of nonoperative treatment, core decompression, arthroscopic dÇ¢ridement, and total knee arthroplasty were evaluated.

Results: There were 106 female patients and thirty male patients, and their mean age was thirty-six years (range, fifteen to fifty-four years) at the time of diagnosis. One hundred and one patients (74 percent) had involvement of other large joints, with eighteen (13 percent) presenting initially with knee symptoms. One hundred and one patients (74 percent) had a disease that affected the immune system; sixty-seven of them had systemic lupus erythematosus. One hundred and twenty-three patients (90 percent) had a history of corticosteroid use. Technetium-99m bone-scanning missed lesions in sixteen (29 percent) of fifty-six knees. Eight (20 percent) of forty-one initially symptomatic knees treated nonoperatively had a successful clinical outcome (a Knee Society score of at least 80 points and no additional surgery) at a mean of eight years. The knees that remained severely symptomatic for three months were treated with either core decompression (ninety-one knees) or total knee arthroplasty (seven knees). Seventy-two (79 percent) of the ninety-one knees treated with core decompression had a good or excellent clinical outcome at a mean of seven years. Efforts to avoid total knee arthroplasty with repeat core decompression or arthroscopic dÇ¢ridement led to a successful outcome in fifteen (60 percent) of twenty-five knees. Thirty-four (71 percent) of forty-eight knees treated with total knee arthroplasty had a successful clinical outcome at a mean of nine years.

Conclusions: Atraumatic osteonecrosis of the knee predominantly affects women, and in our study it was associated with corticosteroid use in 90 percent of the patients. Evaluation should include standard radiographic and magnetic resonance imaging of all symptomatic joints. Prognosis was negatively related to large juxta-articular lesions. Nonoperative treatment should be reserved for asymptomatic knees only. Core decompression was successful (a Knee Society score of at least 80 points and no additional surgery) in 79 percent of the knees in which the disease was in an early stage. Total knee arthroplasty was successful in only 71 percent of the knees.

Figures in this Article

Introduction

Introduction | Materials and Methods | Results | Discussion | References

Although spontaneous osteonecrosis of the distal part of the femur and the proximal part of the tibia has been well described^{2,3,8,13,14,23,29,30,35,43-45,47,59,64,70}, there have been fewer reports of atraumatic osteonecrosis in these areas^{1,9,25,38,44,49,51,54,59}. The knee is the second most common location for osteonecrosis, although it is affected much less often than the hip; the exact prevalence of osteonecrosis of the knee is unknown. As a result of this lower prevalence, most studies have included a limited number of patients from one center and thus have yielded limited information regarding the natural history of this entity and the most effective treatment regimen. As in the hip, dead bone leads to subchondral fracture and later joint collapse, which often results in the need for a total knee arthroplasty.

The diagnosis of osteonecrosis about the knee can be difficult, as knee pain is sometimes thought to be referred pain from concurrent hip disease or may be erroneously considered a symptom of an intra-articular disorder such as a meniscal tear. Atraumatic osteonecrosis can also be confused with various other entities, including posttraumatic osteonecrosis, osteochondritis dissecans, and spontaneous osteonecrosis. In posttraumatic osteonecrosis, there is a history of trauma or surgery leading to bone death, usually in an isolated area of the knee. Osteochondritis dissecans is a condition of unknown origin that is usually found in patients younger than twenty-five years of age and is confined to one knee condyle without other joint involvement⁶⁹. It involves a separation of a segment of articular cartilage and subchondral bone from the articular surface of well vascularized bone. Spontaneous osteonecrosis is usually confined to one femoral condyle or one tibial plateau in patients older than the age of fifty-five years who have no other joint involvement or associated comorbid conditions. It has been described as an entity manifested by a sudden onset of pain of questionable origin. This disease appears similar to an osteonecrotic juxta-articular lesion on radiographic and magnetic resonance imaging, although some authors have questioned whether the disease of osteonecrosis is actually present^44,45. The prognosis associated with these lesions has been related to their size, with smaller lesions being self-limited and larger lesions leading to joint collapse and the need for total knee arthroplasty^{2-4,8,13,14,29,30,35,42,44,45,63,64}.

Atraumatic osteonecrosis of the knee has been called secondary, ischemic, idiopathic, or corticosteroid-associated necrosis and is usually easily differentiated from the previously described disorders. It is commonly found in patients in their mid-thirties, it involves multiple condyles of the knee, it is usually bilateral, and patients often have osteonecrosis of other large joints. The disease is analogous to the similarly named entity found in the hip with many comorbid conditions. It has been reported in association with systemic lupus erythematosus^{1,32,38,46,52,54,55,71}, sickle-cell disease^52,71, alcoholism^{31,38,52,55,71}, and use of corticosteroids^{9,26,36,38,39,51,52,55,57,63,71}. A comparison of atraumatic osteonecrosis and spontaneous osteonecrosis is presented in Table I.

The purposes of the present study were to define the clinical, demographic, and radiographic patterns of atraumatic osteonecrosis of the distal part of the femur and the proximal part of the tibia at presentation and to report the outcome of treatment of this condition.

View Large | Download Slide(.ppt)
Add to My JBJS

Fig. 1:: Figs. 1 and 2: Illustrations of the techniques used to assist in localization of the lesion.
Fig. 1: In the coronal plane, the distal part of the femur and the proximal part of the tibia are divided into four quadrants each.

View Large | Download Slide(.ppt)
Add to My JBJS

Fig. 2: In the sagittal plane, the distal part of the femur is divided into anterior (A), proximal (Pr), posterior (Po), and distal (D) quadrants. The tibia is divided into anterior and posterior columns.

View Large | Download Slide(.ppt)
Add to My JBJS

Fig. 3:: Illustration showing measurement of the combined necrotic angle (the sum of angle A and angle B) for the distal part of the femur. In this example, the angle is 112 degrees (35 + 77 degrees).

View Large | Download Slide(.ppt)
Add to My JBJS

Fig. 4:: Treatment algorithm utilized in this study for patients with atraumatic osteonecrosis of the knee.

View Large | Download Slide(.ppt)
Add to My JBJS

Fig. 5:: Magnetic resonance image demonstrating a lesion involving the femoral epiphysis medially and laterally as well as the metaphysis laterally (arrowheads).

View Large | Download Slide(.ppt)
Add to My JBJS

Fig. 6:: Radiograph showing a stage-II lesion in a thirty-four-year-old woman with systemic lupus erythematosus. Note the sclerotic changes in the distal part of the femur and the proximal part of the tibia (arrowheads).

View Large | Download Slide(.ppt)
Add to My JBJS

Fig. 7:: Radiograph showing a stage-III lesion in a forty-year-old man with a history of alcohol abuse. The arrowheads denote the area of subchondral collapse.

View Large | Download Slide(.ppt)
Add to My JBJS

Fig. 8:: Radiograph showing a stage-IV lesion in a twenty-six-year-old man receiving renal dialysis. There is joint-space narrowing, and lesions (arrowheads) involving the distal part of the femur and the proximal part of the tibia are seen.

