CANCER 176 CYTOPATHOLOGY Fine-Needle Aspiration Biopsy of Skeletal versus Extraskeletal Osteosarcoma Kathleen K. Nicol, M.D.1 William G. Ward, M.D.2 Paul D. Savage, M.D.3 Scott E. Kilpatrick, M.D.1,2 1 Department of Pathology, Wake Forest University School of Medicine, Winston–Salem, North Carolina. 2 Department of Orthopaedics, Wake Forest University School of Medicine, Winston–Salem, North Carolina. 3 Department of Hematology/Oncology, Wake Forest University School of Medicine, Winston–Salem, North Carolina. BACKGROUND. Although fine–needle aspiration biopsy (FNAB) of primary skeletal osteosarcoma (OS) has been described adequately, to the authors’ knowledge, cytologic descriptions of extraskeletal OS appear limited to only rare case reports. METHODS. In an attempt to analyze the utility and accuracy of FNAB in a diagnosis of skeletal versus extraskeletal OS, the authors retrospectively reviewed their 5–year experience. The study sample included 15 skeletal OS specimens (13 primary, 1 local recurrence, and 1 pulmonary metastasis) in 14 patients ages 10 –58 years (mean, 27 years; median, 25 years) and 5 extraskeletal OS specimens (3 primary and 2 metastatic) in 4 patients ages 36, 37, 65, and 79 years, respectively. Based on accepted clinical criteria, two patients (a mother with extraskeletal OS and a daughter with skeletal OS) had Li–Fraumeni syndrome. RESULTS. Of the adequate primary skeletal OS cases analyzed by FNAB, 10 of 12 (83%) were diagnosed correctly and subsequently treated according to a disease specific protocol. One case was considered unsatisfactory. One tumor initially was diagnosed as a giant cell tumor and another was referred to nonspecifically as ‘‘spindle– cell neoplasm.’’ On histologic examination, the former case demonstrated a high grade fibroblastic OS arising within a giant cell tumor. None of the primary extraskeletal OS cases analyzed by FNAB was recognized as OS. One was diagnosed nonspecifically as ‘‘sarcoma’’ and the other was referred to simply as ‘‘atypical mesenchymal cells.’’ A third case was comprised of scant fragments of adipose tissue, fibrous tissue, and cartilage and was considered unsatisfactory. Both examples of metastatic extraskeletal OS were recognized by FNAB. CONCLUSIONS. With appropriate clinicoradiologic correlation, skeletal OS generally is easily diagnosed by FNAB. Because of the older age of most patients with extraskeletal OS and the rather nonspecific radiographic findings (e.g., soft tissue mass), extraskeletal OS may not be recognized easily by FNAB and most likely requires incisional biopsy to establish a definitive diagnosis in most cases. Additional larger series will be required before drawing definite conclusions. Cancer (Cancer Cytopathol) 1998;84:176 – 85. © 1998 American Cancer Society. KEYWORDS: osteosarcoma, extraskeletal, fine–needle aspiration biopsy, cytology, Li–Fraumeni syndrome. Presented in part at the 45th Annual Scientific Session of the American Society of Cytopathology Meeting, Boston, Massachusetts, November 4-8, 1997. Address for reprints: Scott E. Kilpatrick, M.D., Department of Pathology and Laboratory Medicine, The University of North Carolina at Chapel Hill, CB# 7525, Brinkhous–Bullitt Building, Chapel Hill, NC 27599 –7525. Received December 12, 1997; accepted February 13, 1998. © 1998 American Cancer Society F or the diagnosis of bone tumors, radiographs provide a useful ‘‘macroscopic specimen’’ for the pathologist, allowing assessment of such features as geographic versus permeative destruction, periosteal reactions, and soft tissue extension. These features usually correlate with the aggressiveness of the lesion in question. Furthermore, radiographs also may provide information regarding potential tumor matrix production (osteoblastic vs. osteolytic) useful in the differential diagnosis of bone tumors. Indeed, 75 years ago, Dr. Ewing stated, ‘‘The gross anatomy (as revealed in radiographs) is often a safer guide FNAB of Osteosarcoma/Nicol et al. to correct clinical conception of the disease than the variable and uncertain nature of a small piece of tissue.’’1 Not surprisingly, fine–needle aspiration biopsy (FNAB) of primary skeletal osteosarcoma (OS) has been well documented as an effective diagnostic tool, especially when coupled with clinical and radiologic data.2,3 In contrast, clinical and radiologic data (e.g., soft tissue mass) tend to be ‘‘disappointingly nonspecific’’ in soft tissue sarcomas and therefore are of limited diagnostic utility.4 As a result, the diagnosis of extraskeletal OS is much more difficult to establish by FNAB. We should point out that occasional examples of extraskeletal OS will display radiographic evidence of spotty to massive calcifications/ossifications.5 Although such radiologic findings may suggest the diagnosis of extraskeletal osteosarcoma, other soft tissue sarcomas (e.g., synovial sarcomas) may display similar radiologic features.6 Pathologically, this problem is compounded further by the much broader differential diagnosis one must consider in an adult pleomorphic soft tissue sarcoma case (e.g., leiomyosarcoma, malignant fibrous histiocytoma [MFH], liposarcoma). To our knowledge, there is only one adequate FNAB case description of an extraskeletal OS that involved the breast of a 38 –year– old woman.7 With our current understanding, such a tumor arguably may be more appropriately classified as a metaplastic carcinoma. We report herein our 5–year experience with skeletal and extraskeletal OSs evaluated by FNAB. In addition, we discuss the nonspecific features of matrix material, diagnostic criteria, potential cytologic pitfalls, and the relationship between Li–Fraumeni syndrome and OS. MATERIALS AND METHODS We retrospectively reviewed 15 skeletal OS specimens in 14 patients and 5 extraskeletal OS specimens in 4 patients from the cytopathology files at Bowman Gray/Baptist Hospital Medical Center. The study cases were retrieved consecutively during the period from 1992–1997. All cytologic samples were obtained without anesthesia using 23– and 25– gauge needles by the standard manual method (18 cases) and computed tomographic– guided needle placement (2 cases, pulmonary metastases). Aspirated material was expelled onto glass slides and smeared using a second glass slide. Half of the smears were air dried and stained with Diff–Quik. An equal number of slides immediately were immersed in 95% ethanol for staining by the Papanicolaou method. Cell blocks were prepared from ten FNAB specimens, all of which were skeletal osteosarcomas. Hematoxylin and eosin stained histologic slides from resection and/or amputation speci- 177 mens were reviewed in all patients. Clinical data were retrieved from the patients’ medical records. RESULTS Clinical Features The clinical and pathologic features of the study cases are summarized in Table 1. Of the patients with skeletal OSs, ages at diagnosis ranged from 10 –58 years (mean, 27 years; median, 25 years). There were eight men and six women. The aspiration biopsy specimens were obtained from the primary tumors in 13 cases and a local recurrence and pulmonary metastasis in another patient (Case 2). The primary tumors arose within the femur (five patients), the humerus (two patients), the tibia (two patients) and one each in the sacrum, rib, acetabulum, clavicle, and foot (navicular bone). One patient (Case 13) developed an OS approximately 9 years after radiation therapy for Hodgkin’s disease involving a left supraclavicular lymph node. FNAB of this probable postirradiation sarcoma was unsatisfactory. Of the adequate primary skeletal OS cases analyzed by FNAB, 10 of 12 (83%) were diagnosed correctly and subsequently treated according to standard protocol. One tumor (Case 11) was diagnosed initially as a giant cell tumor and a second (Case 8) was referred to nonspecifically as ‘‘spindle– cell neoplasm.’’ On subsequent curettage, the former case represented a high grade fibroblastic OS arising in a giant cell tumor (‘‘malignant giant cell tumor’’). The latter case (Case 8) was the first FNAB specimen from a primary skeletal OS analyzed at our institution. All but 3 of the cases (Cases 1, 8, and 14) were procured during the last 2 years of the period studied (1995– 1997). Of the patients with extraskeletal OS, the ages were 36, 37, 65, and 79 years in 2 women and 2 men. The FNAB specimens were obtained from the primary tumors in three cases (Cases 15, 16, and 17), a pulmonary metastasis in a one case (Case 16), and a lymph node groin metastasis in another case (Case 18). None of the primary extraskeletal OS cases analyzed by FNAB were recognized as OS. One case (Case 15) was diagnosed nonspecifically as ‘‘sarcoma’’ and another (Case 17) simply was referred to as ‘‘atypical mesenchymal cells.’’ FNAB of the primary tumor in Case 16 was considered unsatisfactory for diagnosis. These cases are discussed in more detail separately. Case 15 A 37–year–old woman who underwent excision of a cutaneous leiomyosarcoma of the right eyebrow presented 1 year later with a left breast mass and a large left posterior thigh mass (Fig. 1). An FNAB of the breast mass revealed ‘‘adenocarcinoma’’ whereas an FNAB of the 178 CANCER (CANCER CYTOPATHOLOGY) June 25, 1998 / Volume 84 / Number 3 TABLE 1 Clinical and Pathologic Features of Skeletal and Extraskeletal Osteosarcoma Patient no. Age (yrs) Gender Primary site FNA Cytologic DX Histologic DX 1 2 23 47 F M Right femur Left foot (navicular bone) 16 30 10 10 11 28 13 25 58 F F F M M F M M M Sacrum Right femur Left 4th rib Left humerus Left femur Right acetabulum Right humerus Femur Right tibia Osteosarcoma Osteosarcoma Metastatic osteosarcoma Osteosarcoma Osteosarcoma Osteosarcoma Osteosarcoma Osteosarcoma Spindle cell neoplasm Osteosarcoma Osteosarcoma Giant cell tumor Osteosarcoma Osteosarcoma 3 4 5 6 7 8 9 10 11 Primary Recurrence Pulmonary metastasis Primary Primary Primary Primary Primary Primary Primary Primary Primary 12 13 14 15 16 48 42 14 37 62 M F M F F Right femur Left clavicle Left tibia Thigh Buttock 17 18 79 36 M M Right thigh Left toe Primary Primary Primary Primary Primary Pulmonary metastasis Primary Groin metastasis Osteosarcoma Unsatisfactory Osteosarcoma Sarcoma Unsatisfactory Metastatic osteosarcoma Atypical mesenchymal cells Metastatic osteosarcoma Osteosarcoma Osteosarcoma Osteosarcoma Osteosarcoma Osteosarcoma Osteosarcoma Osteosarcoma Osteosarcoma Osteosarcoma arising in a giant cell tumor Osteosarcoma Osteosarcoma Osteosarcoma Extraskeletal osteosarcoma Extraskeletal osteosarcoma Extraskeletal osteosarcoma Extraskeletal osteosarcoma M: male; F: female; FNA: site from which fine-needle aspiration biopsy was procured; cytologic DX: cytologic diagnosis at the time of fine-needle aspiration biopsy; histologic DX: histologic diagnosis of the primary tumor. thigh mass showed ‘‘sarcoma.’’ Staging evaluation demonstrated multiple bilateral pulmonary nodules consistent with metastases. Subsequently, she received three cycles of preoperative chemotherapy for the sarcoma, followed by a limb salvage resection and a concurrent left breast lumpectomy. Histologic examination revealed extraskeletal OS and intraductal carcinoma, respectively. Bilateral thoracotomies subsequently confirmed the presence of metastatic OS. Clinically, the presence of multiple epithelial and mesenchymal tumors appeared to fulfill the criteria for Li–Fraumeni syndrome. She died 14 months later secondary to widespread metastatic disease. Her family history was significant for cancer, including a father with prostate and colon carcinoma, four paternal aunts with breast carcinoma, and a fifth paternal aunt who reportedly had a brain tumor. Approximately 4 months after her death, the decedent’s daughter (Case 3), who had a prior history of an adrenal adenoma, developed a destructive sacral mass that was proven by FNAB to be osteosarcoma. Case 16 A 62–year– old woman with a previous history of endocervical adenocarcinoma who underwent a hysterectomy presented 3 years later with a large soft tissue mass of the right buttock. An FNAB of the right gluteal mass was interpreted as ‘‘negative’’ but demonstrating rare ‘‘fragments of connective tissue, adipose tissue, and cartilage.’’ Subsequent incisional biopsy was followed by wide resection. Histologic examination revealed a chondroblastic OS. Approximately 27 months after resection, she presented with a left lung mass. A FNAB confirmed the clinical impression of metastatic chondroblastic OS. At last follow– up the patient was alive with evidence of metastatic disease. Case 17 A 79 –year– old man presented with a soft tissue mass within the right thigh, which he first noticed after he attempted to lift a heavy object. There was no prior history of trauma or injury to the affected thigh. An outside FNAB yielded only blood. Radiographically, the lesion appeared cystic and was believed to possibly represent a hematoma. Subsequent FNAB performed at our institution also yielded mostly blood but also contained rare ‘‘atypical mesenchymal cells’’ of uncertain significance. The patient underwent surgical resection of the rectus femoris, which contained a 7.0 cm 3 4.5 cm 3 2.0 cm multiloculated hemorrhagic cyst. Histologic examination revealed an extraskeletal telangiectatic OS. After surgery, this patient received FNAB of Osteosarcoma/Nicol et al. 179 FIGURE 1. (A) Extraskeletal osteosarcoma. Plain film radiography illustrating a large, soft tissue mass involving the posterior thigh (Case 15). Unfortunately, such findings in adult patients are nonspecific. (B) Skeletal osteosarcoma. In contrast, a plain film radiograph of a high grade intramedullary osteosarcoma involving the proximal humerus in a 10 –year– old boy (Case 6) shows features characteristic of osteosarcoma including probable ossified matrix, ill– defined margins and permeative pattern, and periosteal reaction. chemotherapy and radiation therapy and at last follow– up was free of disease with 36 months of clinical follow– up. months after amputation, the patient presented with multiple pulmonary nodules and an enlarged left inguinal lymph node. FNAB of the latter confirmed metastatic OS. Case 18 A 36 –year– old man presented with soft tissue swelling of the left great toe. He reported a history of a fracture of the toe approximately 4 –5 years prior to presentation. Imaging studies including a magnetic resonance imaging scan disclosed a mass within the plantar soft tissues of the flexor hallucis tendon sheath. The clinical impression from the referring hospital was a ganglion cyst. However, incisional biopsy revealed an extraskeletal small cell OS. Subsequently, the patient was referred to our institution for further therapy. Due to the aggressive nature of the sarcoma, a below–the– knee amputation was performed. He did not receive any postoperative chemotherapy. Approximately 5 Pathologic Findings All but one (Case 17) of the adequate cytologic smears ranged from moderately to highly cellular. The predominant cell type was a large polygonal– shaped cell with an ovoid to round nucleus surrounded by scant to abundant dense cytoplasm, often containing minute vacuoles. The nuclear chromatin was coarsely granular; nucleoli occasionally were prominent . Binucleated and multinucleated forms were evident in all cases (Fig. 2). Variable amounts of spindle–shaped cells also were present. One case (Case 18) was comprised predominately of a uniform population of round to ovoid tumor cells 180 CANCER (CANCER CYTOPATHOLOGY) June 25, 1998 / Volume 84 / Number 3 FIGURE 2. Osteosarcoma showing obvious anaplastic, mostly polygonal to spindle–shaped tumor cells with eccentrically located nuclei, coarsely granular chromatin pattern, and an abundant amount of dense cytoplasm. Note the binucleated forms (A:, Diff–Quik stain, 3600; B: Papanicolaou stain, 3400). resembling Ewing’s sarcoma or synovial sarcoma (Fig. 3). Clumps of metachromatic magenta–appearing matrix material, best visualized on the Diff–Quik stain, were observed in 7 of 12 cases of primary skeletal OS (58%) but were not observed definitively in any of the primary cases of extraskeletal OS. The matrix material was tinctorially similar to osteoid, hyaline cartilage, chondroid, or dense collagen (Fig. 4). Because the appearance of the matrix material largely was nonspecific (e.g., osteoid vs. chondroid), histologic subtyping (e.g., osteoblastic vs. chondroblastic) of the OSs generally was not possible. Retrospectively, it appeared that the volume of matrix material was more abundant in cases of chondroblastic OS. Only one extraskeletal OS case was considered unsatisfactory (Case 16, primary tumor), although retrospectively a single fragment of cartilage may have represented tumor. Histopathology On review of the surgical resection specimens, histologic subtyping of the skeletal OSs confirmed six osteoblastic OSs (Cases 1, 6, 7, 10, 13, and 14), five fibroblastic osteosarcomas (Cases 4, 5, 8, 11, and 12), and three chondroblastic OSs (Cases 2, 3, and 9) (Fig. 5). Likewise, the