Optimal Diagnostic Yield Achieved With On-site Pathology Evaluation of Fine-Needle Aspiration Assisted Core Biopsies for Pediatric Osseous Lesions

Transcription

1 Optimal Diagnostic Yield Achieved With On-site Pathology Evaluation of Fine-Needle Aspiration Assisted Core Biopsies for Pediatric Osseous Lesions A Single-Center Experience Kalyani Patel, MD; Darryl Kinnear, PA(ASCP); Norma M. Quintanilla, MD; John Hicks, MD, PhD; Eumenia Castro, MD, PhD; Choladda Curry, MD; John Dormans, MD; Daniel J. Ashton, MD; J. Alberto Hernandez, MD; Hao Wu, MD, PhD Context. Image-guided, fine-needle aspiration assisted core needle biopsy with an on-site evaluation by a pathologist (FNACBP) of osseous lesions is not a common practice in pediatric institutions. Objectives. To evaluate the diagnostic adequacy and accuracy of FNACBP for pediatric osseous lesions and to compare the adequacy with procedures that do not use fine-needle aspiration. Design. Six-year, retrospective review of 144 consecutive children biopsied for osseous lesions with and without fine-needle aspiration assistance. Results. Pathologic diagnosis was achieved in 79% (57 of 72) of the core biopsies without an on-site evaluation, 78% (32 of 41) of the open biopsies (9 with intraoperative consultation), and 97% (30 of 31) of the FNACBPs as the initial diagnostic procedure. Three FNACBP cases were preceded by nondiagnostic open biopsies. Among 34 lesions sampled by FNACBP, 33 (97%) succeeded with diagnostic tissue, with most (30 of 33; 91%) being neoplasms, including 16 malignant (48%), 13 benign (39%), and 1 indeterminate (3%) lesions. The mostcommon diagnoses were osteosarcoma (9 of 33; 27%) and Langerhans cell histiocytosis (7 of 33; 21%). In cases with follow-up information available, 93% (28 of 30) of the FNACBP-rendered diagnoses were clinically useful, allowing initiation of appropriate therapy. The FNACBP procedure had 100% specificity, sensitivity, and positive predictive value for all 14 malignant lesions, with the sensitivity being 88% in benign lesions. Most FNACBP procedures (32 of 34; 94%) yielded adequate material for ancillary testing. A gradual upward trend was observed for the choice of FNACBP as an initial diagnostic procedure for osseous lesions. Conclusions. The FNACBP procedure yields sufficient material for diagnosis and ancillary studies in pediatric, osseous lesions and may be considered an initial-diagnostic procedure of choice. (Arch Pathol Lab Med. 2017;141: ; doi: / arpa oa) No unified, optimal biopsy technique for the diagnosis of osseous lesions has been established, although incisional biopsy is considered the gold standard. Imaging-guided core needle biopsy has become a routine procedure because of its inherent medical and economic advantages. The diagnostic accuracy of this technique significantly increases when paired with fine-needle aspiration (FNA) and on-site evaluation by a pathologist. 1 An FNA is often used in adults to diagnose metastatic osseous lesions, 1 3 and FNA and core biopsy (CB), with or without imaging guidance, has been generally accepted for superficial or deep nonosseous lesions in children. 4 7 However, the use of this technique has been limited for pediatric osseous lesions, which typically have a differential diagnosis that is distinct from that seen in adult patients. The aim of this study was to analyze retrospectively the adequacy and diagnostic accuracy of imaging-guided, FNAassisted core needle biopsies of osseous lesions with the assessment for adequacy by an on-site pathologist (FNACBPs). We also compared diagnostic adequacy of samples obtained through non FNA-assisted biopsies (both open biopsies [OBs] and imaging-guided CBs) during the same period. Accepted for publication September 12, Published as an Early Online Release March 16, From the Departments of Pathology (Drs Patel, Quintanilla, Hicks, Castro, Curry, and Wu and Mr Kinnear), Surgery (Dr Dormans); and MATERIALS AND METHODS Radiology (Drs Ashton and Hernandez), Texas Children s Hospital After approval by the institutional review board, the pathology and Baylor College of Medicine, Houston. The authors have no relevant financial interest in the products or database was searched for the period January 2010 to December companies described in this article. 