n Feature Article
Initial Experiences With a New MRI Scoring System for Differentiating Advanced Femoral Osteonecrosis From Tubercular Arthritis Vasanthakumar Venugopal, MD; Alex Daniel Prabhu, MD; Ishrat Afshan, MD; Mehvash Haider, MD; Ekram Ullah, MD
abstract
The purpose of this study was to formulate a magnetic resonance imaging–based scoring system for differentiating tuberculous arthritis from advanced osteonecrosis of the femoral head. Magnetic resonance imaging findings in 18 hips with tuberculous arthritis and 36 hips with advanced osteonecrosis of the femoral head were reviewed retrospectively. Confirmation of tuberculous arthritis was based on enzyme-linked immunosorbent assay and/or synovial biopsy. Osteonecrosis was confirmed either by histopathology or eventual radiographic evidence on follow-up. The findings were analyzed with an emphasis on the changes in femoral head marrow, joint cavity, synovium, acetabulum, and contrast enhancement patterns. A score of 2 was assigned for the presence of each of the following: T2 hyperintensity of the femoral head, synovial hypertrophy, articular cartilage erosion, unilateral involvement of the femoral head, acetabular edema/sclerosis, and enhancement of the involved head. A score of 1 was assigned for each of the following: joint effusion, edema of adjacent marrow, and enhancement of adjacent soft tissue. A cutoff value of 10 of 15 points was considered to be positive for tuberculous arthritis. Sixteen of 18 cases of tuberculous arthritis were correctly identifiable on the basis of this scoring system. The 2 remaining cases had a score of 9. No case of osteonecrosis of the femoral head scored more than 9. A score of 10 for a positive diagnosis of tuberculous arthritis had a sensitivity of 88.89% and specificity of 100%. Positive and negative predictive values were 1 and 0.94, respectively. Statistical significance for each of the parameters and the entire model was established with logistic regression analysis. This new scoring system is effective in solving the imaging dilemma pertinent to endemic regions.
Figure: Sagittal T2-weighted magnetic resonance imaging sequence in a patient with osteonecrosis of the femoral head showing the classic doubleline sign surrounding a focal hypointense geographical lesion.
The authors are from the Department of Radiodiagnosis (VV, IA, EU), Jawaharlal Nehru Medical College, Aligarh Muslim University, Aligarh; the Department of Radiodiagnosis (ADP), Chettinad Medical College Hospital and Research Centre, Chennai; and the Department of Microbiology (MH), Hamdard Institute of Medical Sciences and Research, New Delhi, India. The authors have no relevant financial relationships to disclose. Correspondence should be addressed to: Vasanthakumar Venugopal, MD, Department of Radiodiagnosis, Jawaharlal Nehru Medical College, Aligarh Muslim University, Aligarh, India 202002 ([email protected]). Received: May 23, 2013; Accepted: March 4, 2014; Posted: November 6, 2014. doi: 10.3928/01477447-20141023-59
e1014
ORTHOPEDICS | Healio.com/Orthopedics
n Feature Article
A
pproximately 2% to 5% of all tuberculous lesions involve the skeletal system.1 Tuberculosis of the hip constitutes approximately 40% of all cases of skeletal tuberculosis.1,2 The hip joint is the second most common site of bone involvement after the spine.1-4 Musculoskeletal spread of tuberculosis is usually hematogenous, secondary to primary focus elsewhere. The classical radiographic presentation of tuberculous hip is described by the Phemister triad, which includes severe periarticular osteoporosis, peripherally located osseous erosions, and gradual joint space narrowing. The current authors have encountered myriad radiological findings due to advanced stages of disease and noncompliance with treatment leading to partially treated cases. Osteonecrosis of the femoral head is a well-described entity that in advanced cases can cause collapse of the femoral head, severe joint space narrowing, and secondary osteoarthritic changes.5,6 In tropical regions, tuberculous arthritis of the hip is a close clinical and radiological mimic of advanced osteonecrosis of the femoral head. Although a sizeable volume of literature describes the role of magnetic resonance imaging (MRI) in the diagnosis of osteonecrosis of the femoral head, to the best of the authors’ knowledge, no study has been published on the differentiation of osteonecrosis of the femoral head from tuberculous arthritis on the basis of MRI findings. The authors devised a scoring system based on certain MRI criteria to differentiate advanced osteonecrosis of the femoral head from tuberculous arthritis of the hip joint. Each of these criteria was allotted a score based on its combined sensitivity, specificity, and statistical significance to differentiate the 2 entities. The pathological basis of choosing these findings as criteria and their individual and combined significance are explored in this study.
