IDc

Ma

b

a

ARR1A

KPECSB

1

oalPpsu

a

H2

h

0h

European Journal of Radiology 83 (2014) 360– 365

Contents lists available at ScienceDirect

European Journal of Radiology

journa l ho me p age: www.elsev ier .com/ locate /e j rad

nitial CT-guided needle biopsy of extremity skeletal lesions:iagnostic performance and experience of a tertiary musculoskeletalenter

ohamed Ragab Nouha,b,∗, Hamdy Mohamed Abu Shadyb,1

Department of Radiology, Faculty of Medicine, Alexandria University, EgyptDepartment of Radiology, Al-Razi Hospital, Sulibikhate 13001, Kuwait

r t i c l e i n f o

rticle history:eceived 18 April 2013eceived in revised form6 September 2013ccepted 12 October 2013

eywords:ercutaneousxtremityT-guidedkeletaliopsy

a b s t r a c t

Introduction: Appendicular long bones are the target for a wide spectrum of bony lesions with variableclinical presentations. Biopsy procedures are needed for subsequent proper patient’s management. Mostof the available literature globally assessed musculoskeletal biopsies with inclusion of repeat biopsyresults.

We thought to retrospectively assess the diagnostic performance of initial CT-guided percutaneouscore needle biopsy (PCNB) of extremity long bone lesions in a tertiary musculoskeletal referral center.Patients and methods: We retrospectively analyzed the outcome of initial CT-guided PCNB of 49 patientswho presented with extremity long bone lesions which were biopsied in our hospital during a 36 months’time period. The diagnostic performance was assessed in terms of diagnostic yield and accuracy.Results: There were 34 males and 15 females with a mean age of 33.69 years (range from 4 to 77 years).The overall diagnostic yield of initial biopsies was 87.75% with a diagnostic accuracy of 82.85% derivedfrom the surgically proven cases. The higher diagnostic yield was recorded with malignancy, presence

of extra-osseous soft-tissue component as well as mixed and sclerotic lesions. The pathologies of thenon-diagnostic biopsies included large-cell lymphoma, giant-cell tumor, langerhans cell histiocytosis,osteoid osteoma and a non-ossifying fibroma.Conclusion: Initial CT-guided PCNB in extremities’ long bones lesions showed high diagnostic performancein malignant, mixed and/or sclerotic lesions as well as lesions with extra-osseous exophytic tissue growth.Lack of extra-osseous components, benign and lytic lesions all had worse diagnostic performance.

. Introduction

Appendicular long bones are the target for a wide spectrumf bony lesions encompassing benign and malignant bone tumorss well as osteomyelitis, reactive focal abnormalities and tumor-ike lesions of developmental and/or metabolic origin [1,2].ain, swelling and/or pathologic fractures are variable clinicalresentations that usually precede the incidental discovery ofuch lesions on routine radiography, the mainstay of their pick

p [1,2].

In spite of recent advances in diagnostic imaging tools andpplications for musculoskeletal lesion characterization [3], these

∗ Corresponding author. Current address: Department of Radiology, Al-Raziospital, Sulibikhate 13001, Kuwait. Tel.: +965 24825909/65099562; fax: +9654825508.

E-mail addresses: mragab73@yahoo.com, mragab73@gmail.com (M.R. Nouh),amdi abushadi@yahoo.com (H.M. Abu Shady).1 Tel.: +965 66199459; fax: +965 24825508.

720-048X/$ – see front matter © 2013 Elsevier Ireland Ltd. All rights reserved.ttp://dx.doi.org/10.1016/j.ejrad.2013.10.012

© 2013 Elsevier Ireland Ltd. All rights reserved.

lesions remain a daily diagnostic challenge for the radiologist.Hence, biopsy procedures are needed to ascertain the histopatho-logic nature of these lesions for proper patient’s management [4–7].