2015, using various combinations of the phrases bone, biopsy, and Reprints: Hao Wu, MD, PhD, Department of Pathology, Baylor fine-needle aspiration found in the final diagnosis section of the College of Medicine, Texas Children s Hospital, 6621 Fannin St., AB195.14, Houston, TX ( hxwu@texaschildrens.org). pathology report. Search results were filtered to enlist all biopsies of osseous lesions. Electronic medical records were reviewed to 678 Arch Pathol Lab Med Vol 141, May 2017 Biopsy for Pediatric Osseous Lesions Patel et al

2 Table 1. Diagnostic Categories Derived From Fine-Needle Aspiration Guided Core Needle Biopsy Pathologic Yield Cases, No. (%), N ¼ 34 Diagnostic 33 (97) Neoplastic 30 (91) a Malignant b 16 (48) c Benign d 13 (39) c Indeterminate 1 (3) c Nonneoplastic 3 (9) c Nondiagnostic 1 (3) a Percentage of neoplastic and nonneoplastic lesions derived from the total number of diagnostic procedure. b Osteosarcoma (n ¼ 9), Ewing sarcoma (n ¼ 2), metastatic rhabdoid tumor (n ¼ 1), metastatic synovial sarcoma (n ¼ 1), spindle cell sarcoma (n ¼ 1), diffuse large B-cell lymphoma (n ¼ 1), and clear cell chondrosarcoma (n ¼ 1). c Percentage of lesions within neoplastic category. d Langerhans cell histiocytosis (n ¼ 7), chondroblastoma (n ¼ 3), enchondroma (n ¼ 1), benign spindle cell lesion with giant cells (n ¼ 1), and benign osteoblastic lesion (n ¼ 1). include basic demographic parameters and clinical details, such as location and size of lesion, radiologic characteristics, type and result of initial diagnostic procedure, type and result of follow up procedures if any, and adequacy of material in terms of diagnosis, prognosis, ancillary testing, and specimen preservation for possible future research studies. At our institution, all FNA-assisted procedures are performed under fluoroscopy, ultrasound, or computed tomography guidance by interventional radiologists with the presence of a pathologist in the interventional radiology (IR) suite for on-site adequacy evaluation and specimen triaging. An orthopedic surgeon is available to mark the biopsy site for suspected sarcomas for future potential limb-salvage surgery and to ensure that the needle track would be within the planned resection bed without traversing unaffected compartments or muscles that would be used in reconstruction. The patient is sedated by an anesthesiologist based on American Society of Anesthesiologists score, location of the lesion, and position of the patient for the procedure. Under imaging guidance, an 11-gauge coaxial guiding needle (Teleflex, Research Triangle Park, North Carolina) is advanced to the edge of the lesion. Fine-needle aspiration using a 22-gauge Chiba needle (Cook Medical, Bloomington, Indiana) or a 25-gauge hypodermic needle (Cook Medical) is performed along the guiding needle. Aspirate smears are made on site, and selected slides from each pass are stained with the Diff-Quik (StatLab, McKinney, Texas) method when still in the IR suite to check for adequacy. Additional passes are obtained until the aspirate appears adequate for diagnostic purpose. Subsequently, tissue cores are obtained at the same site with a 13-gauge Arrow OnControl needle (Teleflex) through the same 11-gauge, coaxial guiding needle. A 16-gauge BioPince Full Core biopsy instrument (Argon Medical Devices, Plano, Texas) is used for juxtacortical or intramedullary soft issue component, if present. Remaining unstained slides are air-dried or preserved in 95% alcohol for subsequent Papanicolaou staining. Based on the clinical suspicion and findings during the on-site evaluation by the pathologist, specimen material is triaged for microbiology culture studies, flow cytometry, and conventional cytogenetics. Afterward, the syringes are rinsed in 10% neutral-buffer formalin for cell block preparation. The FNA materials are also preserved in RNALater Stabilization solution (Thermo Fisher Scientific Life Sciences, Waltham, Massachusetts) for possible future molecular studies. Air-dried, unstained slides from aspirate smears and touch preparations from the needle core biopsies are preserved for possible fluorescence in situ hybridization studies, if indicated. Small portions of the needle core biopsy are saved in 10% glutaraldehyde for potential electron microscopy studies and are cryopreserved for possible molecular studies. Aspirated material is also provided to institutional review board approved institutional research laboratories when the family consents. For the purpose of this study, we defined a procedure as adequate if it generated a diagnosis and defined the clinical utility of the procedure based on its ability to eliminate a subsequent diagnostic/ surgical procedure and to initiate appropriate therapy, whereas the diagnostic accuracy was defined based on confirmation of the precise diagnosis on subsequent resection specimen. In the absence of the resection specimen, the diagnosis was considered accurate when supplemented by ancillary testing, such as immunohistochemistry, cytogenetics and/or molecular analysis, or tumor response after chemotherapy or lesional steroid injection based on pathologic diagnosis. When the FNACBP diagnosis was broadly similar but not precise to that of the resection specimen, the procedure was deemed clinically useful, but not accurate. A 2-tailed Student t test was used to compare the mean size of lesions that were biopsied through CB and through FNACBP and between OB and FNACBP. A 2-tailed Fisher exact test was used to compare the percentage of the diagnostic procedures, the percentage of the neoplastic lesions, the percentage of the malignant lesions between CB and FNACBP and between OB and FNACBP. When the absolute number (outcome) was more than 5 in each group, a 2-tailed v 2 test was used to compare those percentages, and P,.05 was considered statistically significant. RESULTS A total of 34 imaging-guided FNACBPs were performed for osseous lesions in 34 patients from 2010 through 2015 at Texas Children s Hospital (Houston, Texas). Age at the time of the procedure ranged from 6 months to 19 years, with a mean of 10.3 years and a median of 10 years. Male to female ratio was 1.8:1. Initial radiologic measurements were not available in 8 patients (24%); in the remaining 26 patients (76%), the lesions ranged from 1.1 cm to 11.5 cm in greatest dimension (mean, 4.9 cm; median, 3.7 cm). Fifteen of the 26 lesions (58%) appeared heterogeneous by radiology, with both solid and cystic components. The diagnostic impressions, based on clinical information and radiologic characteristics, were malignant lesion in 12 of the 34 patients (35%), benign lesion in 11 patients (32%) (including 7 patients suspected to have Langerhans cell histiocytosis), and indeterminate lesion in 11 patients (32%). Ten patients (29%) had a known history of systemic or malignant lesions at the time of biopsy, including 3 patients (9%) with Langerhans cell histiocytosis, 3 patients (9%) with osteosarcoma, 1 patient (3%) with Ewing sarcoma, 1 patient (3%) with rhabdoid tumor, 1 patient with malignant peripheral neural sheath tumor (MPNST) (3%), and 1 patient (3%) with chondrosarcoma. The most common sites of involvement were long bone (19 of 34; 56%), chest wall and pelvis (n ¼ 5 each; 15% each), scapula (n ¼ 2; 6%), clavicle (n ¼ 1; 3%), foot (n ¼ 1; 3%), and maxillofacial and cranial bones (n ¼ 1; 3%). Other than one procedure (1 of 34; 3%) that generated nondiagnostic tissue with only benign skeletal muscle and adipose tissue from a rib lesion, most procedures (33 of 34; 97%) succeeded in obtaining diagnostic tissue (Table 1). Of the 33 lesions with a tissue diagnosis, 30 were neoplastic (91%), including 16 of 33 malignant tumors (48%), 13 benign tumors (39%), and 1 indeterminate tumor (3%). Three of the 33 lesions (9%) were nonneoplastic. Among the 33 diagnostic procedures, osteosarcoma was the most common malignant neoplasm with 9 cases (27%). Langerhans cell histiocytosis was the most common benign neoplastic diagnosis (7 of 33; 21%) (Table 1). The indeterminate case was a 2.0-cm humerus neoplasm Arch Pathol Lab Med Vol 141, May 2017 Biopsy for Pediatric Osseous Lesions Patel et al 679