Materials and Methods This was a retrospective study and was exempted from institutional review board
NOVEMBER 2014 | Volume 37 • Number 11
Table 1
Distribution of MRI Findings Among Hips With Tuberculous Arthritis and Osteonecrosis of the Femoral Head No. Finding
Tuberculous Arthritis (n=18)
Osteonecrosis (n=36)
P
T2 hyperintensity
16
6
.000
Acetabular involvement
15
14
.000
Synovial hypertrophy
18
15
.002
Articular cartilage erosion
16
9
.000
Unilateral hip involvement
14
8
.000
Enhancement of femoral head lesion
15
1
.000
Joint effusion
8
28
.014
Marrow edema
16
33
.740
Soft tissue enhancement
8
5
.013
Abbreviation: MRI, magnetic resonance imaging.
approval. Using hospital records, the authors reviewed the data from cases of advanced osteonecrosis of the femoral head and tuberculous arthritis treated between December 2008 and December 2009. Included were confirmed cases of tuberculosis and osteonecrosis with chronic hip pain and radiographic evidence of femoral head collapse that had undergone MRI evaluation. Suspected cases of tuberculous arthritis were confirmed by enzymelinked immunosorbent assay (ELISA) or synovial biopsy, and cases of osteonecrosis of the femoral head were confirmed by either histopathology (n=24) or eventual radiographic evidence on follow-up (n=12). Histopathological classification of osteonecrosis was based on Arlet and Durroux classification.7 Only type 2 or higher Arlet and Durroux lesions were included in the study. There were 16 cases of tuberculous arthritis and 22 cases of osteonecrosis of the femoral head fulfilling the inclusion criteria. Two patients with tuberculous arthritis and 14 patients with osteonecrosis of the femoral head had bilateral disease. All patients had undergone MRI examination on a 1.5-T superconduct-
ing system (Magnetom Avanto; Siemens, Erlangen, Germany). The patients’ MRIs were retrieved from the authors’ database. Magnetic resonance image findings in 18 hips with tuberculous arthritis and 36 hips with advanced osteonecrosis of the femoral head were reviewed by 2 radiologists (V.V., I.A.) independently. They were blinded to the histopathological findings in these cases. Certain imaging findings pertinent to the pathogenesis of both entities were identified and evaluated for their weight in disease prediction. These findings included hyperintense areas of the femoral head on T2-weighted sequences, hypertrophy of the synovium, articular cartilage erosion, unilateral involvement of the femoral head, acetabular involvement in the form of edema or sclerosis, joint effusion, and adjacent bone marrow edema. Two findings on the post-contrast sequences— enhancement of the femoral head lesion and enhancement of adjacent soft tissue— were also included.
Results This retrospective study included 54 hips in 38 patients (22 males and 16 fe-
e1015
n Feature Article
A
B
C
Figure 1: Axial T2-weighted magnetic resonance imaging sequence showing a hyperintense lesion surrounded by a sclerotic rim within the right femoral head. The left femoral head is normal (A). Post-contrast fat-saturated T1-weighted image of the same patient showing ipsilateral synovial hypertrophy and enhancement (B). Patchy intralesional enhancement seen at a higher level on post-contrast sequences of the same patient. Synovial biopsy of this patient was suggestive of tuberculous arthritis (C).
