| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Original Research |
1 Department of Radiological Sciences, Division of Diagnostic Imaging, St. Jude
Children's Research Hospital, 332 N Lauderdale St., Memphis, TN 38105.
2 University of Tennessee College of Medicine, Memphis, TN.
Received October 12, 2004;
accepted after revision March 22, 2005.
Supported in part by grants P30 CA-21765 and P01 CA-20180 from the National
Institutes of Health, a Center of Excellence grant from the State of
Tennessee, and by the American Lebanese Syrian Associated Charities
(ALSAC).
Abstract
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
MATERIALS AND METHODS. Osteonecrotic lesion size and location were determined in a retrospective analysis of MR images of the knee obtained in our institution during the past 10 years. Association between MRI findings and clinical symptoms expressed by the knee scores was tested for a subgroup of our patient population who had clinical evaluation of the knees in the orthopedic clinic within 6 weeks of their MRI studies.
RESULTS. In 80% of patients, osteonecrosis was bilateral. Lesions were distributed as follows: femur and tibia, 66% of the knees; femur alone, 26%; and tibia alone, 8%. Clinical symptoms of knee osteonecrosis were associated with lesions involving the articular surface of the distal femur, large lesions, and involvement of any part of the tibia. MR images showed milder osteonecrosis in patients who were less than 10 years old at the time of primary diagnosis.
CONCLUSION. Osteonecrosis affecting the knees of children treated for leukemia and lymphoma is mostly asymptomatic. Children who are 10 years old or older have signs of more advanced osteonecrosis. Clinical symptoms often lag MRI presentation and may develop late in the course of the disease. Routine MRI evaluation is recommended for timely diagnosis of treatment-induced osteonecrosis.
Keywords: knee • leukemia • lymphoma • MRI • musculoskeletal imaging • oncology • osteonecrosis • pediatric imaging
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The association between MRI findings and symptoms or signs of osteonecrosis of the knee, and the clinical significance of the association, have not yet been established for children. Such information could be valuable in assessing the potential seriousness of this relatively frequent complication. In this article we report the demographic, MRI, and clinical findings associated with osteonecrosis of the knee, and the relation between the MRI and clinical findings, in children treated for hematologic malignancy at a pediatric cancer center. To explore this relationship, we tested the association between clinical symptoms and MRI findings of osteonecrosis in a subgroup of patients for whom orthopedic clinical data were available. We also describe MRI patterns of involvement and their influence on clinical manifestations of knee osteonecrosis.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
We divided the knee into eight zones and recorded the presence or absence of an osteonecrotic lesion in each zone: distal femoral diaphysis, distal femoral metaphysis, medial and lateral distal femoral epiphyses, medial and lateral proximal tibial epiphyses, proximal tibial metaphysis, and proximal tibial diaphysis (Figs. 1A and 1B). An osteonecrotic lesion was defined as a geographic area of decreased signal on T1-weighted images and increased signal on STIR images. We recorded additional characteristics of lesions located in epiphyseal zones, including whether the lesion reached the articular surface and what percentage (< 25%, 25-50%, or > 50%) of the articular surface was involved.
|
|
To study the relationship between MRI findings and clinical symptoms, we identified all patients who had knee scores [11] obtained in the orthopedic clinic within 6 weeks of their MRI study. We chose this abbreviated time interval because of the rapid progression of osteonecrosis we have observed in some patients; a longer time interval might have resulted in inaccurate perception of such relationship. Patients were referred for the orthopedic assessment by clinicians if osteonecrosis was discovered on MRI. The knee score was developed by the Knee Society [11] as a measure of clinical symptoms associated with knee arthritis in adults (pain, limitation in the range of motion, and joint instability). The knee score rates only the joint itself and does not involve functional assessment, thus eliminating potentially confounding factors of patient infirmity. Scores range from 0 to 100, with the highest score indicating the absence of pain, perfect alignment and range of motion, and no joint instability [11]. The knee score is combined from a quantitative evaluation of pain, range of motion, and stability. Pain is the major contributor to the score; 50 points are assigned to patients who report no pain, and 45 points are assigned for mild or occasional pain. Range of motion and stability contribute up to 25 points each to the score. Flexion contracture, extension lag, and misalignment are dealt with as deductions. In our group of patients, the score of 95 in all 22 knees was due to mild or occasional pain. The same orthopedic surgeon and nurse obtained the knee scores from all patients. To address the potentially confounding effect of hip osteonecrosis on clinical manifestations of knee osteonecrosis, hip involvement of a subgroup of patients with knee osteonecrosis and knee scores was identified from hip MR images.
Estimation of Disease Severity
We used the estimated percentage of the articular surface involved by an
osteonecrotic lesion as an indicator of disease severity. Several studies of
the natural course of osteonecrosis in adult and pediatric populations
concluded that a worse prognosis was associated with the presence of large
epiphyseal lesions and involvement of the articular surface
[6,
8,
12]. For this analysis, we
defined "large" epiphyseal lesions as those involving more than
25% of the articular surface of the medial or lateral portion of the femoral
or tibial epiphysis. We calculated the prevalence of such lesions according to
age group, sex, and race.
