Benign Recurring Lipoblastoma in an Adult Versus Well Differentiated Subcutaneous Myxoid Liposarcoma

PAlMOLOGY

RESEARCH AND PRAcnCE

4) Urban & Fischer Verlag

http://www.urbanfischer.de/journal5lprp

Benign Recurring lipoblastoma in an Adult Versus Well Differentiated Subcutaneous Myxoid liposarcoma: Clinicopathologic, Immunohistochemical and Molecular Analysis of a Unique Case Jeffrey S. Silverman, Jeffrey Hamilton and Ali Tamsen Department of Pathology and Laboratory Medicine, Southampton Hospital, Southampton, New York, USA

Summary Subcutaneous myxoid liposarcoma (ML) is exceedingly rare, with only two or three cases having been reported. Lipoblastoma (LB), a rare fatty tumor that arises in infants and children, is unknown after adolescence. In contrast to ML, LB is often superficial. The two tumors can be difficult to distinguish due to many histologic similarities. We examined a 0.9 em superficial subcutaneous nodule from the dorsal neck of a 48 year old man that had been growing slowly. Three and one half years later, a 0.4 em palpable recurrent nodule was excised from the scar. The patient is now free of disease at 7.5 years. Because of these unusual features , we performed clinicopathologic, immunohistochemical and molecular analysis of this unusual tumor to decide if this represented a rare cutaneous ML or an unprecedented example of LB in an adult. The primary featured a thick fibrous pseudocapsule with foci of lymphocytes and infiltrating nests of semi mature fetal-appearing adipocytic tissue. This surrounded a more immature cellularbut-cytOlogically-bland myxoid tissue featuring stellate cells and signet lipoblasts. There were fibrous septae at the periphery and the vasculature was rather inconspicuous. The 0.4 em diameter recurrence was distinctly lobular and had minute satellite nodules. It was composed of uniform fetal-appearing bland myxoid lipoblastic tissue featuring signet ring lipoblasts surrounded by a few spindle cells. In both tumors, lipoblasts expressed S-I 00 protein. In the primary, 5% of the lesional cells were FXIIIa+dendritic stromal histiocytes while Pathol. Res. Pra ct. 195: 787-792 (1999)

in the recurrence, 15% of the lesional cells were FXma+ dendritic cells. CD34 stained only scattered small capillaries. The Ki67 proliferation index was I % in the primary and 3% in the recurrence. RT-PCR assay for TLS/FUS-CHOP fusion transcripts was negative despite three repeat tests perfonned on paraffin sections of the primary tumor in the presence of good m-RNA internal controls. We reviewed the clinicopathologic and cytogenetic features of ML and LB. Based on this review and on the growth pattern, anatomic features and molecular data from the present case, we conclude that this tumor may represent the first reported case of adult LB. Key words: Lipoblastoma - Myxoid liposarcoma TLS/FUS-CHOP - RT-PCR

Introduction Vellios and colleagues first coined the term lipoblastomatosis to describe a rare benign adipose tissue neoplasm in infants and chi ldren that has the appearance of embryonal fat [25}. These benign lipoblastic tumors Address for correspondence: Jeffrey S. Silvennan HT, HTL, QIHC (ASCP), Southampton Hospital, Department of Pathology, C/o 5 Bonae Court, East H ampton, New York 11937, USA. Phone: +1 (516)324-8366, Fax: +1(516)726-8256,

E-mail: [email protected]

0344-0338199/195/11 -787 $12 .00/0

788 . J. S. Silvennan et al. occur almost entirely during infancy and childhood and more rarely during adolescence [5, 6, 12, 17,25/. Lipoblastomas variably recapitulate fetal adipogenesis and are known as benign lipoblastoma whcn superficial and circumscribed and as lipoblastomatosis when deeper, infiltrative, or multifocal [5, 6}. These tumors are relatively rare and biologically indolent [6, 12, 17}. The recurrence rate varies from 9-22% in various series and the deep form is said 10 be more likely to recur than the circumscribed superficial form [5, 6}. To OUf knowledge, no cases have been reported in patients over eighteen years of age /5, 6,12,17, 25}. Lipoblastoma can be difficult to distinguish from myxoid liposarcoma [3, 12, 19}. Both lumors can feature a plexiform capillary vasculature and signet ring lipoblasts [3, 6, 17}. Lipoblastomas frequently occur in the hcad and neck and can be subcutaneous [6}. Recurrent lipoblaslomas often show maturation [6}. Myxoid liposarcomas are exceedingly rare in the skin and subcutis [9, 13}. Compared to lipoblaslomas,

