| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1 Division of Diagnostic Radiology, National Cancer Center Hospital and Research
Institute, Tsukiji, Chuo-Ku, Tokyo 104-0045, Japan.
2 Division of Pathology, National Cancer Center Hospital and Research Institute,
Tsukiji, Chuo-Ku, Tokyo 104-0045, Japan.
3 Division of Orthopedics, National Cancer Center Hospital and Research
Institute, Tsukiji, Chuo-Ku, Tokyo 104-0045, Japan.
Received February 3, 2003;
accepted after revision July 11, 2003.
Address correspondence to U. Tateishi.
Abstract
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
MATERIALS AND METHODS. The initial MRI studies of 36 pathologically confirmed myxoid–round cell liposarcomas were retrospectively reviewed, and observations from this review were correlated with the histopathologic features. MR images were evaluated by two radiologists with agreement by consensus, and both univariate and multivariate analyses were conducted to evaluate survival with a median clinical follow-up of 33 months (range, 9–276 months).
RESULTS. Statistically significant MRI findings that favored a diagnosis of intermediate- or high-grade tumor were large tumor size (> 10 cm), deeply situated tumor, tumor possessing irregular contours, absence of lobulation, absence of thin septa, presence of thick septa, absence of tumor capsule, high-intensity signal pattern, pronounced enhancement, and globular or nodular enhancement. Of these MRI findings, thin septa (p < 0.05), a tumor capsule (p < 0.01), and pronounced enhancement (p < 0.01) were associated significantly, according to univariate analysis, with overall survival. Multivariate analysis indicated that pronounced enhancement was associated significantly with overall survival (p < 0.05).
CONCLUSION. Contrast-enhanced MRI findings can indicate a good or adverse prognosis in patients with myxoid–round cell liposarcomas.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The characteristic MRI features of myxoid–round cell liposarcomas are attributable to the predominantly myxoid matrix of the tumor. Tumors appear on T2-weighted MR images as encapsulated tumors with signals that are hyperintense compared with the surrounding structures [10–14]. On contrast-enhanced studies, they often show marked or heterogeneous enhancement with nonenhanced areas corresponding to myxoid material [13, 14]. As expected from the fact that the histopathologic spectrum from myxoid to round cell liposarcomas is continuous, these tumors show considerable diversity on imaging. Therefore, it is important to review the reliability of MRI features for characterizing myxoid–round cell liposarcomas. The objectives of this study were to determine the prognostic significance of MRI findings in patients with myxoid–round cell liposarcomas and to clarify which MRI features best indicate tumors with adverse clinical behavior.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
MRI Studies and Pathologic Correlations
MRI was performed using one of two 1.5-T systems (Horizon, General Electric
Medical Systems, Milwaukee, WI; or Visart, Toshiba Medical Systems, Tokyo,
Japan). Either the spin-echo or the fast spin-echo technique was used to
obtain T1-weighted images (TR range/TE range, 460–720/12–27) in
one or more planes (coronal or axial). T2-weighted images (TR
range/TEeff range, 3,500–6,000/96–112; echo-train
length, 8–12) with flow compensation and presaturation superiorly and
inferiorly were then obtained in one or more planes using a body coil. The
images were obtained with a field of view of 30–40 cm, an image matrix
of 128 x 256, and a slice thickness of 6–10 mm. Gadopentetate
dimeglumine was administered IV, and T1-weighted images were obtained in one
or more planes with (n = 20) or without (n = 16) fat
suppression.
Two radiologists reviewed the MR images, and the findings were reported as
a consensus opinion. The lesions were judged according to size, location,
depth (superficial or deep), type of margin and contours, internal
architecture, presence of a tumor capsule, signal characteristics on T1- and
T2-weighted images, and homogeneity (homogeneous or heterogeneous). A
superficial tumor (dermal or subcutaneous tumor) was located exclusively above
the superficial fascia without invasion of the fascia, whereas a deep tumor
was located either exclusively beneath the superficial fascia or superficial
to the fascia with invasion of or through the fascia. The signal
characteristics were described as isointense or hyperintense relative to the
signal intensity of skeletal muscle. The extent (none and weak or pronounced),
pattern (globular and nodular or diffuse), and homogeneity of gadolinium-based
enhancement were also recorded. Globular and nodular enhancement corresponded
to spotty enhancement (range, 3–10 mm) within the mass on
contrast-enhanced MR images. Septal structures were categorized as thin
(uniform linear structures 
2 mm) or thick (focally thickened linear
structures > 2 mm). Tumors containing areas with high-intensity
characteristics on both T1- and T2-weighted MR images were considered positive
for a high-intensity signal pattern.