Statistical Analysis
For statistical analysis of the association between Knee Society score and
the number of different types of osteonecrotic lesions, we used Fisher's exact
test (for 2 x 2 tables) and Wilcoxon's rank sum test (for 2 x
p ordered tables). The strength of agreement between the two
observers was determined using kappa statistics, as suggested by Landis and
Koch [10].
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
MRI Evaluation
Eighty percent of the affected patients (135/168) had bilateral knee
osteonecrosis. In 66% (200/303) of the affected knees, osteonecrosis was
present in both the femur and the tibia. In 26% (79/303), lesions were present
only in the distal femur; in 8% (24/303), only in the proximal tibia.
The frequency of large epiphyseal lesions (affecting > 25% of articular surface) increased with the patient's age. Approximately 10% (8/84) of patients less than 10 years old at the diagnosis of cancer had large epiphyseal lesions. In contrast, approximately half (41/84) of those 10 years old or older had large epiphyseal lesions (p = 0.0001). Race and sex were not found to be associated with the presence of large epiphyseal lesions (p = 0.6131 and p = 0.4466, respectively).
Children less than 10 years old had bilateral knee involvement slightly less often than did the older group: 76% (64/84) versus 90% (76/84) (p = 0.013). Similarly, they had fewer osteonecrotic lesions overall (463 vs 743; p < 0.0001), fewer epiphyseal lesions (288 vs 381; p = 0.0004), and fewer lesions extending to the articular surface (28 vs 175; p < 0.0001). The two age groups were equal in size (84 patients) and therefore could be directly compared.
Relationship Between Clinical Symptoms and MRI Findings of Knee Osteonecrosis
Clinical evaluation—Knee scores were obtained within 6 weeks
of imaging for 93 knees (of 47 patients) that showed MRI evidence of
osteonecrosis. More than half of the knees (51, 55%) were asymptomatic (knee
score = 100); 22 (24%) had a score of 95; and only 20 (22%) had a score of
less than 95. The median score was 100 (range, 50-100) because of the large
number of asymptomatic knees. Because most knees (51/93) had a perfect knee
score of 100, we used the presence (score < 100) or absence (score = 100)
of knee-related symptoms to test the relationship between the clinical and MRI
findings. Hip osteonecrosis was identified in 20 cases: 16 related to
symptomatic knees and four related to asymptomatic knees. In the subgroup of
patients with no hip osteonecrosis, 64% of knees (47/73) were asymptomatic,
and 36% (26/73) were symptomatic.
Relationship between knee score and frequency of osteonecrotic damage to metaphysis and diaphysis—Osteonecrotic lesions in the distal femoral diaphysis and metaphysis were not found to be associated with symptoms; lesions in the proximal tibial metaphysis or diaphysis were (Table 1).
|
Relationship of knee score to type and frequency of osteonecrotic damage to epiphysis—The distribution of different types of epiphyseal lesions in symptomatic and asymptomatic groups is shown in Table 2. Epiphyseal lesions of any type were associated with symptoms in the femur and the tibia (p < 0.001). Among epiphyses with osteonecrotic lesions, involvement of the articular surface was associated with the presence of symptoms (femur, p < 0.09; tibia, p= 0.034). Among femoral epiphyses with osteonecrotic lesions involving the articular surface, the extent of articular surface involvement was strongly associated with the presence of symptoms (p < 0.001). However, in tibial epiphyses, we found no association between the size of lesions involving the tibial articular surface and symptoms (p= 0.663).
|
Relationship between extent of osteonecrotic damage to epiphysis and knee score— Epiphyseal involvement was highly prevalent in knees with osteonecrosis (Table 3). Osteonecrosis involving at least one epiphyseal zone (in femur or tibia) was about equally prevalent among symptomatic and asymptomatic knees (40/42 vs 44/51, p = 0.1577; Table 3). Symptomatic knees had articular surface involvement significantly more often than did asymptomatic knees (31/42 vs 22/51, p= 0.0092; Table 3). In addition, the size of surface-involving lesions differed between symptomatic and asymptomatic knees (p= 0.0028), with the former tending to have larger lesions. In the symptomatic group, 67% of the knees (28/42) had lesions with substantial (> 25%) articular surface involvement. In the asymptomatic group, 25.5% (13/51) had such lesions (Table 3). The difference is even more striking for lesions involving more than 50% of the articular surface: 12% (6/51) of asymptomatic knees had such lesions compared with 48% (20/42) of symptomatic knees. Three symptomatic and nine asymptomatic knees had lesions involving less than 25% of the surface. It appears that lesions with less than 25% of articular surface involvement are more likely to be asymptomatic.
|
Reliability
Interobserver reliability in using the osteonecrosis classification system
was substantial: 
= 0.66 (95% CI, 0.58-0.75) in locations where only
the presence or absence of a lesion was recorded and weighted = 0.65
(95% CI, 0.59-0.72) in locations where the extent of involvement was recorded.
Additional information regarding the interobserver reliability assessment is
described in part 1 of this series
[13].
