| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Pediatric Imaging |
1 From the Department of Radiology, Children's Hospital and Harvard Medical School, 300 Longwood Ave, Boston, MA 02115. Received September 22, 1998; revision requested November 4; final revision received July 26, 1999; accepted August 30. D.J. was supported in part by National Institutes of Health grant no. AR42396-04. Address reprint requests to D.J. (e-mail: jaramillo@a1.tch.harvard.edu).
 |
Abstract |
|---|
 TOP Abstract IntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
MATERIALS AND METHODS: Sixty-six sagittal T2-weighted or inversion-recovery MR images of the distal femoral epiphysis in children aged 2 months to 5 years 5 months were evaluated. Epiphyses were categorized into five types on the basis of progressive signal intensity changes within the epiphyseal cartilage along the weight-bearing region and posterior condyles. Epiphyseal type was compared with age, sex, and distribution of signal intensity changes within the condyle.
RESULTS: In early infancy, epiphyseal cartilage was homogeneous. During the 2nd year, signal intensity along the weight-bearing region decreased. With further advancing age, signal intensity in the posterior femoral condyles increased and became progressively more focal. The increase in epiphyseal grade correlated with age for both the medial and the lateral femoral condyles (r = 0.71 and r = 0.77, respectively; P < .001). There was no significant difference in epiphyseal changes between boys and girls or between medial and lateral condyles.
CONCLUSION: There is normal age-related variation in MR imaging signal intensity within the cartilaginous epiphysis of the distal femur. This may be related to weight bearing and epiphyseal maturation and should not be confused with disease.
Index terms: Bones, epiphyses, 451.121411, 451.121413, 451.1495, 451.1498, 451.25, 451.811, 452.121411, 452.36 • Bones, growth and development, 451.121411, 451.121413, 451.1495, 451.1498 • Cartilage, MR, 451.121411, 451.121413, 451.1495 • Knee, 451.121411, 451.121413, 451.1495, 451.1498, 451.25, 451.811, 4521.121411, 4521.121413, 452.36 • Knee, MR, 451.121411, 451.121413
|
Introduction |
|---|
|
TOPAbstractIntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
|
MATERIALS AND METHODS |
|---|
|
TOPAbstractIntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
Inclusion criteria included (a) the visualization of the entire distal femoral epiphysis, (b) the availability of T2-weighted or inversion-recovery images in the sagittal plane, and (c) the lack of known epiphyseal tumor, direct traumatic event, infection, arthritis, or dysplasia that involved the knee. In 12 patients, both knees were imaged once during the same examination. Two other patients underwent imaging of both knees on two separate occasions, one patient at 12 months and the other patient at 29 months after the initial imaging examination. Two additional patients underwent repeat MR imaging of a single knee 1 month after initial examination.
The most common indication for undergoing MR imaging in our study, across all age groups, was knee clicking, locking, and pain that was suggestive of a discoid meniscus in 21 of 46 patients (46%). Even in the 1st year of life, the concern for discoid meniscus was the primary indication for imaging. Other indications in this age group included the evaluation of soft-tissue tumors, myositis, and amniotic bands. Common indications for imaging in the older age groups included nonspecific knee pain and tibial growth plate evaluation in patients with remote trauma to the tibia.
Twenty-two studies did not meet the inclusion criteria. Exclusion was most often due to the incomplete depiction of the distal femoral epiphysis in those patients with an examination of the lower extremity that was not focused on the knee.
Examinations were performed by using a 1.5-T MR imaging unit (Signa and Horizon; GE Medical Systems, Milwaukee, Wis). The extremities were imaged by using a linear extremity coil, phased-array extremity coil, head coil, or torso phased-array coil, the use of which was dependent on the patient's size, on the extent of the extremity that required evaluation, and on the need for unilateral versus bilateral imaging. All patients were sedated in accordance with the radiology department sedation committee guidelines, with primarily oral administration of chloral hydrate (50–100 mg per kilogram of body weight; Pharmaceutical Associates, Greenville, SC) in infants less than 1 year of age and intravenous administration of pentobarbital sodium (2–6 mg/kg; Wyeth Laboratories, Philadelphia, Pa) in children 1–5 years old.
