| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1 Both authors: Department of Radiology, Duke University Medical Center, Erwin Rd., DUMC #3808, Durham, NC 27710.
Received July 2, 2003;
accepted after revision September 4, 2003.
Address correspondence to C. M. Gaskin
(cree_gaskin{at}hotmail.com).
Abstract
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
MATERIALS AND METHODS. One hundred twenty-six consecutively imaged grossly fatty masses were retrospectively reviewed. MRI examinations, their prospective interpretations, and their corresponding pathology reports were compared to determine the reliability of MRI in distinguishing simple lipomas, lipoma variants, and well-differentiated liposarcomas.
RESULTS. The success of MRI in identifying well-differentiated liposarcomas among other fatty masses was as follows: sensitivity, 100%; specificity, 83%; accuracy, 84%; positive predictive value, 38%; and negative predictive value, 100%. MRI was 100% specific in the diagnosis of simple lipoma. Sixty-three percent of lesions considered suspicious for well-differentiated liposarcoma were actually simple lipomas (13%) and benign lipoma variants (50%), including chondroid lipoma (13%), osteolipoma (6%), hibernoma (6%), lipoleiomyoma (6%), angiolipoma (6%), and infarcted lipoma (13%).
CONCLUSION. Because of differences in treatment, prognosis, and long-term follow-up, it is important to preoperatively distinguish simple lipomas from well-differentiated liposarcomas. MRI is highly sensitive in the detection of well-differentiated liposarcomas and highly specific in the diagnosis of simple lipomas. However, when an extremity or body wall lesion is considered suspicious for well-differentiated liposarcoma, it is more likely (64%) to represent one of many benign lipoma variants.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Although simple lipomas and well-differentiated liposarcomas are both grossly fatty masses, MRI has been described as useful in attempting to distinguish these two lesions. Several imaging features are more closely associated with well-differentiated liposarcomas, including thickened or nodular septa, associated nonadipose masses, prominent foci of high T2 signal, and prominent areas of enhancement [6–9]. Higher grade liposarcomas generally do not confound the MRI diagnosis of grossly fatty lesions because they typically contain little or no macroscopic fat.
Often when the MRI diagnosis included atypical lipoma, the ultimate pathologic diagnosis was neither simple lipoma nor atypical lipoma, but rather one of many benign lipoma variants. The chief purposes of this study were to evaluate the reliability of MRI in distinguishing lipomas, lipoma variants, and well-differentiated liposarcomas and to identify the types and frequencies of the various benign lesions that mimic well-differentiated liposarcoma on imaging.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Sixty-four of these 126 imaged fatty masses were resected at our institution. We reviewed images from 61 of these studies; the remaining three studies could not be located. We compared the original prospective MRI interpretations with their corresponding pathology reports. MRI interpretations of these grossly fatty masses were in keeping with standard criteria set forth in the literature for distinguishing simple lipomas from well-differentiated liposarcomas.
Simple lipomas may contain a few thin, discrete septa, but they are otherwise homogeneously fatty masses (Fig. 1). Bands of muscle fibers are also acceptable components of intramuscular lipomas (Fig. 2). All other intralesional components raise suspicion for well-differentiated liposarcoma. Thickened or nodular septa (generally > 2 mm) (Fig. 3A), associated nonadipose masses, prominent foci of high T2 signal (Fig. 3B), and prominent areas of enhancement (Fig. 4A, 4B) are all associated with increased risk of well-differentiated liposarcoma [6–12]. In our series, lesions with these features were interpreted as suspicious for, probable, or possible atypical lipoma.
|
|
|
|
|
|
All imaging was performed with 1.5-T magnets (Signa, General Electric Medical Systems, Milwaukee, WI) and commercially available transmit–receive coils when appropriate. Imaging protocols varied, but in general, they included at least two planes of spin-echo T1-weighted images (TR range/TE range, 400–650/8–20) and fast spinecho T2-weighted images (3,000–4,000/60–105) with frequency-selective fat saturation. Occasional studies included STIR sequences (inversion time, 110–150 msec). Most of our examinations were performed without contrast agents, because we have not found contrast administration to be helpful in distinguishing these lesions, and other imaging features were adequate to identify the lesion.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Of the 18 cases considered suspicious for well-differentiated liposarcoma, 16 were resected at our institution (two were lost to follow-up). Of these 16 cases, six (38%) were found to be well-differentiated liposarcoma at pathology (Figs. 3A, 3B, 4A, 4B, 5A, 5B). Six more of these 16 cases were benign lipoma variants: two chondroid lipomas (13%) (Fig. 6), one osteolipoma (6%) (Fig. 7A, 7B), one hibernoma (Fig. 8), one lipoleiomyoma, and one angiolipoma (Fig. 9A, 9B). Two other cases were infarcted, necrotic lipomas (Fig. 10). The remaining two cases were interpreted as simple lipomas at histology (Fig. 11).