View Large | Download Slide(.ppt)
Add to My JBJS

Fig. 9:: Figs. 9, 10, and 11: Kaplan-Meier survival curves. The y axis is the cumulative probability of survival with a Knee Society score of less than 80 points or total knee arthroplasty as the end point. The x axis is the time (in months) from the index clinical assessment until the last follow-up evaluation.
Fig. 9: Curve for nonoperatively treated knees.

View Large | Download Slide(.ppt)
Add to My JBJS

Fig. 10:: Curve for knees treated with core decompression.

View Large | Download Slide(.ppt)
Add to My JBJS

Fig. 11:: Curve for knees treated with total knee arthroplasty.

View Larger

Table I: Comparison of Spontaneous and Atraumatic Osteonecrosis

	Spontaneous Osteonecrosis	Atraumatic Osteonecrosis
Synonyms	Primary necrosis, osteonecrosis, SONK, SPONK	Secondary, ischemic, idiopathic, corticosteroid-associated necrosis
Age (yrs.)	>55	<55
Associated risk factors	None	Corticosteroid use, alcohol use, renal disease, other
Condylar involvement	One	Multiple
Laterality	99% unilateral	>80% bilateral
Other joint involvement	Rare	60-90%
Location	Epiphyseal to subchondral surface	Diaphyseal, metaphyseal, epiphyseal

Add to My JBJS

Table I: Comparison of Spontaneous and Atraumatic Osteonecrosis

	Spontaneous Osteonecrosis	Atraumatic Osteonecrosis
Synonyms	Primary necrosis, osteonecrosis, SONK, SPONK	Secondary, ischemic, idiopathic, corticosteroid-associated necrosis
Age (yrs.)	>55	<55
Associated risk factors	None	Corticosteroid use, alcohol use, renal disease, other
Condylar involvement	One	Multiple
Laterality	99% unilateral	>80% bilateral
Other joint involvement	Rare	60-90%
Location	Epiphyseal to subchondral surface	Diaphyseal, metaphyseal, epiphyseal

Materials and Methods

Introduction | Materials and Methods | Results | Discussion | References

Between July 1, 1974, and September 15, 1998, 1122 patients were managed at our facility because of osteonecrosis of the hip, knee, shoulder, elbow, or ankle. Of these patients, 167 (289 knees) were identified as having osteonecrosis of the distal part of the femur or the proximal part of the tibia, or both. The present study included patients who presented with symptomatic knees that were then found to be affected by osteonecrosis as well as patients with osteonecrosis in other joints who had symptomatic knees evaluated with radiographic or magnetic resonance imaging. Patients with osteochondritis dissecans or spontaneous osteonecrosis were excluded. The study was limited to patients who were younger than the age of fifty-five years, to avoid confusion of the disorder with spontaneous osteonecrosis. No patient in this study had a lesion confined to one condyle. A total of 248 knees in 136 patients with atraumatic osteonecrosis were studied.

All 248 knees were included in our demographic and radiographic evaluation. Sixty-eight knees that were seen after January 1, 1997, did not have the two-year follow-up required for evaluation of the results of treatment. In addition, six patients died of causes unrelated to the osteonecrosis, seven had moved away from the United States and could not be contacted, and seven were lost to follow-up between twelve and twenty-four months after the diagnosis. This left 149 knees for the two-year follow-up evaluation of the results of treatment.

A retrospective clinical and radiographic review was performed. Hospital outpatient records were analyzed with respect to age at presentation, gender, associated diseases, alcohol use (greater than 400 milliliters of 100 percent absolute ethanol per week)²⁸, tobacco use, symptoms at presentation, family history, weight, involvement of other joints, and use of corticosteroids. All of the patients were examined clinically and radiographically by the senior authors (M. A. M. and D. S. H.); the clinical data were obtained at the time of presentation and at the most recent follow-up examination.

Knee Society Score

The objective rating system of the Knee Society was used to define the level of clinical function at presentation and at the most recent follow-up evaluation^15,34. This system is a 100-point scale with 50 points assigned to the pain category, 25 points awarded for range of motion, and 25 points awarded for knee stability. A score of 90 points or more was considered an excellent outcome; a score of 80 to 89 points, a good outcome; a score of 70 to 79 points, a fair outcome; and a score of less than 70 points, a poor outcome. At the time of the final follow-up, all patients with a score of less than 80 points or who had required a total knee arthroplasty were considered to have an unsuccessful clinical outcome. Scores of at least 80 points were considered to indicate a successful clinical outcome.

Radiographic and Magnetic Resonance Imaging

Patients were evaluated with plain radiographs and with magnetic resonance images of the knees. Anteroposterior and lateral radiographs of all knees were used to stage and characterize the osteonecrosis at presentation and at the most recent follow-up examination. Follow-up radiographs were made on a yearly basis. Patients were entered in the study either because they had presented with symptoms in one or both knees and were subsequently diagnosed as having osteonecrosis in those knees (eighteen patients) or, alternatively, because they had known osteonecrosis of other joints and had an evaluation of both knees because one or both were symptomatic. Therefore, not all knees in patients with known osteonecrosis were evaluated, and, consequently, a number of so-called silent knees could have been missed in this study. Whenever a patient had a symptomatic knee, the contralateral knee was evaluated for the presence of osteonecrosis whether or not it was symptomatic. Twenty-four knees were found not to be affected by osteonecrosis as demonstrated by negative findings on the radiographs and magnetic resonance images. Coronal, axial, and sagittal magnetic resonance images through the distal part of the femur and the proximal part of the tibia were made at presentation for ninety-one of the knees.

All of the knees were classified with use of the system of Ficat and Arlet, as modified for the knee^{17,19,20,52,55}. In this radiographic staging system, stage-I knees have a normal appearance. Stage-II knees have cystic or osteosclerotic lesions, or both, with a normal contour of the distal part of the femur or the proximal part of the tibia, or both, and with no subchondral fracture or flattening of the articular surface. Stage-III knees have a crescent sign or subchondral collapse, and stage-IV knees have narrowing of the joint space with secondary changes in the proximal aspect of the tibia (when the disease is confined to the femur), such as cysts, marginal osteophytes, and destruction of cartilage. Each of four areas of the knee (the medial and lateral femoral condyles and the medial and lateral tibial plateaus) were evaluated separately with use of this system.

To assist in localization of the lesion, the distal part of the femur was divided into quadrants. Medial and lateral columns were constructed on the coronal images by drawing a line bisecting the femoral canal and then drawing a line from the medial epicondyle to the lateral epicondyle, thereby creating the quadrants in the distal part of the femur (Fig. 1). The proximal part of the tibia was divided into four quadrants in the same manner.