Histologic slides of all the patients' tumors were reviewed for diagnosis by an expert pathologist. Immunohistochemical staining was performed in all cases to confirm the diagnosis or tumor type according to the classification system described by Enzinger and Weiss [1]. In this study, the histologic grade of a tumor was determined using a three-grade system established by Hasegawa et al. [15–17]. According to this system, myxoid–round cell liposarcomas are assigned a grade of 1, 2, or 3. Grade 1 tumors (n = 12, 33.3%) are considered low-grade tumors, grade 2 tumors (n = 14, 38.9%) are intermediate-grade tumors, and grade 3 tumors (n = 10, 27.8%) are high-grade tumors (n = 24, 66.7%). Excised specimens were available for review or for mapping correlation with images. Pathology reports were reviewed for descriptive comments characterizing the necrosis and myxoid–round cell tumor components of the lesions.
Statistical Analysis
Patients' demographics and imaging characteristics were compared using
Wilcoxon's rank sum test for continuous variables and the chi-square test or
Fisher's exact test for categorized variables. Univariate analysis was
performed by comparing survival curves generated using the Kaplan-Meier method
and carrying out log-rank tests. The relative risk of each variable subjected
to multivariate analysis was estimated using a Cox proportional hazards model.
All analyses were conducted using SPSS software version 11.0J (Statistical
Package for the Social Sciences, Chicago, IL) for Windows (Microsoft, Redmond,
WA). Differences and correlations at a p value of less than 0.05 were
considered statistically significant.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Metastases occurred in 11 (30.6%) of the 36 patients; the location of metastasis was the peritoneal cavity in five patients (13.9%); soft-tissue in five (13.9%); lung in three (8.3%); and bone, liver, retroperitoneum, and mediastinum in one (2.8%). Eight (38.0%) of the 21 patients who received additional treatment had metastasis subsequently. Twelve (33.3%) of the 36 patients developed local recurrences. Three patients (8.3%) with inadequate excision had local recurrence. Four patients (11.1%) with local recurrence underwent additional therapy.
Ten (27.8%) and 26 (72.2%) of 36 tumors had regular and irregular tumor contours, respectively (Figs. 1A, 1B, 2A, 2B, 2C, 3A, 3B, 3C, 4A, 4B, 4C). Sixteen tumors (44.4%) showed lobulated morphology. On MR images, thin and thick septa (Fig. 2A, 2B, 2C) were identified in 31 (86.1%) and 10 (27.8%) tumors, respectively. On T1-weighted MR images the signals of the tumors relative to those of muscle were hyperintense (n = 15), isointense (n = 12), or hypointense (n = 9). Tumors showed predominantly increased signal intensity compared with that of the skeletal muscle on T2-weighted MR images. The images showed the tumor as having a heterogeneous appearance with thin or thick septa of low intensity. High-intensity signals similar to subcutaneous fatty tissue (high-intensity signal pattern) were found in 15 tumors (41.7%) on both T1- and T2-weighted MR images (Figs. 1A, 1B and 4A, 4B, 4C).
|
|
|
|
|
|
|
|
|
|
|
On contrast-enhanced MR images, pronounced enhancement (Fig. 4A, 4B, 4C) located mostly at the peripheries of the lesions was present in 22 tumors (61.1%). Globular and nodular enhancement (Figs. 3A, 3B, 3C and 4A, 4B, 4C) was found mostly in the centers of the lesions of 16 patients (44.4%), whereas diffuse enhancement (Fig. 1A, 1B) was seen in six lesions (16.7%). Contrast-enhanced MR images also revealed that 23 tumors (63.9%) had homogeneously enhanced tumor capsules (Fig. 3A, 3B, 3C).
All tumors were characterized microscopically by a prominent plexiform vascular pattern admixed with an abundant myxoid matrix. The extent of cellularity ranged from slight to moderate, and the lesions were composed of small uniform, round, or spindle-shaped hyperchromatic cells. Tumor necrosis was found on microscopic observation in 12 cases (33.3%). The necrotic areas varied in degree, but most tumors contained only a small amount of necrotic areas that were difficult to identify on MR images.
Statistically significant MRI findings that favored a diagnosis of intermediate- or high-grade tumor were large tumor size (> 10 cm) (p < 0.01), deeply situated tumor (p < 0.05), tumor possessing irregular contours (p < 0.001), absence of lobulation (p < 0.001), absence of thin septa (p < 0.05), presence of thick septa (p < 0.01), absence of tumor capsule (p < 0.001), high-intensity signal pattern (p < 0.01), pronounced enhancement (p < 0.001), and globular and nodular enhancement (p < 0.001). The presence of thin septa or a tumor capsule indicates low-grade tumor. Irregular contours were found in only 10 high-grade tumors (58.8%). All the low-grade tumors had a capsule, thin septa, and a high-intensity signal pattern. The odds ratios for a specific finding favoring a diagnosis of intermediate- or high-grade tumor are shown in Table 1. The multiple logistic regression model showed that irregular contour and thick septa were the most significant predictors of intermediate- or high-grade tumors, with an odds ratio of 13.8 for both (95% confidence interval [CI], 1.5–128.8; p < 0.05).
|
At the last follow-up, 10 (27.8%) of the 36 patients had died of their disease and four (11.1%) were alive with metastatic disease. The 5- and 10-year survival rates were 80.5% and 72.4%, respectively. The univariate analysis showed that thin septa (p < 0.05), tumor capsule (p < 0.01), and pronounced enhancement (p < 0.01) were significantly associated with overall survival (Table 2). The multivariate analysis revealed that pronounced enhancement was the most significant adverse prognostic factor (Fig. 5) with a relative risk of 7.3 (95% CI, 1.5–35.1; p < 0.05).
|
|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The amount of necrosis has been reported to be correlated with clinical outcome [21–25]. Spontaneous tumor necrosis identified in four (4%) of 95 patients with myxoid–round cell liposarcomas was correlated with increased risks of metastasis and death [18]. In our study, we did not evaluate the relationship between the presence of tumor necrosis and patient prognosis, because most tumors accompanied by necrosis in our study contained only a small amount of necrotic areas that were difficult to identify on MR images.