|
|
|
|
|
|
|
|
|
|
MRI examinations of 108 fatty masses were interpreted as diagnostic of simple lipoma (Figs. 1, 2, and 12). Of these 108 masses, 48 were resected at our institution. All 48 cases (100%) were confirmed to be simple lipomas at histology. One of these was an infiltrative intramuscular lipoma (Fig. 2).
|
At our institution, with the experience of seven different musculoskeletal radiologists using standard published criteria, the success of MRI in evaluating a fatty mass for well-differentiated liposarcoma was as follows: sensitivity, 100%; specificity, 83%; accuracy, 84%; positive predictive value (PPV), 38%; and negative predictive value, 100%. MRI was found to be 100% specific for the diagnosis of simple lipoma when standard imaging criteria were met.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Superficial well-differentiated liposarcomas of the extremities have the best prognosis of all sites. Four surgical and pathologic series have reported rates of local recurrence of superficial extremity lesions ranging from 0% to 50% [2, 4, 13, 16]. Results for local recurrence of deep extremity lesions have been more consistent, ranging from 43% to 69% [2, 4, 5]. Studies of retroperitoneal sites have reported rates of local recurrence of 63% [2] and 91% [5]. Lucas et al. [3] reported a 59% rate of local recurrence for all (mostly retroperitoneal) lesions not found in the extremities.
Because of this propensity for local recurrence, wide local excision is advocated for treatment of these lesions. Lucas et al. [3] reported an 11% recurrence rate with wide local excision compared with their 62% recurrence rate with marginal excision. Wide local excision can often be difficult with both deep extremity and retroperitoneal lesions, likely contributing to their higher recurrence rates [3–5].
We are aware of only one superficial lesion of the extremity recurring with dedifferentiated, higher grade histology [13]. Deep lesions of the extremities recur with dedifferentiated histology more frequently, with reported rates ranging from 6% to 13% [3–5]. Retroperitoneal lesions recur with dedifferentiation even more frequently, with reported rates of 17–30% [3, 5]. Metastases from dedifferentiated recurrences have occasionally been reported in both deep extremity and retroperitoneal lesions [2, 3, 5, 17]. Retroperitoneal lesions ultimately cause death (usually because of local extension of disease) in 30–38% of cases [2, 3, 5].
The long-term prognosis and initial operative management of simple lipomas and well-differentiated liposarcomas are different. Simple lipomas are often successfully treated by local or marginal excision, whereas well-differentiated liposarcomas are preferentially treated with wide local excision because of their high rate of local recurrence. Well-differentiated liposarcomas also require long-term clinical follow-up because of their propensity for delayed dedifferentiation, typically occurring 5–10 years after initial resection [2–5]. Because of the differences in treatment and prognosis, it is important to distinguish these lesions preoperatively. The challenge in imaging is that both lesions are grossly fatty masses with significant overlap in their range of appearances [6–12].
Simple lipomas can have a characteristic appearance on MRI. A discrete, encapsulated, homogeneous fatty mass is most certainly a simple lipoma. Simple lipomas, however, may also contain muscle fibers, blood vessels, fibrous septa, and areas of necrosis or inflammation. All these intralesional nonadipose components can confound the correct imaging diagnosis because they can mimic findings associated with well-differentiated liposarcomas.
Thickened or nodular septa (generally > 2 mm thick), associated nonadipose masses, prominent foci of high T2 signal, and prominent areas of enhancement are all findings reported to be suspicious for liposarcoma [6–12]. Unfortunately, all these findings have also been reported in pathologically proven cases of simple lipomas [6, 7, 9, 18–20].