For sagittal plane analysis, the distal part of the femur was divided into anterior, proximal, posterior, and distal quadrants. The quadrants were constructed by drawing a vertical line that bisected the femoral canal. At the central axis of the distal part of the femur, two lines were drawn, one at 45 degrees and one at -45 degrees with respect to the vertical line (Fig. 2). The tibia was divided into anterior and posterior columns by drawing a bisecting line on the sagittal image.

Lesions of the distal part of the femur and the proximal part of the tibia were categorized by the quadrants that they affected, by condylar involvement, and by the presence or absence of diaphyseal involvement. These factors were then analyzed with respect to the clinical outcomes to determine if the type of lesion was a prognosticator of the disease.

The size of the lesion was categorized with two methods. With one method, the height and width of the lesion were approximated on the coronal magnetic resonance image with use of a modification of the method of Lotke and Ecker⁴⁵ (who utilized standard radiographs). When magnetic resonance images were not available, standard radiographs were used. The depth of the lesion was measured on the sagittal view. The volume of the lesion was then approximated by multiplying the height by the width by the depth of the lesion. Femoral and tibial lesions were characterized as small (zero to less than ten cubic centimeters), medium (ten to twenty cubic centimeters), or large (more than twenty cubic centimeters). The size of the lesion was then compared with the clinical outcome.

The size of the lesion was also categorized with use of the method of Kerboul et al.⁴⁰ (Fig. 3). With this method, an outline of the osteonecrotic lesion is made on the anteroposterior and lateral radiographs, and the angle of the arc of the surface involved by the necrosis is measured with a goniometer. These two angles are added together to derive the combined necrotic angle. This method was utilized for the epiphyseal area of the lesion because it was found that the combined necrotic angles of metaphyseal and diaphyseal lesions did not influence prognosis. Lesions of 150 degrees or less were categorized as small, lesions of 151 to 249 degrees were categorized as medium, and lesions of 250 degrees or greater were categorized as large.

Radiographs were made at yearly follow-up examinations for all patients. All 149 knees that were followed clinically for more than two years were followed radiographically for more than two years as well. Radiographic progression was noted if a stage-I or II knee had progressed to stage III or IV or if a stage-III knee had progressed to stage IV.

Technetium-99m bone scans were made for fifty-six knees, and the results of these studies were compared with the findings on radiographs, magnetic resonance images, and histopathological examination. All diagnoses based on bone-scanning, before the advent of magnetic resonance imaging, and all based on magnetic resonance imaging were confirmed by radiographs. In addition, the diagnosis of osteonecrosis was confirmed histologically in all but thirty-nine knees that were treated nonoperatively.

Treatment Methods

All ten asymptomatic knees and all but seven symptomatic knees were initially treated nonoperatively (Fig. 4), with use of analgesics, nonsteroidal anti-inflammatory medications, and protected weight-bearing with crutches or a cane, for a minimum of three months. Seven knees underwent a total knee arthroplasty on presentation because they were severely painful and the disease was radiographic stage IV. All of these patients presented to our institution with symptoms that had lasted for at least six months (mean, eleven months; range, six to eighteen months).

All other patients in the study presented with disease that was radiographic stage III or less severe. When these knees became asymptomatic after nonoperative treatment, surgery was not utilized, but persistent severe pain led to the performance of core decompression^38,52,55. The core track was placed into each condyle or plateau whenever a lesion (femoral or tibial) was located in the epiphysis. This procedure and the rehabilitation protocols have been described elsewhere^18,22,29. If a knee had persistent symptoms after core decompression and no radiographic signs of progression, and the symptoms were described as coming from the bone and there was tenderness on palpation directly over an osseous condyle, it was treated with a repeat core decompression. The placement of the second core track into the epiphyseal lesion was based solely on the patient's description of the location of the pain as condylar or tibial. If a patient had a failed core decompression without radiographic progression and symptoms apparently coming from an intra-articular source (confirmed by a physical examination demonstrating joint-line tenderness and temporary relief of pain with local instillation of an anesthetic), an arthroscopic dÇ¢ridement of any loose chondral flap or torn meniscus was performed^41,50. If a patient had a failed core decompression and evidence of radiographic progression, a total knee arthroplasty was performed^7,36,53,54.

Data Analysis

The data were compiled and tabulated with use of an Access 7.0 database (Microsoft, Redmond, Washington). Descriptive statistics were calculated. Trends concerning the relationship of various demographic variables to radiographic stages were evaluated. Parametric and nonparametric statistical analyses of the results were conducted with use of the Computer Programs for Epidemiologic Analysis (PEPI) Software Package (version 2.03; USD, Stone Mountain, Georgia). Frequencies were analyzed with use of a chi-square distribution with a Yates correction. Nonparametric data were evaluated with use of the Mann-Whitney U test, whereas parametric data were analyzed with analysis of variance.

Results

Introduction | Materials and Methods | Results | Discussion | References

Clinical Evaluation

The mean age of the 106 female patients and thirty male patients was thirty-six years (range, fifteen to fifty-four years). The mean duration of symptoms before the diagnosis was made was six months (range, two weeks to twenty-four months). At presentation, all but ten of the knees (eight of the patients) were symptomatic with either pain or loss of motion. One hundred and one patients (74 percent) had a disease that affected the immune system, including systemic lupus erythematosus (sixty-seven patients), inflammatory bowel disease (eleven), nephrosis requiring a renal transplant (seven), lymphoma (five), idiopathic thrombocytopenic purpura (two), temporal arteritis (two), Sjögren disease (two), polymyositis (one), Wegener granulomatosis (one), multiple sclerosis (one), IgA nephropathy (one), and vasculitis (one). Sixteen patients (12 percent) had a history of moderate alcohol use, and ten patients (7 percent) had a history of heavy alcohol use^26,28,31. Twenty-four patients (18 percent) had a history of moderate tobacco use, and twenty-four additional patients had a history of heavy tobacco use. Twenty-three patients had asthma, two had pneumonitis, and six had a spine injury. Three patients had no identifiable risk factors for osteonecrosis, but laboratory investigation revealed a coagulation disorder (a familial protein-S deficiency) in the one patient who was tested. Starting in 1998, ten patients were tested for a coagulation disorder, and the findings were positive for nine of them. One hundred and twenty-three patients (90 percent) had a history of corticosteroid use.

The 112 patients (82 percent) who had bilateral osteonecrosis of the knee did not differ from the twenty-four patients with unilateral disease with regard to age (mean for the group with bilateral involvement, thirty-five years) or duration of symptoms (mean for the group with bilateral involvement, six months). The two groups also were similar with regard to corticosteroid use. One hundred and three (92 percent) of the 112 patients who had bilateral involvement had a history of corticosteroid use, whereas twenty (83 percent) of the twenty-four patients who had unilateral involvement had used corticosteroids.

Of the 149 knees followed for more than two years, eighty-eight (59 percent) did not have a total knee arthroplasty and had a good or excellent Knee Society score. The number of risk factors did not influence the outcome: the outcome was successful for fifty (60 percent) of the eighty-four knees in patients with one risk factor compared with thirty-eight (58 percent) of the sixty-five knees in patients with more than one risk factor. Other demographic parameters, including alcohol, tobacco, and corticosteroid usage, also did not influence the prognosis.