Fig. 1. The initial tumor has a fibrous pseudocapsule and septae with foci of

chronic inflammation (upper left). Thc capsule is infiltrated by nests of fetal -appearing adipocytic tissue (center). The bulk of the nodule consists of myxoid lipoblastic tissue (bottom right). Hematoxylin and eosin (H&E) xIOO. Inset: The relatively undifferentiated cellular

myxoid lipoblaslic ti ssue is composed of

cytologically bland stellate cells and small signet ring lipoblasts. H&E x400. Fig. 2. The recurrence is circumscribed

and lobular and has small satellite nodule (lower right). H&E x iOO. Inset: The re-

currence consists of more mature lipo-

blasts and a rew delicate spindle cells. H&E x400. Fig. 3. In both tumors, S-IOO protein

stained most of the tumor cell nuclei and

accentuated the Iipoblastic cells; the recurrence id depicted. S-JOO immuno, x 400. Inset: In the recurrence, up to 15% of the spindle cells express factor XlIIa. Faclor XHia immuno. x400.

Unusual Subcutaneous Myxoid Lipoblastic Tumor . 789 myxoid liposarcoma is relatively uncommon in head and neck locations [11,13,15, 16}. In comparison to liposarcoma, the cytomorphology of lipoblastoma is very bland with a lack of anaplasia or hyperchromasia and a distinct maturation of lipoblasts to adipocyte-like forms [I7}. In a recently published immunohistochemical study of subcutaneous lipomatous tumors, we included the present case, which was the only tumor devoid of a CD34-positive and FXlIIa-positive fibrohistiocylic proliferalion and was diagnosed an "unusual " atypical myxoid lipomalous tumor by the reviewers {23}. FurIher study suggested 10 us that this may be the first reported example of lipoblastoma in an adull, however, subsequent review by additional referees and consultants has resulted in mixed opinions with Ihe consensus lying somewhat in favor of myxoid liposarcoma. In an a\tempt 10 further characterize this unusual lesion, we subjected it 10 further clinicopathologic and molecular analysis. Since either diagnostic alternative, subcutaneous myxoid liposarcoma or adult lipoblastoma, is quite rare and worthy of reponing, we herein report in delail the histologic and molecular features of this interesting tumor.

Case report A 48 year old man underwent enucleation of a superficial subcutaneous nodule and excision of the overlying skin ellipse from his dorsal righl neck. The 0.9 em diameler nodule had been present for some time and had been growing slowly. Three and one half years later, a 0.4 cm palpable nodule was excised from the subcutis of the previous excision site. The patient is currently free of disease, 7 years and 5 months from the date of Ihe first excision.

Fig. 4. CD34 stains only scattered small capillaries in the recurrent tumor (cen-

ter) although the tumor pseudocapsule is rich in CD34-positive fibroblasts (lower left). CD34 immuno, x400. Inset: The Ki67 stain shows that cycling

tumor cells are few and far between.

MIB I immuno, x400.

Materials and Methods Both the initial tumor and the recurrent nodule were fixed in 10% buffered formalin and sectioned in paraffin at 4 microns. Hematoxylin and eosin sections were examined and

immunostains for CD34, factor XlIla, S-IOO protein and Ki67 (MlBI) were prepared as previously reported /23/. Subsequently, three sets of five paraffin sections from the primary tumor were cut at 5 microns under stringently clean condi-

tions, placed in three Eppendorf tubes, and submitted to the Department of Pathology and Oncology, School of Medicine, University of Occupational and Environmental Health, Kitakyu shu. Japan for the nesled rev erse transcription-poly-

merase chain reaction (RT-PCR) assay for TLS/FUS-CHOP

fusion lranscripls. Thc details of lhi s assay are reported separately [l4} .

Results The initial tumor was present at the junction of the deep dermis and subcutis. It was a tan, myxoid, semifirm, discreet ovoid nodule that measured 0.9 em in greatest dimension. At low power, there is a thick fibrous pseudocapsule with foci of lymphocytes attd infiltrating nests of semimature fetal-appearing adipocytie tissue (Fig. I). This tissue surrounds a predominantly central immature myxoid tissue, featurittg stellate cells and small, blattd signet ring lipoblasts (Fig. I, inset). There are rare mitoses and apoptotie bodies present in this immature zone. There are some fine fibrous septae at the periphery and the vasculature is rather inconspicuous with widely scattered small vessels. The 0.4 em diameter recurrence is distinctly lobular and shows minute satellitenodules (Fig. 2). It consists of a less cellular organoid, fetal-appearing myxoid lipoblastic tissue featuring signet ring lipoblasts surrounded by a few spindle cells (Fig. 2, inset).

790 . 1. S. Silvennan ct al.

