| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1
Department of Radiology, Children's Hospital, Harvard Medical School, 300
Longwood Ave., Boston, MA 02115.
2
Department of Radiology, Massachusetts General Hospital, Harvard Medical
School, 10 Fruit St., Boston, MA 02215.
Received May 18, 2001;
accepted after revision September 25, 2001.
Supported in part by grant AR 42396-05 from the National Institutes of
Health.
Abstract
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
MATERIALS AND METHODS. We evaluated 111 children with physeal bone bridges (median age, 11.4 years) using MR imaging to analyze bridge size, location in physis, signal intensity, growth recovery lines, avascular necrosis, and metaphyseal cartilage tongues. Fifty-eight patients underwent fat-suppressed 3D spoiled gradient-recalled echo imaging with physeal mapping. The cause, bone involved, radiographic appearance, and surgical interventions (60/111) were also correlated. Data were analyzed with the two-tailed Fisher's exact test.
RESULTS. Posttraumatic bridges, accounting for 70% (78/111) of patients, were most often distal, especially of the tibia (n = 43) and femur (n = 14), whereas those due to the other miscellaneous causes were more frequently proximal (p < 0.0001). The position of the bridge in the physis was related to the bone involved (p < 0.0001). Sixty-five percent of distal tibial bridges involved the anteromedial physis, whereas 60% of the distal femoral arrests were central. Larger bridges had higher T1 signal intensity (p < 0.008). Oblique growth recovery lines were seen exclusively with bridges involving the peripheral physis (p = 0.002) and smaller, more potentially resectable bridges. Metaphyseal cartilaginous tongues were seen with all causes, but avascular necrosis was exclusively posttraumatic (p = 0.03). Signal characteristics and bridge size did not vary with the cause.
CONCLUSION. Premature physeal bony bridging in children is most often posttraumatic and disproportionately involves the distal tibia and femur where bridges tend to develop at the sites of earliest physiologic closure, namely anteromedially and centrally, respectively. MR imaging, especially with the use of fat-suppressed 3D spoiled gradient-recalled echo imaging, exquisitely shows the growth disturbance and associated abnormalities that may follow physeal injury and guides surgical management.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Historically, radiography, tomography, scintigraphy, and CT have been used to evaluate patients with growth arrest. Unfortunately, all of these techniques are limited in their ability to define the relationship of the bone bridge to the cartilaginous physis three-dimensionally. Recently, MR imaging has become the modality of choice to evaluate physeal abnormalities [5]. With the advantage of volumetric gradient-recalled echo physeal mapping techniques [6,7,8], MR imaging can accurately depict bony bridging across the physis and provide all of the necessary details to allow surgical planning. The purpose of our study was to use MR imaging, especially with fat-suppressed three-dimensional (3D) spoiled gradient-recalled echo sequences, to identify patterns of growth arrest in children after physeal insult from a number of causes.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
MR imaging was performed using two 1.5-T systems (Signa releases 5.7 and 5.8 and Horizon releases 8.0-8.3; General Electric Medical Systems, Milwaukee, WI) usually with a receive—transmit surface coil. In all patients, the MR examination included long-axis (sagittal or coronal) fat-suppressed spin-echo or fast spin-echo intermediate-weighted and T2-weighted imaging (TR range/TE range, 2,000-4,000/15-80) and long-axis (usually coronal) spin-echo T1-weighted imaging (TR/TE, 500/15). Coronal gradient-recalled echo imaging (300/13; flip angle, 30°) without volumetric analysis was performed in the initial 53 patients. The subsequent 58 patients were imaged with a fat-suppressed 3D volumetric spoiled gradient-recalled echo sequence that has been shown to be useful for hyaline cartilage evaluation [6]. Images were usually acquired coronally planned off an axial localizer with the following parameters: TR, approximately 21 msec; minimum allowable TE, usually 2 msec; flip angle, 30°; matrix, 256 x 128; excitations, 2; and field of view, variable (depending on patient size). Slice thickness for the 3D sequence was kept to 1 mm or less to allow high-quality multiplanar reformations.