Recent work [8] assessed the results of imaging-guided per-cutaneous core needle biopsy (PCNB) in pathologic fracture ofthe appendicular skeleton. Other common clinical presentationsof bone lesions needing biopsy include the following: a swollenpainful limb in a child or young adolescent, incidental discovery ofa bony lesion in patients presenting with non-specific extremitypain, or patients treated for non-skeletal neoplasia with recentlyevolving extremity pain [1,2]. Moreover, most of the available lit-erature was non-selective including study results for both initialand repeat biopsies as well as both bony and soft-tissue lesions[6,7,9–11]. To our knowledge, no description of diagnostic perfor-mance of first time image-guided PCNB of extremity long bone

lesions is available in the English literature.

Thus, we thought to assess the diagnostic performance of theinitial CT-guided PCNB of extremity long bone lesions through aretrospective audit at our tertiary musculoskeletal referral center.

n Jour

2

2

cllJ

dcsimtTwcp

td

2

tmo

cbh“t

2

psl

2

Miobbmba

2

ai

otv

t

M.R. Nouh, H.M. Abu Shady / Europea

. Patients and methods

.1. Study design and research ethics

We retrospectively analyzed the outcome of initial CT-guidedore-needle biopsies of patients who presented with extremityong bone lesions performed in our hospital, the tertiary muscu-oskeletal referral center in the country, during the period fromanuary 2010 to December 2012.

In our hospital, the decision to biopsy a patient is usuallyiscussed between the referring orthopedic surgeons and the mus-uloskeletal radiologists in view of clinical data, imaging findings,uspected diagnosis and treatment options. We believe that thenterdisciplinary approach will provide optimization of patient

anagement. Hence, the preferred approach, biopsy instrumen-ation and sampling site is individually chosen for each patient.he biopsy procedure indications, benefits and risks are discussedith the patients and/or their guardians and an informed written

onsent is obtained from all patients before commencing biopsyrocedures.

Institutional ethics committee approval was not required forhis study as data used did not breach patient confidentiality orisclose their identifying data.

.2. Study population inclusion and exclusion criteria

Patients presented with (a) pathologic fractures (b) inciden-ally discovered bony lesions suspected to be sarcomatous or (c)

etastatic in nature; either from a known extra-osseous primaryr not, were included in our retrospective audit.

Meanwhile, we excluded patients with (a) lesions completelyonfined to the extremities’ soft-tissues (b) a final diagnosis that haseen reached through repeated image-guided biopsies (c) lesionsaving typical imaging and demographic features consistent withleave me alone lesions”; where a biopsy would not provide addi-ional information or change patient management.

.3. Patients’ demographics and clinic-pathologic data

Medical records and pathologic reports were reviewed to recordatients’ demographics, relevant clinical history, managementtrategies used as well as image-guided and/or surgical histopatho-ogic biopsy results

.4. Lesions’ topography and suggested matrix characteristics

The patient’s imaging workup, including radiography, CT andRI, was used to characterize the lesion to be biopsied categoriz-

ng them according to: (a) location of the host bone, (b) presencer absence of extra-osseous soft-tissue component (mainly definedy MR), and (c) imaging nature of lesion’s matrix (mainly definedy CT) that was described as: (1) lytic (more lysis than normaledullary bone), (2) sclerotic (more dense than normal medullary

one), and (3) mixed lytic and sclerotic (lesions with both densend lucent components) (Fig. 1).

.5. Biopsy guidance and procedures

All biopsies were performed; under CT guidance; by one of theuthors with several years of experience in musculoskeletal imag-ng and procedures.

Prior to the procedure, bleeding parameters as well as other lab-ratory markers were done. An anesthesiologist electively checked

he patient and attended the procedure to monitor the patients’ital signs throughout the procedure.

The majority of patients were in a state of conscious sedation sohat they could lie down comfortably and as pain-free as possible

nal of Radiology 83 (2014) 360– 365 361

during the procedure. Only children and apprehensive patientsrequired general anesthesia.

On procedure start, the patient was usually positioned on CTtable in a practical way to access the lesion to be biopsied andmaximize comfort for both the patient and operator.

Following an initial CT scan through the lesion, the lesion biopsytrajectory was chosen as per the prior interdisciplinary conferencediscussion. This is to ensure that uninvolved compartments wouldnot be violated during the biopsy procedure and biopsy tract exci-sion can be performed during definitive surgery if appropriate.Furthermore, we considered the previously described guidelinesgoverning image-guided core needle biopsies of extremity longbone lesions when appropriate [12,13].