300 200 100

M

1

2

3

4

5

6

M: molecular size marker 1: beta-actin (343 bp) } 2: PBGD (porphobilinogen deaminase, 127 bp) R-1585 3: TLS-CHOP 4: positive control (TLS-CHOPtype I, 379 bp) 5 positive control (TLS-CHOPtype II, 103 bp) 6: negative control (distilled water)

Fig. 5. The results of the nested RT-PCR assay shows positive internal m-RNA controls (lanes I and 2) and absence of type I or type II TLS/FUS-CHOP fusion transcripts (lane 3).

Tn both the primary and the recurrence, the lipoblasts express S-IOO protein (Fig. 3). Tn the primary, 5% of the lesional cells are FXlIIa+ dendritic stromal histiocytes while in the recurrence, 15% of the lesional cells are FXIIla+ dendritic cells (Fig. 3, inset). CD34 stains only a very few scattered small capillaries (Fig. 4). The Ki67 proliferation index is 1 % in the primary and 3% in the recurrence (Fig. 4, inset). Three repeat runs of the RT-PCR assay failed to dctect TLS/FUS-CHOP fusion transcripts in the primary tumor, despite the presence of good internal controls detecting m-RNA for a 343 base pair segment of the beta actin gene and a 127 base pair segment of the gene for porphobilinogen dearninase, indicating excellent RNA integrity in the specimen (Fig. 5).

Discussion Distinction between lipoblastoma and myxoid liposarcoma can be exceedingly difficult in pediatric patients [191. This problem docs not usually present in adult patients because lipoblastoma is not known to occur in adults. Nuclear atypia and hyperchromasia are features of liposarcoma that are absent in lipoblastoma and were absent in our tumor as well. However, myxoid liposarcoma can sometimes lack these features [6J. Tn our case, the focal adipocytic appearance with both synchronous and metachronous evidence of partial maturation, the gross encapsulation and fibrous septation, the superficial, subcutaneous location and occurrence in the neck,

the inconspicuous and sparse, rather than plexiform, vasculature, and the very low proliferation index despite high cellularity in the primary, are all more in keeping with lipoblastoma than with myxoid liposarcoma. The Ki67 index of 1 % in the primary tumor contrasts with much higher Ki67 indices as found in examples of classic myxoid and myxoid round cell liposarcoma (unpublished observations). Additionally, the long interval (3.5 years) and minute size of the recurrence, and the lack of further recurrence or of metastasis are also more in keeping with a benign lesion rather than with myxoid liposarcoma. Myxoid liposarcoma is exceedingly rare in the subcutis, with fewer than five cases on record [9, 13J. Its behavior in skin can be indolent, however {91. The occasional mitoses, prominence of apoptotic bodies, and the predominant pattern of hypercellularity with a compact and immature cell character were troublesome for some reviewers in regarding this tumor as a lipoblastoma. Collins and Chatten state in their review that mitoses can be seen occasionally in lipoblastoma, but did not detect any mitoses in their series of 25 cases [6J. The focal and only limited adipocytic differentiation at the periphery of the primary mitigate against lipoblastoma, which usually shows the full range of adipocytic differentiation. Chung and Enzinger noted that mature fat was more common at the center of the lobules compared with peripheral paraseptal regions, where earlier stages of lipogenesis are located [5J. Our very small tumor showed the reverse of this phenomenon, having a peripheral fibrous capsule and septae with nests of partially mature adipocytogenesis surrounding a central zone of cellular immature myxoid and Iipoblastic tissue. Although both Iipoblastomas and sarcomas are rich in S-100 lipoblasts, myxoid liposarcoma often has mucoid microcystic spaces that are usually lacking in lipoblastoma [17}. Another diagnostic consideration was a benign myxaid lipoma. Myxoid lipoma is rich in CD34+ dendritic fibroblast-like cells, as are other types of cellular benign lipomatous tumors and certain other myxoid soft tissue tumors [22-24j. The present case lacks a CD34+ fibrohistiocytic cell component present in many other benign fatty tumors [23}, however the recurrence did contain up to 15% FXIIIa positive dendritic histiocytes that probably function in stromal organization of the more differentiated recurrence. Modern molecular methods now allow cytogenetic analysis of fatty neoplasms using archival paraffin-cmbedded tissue [2, 14J. Both myxoid liposarcoma and Iipoblastoma have specific genetic alterations that can be analyzed to differentiate between these two tumors that can often be difficult to separate on histologic grounds [7, 8, 10, 14, 19-21}. TLS/FUS-CHOP fusion oncoprotein is found specifically in myxoid and round cell liposarcoma and results from a t(l2; 16) translocation that is seen in most cases. This fusion protein comt