The mean objective Knee Society score at presentation was 75 points (range, 20 to 100 points), and the score did not differ between the patients with bilateral involvement and those with unilateral involvement. The mean pain score was 28 points (range, 0 to 50 points). There was no association between the size of the lesion and the presenting score (p > 0.05).

One hundred and one patients (74 percent) had osteonecrotic involvement of at least one other large joint, and ten of them had involvement of at least three other joints; nine of these ten patients had been taking corticosteroids. Ninety-one patients (67 percent) had involvement of one hip or both, twenty-six patients (19 percent) had involvement of one shoulder or both, and twenty patients (15 percent) had involvement of one ankle or both. Of the 101 patients who had involvement of at least one other joint, eighteen presented with knee symptoms.

Radiographic Evaluation

One hundred and fifty-four knees (62 percent) presented with only distal femoral involvement, four (2 percent) had involvement of only the proximal part of the tibia, and ninety (36 percent) presented with both femoral and tibial involvement. Sixty-nine knees (28 percent) had normal radiographic findings (stage-I disease) at the time of presentation, and the diagnosis was made on the basis of magnetic resonance imaging (Fig. 5). The plain radiographs made at presentation revealed stage-II disease (sclerosis with cystic changes) in 106 knees (43 percent) (Fig. 6). Thirty-eight knees (15 percent) presented with stage-III disease (subchondral collapse) (Fig. 7), and thirty-five knees (14 percent) presented with stage-IV disease (arthritic changes on both sides of the joint) (Fig. 8).

Bone-Scanning

Of the fifty-six knees that were evaluated with bone scans, forty (71 percent) had increased uptake in the distal part of the femur or the proximal part of the tibia, or both. All had either a magnetic resonance image or a core biopsy within three months after the bone scan to confirm the diagnosis of osteonecrosis.

Results of Treatment

We analyzed the results of various treatment methods that were employed for the 149 knees with greater than two years of follow-up. The different treatment options included nonoperative treatment (fifty-one knees); core decompression only (sixty-six knees); core decompression followed by repeat core decompression (fifteen knees); core decompression followed by arthroscopic dÇ¢ridement (ten knees); and total knee arthroplasty as a primary procedure (seven knees), after failed nonoperative treatment (thirty-one knees), and after failed core decompression (ten knees).

Nonoperative Treatment

Ten of the fifty-one knees that were treated nonoperatively were asymptomatic at presentation. Eight of the ten remained asymptomatic, and two became symptomatic at six and eight years. Radiographs of these two knees revealed femoral condylar collapse, and because of radiographic findings of late-stage disease these knees were treated with a total knee arthroplasty rather than a core decompression. Forty-one knees were initially treated nonoperatively because the symptoms had resolved during the first three months after the onset. Twelve of these forty-one knees were not subsequently treated with total knee arthroplasty. None of the twelve knees required a procedure, though only eight were doing well clinically at the time of follow-up, at a mean of eight years (range, two to twenty-three years). The results of clinical survivorship analysis of these nonoperatively treated knees is shown in Figure 9.

Core Decompression Only

Of the ninety-one knees (fifty-seven patients) managed with core decompression, seventy-two (79 percent) had a good or excellent Knee Society score (sixty-one had an excellent score and eleven had a good score) at a mean of seven years (range, two to twenty-four years). Fifteen of these seventy-two knees were doing well after repeat core decompression or arthroscopic dÇ¢ridement, as will be detailed in the next section. Figure 10 shows the survivorship curve for knees treated with core decompression.

Core Decompression Followed by Repeat Core Decompression or by Arthroscopic DÇ¢ridement

Nine of the fifteen knees that underwent repeat core decompression were doing well clinically at a mean of four years (range, two to nine years). After arthroscopic dÇ¢ridement of ten knees, six were clinically satisfactory at a mean of three years (range, two to five years). The ten knees that remained severely symptomatic after either repeat core decompression (six knees) or arthroscopy (four knees) all were treated with a total knee arthroplasty.

Total Knee Arthroplasty

Forty-eight knees underwent total knee arthroplasty (seven, initially; thirty-one, after failed nonoperative treatment; and ten, after failed core decompression). At a mean of nine years (range, two to twenty-four years), thirty-four (71 percent) of these knees had a successful clinical outcome. Of the fourteen knees with an unsuccessful result, nine underwent a revision arthroplasty, two were fused, two were painful and had low scores, and one underwent an above-the-knee amputation after multiple infections. All nine knees in patients who did not have a diagnosis of systemic lupus erythematosus had a successful clinical outcome at a mean of seven years (range, two to fourteen years) after the arthroplasty. Figure 11 shows the survivorship curve for the knees treated with total knee arthroplasty.

Findings on Magnetic Resonance Imaging Related to Outcome

Location of the Lesion

Evaluation of the magnetic resonance images of seventy-three knees with a femoral lesion revealed that the quadrant in which the lesion was located (medial or lateral, or both) did not affect the clinical outcome. Evaluation of the magnetic resonance images to determine if the lesion was located in the diaphysis, the metaphysis, or the epiphysis revealed that lesions involving the epiphyseal periarticular region were associated with a higher failure rate than were those in the other regions (p < 0.01). Of sixty-five knees with a lesion involving the femoral epiphysis, ten (15 percent) had a failure of treatment, whereas none of the eight knees with a lesion that did not extend into this region had a failure.

Forty-four knees with a tibial lesion were evaluated in the same manner. The quadrant of the lesion had no prognostic importance. However, lesions in the epiphyseal region were associated with a 26 percent failure rate (eight failures in thirty-one knees) compared with no failures in the thirteen knees with a lesion that did not extend into this region (p < 0.001).

Size of the Lesion

Seventy-four knees with a femoral lesion were evaluated with magnetic resonance imaging to determine if the size of the lesion could predict the clinical outcome. Thirty-five knees had a small femoral lesion (less than ten cubic centimeters), twelve had a medium lesion (ten to twenty cubic centimeters), and twenty-seven had a large femoral lesion (more than twenty cubic centimeters). As determined with this method, the size of the femoral lesion had no prognostic importance (p > 0.05). When the size of the epiphyseal lesion was determined on the basis of the combined necrotic angle, it was found that large lesions (250 degrees or greater) had a worse prognosis than medium and small lesions (less than 250 degrees) (p < 0.05).

Forty-two knees with a tibial lesion were evaluated with magnetic resonance imaging to determine if the size of the tibial lesion was associated with the clinical outcome. Thirty knees had a small tibial lesion, three had a medium lesion, and nine had a large lesion. There was no significant relationship between the size of the tibial lesion as measured with this method and the clinical outcome (p > 0.05). When the size of the epiphyseal lesion was determined on the basis of the combined necrotic angle, it was found that large lesions were associated with significantly worse outcomes than were medium and small lesions (p < 0.05).