Sagittal imaging allowed good visualization of the distal femoral epiphysis. We limited our evaluation to include T2-weighted (conventional or fast spin-echo) or inversion-recovery images, as these images are known to depict signal intensity changes in the epiphyseal cartilage (1). A conventional spin-echo T2-weighted sequence (1,800–2,500/80–100 [repetition time msec/echo time msec]) was used in 41 examinations, and a fast spin-echo sequence (2,600–4,500/81–108 [effective]; echo train length, eight) was used in 17 examinations. In 13 of these 58 examinations (all fast spin echo), fat saturation was achieved by using frequency-selective presaturation.
Eight examinations included fast inversion-recovery sequences (3,000–4,200/30–54/150 [repetition time msec/echo time msec/inversion time msec]; echo train length, eight) instead of T2-weighted sequences. The appearance of the epiphyseal cartilage on these fast inversion-recovery images very closely resembled that on T2-weighted images. Matrix size varied, and section thickness ranged from 3 to 5 mm.
Image Analysis
During an initial consensus review to establish standards of interpretation and terminology, three pediatric radiologists (L.J.V., T.L., D.J.) with experience in MR imaging analyzed each of the 66 images. Reviewers were blinded to patient identity, age, and sex. The medial and lateral condyles of the distal femoral epiphysis were evaluated separately for signal intensity patterns in both the weight-bearing region and the posterior condyles. The weight-bearing region was defined as the area of epiphyseal cartilage just distal to the ossification center in the central one-third of the epiphysis. Signal intensity within the anterior aspect of each epiphysis served as an internal control for comparison.
During the consensus review, the authors identified five epiphyseal patterns, designated as types 1–5 (Fig 1). Type 1 was a pattern of homogeneous signal intensity throughout the unossified portion of the epiphysis. Type 2 had low signal intensity within the weight-bearing region of the condyle. When such type 2 changes were associated with stippled high signal intensity in the cartilaginous posterior condyle, they were graded as type 3. The epiphyseal changes seen in types 4 and 5 also were similar to those seen in type 2 but with the addition of focally ill-defined or focally well-defined high signal intensity in the cartilaginous posterior condyle for type 4 or 5, respectively.
|
Statistical Analysis
Patients were classified into three age groups: eight "infants," aged less than 1 year; 19 "toddlers," aged 1–3 years; and 19 "preschoolers," aged 3–5 years. Epiphyseal patterns of signal intensity were compared between the three age groups; between sex groups with respect to age groups; and between medial and lateral condyles with respect to age groups.
For the subjective rating of epiphyseal type, interobserver variability was determined by using the 
statistic (STATA; Stata, College Station, Tex). A value of less than 0.40 indicated poor agreement; 0.40–0.59, fair agreement; 0.60–0.74, good agreement; and 0.75–1.00, excellent agreement (5).
The measure of strength of association (r and P values) between epiphyseal type and age was calculated by using the Spearman rank correlation, which is used to measure the relationship between two variables, one of which is ordinal.
The Wilcoxon rank sum test was used to evaluate for statistical differences between epiphyseal patterns in boys and girls with respect to each age group. The Wilcoxon signed rank test was used to analyze differences between patterns in the medial and lateral condyles with respect to each age group.
|
RESULTS |
|---|
|
TOPAbstractIntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
|
|
|
|
|
|
|
|
In the four patients (six knees) who underwent imaging repeatedly, the epiphyseal type either advanced or remained unchanged with increasing patient age. The two single knees of two patients who underwent imaging at a 1-month interval did not change epiphyseal type. In the patient in whom both knees were imaged at a 12-month follow-up examination, one knee had increased by one type and the other had not changed. The patient who underwent imaging of both knees at a 29-month interval had an increase in the type of both knees with age.