Unusual Subcutaneous Myxoid Lipoblastic Tumor· 791 bines the translocated in liposarcoma (TLS/FUS) gene, an RNA binding protein, and the CCAAT/enhancer binding protein (C/EBP) homologous protein (C HOP), fusing the amino-terminal part ofTLS to the entire coding region of CHOP {I, 14}. CHOP is a member of the C/EBP transcription factor family of proteins that comprise part of the adipocyte differentiation machinery. TLS-CHOP blocks adipocyte differentiation by directly preventing C/EBP beta from binding to and transactivating its target gencs /1]. The nested RT-PCR assay utilized in the present case is highly sensitive, detecting the TLS/FUS-CHOP fusion transcripts in 94% of 15 cases tested /14}. Interestingly, the only TLS/FUS-CHOP negative case in the series of Hisaoka and colleagues was a histologically typical my xoid liposarcoma in the subcutis of the instep of a 17 year old male [l4}. In our case, the assay was repeated three times, each time with the presence of both internal positive controls indicating exceUent m-RNA integrity (Fig. 5). However, negative results in this assay do not necessarily rule out the presence of unrecognized variants of the fusion transcript or other chimeric genes. For instance, rare myxoid liposarcomas show a t(J 2;22) leading to fusion between the CHOP and EWS genes [8, 14}. Clinicopathological features of t(l2;22) cases have recently been compared with cases having the more usual t(12;16) and it was not possible to identify any clinical or pathological differences between these molecular genetic subsets of myxoid liposarcoma f8} . This is in contrast to the distinctly unique features of the present tumor. Although only a fcw Iipoblastomas have been investigated cytogenetically, it is increasingly clear through karyOlyping studies that Iipoblastomas generally show varying rearrangements of 8q II--q 13 region but lack the t( 12;16) specific for myxoid liposarcomas [7, 10, 18, 19, 20}. Lipoblastomas with translocations involving chromosome 8 such as t(7;8) and t(3;8) have also been reported [20, 21}. We are unaware at this time of an RTPCR or other assay applicable to paraffin sections for this lipoblastoma-specific genetic aberration. Obviously, such an assay would provide more unequivocal support for the diagnosis of Iipoblastoma in this adult patient. The ncgative results for TLS/FUS-CHOP fusion transcripts mitigates against the diagnosis of myxoid liposarcoma in our case, but is less than conclusive fl4f. However, the slow growth and extremely low prOliferative index , the range of adipocytic maturation that is evident in the periphery of the primary nodule and in the re currence, although somewhat more limited than that seen in many Iipoblastomas, also mitigate against myxoid liposarcoma, since myxoid liposarcomas tend to dedifferentiate on recurrence /18}. [n conclusion, we present clinicopathologic, immunohistochcmical and molecular genetic data on a unique, recurring yet indolent, myxoid lipoblastic

tumor in the subcutis of the neck in an adult male. The differential diagnosis consists of a very rare and atypical example of a subcutaneous my xoid liposarcoma versus the heretofore unprecedented occurrence of Iipoblastoma in an adult. The tumor and its recurrence share some features with lipoblastoma in children and have fewer features, mainly the high cellularity, apoptosis and ocassional mitoses in the primary, that are more consistent with myxoid liposarcoma. We tentatively offer this case as a putative example of benign lipoblastoma in an adult patient in view of the slow growth, clinicopathologic feature s, and the demonstrated lack of TLS/FUS-CHOP fusion transcripts that is found in the overwhelming majority of myxoid liposarcomas. Acknowledgment. The authors are indebted to Drs. Hiroshi Hashimoto and Masanori Hisaoka and their staff at the Department of Pathology and Oncology, School of Medicine, University of Occupational and Environmental Health , Kitakyushu, Japan for kindly performing the nested RT-PCR analysis for TLSIFUS-CHOP fusion transcripts on this case.