extraskeletal OS cases were comprised of fibroblastic (Case 15), telangiectatic (Case 17), chondroblastic (Case 16), and small cell (Case 18) subtypes. DISCUSSION Although still somewhat controversial, the diagnostic accuracy of FNAB in the evaluation of primary bone tumors can approach 95%, especially when correlated with clinical and radiologic data and utilized by those physicians more experienced with orthopedic lesions.8,9 In general, malignant cases FNAB of Osteosarcoma/Nicol et al. 181 FIGURE 3. Case 18, a small cell osteosarcoma, was comprised predominately of mostly uniform, round to ovoid tumor cells closely resembling Ewing’s sarcoma and/or synovial sarcoma (A: Diff–Quik stain, 3100; B: Papanicolaou stain, 3200). are diagnosed more accurately than benign ones.10 More specifically, with regard to skeletal OS, investigators at M. D. Anderson Cancer Center obtained an accuracy rate of 80%.2 In most nondiagnostic cases, the accuracy of the technique was hampered by failure to obtain sufficient diagnostic material and was not related to misinterpretation. Sclerotic and/or heavily ossified OSs also contributed to unsatisfactory FNAB specimens.2 Nevertheless, we believe that a multidisciplinary approach accompanied by an appropriate clinical assessment by the orthopedic surgeon, radiographic interpretation by an experienced radiologist, and ‘‘on site’’ determination of adequacy by a pathologist help reduce the likelihood of obtaining an unsatisfactory or insufficient sample. Utilizing this approach at our institution, we obtained adequate diagnostic material in all but one of our primary skeletal OS cases. Our accuracy rate for rendering a specific diagnosis in adequate primary skeletal OS specimens by FNAB was 83% (10 of 12 cases). One case initially was interpreted as a giant cell tumor and, because subsequent curettage revealed an OS arising in a giant cell tumor, appeared to represent FNAB sampling error and not a ‘‘true’’ misdiagnosis. Another case was referred to nonspecifically as ‘‘spindle– cell neoplasm.’’ The latter represented the first example of FNAB of a primary skeletal OS analyzed at our institution. Retrospectively, the cytologic smears in this case coupled with the characteristic clinical and radiologic data were diagnostic of OS; however, lack of significant experience with musculoskeletal lesions and FNAB likely contributed to a less than specific diagnosis. All but 3 of our cases of skeletal OS analyzed by FNAB were procured during the past 2 years (1995–1997). The Mayo Clinic defines a ‘‘malignant giant cell tumor’’ as either a high grade sarcoma arising 182 CANCER (CANCER CYTOPATHOLOGY) June 25, 1998 / Volume 84 / Number 3 FIGURE 4. Osteosarcoma tumor cells intimately associated with matrix material. Subtyping of the matrix (e.g., osteoid, chondroid, or hyaline cartilage) generally is not possible because they appear tinctorially similar in cytologic preparations. This case (Case 9) was confirmed histologically as a chondroblastic osteosarcoma (Diff–Quik stain, 3200). within a classic giant cell tumor or a high grade sarcoma arising in the location of a previously curettaged or excised giant cell tumor, usually after 11 radiation. Recent evidence suggests that spontaneous malignancy in giant cell tumors most certainly represents a form of dedifferentiation.12,13 When malignant transformation of a benign giant cell tumor occurs, it usually presents as a fibrosarcoma or OS.11 One of our cases initially was diagnosed by FNAB as a giant cell tumor. The subsequent curettage specimen revealed classic giant cell tumor juxtaposed to a high grade fibroblastic OS. Radiographically, the lesion presented as a geographic, lytic, epiphyseal– centered lesion within the right proximal tibia. Retrospective review of the FNAB specimen revealed evidence of characteristic giant cell tumor with scattered, mostly uniform FIGURE 5. Cell block preparations occasionally may be helpful for subtyping osteosarcomas. Case 3 represented a chondroblastic osteosarcoma. In addition to the young age of the patient, the marked cytologic atypia and the tendency toward greater tumor cellularity at the periphery of this cartilaginous nodule are indicative of osteosarcoma and not chondrosarcoma (H & E, 