| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Original Research |
1 Department of Radiology, Ajou University Medical Center, Wonchun Dong,
Yongtong Gu, Suwon 442-721, Republic of Korea.
2 Department of Orthopaedic Surgery, Ajou University Medical Center, Suwon,
South Korea.
3 Department of Radiology, Yonsei University College of Medicine, Seoul,
Korea.
Received December 10, 2006;
accepted after revision February 26, 2007.
Address correspondence to J. H. Cho
(choj{at}ajou.ac.kr).
Abstract
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
MATERIALS AND METHODS. The MRI findings in nine patients with septic arthritis and 11 with transient synovitis were reviewed retrospectively. This study was approved by our institutional review board. The diagnoses were based on findings at physical examination, laboratory studies, and joint aspiration and bacteriologic study. The MRI findings were analyzed with emphasis on the grade of joint effusion, alterations in signal intensity in the soft tissues and bone marrow, and the presence of decreased perfusion at the femoral head.
RESULTS. Low signal intensity on fat-suppressed gadolinium-enhanced T1-weighted coronal MRI suggesting decreased perfusion at the femoral head of the affected hip joint was seen in eight of nine patients with septic arthritis and in two of 11 patients with transient synovitis. Statistically reliable differences (p = 0.005) were found between the two groups. Alterations in signal intensity in the bone marrow were seen in three patients with septic arthritis but in none of the patients with transient synovitis. Decreased perfusion on fat-suppressed gadolinium-enhanced coronal T1-weighted MRI was seen in the six patients with septic arthritis who did not have alterations in signal intensity involving the bone marrow.
CONCLUSION. Decreased perfusion at the femoral epiphysis on fat-suppressed gadolinium-enhanced coronal T1-weighted MRI is useful for differentiating septic arthritis from transient synovitis.
Keywords: hip • MRI • septic arthritis • transient synovitis
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Various clinical, laboratory, and radiographic criteria are used to differentiate septic arthritis from transient synovitis, but no absolute values are sufficient for definitive diagnosis of either condition [2, 5]. MRI has become increasingly important in evaluating musculoskeletal infections in children [14-17]. MRI should aid in the differential diagnosis of these two diseases. While reviewing the MRI results of patients with hip pain, we suspected that a hip affected by sepsis had decreased perfusion in the femoral head, whereas a hip affected by transient synovitis did not. Therefore, we performed a retrospective review of hip MRI of patients with proven septic hip and transient synovitis to evaluate the decrease in perfusion of the femoral head and to find radiologic features that differentiate these two diseases.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Diagnosis
In eight cases of septic arthritis, the diagnosis was made when a bacterial
culture after arthrocentesis had a positive result. In the other case, the
diagnosis was based on the presence of purulent synovial fluid and a high WBC
count [2,
5]. Arthrocentesis was
performed on all patients with septic arthritis. The diagnosis of true septic
arthritis (eight patients) was made when a patient had a positive result of
joint fluid culture or a joint fluid WBC count of at least 50,000 cells/µL
with positive results of blood cultures
[5,
18]. The diagnosis of presumed
septic arthritis (one patient) was made when a patient had a joint fluid WBC
count of at least 50,000 cells/µL with negative results of cultures of
joint aspirate and blood [5,
18].
Synovial fluid was cultured for four of the 11 patients with transient synovitis, and all culture results were negative. Arthrocentesis was not performed in seven patients with transient synovitis. The diagnosis of transient synovitis (11 patients) was made when the patient had a joint fluid WBC count less than 50,000 cells/µL with negative culture results, symptom resolution with conservative treatment, and no further development of disease documented in the medical records.
MRI
One of two 1.5-T MRI units (Signa, GE Healthcare) was used for hip MRI.
Whether a body or a head coil was used depended on the patient's size. The
following imaging parameters were used: TR/TE, 400-800/9-14 for T1-weighted
spin-echo imaging, 3,500-4,000/65-75 (effective) with an echo-train length of
8 for T2-weighted fast spin-echo imaging, and 400-800/9-14 for
gadolinium-enhanced T1-weighted spin-echo imaging. A fat-suppression technique
based on frequency-selective excitation was used for T2-weighted fast
spin-echo imaging and gadolinium-enhanced T1-weighted spin-echo imaging. For
both T1- and T2-weighted images the parameters were a 3.0- to 7.0-mm section
thickness with 0- to 2.5-mm intersection gap, 2-4 signals acquired, and a 256
x 192 or 256 x 256 matrix. The field of view was 180-340 mm
depending on body size and section planes. Axial and coronal images were
obtained with each pulse sequence. Axial and coronal T1-weighted spin-echo
images were obtained after IV administration of gadopentetate dimeglumine
(Magnevist, Schering) 0.1 mmol/L per kilogram of body weight. In our study,
MRI was performed before arthrocentesis in all cases. Contrast-enhanced MRI
was started 1 minute after contrast administration, and acquisition of coronal
T1-weighted spin-echo images was started within 5 minutes (mean, 2.6 minutes)
after contrast administration in all cases.