A
B
C
D
Figure 2: Axial T2-weighted magnetic resonance image of an 18-year-old man showing ill-defined hyperintense areas in the right femoral head extending up to the neck region. Note the opposite femoral head showing hypointense foci (A). Post-contrast axial and coronal images showing patchy intralesional and peripheral enhancement within the right femoral head. The left femoral head shows just peripheral or marginal zone enhancement. This type of enhancement was considered negative according to the authors’ criteria, whereas the patchy enhancement seen on the right side was considered positive. Also note the intensely enhancing hypertrophied synovium on both sides (B). Post-contrast axial and coronal images showing patchy intralesional and peripheral enhancement within the right femoral head. The left femoral head shows just peripheral or marginal zone enhancement. This type of enhancement was considered negative according to the authors’ criteria, whereas the patchy enhancement seen on the right side was considered positive. Also note the intensely enhancing hypertrophied synovium on both sides (C). Histological photograph of Arlet and Durroux type 5 osteonecrosis showing a layer of viable bone on the surface of necrotic bone (hematoxylin and eosin, original magnification ×100) (D).
males). Eighteen hips had tuberculous arthritis and 36 hips had advanced osteonecrosis of the femoral head. Mean patient age was 26.3 years (range, 11-48 years) in the osteonecrosis group and 28 years (range, 14-50 years) in the tuberculous
e1016
arthritis group. Magnetic resonance imaging findings are summarized in Table 1. Signal intensity change pertinent to the diagnosis of tuberculous arthritis was defined as poorly demarcated areas of diffuse or patchy hypointense areas on T1-
weighted images or as hyperintense areas on T2-weighted images in the subchondral region (Figures 1A, 2A, 3A). Such signal intensity changes in the femoral head were seen in 89% (16 of 18) of tuberculous hips and 17% (6 of 36) of osteonecrotic hips. The difference was statistically significant (P
Initial Experiences With a New MRI Scoring System for Differentiating Advanced Femoral Osteonecrosis From Tubercular Arthritis Vasanthakumar Venugopal, MD; Alex Daniel Prabhu, MD; Ishrat Afshan, MD; Mehvash Haider, MD; Ekram Ullah, MD
abstract
The purpose of this study was to formulate a magnetic resonance imaging–based scoring system for differentiating tuberculous arthritis from advanced osteonecrosis of the femoral head. Magnetic resonance imaging findings in 18 hips with tuberculous arthritis and 36 hips with advanced osteonecrosis of the femoral head were reviewed retrospectively. Confirmation of tuberculous arthritis was based on enzyme-linked immunosorbent assay and/or synovial biopsy. Osteonecrosis was confirmed either by histopathology or eventual radiographic evidence on follow-up. The findings were analyzed with an emphasis on the changes in femoral head marrow, joint cavity, synovium, acetabulum, and contrast enhancement patterns. A score of 2 was assigned for the presence of each of the following: T2 hyperintensity of the femoral head, synovial hypertrophy, articular cartilage erosion, unilateral involvement of the femoral head, acetabular edema/sclerosis, and enhancement of the involved head. A score of 1 was assigned for each of the following: joint effusion, edema of adjacent marrow, and enhancement of adjacent soft tissue. A cutoff value of 10 of 15 points was considered to be positive for tuberculous arthritis. Sixteen of 18 cases of tuberculous arthritis were correctly identifiable on the basis of this scoring system. The 2 remaining cases had a score of 9. No case of osteonecrosis of the femoral head scored more than 9. A score of 10 for a positive diagnosis of tuberculous arthritis had a sensitivity of 88.89% and specificity of 100%. Positive and negative predictive values were 1 and 0.94, respectively. Statistical significance for each of the parameters and the entire model was established with logistic regression analysis. This new scoring system is effective in solving the imaging dilemma pertinent to endemic regions.