Discussion

Introduction | Materials and Methods | Results | Discussion | References

This study involved a large series of patients with atraumatic osteonecrosis of the knee. Previous reports have characterized atraumatic osteonecrosis of the distal part of the femur without describing the associated findings in the proximal part of the tibia. To our knowledge, there have been only case studies describing this disease in the proximal part of the tibia. In the present series, ninety-four (38 percent) of 248 knees had tibial involvement. It is important not to neglect this finding when planning treatment of the distal femoral lesion. Isolated proximal tibial involvement with osteonecrosis is rare (a prevalence of 2 percent in the present series).

In 101 patients (74 percent), the osteonecrosis involved at least one other large joint, the most common of which was the hip (67 percent) followed by the shoulder (19 percent) and then by the ankle (15 percent). The prevalence of hip osteonecrosis in patients with knee involvement is greater than the prevalence in patients with ankle involvement (54 percent)¹¹, but the prevalence of hip osteonecrosis in patients with knee involvement is lower than the prevalence in patients with shoulder involvement (greater than 90 percent)^52,54. We recommend a thorough investigation of other symptomatic joints in patients presenting with osteonecrosis of the knee, as eighteen patients with osteonecrosis of the hip, shoulder, or ankle in this study presented with a knee lesion initially. In comparison with patients with osteonecrosis of the hip, patients with knee disease were more likely to be female (an 80 percent prevalence in the present series), to have a history of corticosteroid use (90 percent), and to have bilateral knee involvement (82 percent). In contrast, a review of the cases of 1413 patients with hip osteonecrosis in eight studies^{16,37,60,62,65-68} revealed that 32 percent (456) were female, 31 percent (445) had a history of corticosteroid use, and 39 percent (557) had bilateral involvement.

Various authors have advocated the use of bone-scanning in the early diagnosis of distal femoral osteonecrosis^4,23,25. The advantage of this technique could not be confirmed in the present study, in which only 71 percent of the knees had positive uptake on bone scans, reflecting the lack of sensitivity of this imaging modality with regard to the detection of these lesions. We recommend magnetic resonance imaging^8,27,59 as a screening modality.

Large epiphyseal lesions were associated with a worse prognosis than were small or medium lesions, a finding that has been reported in the hip⁵¹ as well as in knees with spontaneous osteonecrosis^34,45. The location of the lesion (medial or lateral, or both) was not related to the outcome. However, femoral or tibial lesions that involved the periarticular epiphyseal area were found to have a worse prognosis than lesions isolated to the metaphyseal or diaphyseal region (p < 0.05). The finding that the size of the diaphyseal and metaphyseal lesions was not prognostic is not surprising since these lesions are far away from the epiphysis and do not present an untoward influence on the knee joint. It is only when the lesion extends toward the knee joint that it influences the prognosis. These findings, in terms of the size of the lesion, then become comparable with those in spontaneous osteonecrosis.

Treatment options for the knee have included nonoperative methods such as restricted weight-bearing, analgesics, and observation^1,8,30,49,55. Unfortunately, such methods have led to a greater than 80 percent rate of clinical failure (twenty-six of thirty-two knees) in at least one prior report⁵⁵. In the present study, only eight of forty-one symptomatic knees that were treated nonoperatively had a successful clinical outcome; thus, we now recommend the performance of an immediate core decompression rather than waiting for three months to see if symptoms abate. We do recommend observation of asymptomatic knees, as eight of the ten asymptomatic knees in our study had no additional symptoms or signs of radiographic progression.

There have been few reports specifically focusing on the surgical treatment of this problem. Some authors have advocated joint-preserving surgical modalities, including core decompression^38,52,55, high tibial osteotomy⁴², vascularized bone-grafting⁵⁷, and resurfacing with osteoarticular allograft^{5,21,22,24,56}. Core decompression^38,52,55 has been reported to give symptomatic relief. This finding is in agreement with those in the present study, in which seventy-two (79 percent) of ninety-one knees had a successful clinical outcome at a mean of seven years. Compared with the alternatives, this modality appears to be a realistic treatment method for the relief of symptoms in the knee. Recently, arthroscopy has been proposed as a treatment alternative^15,21,41,50. Six of ten knees in the present study were managed effectively with arthroscopy after failed core decompression. The exact role of this modality remains to be defined, but it appears to be appropriate for relief of pain and mechanical symptoms as well as as a means with which to delay arthroplasty in selected patients. Unfortunately, all of these procedures have been most successful for patients with early stages of the disease (before condylar collapse). Consequently, patients who have secondary arthritis from severe disease have few options other than either unicompartmental⁴⁸ or total knee arthroplasty^7,36,54,59 despite their young age.

Total knee arthroplasty is the only choice for the treatment of stage-IV lesions (arthritic changes on both sides of the joint), and this procedure has been used when other treatment regimens have failed. There have been few reports concerning prosthetic replacement for the treatment of this condition. Some studies have included a few patients with this specific diagnosis or have included patients with spontaneous or idiopathic osteonecrosis. Bergman and Rand⁷ reported the early results (at a mean of four years) of total knee arthroplasty in thirty-six patients (thirty-eight knees) with either spontaneous osteonecrosis (twenty-seven knees) or atraumatic osteonecrosis (eleven knees). A good or excellent clinical result was found in thirty-three knees (87 percent). They predicted a five-year implant survival rate of 85 percent using revision as the end point and a 68 percent survival rate using revision because of moderate or severe pain as the end point.

We⁵⁴ previously reported on thirty-one total knee replacements for atraumatic osteonecrosis in patients less than fifty years of age (mean age, thirty-six years), all of whom had a history of corticosteroid use. At a mean of 8.2 years (range, two to sixteen years), 55 percent (seventeen) had a good or excellent result. Thirty-nine percent (twelve) of the knees had to be revised because of aseptic loosening, and 10 percent (three) were revised because of deep infection. All six knees in patients who did not have systemic lupus erythematosus had an excellent clinical result, whereas only 44 percent (eleven) of twenty-five in patients with systemic lupus erythematosus had a good clinical outcome. Factors inherent in any patient with osteonecrosis include areas of poor-quality bone that could compromise prosthetic fixation. Most of these patients were taking corticosteroids, which might have further compromised the bone quality needed for prosthetic fixation. The appropriate use of cement and a long stem might improve results in the future.

In the present study, 29 percent of the knees had a poor result at the time of follow-up at a mean of nine years. These results contrast with those of other studies of total knee arthroplasty in patients younger than fifty years of age, in which the ten-year implant survival rates have been higher than 90 percent^{6,10,12,33,58}. For example, Duffy et al.¹² reported a ten-year implant survival rate of 99 percent and predicted a fifteen-year rate of 95 percent at a minimum ten years (mean, thirteen years) after total knee arthroplasty performed in patients younger than fifty years of age (mean age, forty-three years) at the Mayo Clinic.

