Fax: (614) 293-7626.
Fine-needle aspiration biopsy of high-grade sarcoma
A report of 107 cases
Article first published online: 16 OCT 2007
Copyright © 2007 American Cancer Society
Volume 111, Issue 6, pages 491–498, 25 December 2007
How to Cite
Fleshman, R., Mayerson, J. and Wakely, P. E. (2007), Fine-needle aspiration biopsy of high-grade sarcoma. Cancer, 111: 491–498. doi: 10.1002/cncr.23122
- Issue published online: 17 DEC 2007
- Article first published online: 16 OCT 2007
- Manuscript Accepted: 23 AUG 2007
- Manuscript Revised: 16 AUG 2007
- Manuscript Received: 16 JUL 2007
- fine-needle aspiration biopsy;
- high-grade sarcoma;
- cytopathologic diagnosis
To the authors' knowledge, few studies exist demonstrating the reliability of fine-needle aspiration (FNA) biopsy for high-grade sarcoma (HGS).
In the current study, the authors reviewed their cytopathology database (March 2001 through January 2007) and identified all FNA cases diagnosed as HGS. They also searched their tissue database for all HGS cases that had prior FNA biopsy findings.
A total of 107 FNA samples from 98 patients (age range, 13–90 years, with a male:female ratio of 1:1) had an FNA diagnosis of HGS, or had HGS and a prior FNA diagnosis of another entity. Ten cases were nondiagnostic. Of the 97 remaining samples, 6 were diagnosed as something other than HGS (sensitivity of 94%). The positive predictive value of an FNA diagnosis of HGS was 97% (88 of 91 cases). Fifty-four cases were diagnosed as HGS, not otherwise specified, 8 as myxofibrosarcoma, 8 as osteosarcoma, 5 as malignant peripheral nerve sheath tumor, 5 as leiomyosarcoma, 4 as Ewing sarcoma, 4 as liposarcoma, 2 as epithelioid sarcoma, and 1 as angiosarcoma. Approximately 71% of patients presented with a primary tumor, 23% with disease recurrence, and 7% with metastasis. Sites of disease included the lower extremity (59%), upper extremity (19%), trunk (15%), groin (4%), and head and neck (4%). FNA diagnosis was confirmed histologically in 88% of cases, clinically in 7% of cases, and cytogenetically in 1% of cases; 3% of cases had false-positive results and 1 patient was lost to follow-up. Sixteen of 19 patients received neoadjuvant chemotherapy based on the FNA diagnosis alone.
A cytopathologic diagnosis of HGS was found to be accurate in 88 of 97 cases (91%) with follow-up. A FNA biopsy diagnosis of HGS appears to be clinically reliable in a high percentage of cases when used in close conjunction with the orthopedic team. Cancer (Cancer Cytopathol) 2007. © 2007 American Cancer Society.
Soft tissue sarcomas are uncommon, with approximately 8000 to 9000 new cases diagnosed each year in the U.S.1, 2 Because of this, the consensus among orthopedists, surgical oncologists, and medical oncologists is that these patients should be evaluated in tertiary care centers that are experienced in securing a correct diagnosis and in sarcoma management. The initial assessment of a soft tissue mass includes diagnostic imaging and a variety of biopsy techniques.1 Among the latter are core needle biopsy (CNB), incisional (open) biopsy, and, occasionally, excisional biopsy. Although fine-needle aspiration (FNA) biopsy has been accepted for the evaluation of superficial mass lesions, its use in the arena of soft tissue masses is controversial. Thus, many clinicians resist the idea of requesting FNA when a sarcoma is clinically suspected, let alone permit management based on such a diagnosis.
Over the past several years, close collaboration has existed between orthopedic oncologists and pathologists in the study institution whereby FNA biopsy has become integral in the evaluation of patients presenting with a palpable soft tissue mass. After diagnostic imaging, it is among the first procedures to be performed in a patient presenting with a soft tissue mass. One of the criticisms of soft tissue FNA biopsy is that even if sarcoma is diagnosed, grading is not provided. Therefore, to partially answer this question, we focused on a consecutive sequence of patients treated over the past 6 years who had undergone FNA biopsy of a soft tissue mass, had a diagnosis of high-grade sarcoma (HGS) of any subtype issued based on the FNA, and had clinical or histologic follow-up. Our goal was to determine whether FNA biopsy was capable of making an accurate diagnosis of HGS.
MATERIALS AND METHODS
Case Selection and Grading
We reviewed our cytopathology files for all FNA cases of soft tissue masses diagnosed as HGS (or neoplasms that by definition would be categorized as HGS) between March 2001 and January 2007 in accordance with the approved Ohio State University Institutional Review Board application #2002H0089. This included bone sarcomas presenting as palpable soft tissue masses. There were no cases of gastrointestinal or retroperitoneal HGS. Our surgical pathology files were also searched for any examples of HGS cases with a prior FNA to capture any false-negative cases. The majority of cases that were graded generically as HGS were done so using overall cellularity and cellular pleomorphism, and were typically diagnosed as “pleomorphic sarcoma, high-grade.” Specific subtyping of HGS was accomplished either because the mass was recurrent or metastatic and the primary tissue diagnosis was known, ancillary testing such as immunohistology on a cell block or fluorescence in situ hybridization (FISH) were performed, or, in the case of bone neoplasms, specific knowledge of the bone imaging was known.
Percutaneous FNA biopsy was performed using a standard technique with 21-gauge or 22-gauge, 1.5-inch or 2-inch long needles. A single pathologist performed nearly all aspirations with an average of 3 separate passes into the lesion; no aspirations were performed using radiologic guidance. This was often done after a review of images with the orthopedic surgeon, and occasionally coordinated placement of the needle site. Conventional smears were made. Slides were stained with a Romanowsky stain for the initial assessment. A preliminary interpretation was issued within 20 minutes after the FNA. Remaining slides were later stained using a modified Papanicolaou stain. Cell blocks were attempted by embedding cells that were rinsed from each pass into a balanced salt solution, and then using a thrombin clot method to create a cell clot/block. This was then fixed in 10% buffered formalin and embedded in paraffin. Cell block slides were stained with hematoxylin and eosin.
FISH testing used a commercially available LSI EWSR1 dual-color, break-apart probe set (Vysis Inc, Downers Grove, Ill). This probe searched for translocations involving the EWSR1 gene on chromosome 22q12. An abnormal cell with a translocation reveals 1 pair of separate or “split” signals rather than 2 fused signals as occurs in a normal cell. Split signals in >20% of the cells were considered to be a positive result. FISH testing was performed on cell blocks in only 2 cases, and on both the cell block and cytospin smear in 2 cases.
A total of 107 FNA specimens from 98 patients with HGS were performed. Ten aspirates were found to be nondiagnostic (no malignant cells, necrosis only, or too few cells to issue a diagnosis) (Table 1). Of the 97 cases (91%) deemed satisfactory for interpretation, 6 were diagnosed as something other than HGS (1 malignant lymphoma, 1 hematoma, 1 spindle cell lesion, 1 suspicious, 1 low-grade sarcoma, and 1 sarcoma, not otherwise specified), resulting in a sensitivity of 94%. The remaining 91 HGS aspiration specimens contained 3 false-positive diagnoses (2 low-grade sarcomas and 1 hematoma). The positive predictive value and accuracy of an FNA diagnosis of HGS were 88 of 91 cases (97%) and 88 of 97 cases (91%), respectively.
|Case no.||Age, years/Gender||Site||FNA diagnosis||P/Re/M||Chemotherapy||Specimen||Tissue diagnosis|
|106||53/female||L thigh||Suspicious for sarcoma||P||None||CNB||HGS|
|105||52/male||R thigh||Nondiagnostic||P||None||Open Bx||HGS|
|101||42/male||R thigh||Hematoma||P||None||Open Bx||HGS|
|100||82/female||R shoulder||Nondiagnostic||P||None||CNB||HG LMS|
|99||73/female||R thigh||HG MFS||P||None||WEx||HG LPS|
|96||39/male||R deltoid||Malignant lymphoma||P||None||Open Bx||Ewing sarcoma|
|95||27/female||R arm||Nondiagnostic||P||None||Open Bx||Osteosarcoma|
|94||34/female||R thigh||Nondiagnostic||P||None||Open Bx||SS|
|93||40/female||L thigh||HG LPS||P||None||WEx||Same|
|92||74/female||R thigh||HGS||P||None||WEx||HG LMS|
|90||22/female||R foot||HG LMS||Re||None||Amp||Same|
|89||49/male||R thigh||Spindle cell sarcoma||P||None||CNB||HG LMS|
|88||70/male||R thigh||Nondiagnostic||P||None||CNB||HG LPS|
|84||84/male||R thigh||Nondiagnostic||P||None||Open Bx||HG LMS|
|83||35/female||R leg||Spindle cell lesion||P||None||CNB||Clear cell sarcoma|
|82||53/female||L thigh||Nondiagnostic||P||None||CNB||HG LMS|
|81||76/female||L arm||Myxoid LPS||Re||None||Amp||Round cell LPS|
|80||28/female||R calf||HGS||P||None||CNB||Epithelioid sarcoma|
|78||62/female||L post thigh||HGS||M||None||WEx||Same|
|77||62/female||L ant thigh||HGS||P||None||WEx||Same|
|75||76/male||R thigh||HG LMS||P||None||WEx||Same|
|74||59/male||R thigh||HGS||P||None||WEx||HG LMS|
|73||85/female||L ankle||HGS||Re||None||None||To hospice|
|71||13/male||Back||Ewing sarcoma||P||10/2006||WEx||+ on FISH, same|
|69||85/male||R ant tibia||HGS||P||None||WEx||Same|
|68||51/female||R post thigh||HGS||P||08/2006||WEx||HG LPS|
|67||32/male||R calcaneus||Ewing sarcoma||P||05/2006||Amp||+ on FISH, same|
|66||76/female||L thigh||HGS||P||None||None||To hospice|
|65||66/male||R face||HGS||Re||None||Excision||Epithelioid angiosarcoma|
|59||53/female||R thigh||HGS||P||09/2005||WEx||HG LMS|
|57||85/female||R tibia||HG MFS||P||None||WEx||Same|
|55||51/male||L thigh||HG MFS||P||08/2005||Open Bx||Same|
|53||48/male||L forearm||HG MFS||Re||None||WEx||Same|
|51||25/male||L hip||HG MPNST||Re||None||WEx||Same|
|45||15/male||R leg||Ewing sarcoma||P||03/2005||WEx||+ on FISH, same|
|44||89/female||R calf||HG MPNST||P||None||Amp||HGS|
|43||50/male||L leg||HG LMS||P||None||None||DOD|
|42||49/male||R arm||HG MFS||P||None||Amp||Same|
|41||76/female||R neck||HG LPS||M||None||WEx||Same|
|36||76/female||R thigh||HG MFS||P||None||WEx||LG sarcoma|
|35||85/female||L thigh||HG MFS||P||None||WEx||Same|
|34||48/male||Thigh||Osteosarcoma||Re||None||None||Lost to follow-up|
|33||56/female||R wrist||HGS||P||None||WEx||LG sarcoma|
|32||31/male||L leg||HG MPNST||Re||None||Amp||Same|
|28||73/male||L knee||HG LMS||Re||None||WEx||Same|
|24||56/male||R forearm||HG MPNST||P||None||WEx||Same|
|22||60/male||R knee||HG MFS||P||None||Amp||Same|
|19||76/female||R thigh||HGS||P||None||WEx||HG LPS|
|17||35/male||R hand||Epithelioid sarcoma||Re||None||Amp||Same|
|16||35/male||R hand||Epithelioid sarcoma||Re||None||Amp||Same|
|15||26/male||L scapula||Ewing sarcoma||P||05/2003||None||+ FISH|
|11||81/female||L thigh||HG LPS||Re||None||WEx||Same|
|7||59/male||R calf||HGS||Re||01/2003||WEx||Necrotic sarcoma|
|6||88/male||L buttock||HGS||P||None||Open Bx||Hematoma|
|5||29/male||L thigh||HG MPNST||P||None||Amp||Same|
|4||59/female||L thigh||HG LPS||Re||None||WEx||Same|
|2||60/female||Neck||HG angiosarcoma||Re||None||Punch Bx||Same|
|1||71/male||L back||HGS||Re||None||None||To hospice|
Of the 10 nondiagnostic cases and 6 cases diagnosed as being other than HGS, 9 had a core needle biopsy (CNB) finding that confirmed the diagnosis of HGS, 5 had open biopsy (2 of these had a CNB that was nondiagnostic), and 1 patient diagnosed as having recurrent myxoid liposarcoma (LPS) underwent an electively scheduled amputation. Resection of that tumor demonstrated myxoid LPS with a focus of round cell LPS.
Three false-positive cases (Cases 6, 33, and 36), and 1 case diagnosed as malignant lymphoma (Case 96) did not appear to adversely impact patient outcome. In 2 cases, resection revised the diagnosis of HGS to low-grade spindle cell sarcoma after tissue examination. Neither patient received chemotherapy; each underwent a surgical procedure similar to what would have been performed had a diagnosis of HGS been made. The patient with a hematoma (Case 6) that was misdiagnosed as HGS had a computed tomography (CT) scan finding that was worrisome for sarcoma. The mass decreased in size 1 week after FNA. Incisional biopsy revealed organizing hematoma and exuberant granulation tissue. The patient with a Ewing sarcoma (EWS) that was misdiagnosed as lymphoma underwent open biopsy 6 days later when the flow cytometric findings from the FNA were negative; no interval therapy was instituted. We encountered no cases of metastatic carcinoma that were mistaken for HGS.
Patients ranged in age from 13 to 90 years (mean age, 56 years) with 50 men and 48 women (male: female ratio of 1:1). Six patients underwent 2 FNA biopsies (Cases 16 and 17, 5 and 32, 50 and 56, 30 and 53, 20 and 48, and 37 and 54) because of a local mass or metastasis. One patient underwent 4 FNA biopsies (Cases 47, 77, 78, and 79).
Of the 91 HGS aspiration specimens, 54 (59%) were diagnosed as HGS corresponding to a pleomorphic sarcoma of no specific type, 8 (9%) as high-grade myxofibrosarcoma, 8 (9%) as osteosarcoma, 5 (5%) as high-grade malignant peripheral nerve sheath tumor (MPNST), 5 (5%) as high-grade leiomyosarcoma, 4 (4%) as EWS, 4 (4%) as high-grade LPS, 2 (2%) as epithelioid sarcoma, and 1 (1%) as high-grade angiosarcoma. Seventy-six aspiration specimens (71%) were of a primary soft tissue mass, 24 (22%) were of a locally recurrent mass, and 7 (7%) were presumed metastases. Of the aspiration specimens of a new (primary) soft tissue mass, 58 of 76 (76%) were correctly diagnosed as HGS; 3 were false-positive findings, 10 were nondiagnostic, and 5 were diagnosed as an entity other than HGS. When the 10 nondiagnostic aspiration specimens were discounted, the sensitivity was 92% and the positive predictive value was 95% for recognizing primary HGS by FNA. Only 15% of cell blocks contained sufficient cellular material to aid in the final interpretation.
Sites included the lower extremity in 63 cases (59%), the upper extremity in 20 cases (19%), the trunk in 16 cases (15%), the groin in 4 cases (4%), and the head/neck region in 4 cases (4%). Of 91 HGS diagnoses, 80 (88%) were confirmed histologically, 6 (7%) were confirmed clinically, 1 was confirmed cytogenetically (1%; Case 15), and 3 (3%) were false-positive findings. One patient was lost to follow-up. Tissue specimens after FNA included 56 wide excision/radical resections, 17 amputations, 12 CNBs, 8 open biopsies, 3 excisional biopsies, and 1 punch biopsy. Six clinically confirmed cases included 4 patients who declined further treatment and were sent to hospice for palliative care due to deteriorating clinical status, 1 patient with radiographically demonstrable metastases to several sites, and 1 patient who died with clinical disease but did not undergo an autopsy.
Four of 5 patients diagnosed with MPNST (1 patient with 2 separate FNA biopsies) had a known history and clinical stigmata of NF-1. Sixteen patients had a known bone lesion with soft tissue extension at the time of FNA. No complications was known to occur in any patient as a consequence of FNA biopsy.
Patients Treated With Neoadjuvant Chemotherapy
Nineteen patients received neoadjuvant chemotherapy. Three underwent a subsequent CNB or open surgical biopsy shortly after the FNA diagnosis (Cases 55, 59, and 68) to confirm the cytologic diagnosis. Sixteen patients (6 with HGS, 6 with osteosarcoma, and 4 with EWS) received neoadjuvant chemotherapy based on the FNA diagnosis alone. All 16 patients had subsequent tissue specimens that verified the FNA diagnosis. Four patients with EWS who were treated based on the cytologic diagnosis had positive FISH results for the EWRS-1 translocation. Three patients were diagnosed with osseous EWS with soft tissue involvement, and 1 patient was diagnosed with an extraosseous EWS. Three of 4 patients underwent definitive surgical resection after completing chemotherapy, all of whom demonstrated >90% tumor necrosis. One patient with EWS (Case 15) had such a remarkable response to neoadjuvant chemotherapy that surgery was not pursued, thereby preserving his shoulder function. All 8 soft tissue masses diagnosed as osteosarcoma originated from an underlying bone lesion. Six patients received neoadjuvant chemotherapy before definitive surgical resection based on the FNA diagnosis only. Another patient with osteosarcoma was diagnosed as having HGS, not otherwise specified, by FNA. Subsequent surgical resection demonstrated a sufficient amount of viable tumor to confirm the FNA diagnosis with varying amounts of necrosis noted in each of these cases. One patient with HGS (Case 7) achieved a complete response such that no viable sarcoma remained in the resection; 19 months later, it was found that HGS had metastasized to the small intestine.
Accurate pathologic diagnosis is crucial to the management of patients with soft tissue sarcomas because many will require drastic, debilitating surgery. The most appropriate type of biopsy to establish a pathologic diagnosis has been the subject of many reports. The “gold standard” is incisional/open biopsy because it overcomes (in most instances) the problems of sampling error, histologic heterogeneity, and necrosis within a sarcoma. It often allows for reliable mitotic counts and estimation of the percentage of necrosis, thus permitting accurate grading.3, 4 Nonetheless, because of the wound complications and tumor spillage associated with open biopsy, CNB is currently in vogue and appears to be a more common procedure in many busy orthopedic centers because of substantial savings in time and cost, and minimal morbidity compared with incisional biopsy.5–7 Although some studies quote a lower rate of accuracy when compared with open biopsy, CNB is still considered an acceptable substitute for patients with large soft tissue masses or bone tumors with soft tissue extension.8
Pathologists and even experienced cytopathologists usually approach the FNA interpretation of a possible soft tissue sarcoma with trepidation. Current sentiment is that FNA biopsy can be used to confirm/rule out local disease recurrence or metastasis in a known sarcoma patient, but not for an initial sarcoma diagnosis or to perform a major resection based on FNA diagnosis.9 We agree with others that due to their infrequency, their histologic heterogeneity, their overlapping histomorphologic features with some reactive lesions, the difficulty sometimes associated with the procurement of an ample amount of diagnostic material, and the inherent difficulty associated with the pathologic interpretation of a sarcoma even in tissue specimens, FNA biopsy of these neoplasms should be attempted only in centers associated with a sufficient volume and experience with both performing and interpreting FNA biopsy results.1 Because pathologists in Sweden are reported to have the largest experience with soft tissue aspiration, they have espoused this form of tissue sampling more than others,10 and some have now advocated supplementing FNA with CNB because both techniques appear to complement one another.11 Others also have found the combination of CNB and FNA to be more clinically useful as opposed to FNA alone.12 Ancillary studies should be an important component of soft tissue FNA biopsy whenever possible. Cell blocks made from FNA can act as surrogates for CNB because they allow one to apply immunohistology and FISH to aspiration samples. Unfortunately, during this time period, the quality of our cell block material was poor due primarily to technical factors in the laboratory. We were only able to capture sufficient cells in 15% of our aspiration specimens for immunostaining. FISH testing for chromosomal translocations in select sarcomas is another avenue for confirmatory testing and can be applied to FNA specimens. In all 4 examples of EWS, we were able to apply the EWS gene translocation to support our combined cytologic, radiologic, and clinical impressions. Other sarcomas that are potential targets of FISH testing from FNA specimens include low-grade fibromyxoid sarcoma, alveolar rhabdomyosarcoma, clear cell sarcoma of soft parts, extraskeletal myxoid chondrosarcoma, myxoid liposarcoma, and synovial sarcoma. We either did not possess the probe for these other tumors at the time the case was encountered or did not encounter examples of these tumors in the current study.
To our knowledge, the current study is one of the largest series of FNA biopsy diagnoses of HGS reported to date. The majority of cases represented high-grade pleomorphic sarcoma with no specific differentiation, or what until recently had been termed “malignant fibrous histiocytoma.” An unexplainable fact of this series is that only 1 example of synovial sarcoma was aspirated during this time period, and that aspiration sample was found to be nondiagnostic. After eliminating 10 nondiagnostic samples, we found 91% were adequate for diagnosis; 91 aspiration specimens were diagnosed as some form/subtype of HGS. Six tissue-documented HGS were misdiagnosed by a prior FNA. Three of 91 HGS aspiration specimens represented false-positive diagnoses. During this 6-year interval, we were able to achieve a sensitivity of 94% (excluding the 10 nondiagnostic cases), a positive predictive value of 97%, and an overall accuracy of 91% in our ability to diagnose HGS using FNA biopsy. Nearly 75% of our aspiration specimens were from primary, previously undiagnosed, malignancies. Nearly 90% of our cytologic diagnoses were followed eventually by some form of tissue confirmation and 7% by a subsequent clinical course. This degree of accuracy was accomplished only through a close collaboration of pathologists willing to see the patient (thus creating a clinical impression with review of images when available, and the ability to issue immediate diagnostic feedback), and knowledgeable orthopedic surgeons skilled in radiographic interpretation. As in all musculoskeletal oncology, the clinical and radiologic appearance must correlate with the FNA impression; if it does not, our surgeons will proceed with CNB or open biopsy depending on the nature of the lesion. This is analogous to applying a “triple test” as is done in breast cytopathology so that our 3 false-positive cases did not result in any harm or improper treatment to the patient.
What advantage, if any, has this form of biopsy over more conventional methods? For us, this technique allows the patient to be given a preliminary diagnosis on the day of the initial consultation. Staging studies and medical workup can be expedited and initiated that same day. FNA biopsy tracts cause less contamination because there is not as large an amount of tissue that is passed through the needle tract as in CNB. Therefore, our surgeons have chosen not to remove the biopsy tract with definitive surgical resection. This does not appear to compromise clinical outcome. Therefore, morbidity of resection may be decreased by not resecting the skin with an FNA biopsy, and in some cases it might prevent a split thickness skin graft from having to be performed. There is minimal disruption of the tumor bed leading to a reduced risk of hematoma formation because a smaller needle is used, thereby causing less trauma and risk of injury to a feeder blood vessel. Another reason is practice enhancement for the clinician. A much more rapid turnaround of patients occurs with FNA biopsy, allowing surgeons and oncologists to see more individuals in a single clinical setting. FNA is a lower cost alternative to conventional open biopsy or CNB while still facilitating patient workup and generating sufficient relevant diagnostic information to guide therapeutic management. This latter point was emphasized by Kilpatrick et al, who demonstrated that FNA biopsy was sufficient to begin definitive therapy in 83% of patients with soft tissue sarcoma.13
Grading of soft tissue sarcomas is important in predicting biologic behavior, and involves both tumor subclassification and certain histomorphologic parameters. Although the National Cancer Institute and French grading system are currently the 2 most commonly used grading systems, no grading system fits every type of sarcoma,14 and some centers have employed nomograms to help determine sarcoma survival.15 Furthermore, the histomorphologic criteria used for grading in these 2 schema are not applicable to any form of needle biopsy be it FNA or CNB. Nevertheless, CNB has been shown to accurately grade approximately 85% of sarcomas when it is used by experienced soft tissue pathologists.6 Our practice has been to attempt to determine cytologic grade from sarcoma aspiration specimens whenever possible. This has been used to influence the extent of surgical resection, and decide whether neoadjuvant therapy may be administered to eligible patients. In 16 patients, the clinical and radiographic features were compelling enough that FNA biopsy was the sole pathologic discriminator used to administer neoadjuvant chemotherapy, and in 3 of these patients it was followed by CNB or open biopsy before chemotherapy. The FNA diagnosis was validated in all 16 patients who received neoadjuvant chemotherapy by eventual tissue examination. Previous reports of FNA grading of sarcoma have reported variable success. In a retrospective review of 77 cases using a 3-grade system, 3 different pathologists were able to accurately predict the final histologic grade from FNA biopsy alone in only 43% to 50% of cases,16 whereas Palmer et al were able to achieve 90% concordance in 59 cases of histologically confirmed sarcomas when using a 2-grade system17 and Weir et al reported a 92% grading accuracy rate in 36 cases of spindle cell sarcoma.18
The shortcomings of the current study are that we only included FNA specimens of palpable masses and not those of deep lesions that could have been aspirated using radiologic guidance. Although we used a 1.5-inch and occasionally a 2-inch needle, some of our nondiagnostic aspiration specimens were the result of the lesion being deeper than we could reach with these needles. In addition, we did not attempt during this period to review all soft tissue FNA biopsies diagnosed as sarcoma to determine the overall accuracy of the technique; we focused only on those aspiration samples designated as HGS. Moreover, we included 16 cases that could rightly be interpreted not as soft tissue tumors but as bone sarcomas presenting with soft tissue masses (8 osteosarcomas detected by FNA, 4 osteosarcomas diagnosed as HGS by FNA, and 4 osseous EWS [1 of which was misdiagnosed as lymphoma]). Because the diagnosis of a malignant bone tumor is often reasonably certain from its radiographic appearance alone, one could argue that FNA biopsy is merely a confirmatory test in this situation. As noted, our clinicians were more comfortable administering neoadjuvant chemotherapy without histopathologic confirmation for a bone sarcoma because 10 of 16 patients had either osteosarcoma or EWS diagnoses made by FNA before neoadjuvant chemotherapy.
In the current study, we reported a high success rate in the recognition of HGS using FNA biopsy. Although we have embraced the value of this technique, it should not be overstated. Cytologists without familiarity or the resources to make an accurate diagnosis may wish to avoid FNA in this circumstance or, alternatively, have the patient referred to the appropriate institution. Our experience leads us to believe that patients treated at specialty orthopedic centers with a high volume of these neoplasms can benefit from the incorporation of FNA by orthopedists into their practice.
We thank Dr. Gary Bos, former Chairman of Orthopedics, for his support and encouragement of the fine-needle aspiration service, and Cyril Fisher, MD, FRC Path, for his suggestions and critical review of the article.
- 3FletcherCDM, UnniKK, MertensF, editors. World Health Organization Classification of Tumours. Pathology and Genetics of Tumours of Soft Tissue and Bone. Lyon, France: IARC Press; 2002.
- 4The hazards of the biopsy, revisited. J Bone Joint Surg. 1996; 78-A: 656–663., , .
- 10Critical review on the role of fine needle aspiration in soft tissue tumors. Pathol Case Rev. 1998; 3: 111–117., .
- 16Concordance of histopathologic and cytologic grading in musculoskeletal sarcomas. Can grades obtained from analysis of the fine-needle aspirates serve as the basis for therapeutic decisions? Cancer. 2002; 96: 83–91., , , et al.