References 1. Adelmant G, Gilbert JD, Freytag SO (1998) Human translocation Iiposarcoma-CCAAT/enhancer binding protein (CIEBP) homologous protein (TLS-CHOP) oncoprotein prevents adipocyte differentiation by directly interfering with CIEBPbeta function. J Bioi Chern 273: 15574-15581 2. Aoki T, Hisaoka M, Kouho H, Hashimoto H, Nakata H ( 1997) Interphase cytogenetic analysis of myxoid soft tjssue tumors by fluorescence in situ hybridization and DNA flow cytometry using paraffin-embeddcd tissue. Cancer 79: 284-293 3. Bolen JW, Thoming D (1980) Benign lipoblastoma and myxoid liposarcoma. Am J Surg Pathol4: 163-174 4. Bondar GL, Miner JE, Potter TS (1997) Stump the experts. Case. Myxoid liposarcoma. Dermatol Surg 23: 333: 407-408 5. Chung EB, Enzinger FM (1973) Benign lipoblastomatosis. An analysis of 35 cases. Cancer 32: 482-492 6. Collins MH, Chatten J (1997) Lipoblastoma/lipoblastomatosis: a clinicopathologic slUdy of 25 tumors. Am J Surg Pathol 2I: 1131-1137 7. Dal Cin P. Sciot R, Dc Wever I, Van Damme B, Van den Berghe H (1994) New discriminative chromosomal marker in adipose tissue tumors. The chromosome 8qll-
792 . I. S. Silverman et al. 10. Fletcher CD, Akennan M, Dal Cin P, et al. (1996) Correlation between clinicopathological features and karyotype in lipomatous tumors. A report of 178 cases from the Chromosomes and Morphology (CHAMP) Collaborative Study Group. Am J Pathol 148: 623-{530 11. Golledge J, Fisher C, Rhys-Evans PH (1995) Head and neck liposarcoma. Cancer 76: 1051-1058 12. Grandi E, Trisolini MP (1986) Tumore lipoblastico in giovanc adulto: Liposarcoma mixoidc 0 lipoblastoma bcnigno? Pathologica 78: 269-275 13. Herman 0, Ariely D (1997) Myxoid liposarcoma of the scalp: case report and literature review. Plasl Reconstr S urg 100: 84-85 14. Hisaoka M, Tsuji S, Morimitsu Y, Hashimoto H, Shimajiri S, Komiya S, Ushijima M (1998) Detection of TLS/FUS-CHOP fusion transcripts in myxoid and round cell liposarcomas by nested reverse transcription-poly-

merase chain reaction using archival paraffin-embedded

tissues. Diagn Mol Pathol7: 96-101 15. Kindblom LG, Angervall L, Jarlstedt J (1978) Liposarco-

ma of the neck: a clinicopathologic study of 4 cases. Can-

ccr42: 774-780 16. McCulloch TM, Makiclski KH, McNutt MA(l999) Head and neck liposarcoma. A histopathologic reevaluation of reported cases. Arch Otolatyngol Head Neck Surg II 8: 1045-1049 17. Mentzel T, Calonjc E, Fletcher CD (1993) Lipoblastoma and lipoblastomatosis: a clinicopathological study of 14 cases. Histopathology 23: 527-533 18. Mentzel T, Fletcher CD (1997) Dedifferentiated myxoid liposarcoma: a clinicopathological study suggesting a closer relationship between myxoid and well-differentiated liposarcoma. Histopathology 5: 457-463

19. Miller GO, Yanchar NL, Magee JF, Blair GK (1998) Lipoblastoma and liposarcoma in children: an analysis of 9 cases and a review of the literature. Can] Surg 41: 455-458 20. Ohjimi Y, Iwasaki H, Kaneko Y, Ishiguro M, Ohgami A, Kikuchi M (1992) A case of lipoblastoma with t(3;9)(q 12; 11.2). Cancer Genet Cytogenet 62: 103-105 21. Panarello C, Rosanda C, Morcrio C, Russo I, Dallorso S, Gambini C, Ricco AS, Storlazzi T, Archidiacono N , Rocchi M (1998) Lipoblastoma: a case with t(7;8)(q31;qI3). Cancer Genet Cytogenet 102: 12-14 22. Silverman JS , Albukerk J, Tarnscn A (1997) Comparison of angiomyofibroblastoma and aggressive angiomyxoma in both sexes: four cases composed of bimodal CD34 and factor XIIIa positive dendritic cell subsets. Pathol Res Pract 193: 673-{582 23. Silverman IS, Tamsen A (1997) Fibrohistiocytic differentiation in subcutaneous fatty tumors. Study of spindle cell, pleomorphic, myxoid, and atypical lipoma and dedifferentiated liposarcoma cases composed in part of CD34' fibroblasts and FXIIIa' histiocytes. J Cutan Pathol 24: 484-493. 24. Silverman JS , Tamsen A (1996) Demonstration of CD34 and factor XIIIa-t dendritic cell subsets in intramuscular myxoma and comparison with umbilical cord Wharton 's jelly. A molecular model for myxoid differentiation. Cell Vision 3: 252 (abstract) 25. Vellios G, Baez J, Shumaher HB (1958) Lipoblastomatosis: a tumor of fetal fat different from hibemoma. Am J Patho134: 1149-1159 Received: December 9, 1997 Accepted in revised version: May 22, 1999