In summary, we have reported on a large series of patients with atraumatic osteonecrosis of the knee. Ninety percent had a history of corticosteroid use. Seventy-four percent had involvement of at least one other large joint. Thus, whenever a patient with osteonecrosis of the knee is seen, careful evaluation of the hip is mandatory. Magnetic resonance imaging was useful for diagnosis as well as prognosis (large epiphyseal lesions had worse outcomes), whereas bone-scanning was less effective. Only 31 percent of the patients responded well to nonoperative treatment. Seventy-one percent had a good outcome after total knee arthroplasty; those results are less successful than those of total knee arthroplasty done for the treatment of other diseases.

References

Introduction | Materials and Methods | Results | Discussion | References

Abeles, M.; Urman, J. D.; and Rothfield, N. F.: Aseptic necrosis of bone in systemic lupus erythematosus. Relationship to corticosteroid therapy. Arch. Intern. Med.,138: 750-754, 1978.138750 1978 [PubMed]

Aglietti, P.; Insall, J. N.; Buzzi, R.; and Deschamps, G.: Idiopathic osteonecrosis of the knee. Aetiology, prognosis and treatment. J. Bone and Joint Surg.,65-B(5): 588-597, 1983.65-B(5)588 1983

Ahlbäck, S.; Bauer, G. C.; and Bohne, W. H.: Spontaneous osteonecrosis of the knee. Arthrit. and Rheumat.,11: 705-733, 1968.11705 1968

Al-Rowaih, A.; Björkengren, A.; Egund, N.; Lindstrand, A.; Wingstrand, H.; and Thorngren, K.-G.: Size of osteonecrosis of the knee. Clin. Orthop.,287: 68-75, 1993.28768 1993 [PubMed]

Bayne, O.; Langer, F.; Pritzker, K. P.; Houpt, J.; and Gross, A. E.: Osteochondral allografts in the treatment of osteonecrosis of the knee. Orthop. Clin. North America,16: 727-740, 1985.16727 1985

Bellemans, J.; Victor, J.; Westhovens, R.; Dequeker, H.; and Fabry, G.: Total knee arthroplasty in the young rheumatoid patient. Acta Orthop. Belgica,63: 189-193, 1997.63189 1997

Bergman, N. R.,, and Rand, J. A.: Total knee arthroplasty in osteonecrosis. Clin. Orthop.,273: 77-82, 1991.27377 1991 [PubMed]

Björkengren, A. G.; AlRowaih, A.; Lindstrand, A.; Wingstrand, H.; Thorngren, K. G.; and Pettersson, H.: Spontaneous osteonecrosis of the knee: value of MR imaging in determining prognosis. AJR: Am. J. Roentgenol.,154: 331-336, 1990.154331 1990 [PubMed]

Cruess, R. L.: Steroid-induced osteonecrosis: a review. Canadian J. Surg.,24: 567-571, 1981.24567 1981

Dalury, D. F.; Ewald, F. C.; Christie, M. J.; and Scott, R. D.: Total knee arthroplasty in a group of patients less than 45 years of age. J. Arthroplasty,10: 598-602, 1995.10598 1995 [PubMed]

Delanois, R. E.; Mont, M. A.; Yoon, T. R.; Mizell, M.; and Hungerford, D. S.: Atraumatic osteonecrosis of the talus. J. Bone and Joint Surg.,80-A: 529-536, April 1998.80-A529 1998

Duffy, G. P.; Trousdale, R. T.; and Stuart, M. J.: Total knee arthroplasty in patients 55 years old or younger. 10- to 17-year results. Clin. Orthop.,356: 22-27, 1998.35622 1998 [PubMed]

Ecker, M. L., and Lotke, P. A.: Spontaneous osteonecrosis of the knee. J. Am. Acad. Orthop. Surgeons,,2: 173-178, 1994.2173 1994

Ecker, M. L., and Lotke, P. A.: Osteonecrosis of the medial part of the tibial plateau. J. Bone and Joint Surg.,77-A: 596-601, April 1995.77-A596 1995

Ewald, F. C., on behalf of the Knee Society: The Knee Society total knee arthroplasty roentgenographic evaluation and scoring system. Clin. Orthop.,248: 9-12, 1989.2489 1989 [PubMed]

Fairbank, A. C.; Bhatia, D.; Jinnah, R. H.; and Hungerford, D. S.: Long-term results of core decompression for ischemic necrosis of the femoral head. J. Bone and Joint Surg.,77-B(1): 42-49, 1995.77-B(1)42 1995

Ficat, R. P., and Arlet, J.: Functional investigation of bone under normal conditions. In Ischemia and Necrosis of Bone, pp. 29-52. Edited by D. S. Hungerford. Baltimore, Williams and Wilkins, 1980.

Ficat, R. P., and Arlet, J.: Necrosis of the femoral head. In Ischemia and Necrosis of Bone, pp. 171-182. Edited by D. S. Hungerford. Baltimore, Williams and Wilkins, 1980.

Ficat, R. P.: NÇ£rose aseptique de la tÌ´e femoral. PathogÇ®ie: la thÇ¯rie circulatoire. Acta Orthop. Belgica,47: 198-199, 1981.47198 1981

Ficat, R. P.: Idiopathic bone necrosis of the femoral head. Early diagnosis and treatment. J. Bone and Joint Surg.,67-B(1): 3-9, 1985.67-B(1)3 1985

Flynn, J. M.; Springfield, D. S.; and Mankin, H. J.: Osteoarticular allografts to treat distal femoral osteonecrosis. Clin. Orthop.,303: 38-43, 1994.30338 1994 [PubMed]

Ganel, A.; Israeli, A.; Horoszowski, H.; and Farine, I.: Osteochondral graft in the treatment of osteonecrosis of the femoral condyle. J. Am. Geriat. Soc.,29: 186-188, 1981.29186 1981 [PubMed]

Greyson, N. D.; Lotem, M. M.; Gross, A. E.; and Houpt, J. B.: Radionuclide evaluation of spontaneous femoral osteonecrosis. Radiology,142: 729-735, 1982.142729 1982 [PubMed]

Gross, A. E.; McKee, N. H.; Pritzker, K. P.; and Langer, F.: Reconstruction of skeletal deficits of the knee. A comprehensive osteochondral transplant program. Clin. Orthop.,174: 96-106, 1983.17496 1983 [PubMed]

Gupta, S. M.; Foster, C. R.; and Kayani, N.: Usefulness of SPECT in the early detection of avascular necrosis of the knee. Clin. Nucl. Med.,12: 99-102, 1987.1299 1987 [PubMed]

Havel, P. E.; Ebraheim, N. A.; and Jackson, W. T.: Steroid-induced bilateral avascular necrosis of the lateral femoral condyles. A case report. Clin. Orthop.,243: 166-168, 1989.243166 1989 [PubMed]

Healy, W. L.: Osteonecrosis of the knee detected only by magnetic resonance imaging. Orthopedics,14: 703-704, 1991.14703 1991 [PubMed]