Figure: Sagittal T2-weighted magnetic resonance imaging sequence in a patient with osteonecrosis of the femoral head showing the classic doubleline sign surrounding a focal hypointense geographical lesion.
The authors are from the Department of Radiodiagnosis (VV, IA, EU), Jawaharlal Nehru Medical College, Aligarh Muslim University, Aligarh; the Department of Radiodiagnosis (ADP), Chettinad Medical College Hospital and Research Centre, Chennai; and the Department of Microbiology (MH), Hamdard Institute of Medical Sciences and Research, New Delhi, India. The authors have no relevant financial relationships to disclose. Correspondence should be addressed to: Vasanthakumar Venugopal, MD, Department of Radiodiagnosis, Jawaharlal Nehru Medical College, Aligarh Muslim University, Aligarh, India 202002 ([email protected]). Received: May 23, 2013; Accepted: March 4, 2014; Posted: November 6, 2014. doi: 10.3928/01477447-20141023-59
e1014
ORTHOPEDICS | Healio.com/Orthopedics
n Feature Article
A
pproximately 2% to 5% of all tuberculous lesions involve the skeletal system.1 Tuberculosis of the hip constitutes approximately 40% of all cases of skeletal tuberculosis.1,2 The hip joint is the second most common site of bone involvement after the spine.1-4 Musculoskeletal spread of tuberculosis is usually hematogenous, secondary to primary focus elsewhere. The classical radiographic presentation of tuberculous hip is described by the Phemister triad, which includes severe periarticular osteoporosis, peripherally located osseous erosions, and gradual joint space narrowing. The current authors have encountered myriad radiological findings due to advanced stages of disease and noncompliance with treatment leading to partially treated cases. Osteonecrosis of the femoral head is a well-described entity that in advanced cases can cause collapse of the femoral head, severe joint space narrowing, and secondary osteoarthritic changes.5,6 In tropical regions, tuberculous arthritis of the hip is a close clinical and radiological mimic of advanced osteonecrosis of the femoral head. Although a sizeable volume of literature describes the role of magnetic resonance imaging (MRI) in the diagnosis of osteonecrosis of the femoral head, to the best of the authors’ knowledge, no study has been published on the differentiation of osteonecrosis of the femoral head from tuberculous arthritis on the basis of MRI findings. The authors devised a scoring system based on certain MRI criteria to differentiate advanced osteonecrosis of the femoral head from tuberculous arthritis of the hip joint. Each of these criteria was allotted a score based on its combined sensitivity, specificity, and statistical significance to differentiate the 2 entities. The pathological basis of choosing these findings as criteria and their individual and combined significance are explored in this study.
Materials and Methods This was a retrospective study and was exempted from institutional review board
NOVEMBER 2014 | Volume 37 • Number 11
Table 1
Distribution of MRI Findings Among Hips With Tuberculous Arthritis and Osteonecrosis of the Femoral Head No. Finding
Tuberculous Arthritis (n=18)
Osteonecrosis (n=36)
P
T2 hyperintensity
16
6
.000
Acetabular involvement
15
14
.000
Synovial hypertrophy
18
15
.002
Articular cartilage erosion
16
9
.000
Unilateral hip involvement
14
8
.000
Enhancement of femoral head lesion
15
1
.000
Joint effusion
8
28
.014
Marrow edema
16
33
.740
Soft tissue enhancement
8
5
.013
Abbreviation: MRI, magnetic resonance imaging.