3 Table 2. Clinical Correlation of Fine-Needle Aspiration Guided Core Needle Biopsy Result With Surgical Resection and/or Clinical Course Clinical Utility Cases, No. (%); N ¼ 30 Clinically useful 28 (93) Diagnostically accurate 25 (83) Diagnostically inaccurate 3 (10) Clinically insufficient 2 (6) Inadequate specimen 1 (3) Adequate specimen 1 (3) diagnosed as spindle cell neoplasm. Three lesions were nonneoplastic, including 1 acute osteomyelitis and 2 reactive bone components. All lesions, except metastatic rhabdoid tumor and synovial sarcoma, were considered primary osseous lesions. Four patients (12%) did not have follow-up treatment at our institution (1 with enchondroma, 2 with osteosarcoma, and 1 with chondroblastoma) to compare the diagnostic accuracy of the FNACBP procedure. Of the remaining 30 FNACBP procedures, 93% (28 of 30) were clinically useful, and 83% (25 of 30) were diagnostically accurate (Table 2). Three of the 30 cases (10%) were clinically useful but not accurate: (1) a 4.6-cm tibial lesion diagnosed as spindle cell sarcoma by FNACBP with a subsequent diagnosis of MPNST on resection, (2) a 1.9-cm benign spindle cell acetabular lesion with a subsequent diagnosis of benign bone cyst on excision, and (3) a 3.7-cm frontal sinus, benign osteoblastic lesion subsequently diagnosed as a craniofacial osteoma with osteoid osteoma-like nidus on excision. Two of the 30 procedures (7%) lacked clinical utility and required a subsequent diagnostic/surgical procedure. One of them was an inadequate specimen from a rib lesion that demonstrated Langerhans cell histiocytosis on resection. The other case lacking clinical utility and considered diagnostically inaccurate was a humerus lesion, classified as indeterminate (spindle cell neoplasm), which, on subsequent curettage of the lesion, demonstrated an osteoid osteoma. All 14 clinically and surgically proven malignant lesions (including the MPNST) were diagnosed as such by FNACBP, thus, giving a sensitivity, specificity, and positive predictive value of 100% for malignancy. Among the 16 benign/nonneoplastic entities confirmed upon resection, the diagnosis was also benign by FNACBP in 14 cases (sensitivity, 88%). One of the 2 discordant cases was inadequate, and the other was indeterminate (spindle cell neoplasm). Most of the procedures (32 of 34; 94%) yielded adequate material for ancillary studies (immunohistochemistry, conventional cytogenetics, fluorescence in situ hybridization, electron microscopy, cryopreserved tissue for possible molecular studies). Based on the on-site evaluation, FNACBP material from the acute osteomyelitis case was triaged for microbiology culture studies, which yielded colonies of methicillin-sensitive Staphylococcus aureus. Nine procedures also provided adequate material for institutional review board approved institutional research laboratories. Of the 2 procedures that were apparently inadequate with respect to material available for ancillary testing, 1 case was a metastatic Ewing sarcoma without any tumor cells by FNACBP and upon subsequent surgical resection as well. The other case without material for ancillary testing was considered inadequate for diagnosis. A gradual upward trend was observed for the choice of performing FNACBP as an initial diagnostic procedure for osseous lesions when compared with other modalities (Figure 1) during this study period. Among the 113 non- FNA procedures, 72 imaging-guided CBs (64%) and 41 OBs (36%) (including 31 incisional biopsies [76%], 4 curettages [10%], and 6 unspecified procedures [15%]) were performed as an initial diagnostic procedure for osseus lesions during the study period. In comparison, 31 primary FNACBP and 3 follow-up FNACBP were performed during the same period. Size of the lesion by imaging, when available, was significantly smaller in patients who underwent CBs (35 of 72; 49%) than in those who underwent FNACBP (P ¼.01 by 2-tailed student t test). However, there was no significant radiologic size difference among the lesions undergoing OBs (19 of 41; 49%) versus FNACBP (P ¼.12) (Table 3). Overall, pathologic diagnosis was achieved in 79% (57 of 72) of the CB procedures and 78% (32 of 41) of the OB procedures, significantly lower than that of FNACBP (30 of 31; 96.8%) (P ¼.03 and.04 by Fisher exact test, respectively). Figure 1. Initial diagnostic procedures performed from 2010 to 2015 by year. A trend toward increased use of fine-needle, aspiration-assisted core biopsy with on-site pathologist evaluation (FNACBP) as the initial diagnostic procedure (solid bar), compared with core biopsy (white bar) and open biopsy (hatched line), was noted over time. Numbers over each bar indicate the actual number of procedures performed in that year. 680 Arch Pathol Lab Med Vol 141, May 2017 Biopsy for Pediatric Osseous Lesions Patel et al

4 Table 3. Procedure Type Demographic and Radiologic Characteristics of Patients With Various Initial Diagnostic Procedures M:F Ratio Median Age, y Median Size, cm Diagnostic Procedure, No. (%) Neoplastic, No. (%) a Malignant, No. (%) b Benign, No. (%) b Indeterminate, No. (%) a CB, n ¼ c 57 d (79) 36 (63) 15 (42) 20 (56) 1 (1) OB, n ¼ e (78) 18 (56) 4 (22) 14 (78) 0 FNACBP, n ¼ (97) 27 (90) 14 (51) 12 (44) 1 (3) Abbreviations: CB, core biopsy performed by interventional radiologists; cm, centimeter; F, female; FNACBP, fine-needle aspiration guided core needle biopsy with pathologist adequacy evaluation; M, male; OB, open biopsy performed by orthopedic surgeon. a Percentage derived from the diagnostic procedure. b Percentage derived from the number of neoplastic cases. c Lesions sampled with CB were significantly smaller than those sampled by FNACBP (P ¼.01 by 2-tailed Student t test). d FNACBP yielded more-frequent diagnostic procedures compared with CB (P ¼.03 by Fisher exact test). e FNACBP yielded more-frequent diagnostic procedures compared with OB (P ¼.04 by Fisher exact test). In cases with pathologic diagnosis, neoplastic entities were diagnosed more frequently by FNACBP (27 of 30; 90%), compared with either CB (36 of 57; 63%) or OB (18 of 32; 56%) (P ¼.01 and.004, respectively, by Fisher exact test). There was a trend for a greater proportion of malignancies diagnosed by FNACBP (14 of 27; 51%) compared with OB (4 of 18; 22%) (P ¼.07 by 2-tailed Fisher exact test) or CB (15 of 36; 42%) (P ¼.42 by 2-tailed v 2 test). Intraoperative consultation for specimen adequacy or diagnosis was performed on 9 of 41 OB cases (22%), including 8 of 18 (44%) with neoplastic diagnoses and 1 of 9 nondiagnostic case (11%). Two nondiagnostic CB procedures were followed by open surgery. Chronic osteomyelitis was seen in curettage from a vertebral lesion that only showed cortical bone with CB. The other nondiagnostic CB was an ulna lesion demonstrated by surgical excision as desmoid fibromatosis. This was a sampling problem with CB because only cortical bone was obtained at CB. In the indeterminate category (Table 3), a cartilaginous lesion was diagnosed as grade 1 chondrosarcoma by excision. Three nondiagnostic OB procedures were followed by FNACBP, including diffuse large B-cell lymphoma involving the left tibia (Figure 2, A through G), which also had an intraoperative frozen section taken to check for specimen adequacy, and a parosteal osteosarcoma involving the right femur (Figure 3, A through H) after 2 prior nondiagnostic OBs. The third case was a welldemarcated lytic acetabular lesion. The FNACBP identified benign spindle cells, and the subsequent additional OB remained nondiagnostic with benign bone and marrow elements only. A later resection at another institution identified the lesion as a benign bone cyst. Figure 2. Tibial lesion in a 14-year-old boy. A, X-ray of the left knee with a metaphyseal lytic lesion. B, Magnetic resonance imaging showing a heterogeneous tumoral mass centered in the proximal tibial metadiaphysis with extension into the proximal tibial epiphysis. There was extensive cortical breach with a large, posterior, soft tissue component (right). Open biopsy of left tibia lesion demonstrated benign fibrovascular connective tissue (C) and benign cortical bone (D). E, Aspirate smear was cellular, with individual and clusters of discohesive, intermediate to large cells with irregular nuclear contours, frequent distinct nucleoli, and high nuclear to cytoplasmic ratio. Core needle biopsy with a monomorphic solid pattern of predominantly large-sized lymphoid cells (F), which were immunopositive for PAX-5 (G), confirming the diagnosis of a diffuse large B-cell lymphoma (hematoxylin-eosin, original magnifications 340 [C and D] and 3200 [F]; Diff-Quick, original magnification 3200 [E]; original magnification 3200 [G]). Arch Pathol Lab Med Vol 141, May 2017 Biopsy for Pediatric Osseous Lesions Patel et al 681

5 Figure 3. Femur lesion in a 14-year-old boy. A, X-ray of the lateral right knee with aggressive periosteal reaction, Codman triangle, and bone destruction along the posterior cortex of the metaphysis of the distal femur. B, Magnetic resonance imaging and sagittal T1-signal fat saturated image after contrast administration showing intermediate signal intensity of the tumor with enhancement of the posterior soft tissue component. Open biopsy showed benign fibroadipose tissue (C) and benign lamellar bone (D). E, Aspirate smear with rare clusters of spindle cells (inset) and large, atypical osteoblastic cells (white circles) in a background of peripheral blood. F, Core needle biopsy demonstrated well-formed, parallel bony trabeculae in a hypocellular stroma. G, The hypocellular stroma invaded into skeletal muscle (black arrow head). Circled area is shown in higher magnification in panel H. H, High-power view of the circled area in panel G, showing low- to intermediate-grade spindle cell lesion with lacy osteoid (arrow), characteristic of osteosarcoma, parosteal variant. Black arrowheads highlight entrapped skeletal muscle fibers (hematoxylin-eosin, original magnifications 340 [C, D, F, and G] and 3400 [H]; modified Giemsa, original magnification 3600 [E]). DISCUSSION Most studies evaluating the utility of FNA as a diagnostic procedure for osseous lesions have involved predominantly adult patients, in which secondary (metastatic) lesions are more common. In prior studies, FNA alone is reported to have a diagnostic accuracy ranging from 80% 2,8,9 to close to 90% with osseous lesions. Accuracy is reported as being greater for bone metastases possibly because of prediagnostic clinical suspicion and characteristic radiologic manifestations. Additionally, histomorphologically, tumor cells tend to have a distinct cytomorphology (often epithelial neoplasm in adults) that differs from the native bone elements. Compared with adults, the pediatric population tends to have vastly different disease categories with respect to osseous lesions. In this regard, the optimal initial diagnostic procedure has not been studied extensively. 1,13 We reviewed the initial sampling procedures retrospectively for pediatric osseous lesions and focused on the unique combination of FNACBP. We also addressed the adequacy and diagnostic accuracy, clinical utility, and predictive value of FNACBP. In our study, FNACBP had an overall clinical utility of 93%. The sensitivity, specificity, and positive predictive value were 100% for malignant osseous lesions in this study with a limited sample size. In contrast, the sensitivity was 88% for benign lesions. Of the clinically useful, but not accurate, cases, the patient with a tibial lesion had a prior sciatic nerve biopsy from that area showing MPNST. The clinical and radiologic differential diagnoses for the tibial lesion included direct extension of the adjacent MPNST, metastatic MPNST, and a primary osseous malignancy, such as osteosarcoma. Hence, a guarded FNACBP diagnosis of spindle cell sarcoma was rendered. The other 2 cases were benign entities, a bone cyst and an osteoid osteoma, which were diagnosed as a benign spindle cell lesion and a benign osteoblastic lesion, respectively. Most of the published literature considers the FNA diagnosis of small round cell sarcoma or spindle cell sarcoma as accurate when the resection specimen shows Ewing sarcoma or MPNST, respectively. Similarly, a diagnosis of benign osteoblastic lesion would be considered accurate for an osteoid osteoma. However, we considered cases accurate only when the FNACBP diagnosis matched exactly with that obtained at resection. Despite using such a stringent criterion, our accuracy rate at 83% was comparable to the published literature. 