Hirota, Y.; Hirohata, T.; Fukuda, K.; Mori, M.; Yanagawa, H.; Ohno, Y.; and Sugioka, Y.: Association of alcohol intake, cigarette smoking, and occupational report status with the risk of idiopathic osteonecrosis of the femoral head. Am. J. Epidemiol.,137: 530-538, 1993.137530 1993 [PubMed]

Houpt, J. B.; Alpert, B.; Lotem, M.; Greyson, N. D.; Pritzker, K. P.; Langer, F.; and Gross, A. E.: Spontaneous osteonecrosis of the medial tibial plateau. J. Rheumatol.,9: 81-90, 1982.981 1982 [PubMed]

Houpt, J. B.; Pritzker, K. P.; Alpert, B.; Greyson, N. D.; and Gross, A. E.: Natural history of spontaneous osteonecrosis of the knee (SONK): a review. Sem. Arthrit. and Rheumat.,13: 212-227, 1983.13212 1983

Hungerford, D. S., and Zizic, T. M.: Alcoholism associated ischemic necrosis of the femoral head. Early diagnosis and treatment. Clin. Orthop.,130: 144-153, 1978.130144 1978 [PubMed]

Hungerford, D. S.,, and Zizic, T. M., II: The treatment of ischemic necrosis of bone in systemic lupus erythematosus. Medicine,59: 143-148, 1980.59143 1980 [PubMed]

Insall, J. N., and Kelly, M.: The total condylar prosthesis. Clin. Orthop.,205: 43-48, 1986.20543 1986 [PubMed]

Insall, J. N.; Dorr, L. D.; Scott, R. D.; and Scott, W. N.: Rationale of the Knee Society clinical rating system. Clin. Orthop.,248: 13-14, 1989.24813 1989 [PubMed]

Insall, J. N.; Aglietti, P.; Bullough, P. G.; and Windsor, R. E.: Osteonecrosis. In Surgery of the Knee, edited by J. N. Insall, R. E. Windsor, W. N. Scott, M. A. Kelly, and P. Aglietti. Ed. 2, pp. 609-633. New York, Churchill Livingstone, 1993.

Isono, S. S.; Woolson, S. T.; and Schurman, D. J.: Total joint arthroplasty for steroid-induced osteonecrosis in cardiac transplant patients. Clin. Orthop.,217: 201-208, 1987.217201 1987 [PubMed]

Jacobs, B.: Alcoholism-induced bone necrosis. New York State J. Med.,92: 334-338, 1992.92334 1992

Jacobs, M. A.; Loeb, P.; and Hungerford, D. S.: Core decompression of the distal femur in the treatment for avascular necrosis of the knee. J. Bone and Joint Surg.,71-B(4): 583-587, 1989.71-B(4)583 1989

Kelman, G. J.; Williams, G. W.; Colwell, C. W., Jr.; and Walker, R. H.: Steroid-related osteonecrosis of the knee. Two case reports and a literature review. Clin. Orthop.,257: 171-176, 1990.257171 1990 [PubMed]

Kerboul, M.; Thomine, J.; Postel, M.; and Merle d'AubignÆ¬ R.: The conservative surgical treatment of idiopathic aseptic necrosis of the femoral head. J. Bone and Joint Surg.,56-B(2): 291-296, 1974.56-B(2)291 1974

Koshino, T.; Okamoto, R.; Takamura, K.; and Tsuchiya, K.: Arthroscopy in spontaneous osteonecrosis of the knee. Orthop. Clin. North America,10: 609-618, 1979.10609 1979

Koshino, T.: The treatment of spontaneous osteonecrosis of the knee by high tibial osteotomy with and without bone-grafting or drilling of the lesion. J. Bone and Joint Surg.,64-A: 47-58, Jan 1982.64-A47 1982

LaPrade, R. F., and Noffsinger, M. A.: Idiopathic osteonecrosis of the patella: an unusual cause of pain in the knee. A case report. J. Bone and Joint Surg.,72-A: 1414-1418, Oct 1990.72-A1414 1990

Lotke, P. A., and Ecker, M. L.: Osteonecrosis-like syndrome of the medial tibial plateau. Clin. Orthop.,176: 148-153, 1983.176148 1983 [PubMed]

Lotke, P. A., and Ecker, M. L.: Current concepts review. Osteonecrosis of the knee. J. Bone and Joint Surg.,70-A: 470-473, March 1988.70-A470 1988

Mahood, J.; Bogoch, E.; Urowitz, M.; Hastings, D.; and Gladman, D.: Osteonecrosis of the knee in systemic lupus erythematosus. In Proceedings of the Canadian Orthopaedic Association. J. Bone and Joint Surg.,72-B(4): 541, 1990.72-B(4)541 1990

Mankin, H. J.: Nontraumatic necrosis of bone (osteonecrosis). New England J. Med.,326: 1473-1479, 1992.3261473 1992

Marmor, L.: Unicompartmental arthroplasty for osteonecrosis of the knee joint. Clin. Orthop.,294: 247-253, 1993.294247 1993 [PubMed]

Meyers, M. H.; Akeson, W.; and Convery, F. R.: Resurfacing of the knee with fresh osteochondral allograft. J. Bone and Joint Surg.,71-A: 704-710, June 1989.71-A704 1989

Miller, G. K.; Maylahn, D. J.; and Drennan, D. B.: The treatment of idiopathic osteonecrosis of the medial femoral condyle with arthroscopic debridement. Arthroscopy,2: 21-29, 1986.221 1986 [PubMed]

Mont, M. A., and Hungerford, D. S.: Current concepts review. Non-traumatic avascular necrosis of the femoral head. J. Bone and Joint Surg.,77-A: 459-474, March 1995.77-A459 1995

Mont, M. A., and Hungerford, D. S.: Osteonecrosis of the shoulder, knee, and ankle. In Osteonecrosis: Etiology, Diagnosis, and Treatment, pp. 429-436. Edited by J. R. Urbaniak and J. P. Jones. Rosemont, Illinois, American Academy of Orthopaedic Surgeons, 1997.