Two musculoskeletal radiologists blinded to the diagnoses retrospectively evaluated the MRI results in consensus regarding the grade of joint effusion, presence of synovial thickening, alterations in signal intensity of soft tissue and bone marrow in the affected hip joint, and decreased femoral head perfusion in each pulse sequence. Joint effusion was classified according to the system of Lee et al. [17] and Mitchell et al. [19]: 0, no effusion; 1, minimal effusion; 2, effusion surrounding the femoral neck; 3, distention of the capsular recesses. Synovial thickening and decreased femoral head perfusion of the affected hip were compared with the findings in the contralateral hip.
Statistical Analysis
Fisher's exact test and the Mann-Whitney U test were used for the
statistical analysis. Receiver operating characteristic curves were obtained
for decreased perfusion (SPSS 12.0.0).
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Altered Bone Marrow Signal Intensity
Low signal intensity was seen in the bone marrow of the affected hip on
unenhanced T1-weighted images, whereas high signal intensity was seen on
fat-suppressed T2-weighted images and on contrast-enhanced T1-weighted images.
In addition to the change in signal intensity at the femoral head, altered
signal intensity was seen in the bone marrow of the proximal femur or pelvic
bone of the affected hip in three of the nine patients with septic arthritis
(Table 1). No alteration in the
signal intensity of bone marrow was seen in the patients with transient
synovitis, although the difference between the two groups was not significant
(p = 0.074).
Altered Soft-Tissue Signal Intensity
The soft tissues around the affected hip were seen as poorly defined areas
of high signal intensity on fat-suppressed T2-weighted images and on
contrast-enhanced T1-weighted images. Alterations in soft-tissue signal
intensity occurred mainly within the muscles around the affected hip and were
seen in six of the nine patients with septic arthritis and in four of the 11
patients with transient synovitis. The difference between the groups was not
significant (p =0.370).
Grade of Joint Effusion
Joint effusion was seen on coronal fat-suppressed contrast-enhanced images
and was easier to grade on those images (Figs.
4A,
4B,
4C,
4D and
5A,
5B,
5C). Four of the patients with
septic arthritis had grade 3 and five had grade 2 joint effusion. Seven of the
patients with transient synovitis had grade 3 and four had grade 2 effusion of
the affected hip. The difference between the two patient groups was not
significant (p = 0.403).
|
|
|
Laboratory Studies
In the nine patients with septic arthritis, the mean peripheral WBC count
was 14.32 x 109/L (range, 3.44-22.78 x
109/L), the erythrocyte sedimentation rate was 34.3 mm/h (range,
2-74 mm/h), and the C-reactive protein concentration was 6.05 mg/dL (range,
1.02-19.66 mg/dL). In 10 of the patients with transient synovitis, the
respective values were 9.2 x 109/L (range, 3.74-19.42 x
109/L), 31.2 mm/h (range, 7-75 mm/h), and 3.07 mg/dL (range,
0.28-12.9 mg/dL). No laboratory examination was performed for the last patient
with transient synovitis, who recovered with conservative treatment.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
An interesting finding in our patients was decreased perfusion at the femoral epiphysis on fat-suppressed contrast-enhanced T1-weighted MRI. These perfusion defects were more prominent in the epiphyseal region and were easier to identify on coronal than on axial images. Lee et al. [17] reported signal intensity alterations in the bone marrow of the affected hip, especially high signal intensity on fat-suppressed contrast-enhanced T2-weighted MRI. In retrospect, the images in that study also showed decreased perfusion at the femoral head, but the authors did not mention this feature. We believe it might have been difficult for the investigators to recognize the finding because only axial plane images were obtained or because the hip joint was evaluated unilaterally.