approval. Using hospital records, the authors reviewed the data from cases of advanced osteonecrosis of the femoral head and tuberculous arthritis treated between December 2008 and December 2009. Included were confirmed cases of tuberculosis and osteonecrosis with chronic hip pain and radiographic evidence of femoral head collapse that had undergone MRI evaluation. Suspected cases of tuberculous arthritis were confirmed by enzymelinked immunosorbent assay (ELISA) or synovial biopsy, and cases of osteonecrosis of the femoral head were confirmed by either histopathology (n=24) or eventual radiographic evidence on follow-up (n=12). Histopathological classification of osteonecrosis was based on Arlet and Durroux classification.7 Only type 2 or higher Arlet and Durroux lesions were included in the study. There were 16 cases of tuberculous arthritis and 22 cases of osteonecrosis of the femoral head fulfilling the inclusion criteria. Two patients with tuberculous arthritis and 14 patients with osteonecrosis of the femoral head had bilateral disease. All patients had undergone MRI examination on a 1.5-T superconduct-
ing system (Magnetom Avanto; Siemens, Erlangen, Germany). The patients’ MRIs were retrieved from the authors’ database. Magnetic resonance image findings in 18 hips with tuberculous arthritis and 36 hips with advanced osteonecrosis of the femoral head were reviewed by 2 radiologists (V.V., I.A.) independently. They were blinded to the histopathological findings in these cases. Certain imaging findings pertinent to the pathogenesis of both entities were identified and evaluated for their weight in disease prediction. These findings included hyperintense areas of the femoral head on T2-weighted sequences, hypertrophy of the synovium, articular cartilage erosion, unilateral involvement of the femoral head, acetabular involvement in the form of edema or sclerosis, joint effusion, and adjacent bone marrow edema. Two findings on the post-contrast sequences— enhancement of the femoral head lesion and enhancement of adjacent soft tissue— were also included.
Results This retrospective study included 54 hips in 38 patients (22 males and 16 fe-
e1015
n Feature Article
A
B
C
Figure 1: Axial T2-weighted magnetic resonance imaging sequence showing a hyperintense lesion surrounded by a sclerotic rim within the right femoral head. The left femoral head is normal (A). Post-contrast fat-saturated T1-weighted image of the same patient showing ipsilateral synovial hypertrophy and enhancement (B). Patchy intralesional enhancement seen at a higher level on post-contrast sequences of the same patient. Synovial biopsy of this patient was suggestive of tuberculous arthritis (C).
A
B
C
D
Figure 2: Axial T2-weighted magnetic resonance image of an 18-year-old man showing ill-defined hyperintense areas in the right femoral head extending up to the neck region. Note the opposite femoral head showing hypointense foci (A). Post-contrast axial and coronal images showing patchy intralesional and peripheral enhancement within the right femoral head. The left femoral head shows just peripheral or marginal zone enhancement. This type of enhancement was considered negative according to the authors’ criteria, whereas the patchy enhancement seen on the right side was considered positive. Also note the intensely enhancing hypertrophied synovium on both sides (B). Post-contrast axial and coronal images showing patchy intralesional and peripheral enhancement within the right femoral head. The left femoral head shows just peripheral or marginal zone enhancement. This type of enhancement was considered negative according to the authors’ criteria, whereas the patchy enhancement seen on the right side was considered positive. Also note the intensely enhancing hypertrophied synovium on both sides (C). Histological photograph of Arlet and Durroux type 5 osteonecrosis showing a layer of viable bone on the surface of necrotic bone (hematoxylin and eosin, original magnification ×100) (D).
males). Eighteen hips had tuberculous arthritis and 36 hips had advanced osteonecrosis of the femoral head. Mean patient age was 26.3 years (range, 11-48 years) in the osteonecrosis group and 28 years (range, 14-50 years) in the tuberculous
e1016
arthritis group. Magnetic resonance imaging findings are summarized in Table 1. Signal intensity change pertinent to the diagnosis of tuberculous arthritis was defined as poorly demarcated areas of diffuse or patchy hypointense areas on T1-
weighted images or as hyperintense areas on T2-weighted images in the subchondral region (Figures 1A, 2A, 3A). Such signal intensity changes in the femoral head were seen in 89% (16 of 18) of tuberculous hips and 17% (6 of 36) of osteonecrotic hips. The difference was statistically significant (P
![[Initial experiences with a new cerebrospinal fluid collection system].](/assets/img/hold.png)