On contrast-enhanced MR images, pronounced enhancement was located mainly at the periphery of the lesion in 61.1% of the patients, and globular and nodular enhancement occurred at the lesion center in 44.4%. These two patterns of enhancement were characteristic of intermediate- or high-grade tumors. Round cell components were reported to be located at the peripheries of lobules; adjacent to fibrous septa extending through the tumor; and surrounding large vessels, particularly in tumors with only a small amount of round cell components [7, 8]. Thus, these two enhancement patterns may be reliable imaging findings for detecting round cell components within tumors. In one study, despite a small sample size, researchers showed that patients (n = 5) who initially had a tumor with 5% or greater round cell components had a significantly higher incidence of metastasis or death from disease than those (n = 7) who initially had a tumor with less than 5% round cell components [18]. In a study of 24 patients with round cell components composing 25% or more of the tumor, round cell components were associated significantly with a lower survival rate [19]. However, the correlation between the quantity of round cell components and the clinical outcome may depend on the difficulty in quantifying the round cell components at transitional areas at microscopic observation.
There was no significant difference between the risks of an adverse outcome in patients with myxoid and transitional areas without round cell components and those with myxoid areas alone [19]. The pathologic variables responsible for differences among observers in identifying round cell components are considered to be numerous and include inaccurate criteria for tissue processing and selection of the assessment area within the spectrum of myxoid–round cell liposarcomas [20]. Our results suggest that contrast-enhanced MR images can assist in detecting round cell component content within the entire tumor and assist in the distinction of low-grade and of intermediate- or high-grade myxoid–round cell liposarcomas.
In previous reports [26–29], the descriptions of the enhancement patterns identified on MR images included little enhancement or a few patterns (i.e., heterogeneous, homogeneous, no enhancement). However, the end points selected in these prior studies depended simply on the pathologic diagnosis of "myxoid liposarcoma," and the investigators were unaware of the lineage of "myxoid–round cell liposarcoma" as a disease entity. The results of our study are based on a definite diagnosis of myxoid–round cell liposarcoma, and we stress that the presence of globular and nodular or pronounced enhancement identified on MRI is a finding suggestive of intermediate- or high-grade tumor and reflects the amount of round cell components in the tumor, which strongly affects patient outcome.
The presence of linear or amorphous hyperintense foci behaving like fatty tissue on T1-weighted MR images has been reported to be a pattern suggestive of myxoid liposarcoma [27]. Myxoid–round cell liposarcoma often consists of multiple histologic subtypes in the same lesion. We observed a high-intensity signal pattern in 15 low-grade tumors, and this finding was consistent with immature fatty tissue or the fat components of the tumors. Immature spindle cells lacking obvious fat genesis may be seen next to multivacuolated lipoblasts. Although MRI is sensitive enough to detect minute fat deposits or immature fatty components, our univariate analysis showed no significant association between high-intensity signal pattern on MR images and survival [28, 29].
In summary, the spectrum of MRI findings in myxoid–round cell liposarcomas is continuous. MRI findings can assist in the distinction between low-grade and intermediate- or high-grade myxoid–round cell liposarcomas. MRI findings that favored a diagnosis of intermediate- or high-grade tumor included large (> 10 cm) size of tumor, deeply situated tumor, tumor possessing irregular contours, absence of lobulation, absence of a tumor capsule, absence of thin septa, presence of thick septa, high-intensity signal pattern, pronounced enhancement, and globular and nodular enhancement. The presence of thin septa or a tumor capsule indicates low-grade tumor. Imaging features associated with overall survival were thin septa, a tumor capsule, and pronounced enhancement. Multivariate analysis showed that pronounced enhancement on MRI is the most significant factor in predicting an adverse prognosis for patients with myxoid–round cell liposarcoma.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  M. D. Murphey, L. K. Arcara, and J. Fanburg-Smith From the Archives of the AFIP: Imaging of Musculoskeletal Liposarcoma with Radiologic-Pathologic Correlation RadioGraphics, September 1, 2005; 25(5): 1371 - 1395. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 U. Tateishi, T. Hasegawa, H. Onaya, M. Satake, Y. Arai, and N. Moriyama Myxoinflammatory Fibroblastic Sarcoma: MR Appearance and Pathologic Correlation Am. J. Roentgenol., June 1, 2005; 184(6): 1749 - 1753. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