Maximum intensity projections of the juxtaphyseal area were then obtained to provide an axial map of the physis using a workstation (Advantage Windows, versions 2.0 and 3.1; General Electric Medical Systems). The workstation software allows a rapid maximum intensity projection constructed from a 5- to 10-mm strip centered on the growth plate (Fig. 1A,1B,1C,1D). This technique, which generated a physeal map in less than 2 min, sufficed for most physes. In the minority of cases in which physeal irregularity would not allow a simple maximum intensity projection, manual segmentation of a periphyseal volume was performed. Although this was more laborious, the reconstructions were completed in less than 5 min by experienced users. In the map obtained, the physeal cartilage was of high signal intensity with lower signal intensity ridges coursing through the cartilage in a radial fashion. Bone bridges were easily identified as low-signal-intensity interruptions in the overall high-signal-intensity physis (Fig. 1A,1B,1C,1D).
|
|
|
|
The MR images were evaluated for several features. The size of the bridge was recorded as less than or equal to 25%, 26-50%, 51-75%, or greater than 75% of the total physeal area. In addition to the anatomic site involved, the location of the bridge in the physis was classified as central, peripheral, or mixed (both central and peripheral). Additionally, the physis was segmented into a central core and four peripheral quadrants (Fig. 2), and the bridge was assigned to one or more segments on the basis of location. The signal intensity of the bridge on T1-weighted images was defined as hypointense, isointense, or hyperintense relative to adjacent metaphyseal marrow. If metaphyseal growth recovery lines were identified, they were classified as either parallel to the physis or tethered at the physeal bridge. Associated metaphyseal abnormalities, namely extensions of physeal cartilage and areas of necrosis, were also noted. Linear or bandlike areas of metaphyseal high signal intensity on T2-weighted or gradient-recalled echo images were considered persistent cartilage [9]. Areas of increased marrow signal intensity with hypointense margins on T1-weighted images were considered to represent avascular necrosis.
|
Radiographs obtained concurrent to the MR examinations were evaluated for the presence of a visible bony bridge and associated growth recovery lines. From the clinical records, we obtained data regarding the cause of the growth arrest, the Salter-Harris fracture type in posttraumatic cases, the time from injury to MR imaging, the age of the patient at the time of injury or insult, and any surgical intervention following MR imaging.
Associations between the 111 patients and their bridge characteristics and features were identified using the two-tail Fisher's exact test.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Radiographs either identified or suggested the presence of a bone bridge in 72 of 111 patients and were, in fact, the impetus for the MR evaluation in most of the patients. Surgical intervention was performed on the basis of the MR findings in 60 of 111 patients. Surgical findings were consistent with bony bridging in all 60 patients.
Physeal bone bridges were most easily visualized on the fat-suppressed 3D spoiled gradient-recalled echo sequence. On this fat-suppressed image, the low-signal-intensity bridge, isointense to suppressed fatty marrow, was easily contrasted with the high-signal-intensity physeal cartilage (Figs. 1A,1B,1C,1D and 3A,3B,3C,3D). Successful physeal maximum-intensity-projection maps and bridge or physis area ratios were obtained in all 58 patients who underwent the fat-suppressed 3D spoiled gradient-recalled echo sequence.
|
|
|
|
We identified a significant association between the cause of the growth arrest and the anatomic site involved (p < 0.0001). The posttraumatic bridges tended to be located distally in a bone (71/78), particularly the distal tibia, femur, and radius. Bridges that were due to other causes occurred more frequently at the proximal ends of bones (24/33), with the proximal tibia and femur as the most common sites. Regardless of cause, growth arrest with bony bridging was far more common in the lower extremities (93/111) than the upper extremities.
The location of the bridges in the physis varied with the anatomic site involved (p < 0.0001). Sixty-five percent (30/46) of distal tibial bone bridges involved the anteromedial aspect of the physis that includes the superiorly convex undulation known as Kump's bump (Fig. 3A,3B,3C,3D). The central aspect of the physis was involved in 60% (12/20) of the distal femoral bridges (Fig. 1A,1B,1C,1D). Lesions of the proximal femur were more often peripheral (6/7), as were those of the proximal tibia (13/17).