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
MRI evidence of milder osteonecrosis in children younger than 10 years may be associated with fewer patients with clinical symptoms. This finding could lead to the underdiagnosis of osteonecrosis in this age group. It is also more difficult to elicit information about clinical symptoms, especially very mild ones, from younger patients. Studies of the natural course of steroid-induced osteonecrosis showed that large lesions close to the articular surface were associated with a progressive course of osteonecrosis and with a higher rate of treatment failure [6, 8]. The higher prevalence of large epiphyseal lesions that we found in children 10 years or older may indicate that osteonecrosis of the knee leads to a worse outcome in this age group.
Association Between Clinical Symptoms and MRI Findings of Knee Osteonecrosis
Of the 93 knees in our study with available knee scores, more than half
(51) were asymptomatic (knee score = 100); 22 had a score of 95, and only 20
knees had a score of less than 95. Only three of 93 cases revealed isolated
tibial involvement, which did not allow us to separate the influence of tibial
versus femoral lesions on clinical symptoms. Clinical symptoms were associated
with lesions in any part of the tibia. In the case of femoral lesions,
clinical symptoms were associated with large lesions involving the articular
surface. Epiphyseal lesions that did not extend to the articular surface were
not found to be associated with symptoms. Although the presence of large
surface lesions was associated with symptoms, a substantial number of the
knees with large surface lesions (13/41, 32%) were asymptomatic. Furthermore,
even when the lesions involved more than 50% of the articular surface, six
(23%) of 26 knees were asymptomatic (Table
3). This finding may mean that therapy-induced osteonecrosis of
the knee becomes symptomatic late in its progression when the articular
surface is substantially involved. Even then, clinical symptoms may be absent
in many cases. Therefore, clinical symptoms are not a reliable diagnostic sign
of osteonecrosis of the knees in children.
Asymptomatic osteonecrosis is common. In two prospective screening studies of ALL patients, osteonecrosis was diagnosed in more than 30% of patients, and one half to two thirds of those with a diagnosis of osteonecrosis were asymptomatic [2, 16]. Our data closely agree with those previously reported. In our group of 93 knees with available function scores, 51 (55%) were completely asymptomatic and 22 (24%) had scores of 95 due to mild or occasional pain. Considering the somewhat subjective nature of pain evaluation, which is especially challenging in the pediatric population, some patients with a knee score of 95 might have been considered asymptomatic at a repeated evaluation. Elimination of hip osteonecrosis as a confounding factor for clinical manifestations of knee osteonecrosis increases the proportion of asymptomatic knees: 64% (47/73) versus 55% (51/93). When it is asymptomatic, osteonecrosis may not be diagnosed unless patients are prospectively monitored with imaging. Undiagnosed, asymptomatic osteonecrosis does not present a problem when it resolves without the need for treatment [6, 16]. However, osteonecrosis that is initially asymptomatic may progress to become symptomatic later [8, 17-19]. A significant percentage of patients with ALL and symptomatic osteonecrosis require surgical treatment [3]. The most significant morbidity ensues when osteonecrosis develops in a weight-bearing joint, such as the hip or knee, and when lesions are located in the epiphysis. Longitudinal studies of osteonecrosis in adults show improved results when treatment is initiated at an early stage [20-22]. However, the benefit of early intervention for knee osteonecrosis, such as cessation of glucocorticoid treatment or core decompression, remains to be shown in a pediatric population.
The limitations of this study include its retrospective nature. Clinical orthopedic data were available for only a modest number of patients. This limitation prompted us to use qualitative (asymptomatic vs symptomatic) instead of quantitative measurements of clinical symptoms. Also, we assumed that an MRI pattern (a large lesion extending to the articular surface) was an indicator of a progressive course of osteonecrosis. A longitudinal study of knee osteonecrosis in this population is needed to fully understand the clinical implications of MRI findings indicative of osteonecrosis. A prospective study would also address a potential limitation of the current study—namely, it would provide information about the protractability and timing of symptom development during the course of treatment.
In conclusion, asymptomatic osteonecrosis of the knee is more common than the symptomatic type among children treated for ALL. Although clinical symptoms usually signify substantial damage, they cannot be relied on as an indicator of the presence or absence of osteonecrosis of the knees. Clinical symptoms often lag MRI presentation and may occur late in the course of the disease. Because it is not known how often patients with asymptomatic knee osteonecrosis subsequently experience symptomatic or progressive joint destruction, further longitudinal investigation is needed. Children 10 years or older at the time of the primary diagnosis of hematologic malignancy may be at greater risk than those who are younger for further progression of osteonecrosis.
Acknowledgments
We thank Margaret Carbaugh for her editorial assistance and manuscript
preparation.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  E. J. Karimova, S. N. Rai, X. Deng, D. J. Ingle, A. C. Ralph, M. D. Neel, and S. C. Kaste MRI of Knee Osteonecrosis in Children with Leukemia and Lymphoma: Part 1, Observer Agreement Am. J. Roentgenol., February 1, 2006; 186(2): 470 - 476. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