Epiphyseal Type and Patient Sex
The Wilcoxon rank sum test results demonstrated no significant differences in age-related signal intensity changes between boys and girls (medial condyles, P 
.78; lateral condyles, P .29).
Epiphyseal Type and Distribution
No differences were identified between medial and lateral condylar signal intensity changes within each age-related group by using the Wilcoxon signed rank test (P .27).
Epiphyseal Type and Interobserver Agreement
By using the statistic, we demonstrated fair interobserver agreement of 0.48 and 0.52 for the medial and lateral condyles, respectively. There was substantial agreement in classifying the type 1 pattern (medial, 0.80; lateral, 0.78). The highest degree of interobserver variability was seen in classifying the type 3 pattern (medial, 0.39; lateral, 0.51) and the type 4 pattern (medial, 0.28; lateral, 0.49). Disagreements in categorization occurred in 39% (36 of 92) of observations between type 3 and type 4 groups.
|
DISCUSSION |
|---|
|
TOPAbstractIntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
Epiphyseal types were found to be consistent between the knees of the same patient, which helped to substantiate our hypothesis that changes occur in an age-related pattern. In the four patients who underwent repeat imaging, epiphyseal type either increased or remained unchanged. Although the number of patients in this group was small, this finding adds support to our hypothesis that advancing pattern occurs with advancing age.
Epiphyseal Type, Age, and Weight Bearing
Our finding of homogeneous cartilaginous distal femoral epiphyseal signal intensity (type 1 epiphyseal pattern) correlates with the findings in published reports in which MR images of the epiphysis in newborn piglet hips and in newborn lamb knees (6,7) were analyzed. Homogeneous signal intensity in the human cartilaginous epiphysis therefore may be related to the persistence of the neonatal pristine state prior to normal weight-bearing stress and prior to the period of rapid ossification.
At the beginning of the 2nd year of life, when children start walking, weight-bearing stress begins. The appearance of low-signal-intensity changes along the inferior portion of the condyles (type 2 epiphyseal pattern) likely is related to pressure changes that occur with the normal upright position. It is known that glycosaminoglycans, an important constituent of hyaline cartilage, release water in response to applied pressure and recapture water when pressure is released. In vitro study results have shown that compression of the hyaline articular cartilage causes emigration of fluid from the compressed areas and therefore causes a decrease in T2-weighted signal intensity (8,9). We believe that pressure-associated changes similar to those seen in articular hyaline cartilage can occur within epiphyseal hyaline cartilage with weight bearing.
In the preschool group, high-signal-intensity change on MR images was seen within the posterior condyles (types 3–5 epiphyseal patterns). There was difficulty in differentiating the types 3–5 epiphyseal patterns. We believe that these three types represent a continuum along which epiphyseal type increases with advancing age.
High-signal-intensity changes may relate to more active ossification within the posterior condyles as the secondary ossification center enlarges. Increased signal intensity on T2-weighted images seen in the process of normal endochondral ossification may occur when cartilage cells undergo hypertrophy (10). Areas of hypertrophic high-signal-intensity cartilage are seen initially in the epiphyseal cartilage that surrounds the secondary center of ossification. Later, similar changes are seen in a patchy distribution along the posterior and lateral aspects of the femoral condyles. This correlates well with what is seen on radiographs in children, in whom several small centers of ossification often develop along the posterior condyles. This irregularity of ossification is not seen anteriorly. Therefore, the hypertrophic changes of high signal intensity on T2-weighted images are confined to the posterior condyles.
High-signal-intensity change has been described previously within articular cartilage as an artifact related to radially oriented collagen fibers at a 55° angle to the main magnetic field (the "magic-angle" effect) (11,12). To our knowledge, no similar effect has been described within nonarticular hyaline cartilage, which contains a less organized distribution of collagen fibers without radial orientation. Therefore, a substantial magic-angle effect is less likely to be seen in this location.