Uren and Howman-Giles [21] reported that the cold hip sign is often seen on bone scans of patients referred with irritable hip and that this sign is present when fluid under pressure in the hip joint impairs perfusion of the structures in the joint capsule. Avascular necrosis of the femoral head is a well-known complication of septic arthritis. Vidigal et al. [22] found that early diagnosis and management of infection are mandatory in the care of patients with pyogenic arthritis and increased intraarticular hydrostatic pressure and in those with septic thrombosis of the epiphyseal vessels, which can cause avascular necrosis. In evaluation of bone scans, Williamson and Sistrom [23] found femoral and acetabular photopenia associated with septic arthritis of the hip.
Yang et al. [24] reported that decreased enhancement of the femoral head is frequently seen on images of patients with septic arthritis, but they did not find a statistical difference. In our study, eight patients with septic arthritis had decreased perfusion at the femoral head on fat-suppressed contrast-enhanced T1-weighted MRI. Only two of 11 patients with transient synovitis had decreased perfusion. In our study, MRI was started 1 minute after contrast administration, and acquisition of coronal T1-weighted spin-echo images was started within 5 minutes after contrast administration in all cases. All MRIs in our study were performed before arthrocentesis.
The area under the receiver operating characteristic curve for predicting septic arthritis was 0.854. Although two patients in the transient synovitis group had this finding, decreased perfusion at the femoral head is still useful for differentiating septic arthritis and transient synovitis.
Some investigators [25, 26] have reported a relation between Legg-Calvé-Perthes disease and the incidence and causation of transient synovitis. Hochbergs et al. [27] found that the degree and persistence of synovitis on MRI correlated with the extent of epiphyseal necrosis in patients with Legg-Calvé-Perthes disease. Mahnken et al. [28] found various MRI signal intensities in different phases of Legg-Calvé-Perthes disease and found low signal intensity on contrast-enhanced T1-weighted MRI. We postulate that there is a relation between the decreased signal intensity with enhancement at the femoral epiphysis that we observed and perfusion of the femoral head. Kesteris et al. [29] and Wingstrand et al. [30] reported the significance of increased intracapsular pressure related to hip joint effusion in patients with transient synovitis. Those investigators found that radionuclide uptake returned to normal depending on the amount of hip effusion after arthrocentesis. We recommend performing MRI before arthrocentesis in patients with irritable hip because reports [21, 29, 30] suggest that intracapsular pressure, effusion, and impaired perfusion of the hip are related to one another.
Caird et al. [18] and Luhmann et al. [2] isolated various organisms from cultures of patients with confirmed septic arthritis of the hip. These organisms included coagulase-negative staphylococci, Staphylococcus aureus, Staphylococcus epidermidis, Streptococcus pneumoniae, Streptococcus viridans, Haemophilus influenza type b, Pseudomonas aeruginosa, Micrococcus species, Corynebacterium species, and Abiotrophia species. In our study, one adult with septic arthritis of the hip had decreased perfusion of the femoral epiphysis on MRI, and Candida tropicalis grew on culture of aspirate. C. tropicalis arthritis has been found in patients with hematologic malignant disease, and septic arthritis can be the first sign of C. tropicalis infection [31-34]. Our patient later proved to have hemophagocytic lymphohistiocytosis.
Lee et al. [17] found altered bone marrow signal intensity in six of nine patients with septic arthritis. Although the difference in bone marrow signal intensity between the two groups in our study was not significant (p = 0.074), we observed signal intensity alteration in three of nine patients with septic arthritis. The observed differences between the two groups in numbers of patients with altered soft-tissue signal intensity, grade of effusion, and synovial thickening of the affected hip were not statistically significant in our study, as was true in the study by the Lee et al.
Our study had limitations. It was a retrospective study of a small number of patients. In addition, because our hospital is a tertiary referral center, it was impossible to confirm whether our patients had been treated with oral antibiotics before their arrival at our hospital. Thus we could not exclude the possibility that the culture results were influenced by previous treatment elsewhere.
In conclusion, our study showed the potential usefulness of decreased perfusion at the femoral epiphysis in differentiating septic arthritis of the hip from transient synovitis. This sign may be useful for predicting decreased blood perfusion in patients with irritable hip. We recommend performing coronal T1-weighted MRI immediately after contrast administration to evaluate the difference in blood perfusion between the hips before arthrocentesis. The observation of decreased perfusion at the femoral epiphysis on MRI suggests a need for further evaluation and treatment, including arthrocentesis.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  S.-N. Kang, T. Sanghera, J. Mangwani, J. M. H. Paterson, and M. Ramachandran The management of septic arthritis in children: SYSTEMATIC REVIEW OF THE ENGLISH LANGUAGE LITERATURE J Bone Joint Surg Br, September 1, 2009; 91-B(9): 1127 - 1133. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