The T1 signal intensity of the bone bridges depended on size (p < 0.008). Larger bridges comprising greater than 25% of the physeal area were more frequently isointense or hyperintense to metaphyseal marrow (33/45), whereas smaller bridges that had variable signal intensity were often hypointense (32/65). A particularly strong association was noted between the location of the bone bridge in the physis and the appearance of growth recovery lines (p <0.0002). Only those bridges with a peripheral component were associated with tethered growth recovery lines (Fig. 3B). No oblique growth recovery lines were seen in cases of bridges exclusively involving the central aspect of the physis. Furthermore, most of the bridges associated with tethered growth recovery lines (21/29, 72%) were small, comprising less than 25% of the total physeal area.
Metaphyseal extensions of the physis were of high signal intensity on T2-weighted, intermediate-weighted, gradient-recalled echo, and fat-suppressed 3D spoiled gradient-recalled echo images (Fig. 4). There was no association between the prevalence of metaphyseal extensions or the cause: 13 were posttraumatic and 12 had other causes. Segments of avascular necrosis, however, only occurred after trauma (11/78) and were limited to the distal femur (n = 6), distal tibia (n = 3), and distal humerus (n = 2). Metaphyseal areas of avascular necrosis were evident as high T1 signal intensity surrounded by a low-signal-intensity periphery (Fig. 5). Segments of avascular necrosis were typically of low to intermediate signal intensity on gradient-recalled echo and fat-suppressed 3D spoiled gradient-recalled echo images.
|
|
There were no statistically significant associations between the bony bridge characteristics, such as size, location, T1 signal intensity, Salter-Harris fracture type, or patient sex.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Certain anatomic sites seem to be particularly prone to growth arrest. In the large series of physeal fractures by Peterson [2], the distal femur was infrequently injured (1.4% of all physeal fractures) but had a disproportionately high incidence of posttraumatic bridge formation, accounting for 35% of all premature physeal bridges. The explanation for this predilection, also seen in our patients, probably relates to a central physeal undulation, which is the region of earliest physiologic closure of the distal femoral physis [12] and the most common site of premature bridge formation. Sixty percent of our distal femoral bone bridges involved the central physeal undulation (Fig. 1A,1B,1C,1D). In a recent study of transverse physeal fractures in rabbits, 60% of the distal femoral fractures involved the juxtaepiphyseal layers of the physis centrally and the juxtametaphyseal region peripherally [13]. This injury to the germinal and proliferative physeal layers at the central undulation likely explains the high incidence of posttraumatic distal femoral bone bridge formation. Additionally, the forces required to create a distal femoral physeal fracture are severe abduction, adduction, hyperextension, hyperflexion, or a combination of these [14]. Distal femoral injuries are often severe with a high incidence of associated ligamentous injury. Although our study was not designed to evaluate mechanism of injury as a risk factor for growth disturbance, Peterson has described that the severity of forces and the subsequent complex physeal injury relate to the development of growth arrest.
The proximal tibia and distal femur contribute 55-70% to the growth of their respective bone [15], and thus, growth disturbance at these locations results in substantial shortening or deformity. Similarly, 60% of the distal tibial bridges involved the anteromedial quadrant of the physis that encompasses the undulation known as Kump's bump (Fig. 3A,3B,3C,3D). Normal distal tibial physeal closure begins at this point [16, 17].
Peripheral bone bridges in the distal tibia frequently caused angular deformity and, not surprisingly, were strongly associated with tethered growth recovery lines. Growth recovery lines, also called Park or Harris lines, represent disks of transversely oriented, rather than the normal longitudinally oriented, bony trabecula [18]. These disks form at the physis during slowed growth because of injury, immobilization, or illness. As growth resumes, the physis migrates away from the line that remains in the metaphysis. A growth recovery line angled relative to the physis indicates tethering of physeal migration by a bony bridge [19]. Growth recovery lines were best identified on T1-weighted images as low-signal-intensity bands coursing through the high-signal-intensity metaphyseal marrow (Fig. 3B). When the peripheral bridges were particularly small, especially those involving the perichondrium, tethered growth recovery lines provided strong evidence of physeal growth disturbance, whereas lines parallel to the physis confirmed uniform physeal growth. The corollary is that a tethered growth recovery line suggests that the bridge is small (likely <25% of the physeal area) and, thus, surgically resectable.
The metaphyseal abnormalities associated with bone bridge formation included segments of avascular necrosis and cartilaginous rests. Avascular necrosis was an exclusively posttraumatic phenomenon in our patients. Devascularized metaphyseal fracture fragments typically appeared as wedge-shaped regions of high T1-signal-intensity fatty marrow surrounded by peripheral low-signal-intensity sclerosis located between the fracture line and the avascular physis (Fig. 5). Similar findings have been mentioned in prior descriptions of physeal fractures using MR imaging [3, 5]. These fragments have not, however, been emphasized on radiographic descriptions because they can be difficult to identify on radiographs and tomograms. We identified them in nearly 15% of the patients with posttraumatic growth arrest, primarily involving the distal femur (6/11) and distal tibia (3/11). The presence of avascular fragments did not vary substantially with the size of the bone bridge.
High signal extensions of the physis into the metaphyseal marrow, likely representing tongues of metaphyseal cartilage, were present regardless of the cause of growth arrest (Fig. 4). Cartilage persists in the metaphysis when enchondral ossification is disrupted as the result of metaphyseal vascular disruption [9, 20]. This may be related to many metaphyseal insults, including acute physeal fracture (especially Salter-Harris types 2 and 4), chronic metaphyseal injury such as Blount disease and Madelung deformity, and osteomyelitis and septic arthritis.
The results of the 111 patients in our study show that premature physeal bone bridges have a characteristic MR imaging appearance. Larger bridges tend to be of high signal intensity on T1-weighted images, whereas smaller bridges have variable signal intensity. T1-weighted images also ideally reveal growth recovery lines that indicate differential physeal growth. Intermediate- and T2-weighted images best reveal associated metaphyseal cartilage extensions. Gradient-recalled imaging optimally depicts the bridge as a low-signal-intensity interruption in otherwise high-signal-intensity physeal cartilage. The volumetric fat-suppressed 3D spoiled gradient-recalled echo sequence allows accurate mapping of the physeal bone bridges that appear as low-signal-intensity areas within the otherwise high-signal-intensity physis. The low-signal-intensity radial ridges seen in the physeal maps, which likely correspond to the normal small mamillary physeal undulations, rarely cause difficulty in interpretation (Figs. 1D and 3D). The accurate visualization of the bridge size and of the location relative to the entire physis is critical to the surgical treatment of patients with growth disturbance. Physes with complex geometry, especially the distal femur and proximal and distal tibia, are particularly vulnerable to bridge formation. Bridging usually occurs at the undulations where physiologic closure begins.
The limitations of this study stem from its retrospective nature. We have focused only on the patients who had physeal abnormalities on MR imaging, which precludes determination of the sensitivity and specificity of the technique. To identify sensitivity and specificity accurately, researchers need to perform a prospective study. A prospective study is currently underway to identify prognostic MR imaging features that reveal growth disturbance, possibly allowing intervention before a bone bridge develops.
In summary, we identified associations between imaging characteristics, anatomic site, and cause of growth arrest. MR imaging, especially with the use of fat-suppressed 3D spoiled gradient-recalled echo sequences, exquisitely reveals growth disturbance and associated abnormalities that may follow physeal injury in children. MR images in these patients provide information to the treating orthopedist who guides surgical management.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  K. I. Khoshhal and G. N. Kiefer Physeal Bridge Resection J. Am. Acad. Ortho. Surg., January 1, 2005; 13(1): 47 - 58. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Guermazi, A. Mohr, P. Henry, H. K. Genant, K. Ecklund, and D. Jaramillo Does Blount Disease Alter the Growth of the Distal Femur? Am. J. Roentgenol., January 1, 2003; 180(1): 282 - 284. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