Epiphyseal Type and Patient Sex
A comparison between boys and girls did not help to identify differences in epiphyseal pattern within age groups. Therefore, the known osseous maturational differences between boys and girls were not demonstrated in our study by using MR imaging. Although the small sample size may be contributory, we saw no trend in our statistical analysis. This finding is contrary to the known greater rate of maturation in girls (13).
Epiphyseal Type and Distribution
No differences were seen between medial and lateral condylar signal intensity changes among age-related groups. We did not detect any differences that might have resulted from the postural change (genu varum to genu valgum) that occurs in early childhood (14).
A limitation of our study was that all infants and children who underwent MR imaging had clinical symptoms or a nonepiphyseal abnormality referable to the imaged extremity. This limitation was likely small as, despite the broad range of clinical indications for imaging, the findings were consistent within each age group. A prospective study of normal knees in asymptomatic children may be performed to further address this issue.
The inclusion of both knees within the same patient could weight the results toward the patient's age group. Because there was an even distribution of this subset of patients between the three age groups, this limitation was unlikely to be significant.
Another limitation may have been the inclusion of both T2-weighted and inversion-recovery sequences in the analysis of signal intensity. Although these techniques are markedly different, they were comparable in depicting age-related signal intensity changes within the epiphyseal cartilage.
Our results demonstrate that during the normal development of the cartilaginous distal femoral epiphysis, signal intensity decreases within the weight-bearing region and becomes progressively more heterogeneous within the posterior condyles. We believe that the signal intensity changes can be explained (a) by cartilage response to pressure incurred with weight bearing and (b) by advancing ossification within the cartilaginous epiphysis as a child develops. It is important to recognize the normal signal intensity changes that occur with development to differentiate them from epiphyseal disease.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
3 Department of Radiology, Children's Hospital Medical Center, Cincinnati, Ohio.
Author contributions: Guarantors of integrity of entire study, L.J.V., D.J., T.L.; study concepts, D.J., T.L.; study design, L.J.V., D.J., T.L.; definition of intellectual content, L.J.V., T.L., D.J.; literature research, L.J.V., D.J.; clinical studies, L.J.V., D.J., T.L.; data acquisition and analysis, L.J.V., D.J., T.L.; statistical analysis, D.J.; manuscript preparation, editing, and review, L.J.V., D.J., T.L.
|
References |
|---|
|
TOPAbstractIntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  T. Laor and D. Jaramillo MR Imaging Insights into Skeletal Maturation: What Is Normal? Radiology, January 1, 2009; 250(1): 28 - 38. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
N. M. Menezes, S. A. Connolly, F. Shapiro, E. A. Olear, R. M. Jimenez, D. Zurakowski, and D. Jaramillo Early Ischemia in Growing Piglet Skeleton: MR Diffusion and Perfusion Imaging Radiology, January 1, 2007; 242(1): 129 - 136. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. A. Connolly, D. Jaramillo, J. K. Hong, and F. Shapiro Skeletal Development in Fetal Pig Specimens: MR Imaging of Femur with Histologic Comparison Radiology, November 1, 2004; 233(2): 505 - 514. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
B. J. Dardzinski, T. Laor, V. J. Schmithorst, L. Klosterman, and T. B. Graham Mapping T2 Relaxation Time in the Pediatric Knee: Feasibility with a Clinical 1.5-T MR Imaging System Radiology, October 1, 2002; 225(1): 233 - 239. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. Laor, G. F. H. Chun, B. J. Dardzinski, J. A. Bean, and D. P. Witte Posterior Distal Femoral and Proximal Tibial Metaphyseal Stripes at MR Imaging in Children and Young Adults Radiology, September 1, 2002; 224(3): 669 - 674. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |
