10.1245/s10434-006-9306-3
Annals of Surgical Oncology 14:1507-1514 (2007)
© 2007 Society of Surgical Oncology
Skeletal Metastases in Myxoid Liposarcoma: An Unusual Pattern of Distant Spread
Joseph H. Schwab, MD1,
Patrick Boland, MD1,
Tianhua Guo, MD2,
Murray F. Brennan, MD1,
Samuel Singer, MD1,
John H. Healey, MD1 and
Cristina R. Antonescu, MD2
1 Department of Surgery, Memorial Sloan-Kettering Cancer Center, New York, NY 10021, USA
2 Department of Pathology, Memorial Sloan-Kettering Cancer Center, 1275 York Ave, New York, NY 10021, USA
Correspondence: Address correspondence and reprint requests to: Cristina R. Antonescu, MD; E-mail: antonesc{at}mskcc.org
 |
ABSTRACT
|
|---|
Background: Myxoid liposarcoma (MLS), the second most common subtype of liposarcoma, occurs predominantly in the extremities of young adults and has a disproportionately high tendency to metastasize to unusual soft tissue locations, before disseminated spread or pulmonary metastases. Anecdotal evidence, mainly supported by isolated case reports, suggests that a subset of these patients also develop bone metastasis, especially within the spine, which was previously under-appreciated.
Study Design: In this study we investigate the incidence of osseous metastases in a well-annotated sarcoma database and correlate this endpoint with clinicopathologic and molecular findings.
Results: From a total of 230 patients with MLS diagnosis confirmed histologically, who were managed and followed prospectively at MSKCC, 40 (17%) developed skeletal metastases, comprising 56% of all metastatic events. A significant number of these bone metastases were identified early in the disease course, before the manifestation of disease in sites where sarcomas usually metastasize, such as lung. From the time of 1st metastasis, the 5 years median survival was 16%. The majority (78%) of MLS patients developing bone metastases had a histologic high grade primary tumor. The median overall survival for the high grade tumors was 55 months, as compared to 105 months for low grade cases. Eleven (84%) of 13 cases tested by RT-PCR demonstrated a type II TLS-CHOP fusion transcript.
Conclusion: These findings suggest that MLS has a high incidence of osseous metastases, with predilection to spine, and often associated with the most common type of TLS-CHOP transcript. Screening should include images of the spine in high-risk MLS patients to exclude spinal metastases.
Key Words: Liposarcoma • Myxoid • Metastatic • Bone • TLS-CHOP
|
INTRODUCTION
|
|---|
Liposarcoma is the most common soft tissue sarcoma and can be classified in three distinct histologic types (well-differentiated, myxoid and pleomorphic), based on morphology and genetics. Myxoid liposarcoma (MLS) is the second most common subtype of liposarcoma, accounting for 10% of all adult soft tissue sarcomas.1 In contrast with other types of liposarcomas, MLS typically occurs in younger individuals, with a peak in the 4th and 5th decade of life, and represents the most common type of liposarcoma in patients younger than 20-year old. It characteristically occurs in the deep-seated soft tissues of the extremities, and in more than two-thirds of cases arises in the musculature of the thigh.
Another important distinction from other types of liposarcomas or myxoid sarcomas of the extremities is that MLS is associated with an unusual pattern of metastasis to somatic soft tissue sites, including retroperitoneum, opposite extremity, axilla, etc, even before it metastasized to the lung.2–4 Soft tissue sarcomas in general tend to metastasize to the lungs, while other sites are typically involved in advanced stages of disease. Furthermore, MLS patients present more commonly with multifocal disease, either synchronous or metachronous, than any other soft tissue sarcoma.2 Our group previously determined the monoclonality of these multifocal implants, establishing the metastatic nature of the distant soft tissue lesions.3 Reportedly, MLS metastasize in about one out of three cases. The propensity to metastasis has been correlated with larger, higher grade tumors and the degree of necrosis.4
A specific TLS-CHOP fusion gene resulting from t(12;16)(q13;p11) is present in >95% of MLS. Three main types of TLS-CHOP fusion have been described, differing by the presence or absence of TLS exons 5, 7 or 8 in the fusion product.4 In contrast with other translocation-associated sarcomas, the molecular variability of TLS-CHOP fusion transcript structure does not appear to have a significant impact on clinical outcome.4 However, the effect of the TLS-CHOP transcript heterogeneity on the pattern and distribution of metastatic sites in MLS was not previously investigated.
The degree to which MLS spreads to bone has not been specifically studied. It is unclear whether skeletal metastasis represents the usual pattern of spread in MLS or whether it is the mark of a specific molecular subset of MLS. In this study we sought to evaluate the incidence and clinicopathologic features associated with MLS metastatic to bone and investigate the whether the TLS-CHOP fusion transcript structure correlates with the metastatic propensity to bone.
|
MATERIALS AND METHODS
|
|---|
Study Group
Our prospective sarcoma database (1982–2005) was searched for MLS patients who developed bony metastasis. The diagnosis of MLS was confirmed by review of the histologic slides in all cases, including primary and metastatic sites. In order to determine an accurate incidence of bone metastasis in MLS, we also reviewed the remaining MLS cases without skeletal spread. Cases in which the pathologic material was not available for review were excluded. The diagnosis of MLS was based on the presence of uniform round to oval shaped cells in a myxoid background, associated with a plexiform capillary network. The primary tumors were scored for the percentage of round cell component and tumor necrosis present. The percentage of round cells was estimated by scanning all individual sections (average 1 block/1 cm3 of tumor) of the primary specimen. Tumors demonstrating a round cell component of at least 5% were designated as high grade, and tumors with less than 5% were deemed low grade as described previously.4,5
All primary tumors were imaged with magnetic resonance imaging. Myxoid liposarcomas are typically heterogenous on T1 weighted images. Areas of low signal intensity correspond to the myxoid component and high signal intensity correspond to the lipid component. The areas of high signal intensity seen on T1 will drop out with fat saturation, which is also characteristic (Fig. 1
). Myxoid liposarcomas have heterogenous signal characteristics with T2 weighted images.
View larger version (49K):
[in this window]
[in a new window]
|
FIG. 1. Primary myxoid liposarcoma of the posterior thigh. A T1 weighted images demonstrating heterogeneous signal within a lobulated mass. B T1 with fat saturation demonstrating decreased signal within the mass.
|
|
All primary tumors were managed by surgical resection with a curative intent. Radiation therapy and chemotherapy were utilized on a case-by-case basis. All patients were followed clinically and with radiographic (chest X-ray and/or a CT of the chest) surveillance for metastasis. Local recurrence was monitored by physical exam and MRI. Additional screening tests, ordered at the discretion of the physician to rule out skeletal metastases, included bone scan, PET scan, and CT of abdomen/pelvis.
Molecular Analysis: RT-PCR for TLS-CHOP and EWS-CHOP Fusion Transcripts
Samples were snap-frozen in liquid nitrogen and stored at –70°C. Extraction of total RNA was based on the guanidinium isothiocynate–phenol chloroform method using the RNA Wiz reagent (Ambion, Inc, Austin, TX, USA). Analysis by reverse-transcriptase polymerase chain reaction (RT-PCR) for TLS-CHOP and EWS-CHOP transcripts was performed. The adequacy of the extracted RNA was assessed by RT-PCR, using primers for PGK (phosphoglycerate kinase) transcripts. Negative controls that lacked either tumor RNA or RT were routinely used. Three micrograms of total RNA were subjected to RT-PCR using Qiagen 1 Step RT-PCR kit (Qiagen, Inc., Valencia, CA, USA), using a forward primer within exon 5 of TLS (5'-CAG CCA GCA GCC TAG CTA TG-3') or in exon 7 of EWS (5'-CTG GAT CCT ACA GCC AAG CTC CAA G-3') and a reverse primer in exon 3 of CHOP (5'-TGT CCC GAA GGA GAA AGG CAA TG-'3). The Qiagen 1 Step RT-PCR conditions included: (1) the RT step: 50°C for 30 min; (2) 95°C for 15 min; (3) 95°C for 45 s, 64°C for 45 s, 72°C for 1 min (35 cycles); (4) 72°C for 7 min. The amplified RT-PCR products were identified by agarose gel electrophoresis. The expected sizes for TLS-CHOP fusion transcripts were 526 bp for type 7-2, 238 bp for type 5-2, and 625 bp for type 8-2 (Fig. 2). The expected size for the EWS-CHOP fusion transcript was 833 kb.
Clinical Follow-up and Statistical Analysis
Statistical analysis was performed using SPSS software (SPSS Chicago, IL, USA). Time to event end points were analyzed using Kaplan Meier survival plots analyzed by the log rank test. Clinical endpoints analyzed were time to distant disease specific survival (DDSS) and disease specific survival (DSS). The factors analyzed included: age, location of bony metastasis, site of first metastasis, histologic grade, and the TLS-CHOP fusion transcript type.
Clinical follow up was available in all patients, and was achieved by reviewing medical histories, histologic specimen and imaging studies (plane X-ray, MRI, CT, bone scan, FDG-PET). Patients were followed for a mean of 57.8 months (median 38 months) or until death (range 7–408 months).
|
RESULTS
|
|---|
Patients
Two hundred and thirty patients with confirmed MLS diagnosis after pathologic review were selected from our prospective sarcoma database. Seventy-two (31%) patients developed metastatic disease at last follow-up and among these 40 (17%) patients were diagnosed with bony metastasis (Table 1). Twenty-nine (72%) of the 40 patients with skeletal metastasis were male and 11 were female (28%). The median age at diagnosis of the primary was 47 (range 29–75). Thirty-one (78%) of the primary tumors were designated as high grade, and nine were low grade (22%). Two (5%) of the primary tumors were smaller than 5 cm, 6 (15%) were between 5 and 10 cm, and 28 (70%) were greater than 10 cm. All the tumors were located deep to the overlying fascia except one.
Thirty-two (14%) of the initial 230 patients developed pulmonary metastasis, of which only 14 (6%) were associated with bony metastasis. Fourteen (6%) of the initial 230 patients with distant metastases did not have pulmonary or bony metastases. The most common sites of metastases in this subgroup were the abdomen in eight cases and retroperitoneum in six cases. Thirty-seven (93%) of 40 patients with osseous metastases presented with localized disease, while 3 (7%) patients had distant metastases at diagnosis.
Treatment of Primary Tumor
Thirty-eight of the 40 patients with bony metastasis were treated surgically at MSKCC for either the primary disease (28 patients, 70%) or for the metastatic disease (10 patients, 25%). Of the 20 patients presenting to our institution after having had surgery on their primary tumor at an outside hospital, 18 underwent further surgery for their primary tumor after referral. Two patients had received adequate wide-resection at an outside hospital and were referred only for management of metastatic disease, which included radiation and in one case chemotherapy. Eight patients had surgical decompression of their spinal metastasis after having their initial surgery elsewhere.
Overall, thirty-two (80%) patients were treated with a wide-resection of the primary tumor. One patient was treated with an amputation and one patient was treated with marginal excision. Five (12.5%) patients had a widely contaminated margin. Two of these five underwent subsequent surgery including re-excision in one and amputation in the other. The other three were treated with adjuvant radiation therapy. In one patient a wide margin was not possible, and she was treated with a gross total resection followed by external beam radiation.
Radiation
Twenty-eight (70%) patients underwent radiation therapy to their primary tumor site. Twenty-one (53%) patients received external beam radiation therapy with a median dose of 6,000 cGy (range 3,000–6,600 cGy). Seven (17.5%) patients received brachytherapy with four of those seven receiving external beam boosts of 3,000 cGy.
Local Recurrence
Eight (20%) patients developed a local recurrence to their primary tumor site. Five of the eight had surgery elsewhere prior to presenting to our institution. Six of the eight patients had further surgery on their primary upon presentation to our institution. Three wide resections were performed, two of which occurred after an open biopsy had been obtained elsewhere. One of these three wide resections was performed on a local recurrence that had been initially resected elsewhere. Two patients presented with local recurrences after having had wide resections performed elsewhere. These two patients had wide spread metastasis and were treated with radiation therapy rather than further surgery to their primary site. Two of the eight patients developed recurrence after being treated with wide contaminated margins at our institution. One patient was treated with intralesional resection of a retroperitoneal lesion that recurred. Five of the eight recurrences were high grade at initial presentation and three were low grade. The average time to local recurrence was 16 months (range 3–28). Local recurrence preceded distant metastasis in all but one case.