In our series, MRI was 100% specific in the diagnosis of simple lipoma when a grossly fatty mass had few or no thin septa and minimal or no areas of enhancement or high T2 signal. Infiltrative intramuscular lipomas are the exception to this description of lipoma. Although some intramuscular lipomas are homogeneous fatty masses (Fig. 12), others are heterogeneous lesions with infiltrative margins and intermingled muscle fibers. Two prior studies have shown that infiltrative margins suggest the diagnosis of benign intramuscular lipoma rather than that of well-differentiated liposarcoma [12, 18]. One of our cases had this appearance and was correctly interpreted prospectively as an infiltrating intramuscular lipoma (Fig. 2). Other than this single case of infiltrating intramuscular lipoma, we did not find the margins to be critical to our interpretations. When a grossly fatty mass did not match either of these descriptions, it was considered suspicious for well-differentiated liposarcoma. With these criteria, we found MRI to be 100% sensitive for well-differentiated liposarcoma among 126 consecutive grossly fatty masses.
In spite of 100% sensitivity and 83% specificity, our PPV was only 38%. This is largely because 10 (63%) of 16 of our suspicious masses actually had alternative benign histology—what we call benign lipoma variants. These lesions were neither simple lipomas nor well-differentiated liposarcomas but were a mix of chondroid lipoma, osteolipoma, hibernoma, angiolipoma, lipoleiomyoma, and necrotic lipoma.
Chondroid lipoma was first described in the pathology literature in 1993 [21]. It is considered a benign lesion that is treated successfully with local excision. Despite its benign nature and distinctive pathologic features, it has some findings that mimic liposarcoma and myxoid chondrosarcoma at pathology [21–23]. We show that this tumor can also mimic liposarcoma on MRI (two of 16 fatty tumors suspicious for liposarcoma) (Fig. 6). At least one other case of MRI of chondroid lipoma has been reported; however, this mass was predominantly low in signal on T1-weighted images with a few thin strands of high signal [24].
Hibernoma is an uncommon tumor of brown fat thought to morphologically mimic the fat of hibernating animals. The first large pathology series [25] evaluating this tumor was published in 2001, reviewing 170 cases. That series showed the benign nature of this tumor; none of the cases recurred during a mean follow-up period of 7.7 years. In this series and a review of the literature, the authors described no cases of metastasis due to hibernoma. Despite their benign behavior, some variants of hibernoma can be confused histologically with liposarcoma. Similarly, five independent reports describe isolated cases of hibernoma mimicking liposarcoma on MRI [26–30]. In our series, one of 16 fatty masses considered suspicious for liposarcoma was actually a case of hibernoma (Fig. 8).
Simple lipomas may undergo necrosis or infarction, producing heterogeneity and masslike regions with imaging features similar to those of nonfatty tissue. At least three cases of fat necrosis mimicking liposarcoma on CT or MRI have been reported [6, 31, 32]. In our series, two simple lipomas with areas of infarction were interpreted prospectively as suspicious for possible liposarcoma (Fig. 10).
Two more cases of false-positive results for MRI examinations interpreted as suspicious for liposarcoma were actually found to represent simple lipomas at histology (without mention of areas of necrosis or nonfatty elements). One of these was a diffuse infiltrating lipomatosis of the extremity that had undergone multiple resections before being imaged at our institution. The other lesion was a grossly fatty mass that clearly, even in retrospect, had multiple small nodular areas with signal characteristics other than those of fat (Fig. 11). Although we considered this a false-positive finding for statistical purposes, we speculate that the histologic diagnosis of simple lipoma was due to either sectioning bias (sampling error) or the presence of benign tissue elements (e.g., fibrous tissue or fat necrosis) that the pathologist chose to ignore as incidental.
In our series, one osteolipoma (Fig. 7A, 7B), one lipoleiomyoma, and one angiolipoma (Fig. 9A, 9B) also mimicked well-differentiated liposarcomas. Galant et al. [6] reported a similar case of an osteolipoma of the extremity presenting as a fatty mass with nodular nonfatty foci depicted on MRI. Osteolipoma, lipoleiomyoma, and angiolipoma are all thought to be benign, but close and long-term follow-up may be prudent because of a paucity of pertinent clinical information.