2,3,9,11,12 A recent meta-analysis 14 of both bone and soft tissue tumors suggested that CB was more accurate than FNA, and incisional (open) biopsy appeared to be more accurate than both of those techniques; however, the differences did not reach statistical significance. Our results support the findings of the study by Pohlig et al, 15 which found percutaneous biopsy techniques (fine-needle aspiration and core needle biopsy) yielded slightly superior, but not statistically significant, results compared with OB. With the 31 lesions sampled by FNACBP as the initial diagnostic procedure in this study, 97% were diagnostic, compared with 79% and 78% for CB and OB, respectively. This might be partially due to neoplastic lesions more-frequently undergoing FNACBP, along with a relatively large propor- 682 Arch Pathol Lab Med Vol 141, May 2017 Biopsy for Pediatric Osseous Lesions Patel et al

6 tion of malignant entities (51% versus 42% and 22% for CB and OB, respectively) that are often hypercellular and morphologically distinct. In the current study, FNACBP had slightly less sensitivity for benign lesions at 88% compared with malignant lesions. Additionally, smallersized lesions sampled with CB may also partially explain the less-desirable diagnostic rate of CB. The trend toward increasing FNACBP use at our institution in the past 6 years is a testament to the efficacy of FNACBP. Moreover, accurately diagnosing 2 lesions (diffuse large B-cell lymphoma and parosteal osteosarcoma) that were nondiagnostic by OB further supported the utility of FNACBP in diagnosing pediatric osseous lesions. Because of radiologic heterogeneity that resulted from frequent necrosis in malignant bony lesions and hemorrhage and cystic degeneration in benign lesions, CB or FNA alone may not be diagnostic. However, when combined, they act synergistically to increase the yield and overall efficacy. This effect is markedly enhanced when the procedure is associated with on-site adequacy evaluation by a pathologist to ensure that adequate diagnostic and prognostic material has been obtained. As shown in this study, only 2 FNACBP procedures required follow-up surgical procedures for diagnosis. The utility of concurrent FNA with core biopsy in improving the diagnostic yield was also shown in a prior study of 144 skeletal lesions, including children. In that study, 24% of core biopsies were nondiagnostic by themselves but, when accompanied by FNAs, provided diagnostic samples. 1 Complications from percutaneous techniques for musculoskeletal lesions have a reported incidence of about 1%, less than that of OB Importantly, the rates of altered treatment and altered outcome as a result of needle biopsy were considerably less than those for OB. 14,19 In our institution, IR generally has greater availability (physician, IR room time), decreased procedure duration, and less nurse staffing requirement, when compared with surgical incisional or excisional biopsy requiring the use of an operating room. Most of our CBs or FNACBPs were performed with ultrasound and/or fluoroscopy guidance, which has less radiation exposure than computerized tomography guidance does. Active discussions between the pathologist and the interventional radiologist regarding the differential diagnoses based on clinical and radiologic features, the consistency of the lesion as determined during the aspiration/biopsy needle procedure, and the real-time aspirate smear morphology assessment by the pathologist narrow the differential diagnosis and limit unnecessary testing. Typically, an orthopedic surgeon would mark the biopsy site in the IR suite, which avoids violating tissue planes and ensures inclusion of the biopsy needle tract when the patient undergoes a subsequent resection of the lesion. An important highlight of this study is the availability of adequate material for routine and advanced ancillary testing including cytogenetic and molecular analysis. To our knowledge, this is the first study focusing on the use of FNACBP in the diagnosis of pediatric osseous lesions. The current study highlights the clinical utility of FNACBP as the initial diagnostic procedure of choice in pediatric osseous lesions, particularly in presumably malignant lesions. In our experience, imaging guidance combined with on-site evaluation increases the adequacy and diagnostic accuracy of bone biopsies and, therefore, reduces the potential risk of complications and repeat biopsies. Overall, a team approach that integrates efforts from pathologists, interventional radiologists, and orthopedic surgeons provides an optimal diagnosis and management of pediatric osseous lesions. We thank Karen Prince for her help with graphic work. References 1. Koscick RL, Petersilge CA, Makley JT, Abdul-Karim FW. CT-guided fine needle aspiration and needle core biopsy of skeletal lesions: complementary diagnostic techniques. Acta Cytol. 1998;42(3): El-Khoury GY, Terepka RH, Mickelson MR, Rainville KL, Zaleski MS. Fineneedle aspiration biopsy of bone. J Bone Joint Surg Am. 1983;65(4): Treaba D, Assad L, Govil H, et al. Diagnostic role of fine-needle aspiration of bone lesions in patients with a previous history of malignancy. Diagn Cytopathol. 2002;26(6): Cole CD, Wu HH. Fine-needle aspiration in pediatric patients 12 years of age and younger: a 20-year retrospective study from a single tertiary medical center. Diagn Cytopathol. 2014;42(7): D Anza B, Kraseman SJ, Canto-Helwig C, Greene JS, Wood WE. FNA biopsy of pediatric cervicofacial masses and validation of clinical characteristics of malignancy. Int J Pediatr Otorhinolaryngol. 2015;79(8): Eisenhut CC, King DE, Nelson WA, Olson LC, Wall RW, Glant MD. Fineneedle biopsy of pediatric lesions: a three-year study in an outpatient biopsy clinic. Diagn Cytopathol. 1996;14(1): Silverman JF, Gurley AM, Holbrook CT, Joshi VV. Pediatric fine-needle aspiration biopsy. Am J Clin Pathol. 1991;95(5): Kabukçuoğlu F, Kabukçuoğlu Y, Kuzgun U, Evren I. Fine needle aspiration of malignant bone lesions. Acta Cytol. 1998;42(4): Layfield LJ, Schmidt RL, Sangle N, Crim JR. Diagnostic accuracy and clinical utility of biopsy in musculoskeletal lesions: a comparison of fine-needle aspiration, core, and open biopsy techniques. Diagn Cytopathol. 2014;42(6): Agarwal PK, Goel MM, Chandra T, Agarwal S. Predictive value of fine needle aspiration cytology of bone lesions. Acta Cytol. 1997;41(3): 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(3): Jorda M, Rey L, Hanly A, Ganjei-Azar P. Fine-needle aspiration cytology of bone: accuracy and pitfalls of cytodiagnosis. Cancer. 2000;90(1): Handa U, Bal A, Mohan H, Bhardwaj S. Fine needle aspiration cytology in the diagnosis of bone lesions. Cytopathology. 2005;16(2): Traina F, Errani C, Toscano A, et al. Current concepts in the biopsy of musculoskeletal tumors: AAOS exhibit selection. J Bone Joint Surg Am. 2015; 97(2):e Pohlig F, Kirchhoff C, Lenze U, et al. Percutaneous core needle biopsy versus open biopsy in diagnostics of bone and soft tissue sarcoma: a retrospective study. Eur J Med Res. 2012;17: Moore TM, Meyers MH, Patzakis MJ, Terry R, Harvey JP Jr. Closed biopsy of musculoskeletal lesions. J Bone Joint Surg Am. 1979;61(3): Welker JA, Henshaw RM, Jelinek J, Shmookler BM, Malawer MM. The percutaneous needle biopsy is safe and recommended in the diagnosis of musculoskeletal masses. Cancer ;89(12): Skrzynski MC, Biermann JS, Montag A, Simon MA. Diagnostic accuracy and charge-savings of outpatient core needle biopsy compared with open biopsy of musculoskeletal tumors. J Bone Joint Surg Am. 1996;78(5): Mankin HJ, Mankin CJ, Simon MA; for Members of the Musculoskeletal Tumor Society. The hazards of the biopsy, revisited. J Bone Joint Surg Am. 1996; 78(5): Arch Pathol Lab Med Vol 141, May 2017 Biopsy for Pediatric Osseous Lesions Patel et al 683