Distant Metastasis
All 40 patients in our study had metastases to osseous structures to be included. Thirty-two (80%) patients were diagnosed with bony metastases after the patient presented with signs or symptoms of metastasis. The most common complaint was back pain which occurred in 27 (68%) patients. The osseous metastases commonly presented as lytic, destructive lesions without significant sclerosis (Fig. 3). The median time to first metastasis was 23 months (range 0–130). Twenty-seven (68%) patients had their first metastasis to bone. The lung was the first site of metastasis in two (5%) patients. The most common site of metastasis was the spine occurring in 33 (83%) patients, followed by the ribs (16%), lung parenchyma (14%), and subcutaneous tissues (14%) (Table 2). The spine metastases demonstrated the typical MRI characteristics found in myxoid liposarcoma. T1 weighted images appeared heterogenous with areas of higher signal intensity corresponding to the lipid component and low signal to the myxoid component. T2 weighted images were also heterogenous with areas of high and low signal intensity (Fig. 4). The treatment of metastasis was individualized to each patient. Chemotherapy was given to 32 (80%) patients, with Adriamycin being utilized in 19 (48%) and Adriamycin and ifosfamide in 13 (32%). Radiation therapy was utilized most commonly for metastatic disease. All patients were treated with external beam radiation for at least one of their metastasis.
View larger version (103K):
[in this window]
[in a new window]
|
FIG. 4. Magnetic resonance imaging demonstrating spine metastases from Myxoid Liposarcoma. A T1 weighted image demonstrating low signal intensity in the sacral spine. B T2 weighted image demonstrating increased signal intensity in the lumbar spine. C T2 STIR image demonstrating increase signal intensity in the third lumbar vertebrae with anterior soft tissue extension.
|
|
Molecular Analysis
Of the 18 (45%) cases with available frozen tissue for molecular analysis, 13 had adequate quality RNA for RT-PCR analysis. In twelve of the 13 cases a TLS-CHOP fusion transcript was found. A type II fusion (exon 5 of TLS fused to exon 2 of CHOP) was found in all except one tumor, which was a 7-2 TLS-CHOP fusion transcript (Fig. 2
). An EWS-CHOP transcript was found in the remaining case.
Survival
Twenty-five (63%) patients died of their disease. During the follow-up, 12 (30%) patients were alive with disease and 3 (7%) patient were alive with no evidence of disease. The median 5-year disease specific survival was 42% (CI 38–71), while the median 5-year disease specific survival from the time of first metastasis was 16% (CI 18–28). The median survival was 105 (CI 15–196) months for patients with low-grade malignancy and 55 (CI 27–83) months for high-grade malignancy (P = 0.26) (Fig. 5).
View larger version (6K):
[in this window]
[in a new window]
|
FIG. 5. Overall disease specific survival in patients with extremity myxoid liposarcoma metastatic to bone. The dashed line represents the patients that presented with a low grade tumor. The solid line represents patients who presented with high grade tumors. No significant difference was found between the two groups (P = 0.26).
|
|
|
DISCUSSION
|
|---|
In contrast with other extremity soft tissue sarcomas, which metastasize primarily to the lung, MLS is associated with an unusual pattern of metastasis. Previous reports have shown metastases to extra-pulmonary sites, including the retroperitoneum, pericardium, subcutaneous and bone.4,6,7 Our series focuses only on those patients with metastasis to bone, since more than half (56%) of the total metastatic sites represent skeletal metastases, 70% in the absence of pulmonary spread, and a high incidence of metastasis to the spine.
Most (85%) patients who developed bone metastases showed a type II TLS-CHOP fusion transcript. In a prior study from our group, the type II TLS-CHOP fusion comprised 67% of 82 MLS patients. The difference between the incidence of the type II TLS-CHOP fusion transcript between these two groups was not significant (P = 0.2).4 In that same series, which included only molecularly confirmed MLS, a distant metastasis was found in 30 out of 79 (40%) patients with available follow-up, with 63% of the total distant metastases were somatic soft tissue sites. Osseous metastasis was found in 37% of cases, while lung metastasis was noted in 33%.
In the present series, 31% of patients with MLS developed distant metastases. Fifty-six percent of those patients developed bony metastases. Forty-six percent of those patients with metastatic MLS develop spinal metastasis, and 44% will develop pulmonary metastasis. This has implications on screening for metastasis. Most soft tissue sarcomas are screened by physical examination and radiographic images of the lung and the primary site. Some of the metastatic lesions found in our study were palpable, while others, such as the retrobulbar orbital metastasis would have been picked up by history and physical examination. In the absence of pain it may be difficult to detect the osseous metastasis. Bone scans are not routinely ordered when screening for metastasis in patients with soft tissue sarcomas, and bone scan has been reported to be less sensitive in identifying bony metastasis from myxoid liposarcoma.8–10 CT scan of chest abdomen and pelvis should certainly be considered in this setting. MRI of the spine is likely the most sensitive test available at present for identifying spinal metastasis.