Mont, M. A.; Tomek, I. M.; and Hungerford, D. S.: Core decompression for avascular necrosis of the distal femur. Long-term follow-up. Clin. Orthop.,334: 124-130, 1997.334124 1997 [PubMed]

Mont, M. A.; Myers, T. H.; Krackow, K. A.; and Hungerford, D. S.: Total knee arthroplasty for corticosteroid associated avascular necrosis of the knee. Clin. Orthop.,338: 124-130, 1997.338124 1997 [PubMed]

Motohashi, M.; Morii, T.; and Koshino, T.: Clinical course and roentgenographic changes of osteonecrosis in the femoral condyle under conservative treatment. Clin. Orthop.,266: 156-161, 1991.266156 1991 [PubMed]

Ochi, M.; Kimori, K.; Sumen, Y.; and Ikuta, Y.: A case of steroid-induced osteonecrosis of femoral condyle treated surgically. Clin. Orthop.,,312: 226-231, 1995.312226 1995

Pollack, M. S.; Dalinka, M. K.; Kressel, H. Y.; Lotke, P. A.; and Spritzer, C. E.: Magnetic resonance imaging in the evaluation of suspected osteonecrosis of the knee. Skel. Radiol.,16: 121-127, 1987.16121 1987

Ranawat, C. S., and Boachie-Adjei, O.: Survivorship analysis and results of total condylar knee arthroplasty. Eight- to 11-year follow-up period. Clin. Orthop.,226: 6-13, 1988.2266 1988 [PubMed]

Ritter, M. A.; Eizember, L. E.; Keating, E. M.; and Faris, P. M.: The survival of total knee arthroplasty in patients with osteonecrosis of the medial condyle. Clin. Orthop.,267: 108-114, 1991.267108 1991 [PubMed]

Saito, S.; Ohzono, K.; and Ono, K.: Joint-preserving operations for idiopathic avascular necrosis of the femoral head. J. Bone and Joint Surg.,70-B(1): 78-84, 1988.70-B(1)78 1988

Sasaki, T.; Yagi, T.; Monji, J.; Yasuda, K.; Masuda, T.; Fukazawa, M.; and Konno, H.: Steroid-induced osteonecrosis of the femoral condyle: a clinical study of eighteen knees in ten patients. Nippon Seikeigeka Gakkai Zasshi,60: 361-372, 1986.60361 1986 [PubMed]

Sotereanos, D. G.; Plakseychuk, A. Y.; and Rubash, H. E.: Free vascularized fibula grafting for the treatment of osteonecrosis of the femoral head. Clin. Orthop.,344: 243-256, 1997.344243 1997 [PubMed]

Soucacos, P. N.; Beris, A. E.; Xenakis, T. H.; and Malizos, K. N.: Knee osteonecrosis: distinguishing features and differential diagnosis. In Osteonecrosis: Etiology, Diagnosis, and Treatment, pp. 413-424. Edited by J. R. Urbaniak and J. P. Jones. Rosemont, Illinois, American Academy of Orthopaedic Surgeons, 1997.

Soucacos, P. N.; Xenakis, T. H.; Beris, A. E.; Soucacos, P. K.; and Georgoulis, A.: Idiopathic osteonecrosis of the medial femoral condyle. Classification and treatment. Clin. Orthop.,341: 82-89, 1997.34182 1997 [PubMed]

Steinberg, M. E.; Larcom, P. G.; Stafford, B. B.; Horick, W. B.; Corces, A.; Bands, R. E.; and Hartman, K. M.: Treatment of osteonecrosis of the femoral head by core decompression, bone grafting, and electrical stimulation. In Osteneocrosis: Etiology, Diagnosis, and Treatment, pp. 293-299. Edited by J. R. Urbaniak and J. P. Jones. Rosemont, Illinois, American Academy of Orthopaedic Surgeons, 1997.

Stulberg, B. N.; Bauer, T. W.; Belhobek, G. H.; Levine, M.; and Davis, A.: A diagnostic algorithm for osteonecrosis of the femoral head. Clin. Orthop.,249: 176-182, 1989.249176 1989 [PubMed]

Takatori, Y.; Kokubo, T.; Ninomiya, S.; Nakamura, S.; Morimoto, S.; and Kusaba, I.: Avascular necrosis of the femoral head. Natural history and magnetic resonance imaging. J. Bone and Joint Surg.,75-B(2): 217-221, 1993.75-B(2)217 1993

Tooke, S. M.; Nugent, P. J.; Bassett, L. W.; Nottingham, P.; Mirra, J.; and Jinnah, R.: Results of core decompression for femoral head osteonecrosis. Clin. Orthop.,228: 99-104, 1988.22899 1988 [PubMed]

Williams, J. S., Jr.; Bush-Joseph, C. A.; and Bach, B. R., Jr.: Osteochondritis dissecans of the knee. Am. J. Knee Surg.,11: 221-232, 1998.11221 1998 [PubMed]

Zizic, T. M.,, and Hungerford, D. S.: Avascular necrosis of bone. In Textbook of Rheumatology, edited by W. N. Kelley, E. D. Harris, Jr., S. Ruddy, and C. B. Sledge. Ed. 2, vol. 2, pp. 1689-1710. Philadelphia, W. B. Saunders, 1985..

Zizic, T. M.; Marcoux, C.; Hungerford, D. S.; Dansereau, J. U.; and Stevens, M. B.: Corticosteroid therapy associated with ischemic necrosis of bone in systemic lupus erythematosus. Am. J. Med.,,79: 596-604, 1985.79596 1985

Topics

View Large | Download Slide(.ppt)
Add to My JBJS

Fig. 4:: Treatment algorithm utilized in this study for patients with atraumatic osteonecrosis of the knee.

View Large | Download Slide(.ppt)
Add to My JBJS

Fig. 5:: Magnetic resonance image demonstrating a lesion involving the femoral epiphysis medially and laterally as well as the metaphysis laterally (arrowheads).

View Large | Download Slide(.ppt)
Add to My JBJS

Fig. 7:: Radiograph showing a stage-III lesion in a forty-year-old man with a history of alcohol abuse. The arrowheads denote the area of subchondral collapse.

View Large | Download Slide(.ppt)
Add to My JBJS

Fig. 10:: Curve for knees treated with core decompression.

View Large | Download Slide(.ppt)
Add to My JBJS

Fig. 11:: Curve for knees treated with total knee arthroplasty.

View Larger

Table I: Comparison of Spontaneous and Atraumatic Osteonecrosis

	Spontaneous Osteonecrosis	Atraumatic Osteonecrosis
Synonyms	Primary necrosis, osteonecrosis, SONK, SPONK	Secondary, ischemic, idiopathic, corticosteroid-associated necrosis
Age (yrs.)	>55	<55
Associated risk factors	None	Corticosteroid use, alcohol use, renal disease, other
Condylar involvement	One	Multiple
Laterality	99% unilateral	>80% bilateral
Other joint involvement	Rare	60-90%
Location	Epiphyseal to subchondral surface	Diaphyseal, metaphyseal, epiphyseal

Add to My JBJS

Table I: Comparison of Spontaneous and Atraumatic Osteonecrosis

	Spontaneous Osteonecrosis	Atraumatic Osteonecrosis
Synonyms	Primary necrosis, osteonecrosis, SONK, SPONK	Secondary, ischemic, idiopathic, corticosteroid-associated necrosis
Age (yrs.)	>55	<55
Associated risk factors	None	Corticosteroid use, alcohol use, renal disease, other
Condylar involvement	One	Multiple
Laterality	99% unilateral	>80% bilateral
Other joint involvement	Rare	60-90%
Location	Epiphyseal to subchondral surface	Diaphyseal, metaphyseal, epiphyseal