Our imaging protocol does not typically include contrast administration for the evaluation of fatty masses. In our opinion, contrast administration does not contribute significantly to the diagnosis (given other imaging features) and subjects the patient to additional cost and risk. Hosono et al. [7] have shown that well-differentiated liposarcomas have septa that enhance more dramatically than those of simple lipomas. We agree with this finding but maintain that the diagnosis can be made on the basis of other imaging features. For example, dramatically enhancing septa are typically thick, nodular, and abnormal in signal on unenhanced sequences. These other imaging features would suggest the correct diagnosis. Thick septa that enhance poorly would favor a benign diagnosis but would not eliminate uncertainty about a possible malignant nature in light of the other imaging features. In our series, benign fatty tumors, other than simple lipoma, confounded the diagnosis of well-differentiated liposarcoma much more frequently. There is a paucity of information in the literature regarding the enhancement patterns of these other benign fatty lesions. More work needs to be done to define the role of contrast agents in distinguishing these benign lipoma variants from well-differentiated liposarcomas.
In conclusion, MRI is useful, but imperfect, in distinguishing lipomas, lipoma variants, and well-differentiated liposarcomas. An MRI examination showing a discrete, homogeneous, fatty mass can be considered diagnostic of a simple lipoma. In our series, MRI was 100% specific for the diagnosis whenever a grossly fatty mass had few or no thin septa; mild, linear, or no enhancement; and minimal or no areas of high T2 signal.
Well-differentiated liposarcoma should be considered in the differential diagnosis of a grossly fatty mass that does not meet these criteria. In our series, not meeting these criteria was 100% sensitive and 83% specific for the diagnosis of well-differentiated liposarcoma among 126 consecutive grossly fatty masses. When the MRI appearance of an extremity or body wall tumor suggests the diagnosis of a well-differentiated liposarcoma, the lesion is actually more likely to be one of the benign lipoma variants. In our series, these tumors included chondroid lipoma, osteolipoma, angiolipoma, hibernoma, lipoleiomyoma, and necrotic lipoma.
It is important for the radiologist to preoperatively suggest the diagnosis of a well-differentiated liposarcoma, rather than a simple lipoma, because of the differences in prognosis, initial treatment, and long-term care. Wide local excision and long-term clinical follow-up (> 5 years) are advocated for well-differentiated liposarcomas because of their high rate of local recurrence and potential for delayed dedifferentiation (and subsequent risk for metastasis). Although we show that lesions suspicious for well-differentiated liposarcoma on MRI are actually more likely to be benign (64%), we still believe that these lesions should be treated with wide local excision if feasible (given the 36% risk of well-differentiated liposarcoma).
The terms "atypical lipoma," "atypical intramuscular lipoma," and "atypical lipomatous tumor" have all been introduced to provide less menacing names for well-differentiated liposarcomas of the extremities [2, 33]. We believe that these terms are appropriate for well-differentiated liposarcomas of the extremities as long as there is mutual understanding between the radiologist and the referring clinician as to their meaning.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  N. M. Blitman, D. Pandya, B. A. Thornhill, D. M. Barlev, and T. L. Levin Pain in the Butt: Spectrum of Gluteal Region Abnormalities in Children Am. J. Roentgenol., May 1, 2009; 192(5): 1286 - 1291. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 O. M. Navarro, E. E. Laffan, and B.-Y. Ngan Pediatric Soft-Tissue Tumors and Pseudo-tumors: MR Imaging Features with Pathologic Correlation: Part 1. Imaging Approach, Pseudotumors, Vascular Lesions, and Adipocytic Tumors RadioGraphics, May 1, 2009; 29(3): 887 - 906. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 I-C Chiang, M-Y Jang, K-B Tsai, and T-J Hsieh Huge renal lipoma with prominent hypervascular non-adipose elements. Br. J. Radiol., October 1, 2006; 79(946): e148 - e151. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Lopez Soriano, A. Tomasello, P. Luburich, and A. Noel Fat Necrosis in a Chest Wall Lipoma Am. J. Roentgenol., September 1, 2004; 183(3): 866 - 866. [Full Text] [PDF]
|
|
|
|
|
|
M. D. Murphey, J. F. Carroll, D. J. Flemming, T. L. Pope, F. H. Gannon, and M. J. Kransdorf From the Archives of the AFIP: Benign Musculoskeletal Lipomatous Lesions RadioGraphics, September 1, 2004; 24(5): 1433 - 1466. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