A strictly aseptic technique is followed as in any invasive proce-dure. Local anesthetic infiltration; with 1% lidocaine; was deliveredalong the recommended biopsy needle trajectory to ensure bet-ter post-procedure tolerance by the patient. The biopsy procedureswere performed using standard coaxial technique to allow multiplepasses through a single skin puncture, and a single track.

A 12–14 gauge bone biopsy needle (TrueGuid® disposable coax-ial needle/AngioMed/Bard, Karlsruhe, Germany) was advancedthrough the planned biopsy trajectory to the targeted site. Afterthat the sharp introducer was removed leaving the needle in placeto act as a monorail for further biopsy instruments passage.

A 16- or 18-gauge automated biopsy gun system (Monoptybiopsy system, Bard, Temp, AZ, USA) is sufficient in obtaining biop-sies of lesions without intact cortex and lesions with exophyticgrowth into the surrounding soft tissues.

Lesions with intact bony cortex were accessed by the drillingaction of a 12–15 gauge Ostycut (OstycutTM/AngioMed/Bard, Karl-sruhe, Germany) biopsy needle with sharp introducer. Sometimes,gentle tapping with an orthopedic hammer over the Ostycut wasneeded for more sclerotic lesions.

If a lesion was predominately lytic in nature, after passingthrough the cortex, aspiration samples were gathered for cytol-ogy, biochemistry and bacteriology evaluation. After that, a 16–18biopsy needle or an Ostycut bone needle was passed through thesheath to curette the wall of the lesion and obtain tissue samples.

In general, our standard practice was to gather three to sixcore-biopsy specimens per procedure except in small lesionsand osseous lesions at risk for fracture with excess manipula-tion which was left for the judgment of the interventionalist. Asubjective assessment of sample adequacy was made by visualinspection at the time of biopsy. Biopsied tissue fragmentswere placed in 10% buffered formaldehyde solution and sent tohistopathology accompanied by a pathology request form provid-ing relevant clinical information including age, sex, involved bone,number of lesions, presence of a pre-existing lesion and provi-sional imaging differential diagnosis concluded in pre-procedurediscussions.

Post-procedural CT scans were acquired to ensure absence ofcollections or vascular injuries. The patient was then kept underobservation for a median of 2 h before discharge, for any potentialcomplications.

2.6. Biopsy results terminologies/evaluation

The initial CT biopsy sample was considered diagnostic whencomplied with one of the following criteria:

(a) Concordant with the results of a post-surgical resection speci-

men; or

(b) Pathologic examination resulted in a distinctive pathologicdiagnosis revealing unchanged imaging features and clinicalstatus on a 12–24 months follow-up.

362 M.R. Nouh, H.M. Abu Shady / European Journal of Radiology 83 (2014) 360– 365

F owing a lytic proximal humerus lesion in (a), a sclerotic tibial lesion in (b) and a mixedl

n

(

bd

b

Oiopsieerfor

bt

De bioed in

3

igbs

Table 1Clinical presentation among current study population presented for CT-guidedbiopsy of their extremities’ long bones lesions.

Salient clinical presentation Patients number Percentage

Pain 31 63.27%Swelling 8 16.33%Incidental discovery 6 12.24%

ig. 1. Intra-procedural axial CT images of different extremity long bone lesions shytic/sclerotic distal femur lesion in (c).

On the other hand, the initial CT biopsy sample was defined ason-diagnostic (ND) when:

(a) The histological diagnosis was discordant with the results of apost-surgical biopsy, or

b) The specimen volume was judged inadequate by the patholo-gist to reach a specific diagnosis.

The diagnostic performance; of initial CT-guided percutaneousiopsy of extremity long bone lesions; was assessed in terms ofiagnostic yield and accuracy.

Diagnostic yield was defined as the percentage of true positiveiopsies of the total number of biopsies.

verall diagnostic yield = True positive CT-guided percutaneous needle bTotal number of p

Diagnostic accuracy was defined as the concordant histologyetween both needle and surgical biopsy after exclusion of caseshat had been managed non-surgically following needle biopsy.

iagnostic accuracy = True positive CT-guided percutaneous needlTotal number of performed biopsies includ

. Results

As per our inclusion and exclusion criteria, the current study

ncluded 49 subjects (34 males and 15 females) who underwent CT-uided needle biopsy of their extremity long bone skeletal lesionsetween January 2010 and December 2012. The mean age of ourtudy population was 33.69 years (range from 4 to 77 years).

s confirmed surgically and with non-surgical 1–2 years follow-upmed biopsies included in the study

× 100

psies concordant with the post-surgical biopsy results the study − the non-surgically confirmed biopsies

× 100

Pathologic fracture 4 8.16%Total 49 100.00%

The clinical presentations in current study were pain (63.27%),swelling of the affected site (16.33%), incidental discovery on imag-ing workup (12.24%) and pathologic fractures (8.16%) (Table 1).

Cross sectional imaging studies revealed an extra-osseous com-ponent in (26/49) patients. This facilitated biopsy and obviatedthe need for cortical breach. The remaining 23 patients had acompletely intra-osseous lesion mandating a more aggressive pro-

cedure to reach the lesion.

No biopsy-related complications (e.g. hematoma or infection)were recorded among our current study population. The most com-mon long bone to be biopsied in our series was the femur (37%)

M.R. Nouh, H.M. Abu Shady / European Journal of Radiology 83 (2014) 360– 365 363

Table 2Site-distribution and diagnostic yield of initial CT-guided biopsied lesions of extremities’ long bones.

Extremity Long bone No. of patients Diagnostic samples ND samples Diagnostic yield

UpperHumerus 14 (29%) 14 0 100.00%Radius 3 (6%) 2 1 66.67%

LowerFemur 18 (37%) 15 3 83.33%Tibia 11 (22%)

Fibula 3 (6%)

Total 49 (100%) 4

Table 3Initial pathologic results of diagnostic CT-guided PCNB procedures of extremities’long bones.

Tumor type Number

Malignant lesions; n = 33Marrow cells neoplasia and metastasis

Plasmacytoma 1Lymphoma 1Thyroid 1Lung 2Breast 2Kidneys 1

Primary malignant tumors of the boneOsteosarcoma 7Chondrosarcoma 6GCT 6Round cell tumor (PNET/Ewing’s) 6

Benign lesions; n = 10Simple cyst 3Chondroblastoma 1Osteochondroma 2Gouty Cyst 1Langerhans cell histiocytosis 2

fw

3b

oy

sdoep

TR

Normal bone 1Total 43

ollowed by the humerus (29%) and tibia (22%). No ulnar lesionsere detected in our patients (Table 2).

.1. Overall diagnostic yield of initial CT-guided percutaneousiopsy

Initial percutaneous biopsy provided diagnostic specimens in 43f 49 CT-guided percutaneous biopsies with an overall diagnosticield of 87.75% (Table 3).

Following the CT-guided percutaneous biopsy, 6 cases were con-idered non-diagnostic (Table 4); in five of them the pathologists

escribed the provided specimen composition as areas of hem-rrhagic components, necrotic tissues and/or abundant in fibrouslements. Hence, they considered the specimen inadequate torovide specific tissue diagnosis and recommended a larger tissue

able 4esults of the ND cases following CT-guided PCNB.

Post-percutaneous histopathologycomment

Post-surgicalhistopathologydiagnosis

Number ofcases

Abundant fibrous stroma with foci ofhemorrhage

GCT 1

Disorganized trabecular bones withfew necrotic cells

Osteoid osteoma 1

Disorganized trabecular bones withhemorrhagic clots and necrotic tissues

Lymphoma (large-celltype)

1

Trabecular bones with fibrous stromaand scanty fibroblasts

Non-ossifying fibroma 1

Aggregates of inflammatory cells withscanty spindle cells

Langerhans cellhistiocytosis

1

Fibrous stroma with scanty spindle andcells. . .suggestive of fibrous dysplasia

Adamantinoma 1

Total 6

9 2 81.82%3 0 100.00%3 6 87.76%

core from a surgical specimen. These cases were finally diagnosedas: large-cell lymphoma, giant-cell tumor, langerhans cell his-tiocytosis, osteoid osteoma and a non-ossifying fibroma. In theremaining case, a tibial lesion was diagnosed as fibrous dyspla-sia following CT-guided percutaneous biopsy which proved to beadamantinoma on the post-surgical pathology report.