There was a trend towards improved survival in patients with a low-grade MLS. Three patients treated for metastatic disease are currently alive without evidence of disease. One 54-year old female was diagnosed with high grade MLS to her right thigh in 1993. She was treated with wide resection followed by 6,500 cGy of radiation. She developed a distant metastasis to her 7th rib in 1997. This was widely excised, and she remains alive without further evidence of disease.
A second patient, who presented at age 31 to an outside hospital, was treated with wide excision followed by radiation therapy for low grade MLS around the ankle. He developed a retroperitoneal metastasis with extension into the 4th lumbar vertebrae diagnosed in May of 2004. He received five cycles of Adriamycin and ifosfamide which reduced the tumor size by approximately 30%. Pathology from the metastatic lesion demonstrated low grade MLS. He subsequently underwent wide resection and remains without evidence of disease.
A third patient was diagnosed with low grade MLS to the thigh. This was excised in 1972. He developed a metastatic lesion to the subcutaneous tissues overlying the abdomen in 1982, which was widely excised. In 1984, he developed a metastatic lesion to the abdominal wall, which was excised and reconstructed. In 1985, he had an osseous metastasis to the sterno-clavicular region. This was intralesionally excised and radiated with 5,600 cGy. In 1986, he developed a posterior chest wall metastasis to the 8th and 9th ribs. The lesion extended to the 8th vertebral body, but it did not invade the bone. This lesion was widely excised and I-125 was implanted along the vertebral body. His pathology remained low grade. He was without evidence of disease at his last follow up in 2005.
In conclusion, bony metastasis of MLS is relatively common and it is associated predominantly with type II (5–2) TLS-CHOP fusion transcript. Images of the spine should be obtained when one is screening for metastasis.
Received for publication March 21, 2006.
Accepted for publication July 13, 2006.
|
REFERENCES
|
|---|
- Fletcher CD, Unni K. Pathology and genetics of tumours of soft tissue and bone. Lyon: IARC Press 2002.
- Blair SL, Lewis JJ, Leung D, et al. Multifocal extremity sarcoma: an uncommon and controversial entity. Ann Surg Oncol 1998; 5:37–40.[Abstract]
- Antonescu CR, Elahi A, Healey JH, et al. Monoclonality of multifocal myxoid liposarcoma: confirmation by analysis of TLS-CHOP or EWS-CHOP rearrangements. Clin Cancer Res 2000; 6:2788–93.[Abstract/Free Full Text]
- Antonescu CR, Tschnernyavsky SJ, Decuseara R, et al. Prognostic impact of p53 status, TLS-CHOP fusion transcript structure, and histologic grade in myxoid liposarcoma. Clin Cancer Res 2000; 6:2788–93.[Abstract/Free Full Text]
- Smith TA, Easley KA, Goldblum JR. Myxoid/round cell liposarcoma of the extremities. A clinicopathologic study of 29 cases with particular attention to extent of round cell liposarcoma. Am J Surg Pathol 1996; 20:171–80.[CrossRef][Medline]
- Estourgie SH, Nielsen GP, Ott MJ. Metastatic patterns of extremity myxoid liposarcoma and their outcome. J Surg Oncol 2002; 80:89–93.[CrossRef][Medline]
- Kilpatrick SE, Doyon J, Choong PF, et al. The clinicopathologic spectrum of myxoid and round cell liposarcoma. A study of 95 cases. Cancer 1996; 77:1450–8.[CrossRef][Medline]
- Khurana JS, Rosenthal DI, Rosenberg AE, Mankin HJ. Skeletal metastases in liposarcoma detectable only by magnetic resonance imaging. Clin Orthop Relat Res 1989; 243:204–7.[Medline]
- Kattapuram SV, Khurana JS, Scott JA, el-Khoury GY. Negative scintigraphy with positive magnetic resonance imaging in bone metastases. Skeletal Radiol 1990; 19:113–6.[Medline]
- Ishii T, Ueda T, Myoui A, et al. Unusual skeletal metastases from myxoid liposarcoma only detectable by MR imaging. Eur Radiol 2003; 13(4):L185–L91.[CrossRef][Medline]
TOP
ABSTRACT
INTRODUCTION