3100). ovoid to spindled stromal cells admixed with benign–appearing osteoclast–type giant cells. Although a few slightly atypical mesenchymal cells were observed, no evidence of high grade sarcoma was observed in the FNAB specimen. Cytologically, skeletal OSs are characterized by hypercellular smears comprised of mostly individually dispersed and small clusters of neoplastic cells.2,3 The tumor cells vary considerably in size and shape, ranging from round and polygonal to spindled. Likewise, significant nuclear variation and pleomorphism also are observed, accompanied by a coarsely granular chromatin pattern and usually one to two small nucleoli.3 Because the majority of OSs are high grade, most are easily recognized as malignant. In up to 50% of aspirate specimens, ma- FNAB of Osteosarcoma/Nicol et al. trix material consistent with osteoid also may be identified.2 Matrix material was recognized in 58% of our primary skeletal OS cases but was not identified in any of the primary extraskeletal OS cases analyzed by FNAB. Furthermore, in our experience and others, the cytomorphologic features of the matrix material in both Diff–Quik and Papanicolaou preparations are relatively nonspecific.2,14 Dense collagen, osteoid, chondroid, and even hyaline cartilage appear tinctorially similar.14 As a result, we do not believe that the matrix material can be specifically and reliably subclassified by FNAB. Meticulous attention to the ’’background’’ neoplastic cells and pertinent clinicoradiologic features is of more importance in establishing a diagnosis of OS. At our institution, the finding of a pleomorphic sarcoma of bone in a young patient with characteristic radiographic features is interpreted as an OS regardless of the presence or absence of matrix material. In these patients, the differential diagnosis of a pleomorphic sarcoma of bone includes MFH. However, at our institution and others, primary MFH of bone and skeletal OSs are treated utilizing a similar protocol (e.g., preoperative chemotherapy followed by surgical resection).2 A recent review of MFHs of bone found no significant difference in overall prognosis between it and OS.15 We believe that most examples of MFH of bone cannot be separated reliably from fibroblastic OS in FNAB and/or incisional biopsy specimens and thus tend to regard (therapeutically) all such tumors as OSs. All but one (Case 13) of our FNAB specimens clearly were recognizable as sarcomas. When coupled with the clinical and radiologic data, a specific diagnosis of OS was rendered in 83% of the primary skeletal cases. Subsequent biopsy/resection specimens confirmed that all primary skeletal cases represented high grade OSs. In contrast to its skeletal counterpart, the diagnosis of extraskeletal OS by FNAB is far more problematic. First, the radiographic features (e.g., soft tissue mass) are relatively nonspecific. In fact, there appears to be no reliably specific clinical and radiologic features for soft tissue sarcomas.4 Because the majority of examples of extraskeletal OS occur in older patients, the differential diagnosis tends to be broader. For example, the diagnosis of a pleomorphic soft tissue sarcoma in an older adult patient encompasses at least MFH, pleomorphic leiomyosarcoma, pleomorphic liposarcoma, malignant peripheral nerve sheath tumor, and pleomorphic rhabdomyosarcoma. Malignant melanoma and metastatic carcinoma, tumors usually not observed in young patients, also must be considered. None of our primary extraskeletal OS cases analyzed by 183 FNAB was recognized as OS, although two were believed to be malignant. All necessitated excisional biopsy to establish a specific diagnosis of extraskeletal OS. We are aware of only one case report describing the FNAB findings in extraskeletal OS.7 The tumor involved the breast of a 38 –year– old woman and may be classified more appropriately as a metaplastic carcinoma.7 Greenwood and Meschter16 also described the cytologic findings of a case of extraskeletal OS arising in the mediastinum. However, their cytologic preparations were obtained from scrapings of the cut surface of the resected specimen and not from FNAB.16 It may be argued that histologic subtyping of soft tissue sarcomas is of less importance therapeutically than establishing the lesion in questions as a ‘‘sarcoma.’’ Such a philosophy is related in part to the fact that therapy for most soft tissue sarcomas is driven by tumor stage, which incorporates histologic grade, and anatomic site. However, some pertinent points are worth consideration. First, accurate histologic grading of soft tissue sarcomas requires accurate histologic subtyping17,18 Second, at our institution and others, Ewing’s sarcoma (skeletal and extraskeletal), OS (skeletal), and rhabdomyosarcoma are treated by potentially curative, histogenetic specific protocols, necessitating the need for accurate histologic subtyping.19 –21 Whether specific chemotherapeutic protocols will become standard for other soft tissue sarcomas remains to be seen. Nevertheless, it is clearly important that, when possible, histologic subtyping of a soft tissue sarcoma be attempted to ensure proper therapeutic management. Both skeletal and extraskeletal OS may exist as multiple subtypes including osteoblastic, chondroblastic, fibroblastic, telangiectatic, and even small cell forms. Although various combinations of these subtypes also may be observed, histologic subtyping is based on the predominant matrix pattern and/or cell type. Accurate and reliable histologic subtyping of OS usually is not possible by FNAB but most likely is of limited importance because therapy predominately is based on the grade of the OS. A recent study of extraskeletal OS, utilizing univariate analysis, found that patients with predominately chondroblastic tumors fared better (prognostically) than those with predominant osteoblastic differentiation; however, this significance was not maintained with multivariate analysis.22 Extraskeletal telangiectatic OS represents one of the rarest subtypes.22,23 Morphologically, the tumors are hemorrhagic and comprised of aneurysmal bone cyst–like septa containing frankly anaplastic tumor cells and infrequent deposits of osteoid. One of our cases of extraskeletal 184 CANCER (CANCER CYTOPATHOLOGY) June 25, 1998 / Volume 84 / Number 3 OS represented the telangiectatic subtype. Clinically, the tumor initially was believed to represent an organizing hematoma, despite the fact that the patient had no prior history of injury. The FNAB specimen was comprised mostly of blood and scattered, markedly atypical mesenchymal cells. Such lesions, especially in the absence of a conclusive FNAB specimen, necessitate further investigation. The Li–Fraumeni cancer syndrome is an autosomal dominant disorder that predisposes affected individuals to develop multiple forms of cancer including, but not limited to, breast carcinoma, soft tissue sarcoma, OS, adrenocortical carcinoma, and leukemia.24,25 The syndrome has been linked to germ line mutations of the tumor suppressor gene p53. Testing of suspected individuals may be performed through analysis of genomic DNA from peripheral blood. Two of our patients, a mother (Case 15) and a daughter (Case 3), developed multiple epithelial and mesenchymal tumors that, when coupled with their family history, are consistent with the clinical diagnosis of Li–Fraumeni syndrome. Molecular analysis of p53 was not performed in either patient. We are aware of only one case report describing the FNAB findings in a patient with Li– Fraumeni syndrome.26 When correlated with clinical and radiologic data, FNAB represents a valuable tool for rendering a specific diagnosis for primary skeletal OS. Because of the older age of most patients with extraskeletal OS and the rather nonspecific radiographic findings, extraskeletal OS is not easily recognized by FNAB. In our limited experience, most examples require incisional, and in some cases, excisional biopsy specimens to establish a definitive diagnosis. Additional larger series are needed to fully establish the role and utility of FNAB for the diagnosis of extraskeletal osteosarcoma. The possibility of Li–Fraumeni syndrome must be considered in any newly diagnosed case of OS, especially in patients with prior history of cancer and/or positive family histories. REFERENCES 1. Ewing J. A review and classification of bone sarcomas. Arch Surg 1922;4:485–533. 2. White VA, Fanning CV, Ayala AG, Raymond AK, Carrasco CH, Murray JA. Osteosarcoma and the role of fine–needle aspiration: a study of 51 cases. Cancer 1988;62:1238 – 46. 3. Walaas L, Kindblom L–G. Light and electron microscopic examination of the fine–needle aspirates in the preoperative diagnosis of osteogenic tumors: a study of 21 osteosarcomas and two osteoblastomas. Diagn Cytopathol 1990;6:27–38. 4. Moser RP Jr., Madewell JE. Radiologic evaluation of soft 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. tissue tumors. In: Enzinger FM, Weiss SW, editors. Soft tissue tumors. 3rd edition. St. Louis: C.V. Mosby, 1995:39 – 88. Chung EB, Euzinger FM. Extraskeletal osteosarcoma. Cancer 1987;60:1132– 42. Koplyay PD, Lenchik L, Pope TL Jr., Kilpatrick SE. Imaging of synovial sarcomas [abstract]. Radiology 1997;206:636. Mertens HH, Langnickel D, Staedtler F. Primary osteogenic sarcoma of the breast. Acta Cytol 1982;26:512– 6. Kumar RV, Rao CR, Hazarika D, Mukherjee G, Gowda BM. Aspiration biopsy cytology of primary bone lesions. Acta Cytol 1993;37:83–9. Bommer KK, Ramzy I, Mody D. Fine–needle aspiration biopsy in the diagnosis and management of bone lesions: a study of 450 cases. Cancer 1997;81:148 – 6. Layfield LJ, Armstrong K, Zaleski S, Eckardt J. Diagnostic accuracy and clinical utility of fine–needle aspiration cytology in the diagnosis of clinically primary bone lesions. Diagn Cytopathol 1993;9:168 –3. Unni KK. Malignancy in giant cell tumor of bone. In: Dahlin’s bone tumors: general aspects and data on 11,087 cases. 5th edition. Philadelphia: Lippincott–Raven, 1996: 285–9. Brien EW, Mirra JM, Kessler S, Suen M, Ho JKS, Yang WT. Benign giant cell tumor of bone with osteosarcomatous transformation (‘‘dedifferentiated’’ primary malignant GCT): report of two cases. Skeletal Radiol 1997;26: 246 –55. Meis JM, Dorfman HD, Nathanson D, Haggar AM, Wu KK. Primary malignant giant cell tumor of bone: ‘‘dedifferentiated’’ GCT. Mod Pathol 1989;2:541– 6. Kilpatrick SE, Pike EJ, Geisinger KR, Ward WG. Chondroblastoma of bone: use of fine–needle aspiration biopsy and potential diagnostic pitfalls. Diagn Cytopathol 1997;16:65– 71. Nishida J, Sim FH, Wenger DE, Unni KK. Malignant fibrous histiocytoma of bone: a clinicopathologic study of 81 patients. Cancer 1997;79:482–93. Greenwood SM, Meschter SC. Extraskeletal osteogenic sarcoma of the mediastinum. Arch Pathol Lab Med 1989;113: 430 –3. Costa J, Wesley RA, Glatstein E, Rosenberg SA. The grading of soft tissue sarcomas: results of a clinicohistopathologic correlation in a series of 163 cases. Cancer 1984;53: 530 – 41. Gulliou L, Coindre JM, Borichon F, Binh BN, Terrier P, Collin F, et al. Comparative study of the National Cancer Institute and French Federation of Cancer Centers Sarcoma Group grading systems in a population of 410 adult patients with soft tissue sarcoma. J Clin Oncol 1997;15:350 – 62. Kilpatrick SE, Ward WG, Savage PD. Fine needle aspiration biopsy of mesenchymal lesions: the value of clinicoradiologic correlation and the importance of histogenetic subtyping [letter]. Am J Clin Pathol 1997;108:237. Raymond AK, Chawla SP, Currasco CH, Ayala AG, Fanning CV, Grice B, et al. Osteosarcoma chemotherapy effect: a prognostic factor. Semin Diagn Pathol 1987;4:212– 36. Kinsella TJ, Triche TJ, Dickman PS, Costa J, Tepper JE, Glanbiger D. Extraskeletal Ewing’s sarcoma: results of combined modality treatment. J Clin Oncol 1983;1:489 – 95. FNAB of Osteosarcoma/Nicol et al. 22. Lee JSY, Fetsch JF, Wasdhal DA, Lee BP, Pritchard DJ, Nascimento AG. A review of 40 patients with extraskeletal osteosarcoma. Cancer 1995;76:2253–9. 23. Mirra JM, Fain JS, Ward WG, Eckardt JJ, Eilber F, Rosen G. Extraskeletal telangiectatic osteosarcoma. Cancer 1993;71:3014–9. 24. Porter DE, Holden ST, Steel CM, Cohen BB, Wallace MR, Reid R. A significant proportion of patients with osteosarcoma may belong to Li–Fraumeni cancer families. J Bone Joint Surg Br 1992;74:883– 6. 185 25. Garber JE, Goldstein AM, Kantor AF, Dreyfus MG, Fraumeni JF Jr., Li FB. Follow– up study of twenty–four families with Li–Fraumeni syndrome. Cancer Res 1991;51: 6094 –7. 26. Peréz–Guillermo M, Bonmati–Limorte C, Garcia–Rojo B, Hernandez–Gil A. Infantile cutaneous rhabdomyosarcoma (Li–Fraumeni syndrome): cytological presentation of fine– needle aspirate biopsy, report of a case. Diagn Cytopathol 1992;8:621– 6.