3.2. Diagnostic yield according to the site of the lesion

Stratifying the CT-guided percutaneous biopsy results of sam-pled extremity long bone lesions by the bone hosting the lesion,the highest initial diagnostic yield was for the humerus and fibulawhile the lowest diagnostic yield was for radius (Table 2).

3.3. Diagnostic yield according to the major diagnostic category

For the metastatic bony lesions, 6/6 cases were confirmed on ini-tial percutaneous biopsies and histological features suggested theirprimary site of origin with a diagnostic yield of 100%. The initial per-cutaneous biopsies successfully obtained the correct diagnosis in:2/3 cases with hematologic malignancies, 29/32 cases of primarymalignant bone tumors and 10/12 of non-malignant bone processesin our study resulting in diagnostic yields of 66.6%, 90.6% and 83.3%;respectively.

3.4. Diagnostic yield according to imaging study findings

Stratifying the CT-guided percutaneous biopsy results of sam-pled extremities long bone lesions by the descriptive X-ray andCT nature of the lesion, the initial diagnostic yield was higherfor the mixed lesions (16/17 = 94.12%) followed by the sclerotic(8/9 = 88.89%) and lytic (19/23 = 82.61%) lesions.

Samples from completely intra-osseous lesions gave a diag-nostic yield of 73.91% (17/23). On the contrary, presence ofextra-osseous soft-tissue components raised the diagnostic yieldto 100% (26/26).

3.5. Diagnostic accuracy of initial CT-guided percutaneous biopsy

Of the 43 cases with diagnostic samples, the biopsy results alongwith the clinical and unchanged imaging features over a 18–24months follow-up mitigated the need for further investigation in14 patients. Hence, these cases were managed non-surgically asappropriate. This included 6 cases of metastatic bone disease, 2cases of langerhans cell histiocytosis and one case for each of thefollowing diagnoses: lymphoma, plasmacytoma, simple bone cyst,sessile osteochondroma, stress fracture reparative response andgouty cyst.

Following exclusion of these non-surgically managed 14patients, 29 cases, in addition to the 6 ND cases, underwent surgi-cal management and open biopsy procedures. In those 29 patients,the pathologic tissue type was concordant with the CT-guided per-

cutaneous biopsy. Five ND cases with pathologic descriptions ofspecimen inadequacy had the diagnoses of: large-cell lymphoma,giant-cell tumor, langerhans cell histiocytosis, osteoid osteoma anda non-ossifying fibroma. In the remaining case, a tibial lesion was

3 n Jour

dbp

d(

4

moo

rbsbl

bts

oppa

flap

cfi

s

mh[sb

pwasf

bghmrLn

ctap[

64 M.R. Nouh, H.M. Abu Shady / Europea

iagnosed as fibrous-dysplasia following CT-guided percutaneousiopsy while it had proved to be adamantinoma on post-surgicalathology report.

Taking post-surgical biopsy results as our gold standard, theiagnostic accuracy of initial CT-guided percutaneous biopsy was29/35) 82.85%.

. Discussion

The image-guided percutaneous biopsy has proved to be a safe,inimally invasive and cost-effective procedure for the diagnosis

f skeletal lesions obviating the need for the more risky and invasivepen surgical biopsy in most patients [4–11].

Meticulous technique regarding compartmental barriers, neu-ovascular structures and seeding of malignant cells along theiopsy needle track are pre-requisites for optimized treatmenttrategies with limb-salvage procedures [12,13]. CT has proven toe a recognized modality for guiding PCNB of deep-seated muscu-

oskeletal lesions [10,14,15].Pain, hematoma, and infection are potential complications for

iopsy procedures [4,5,9,16,17]. Tumor seeding along the biopsyract is another possible complication. However; a recent reportuggested that the potential is very low or negligible [18].

In agreement with previous literature data [19–21], none ofur study population showed biopsy-related complications. Properatient selection and attentive adherence to established princi-les of musculoskeletal biopsy [12,13] made this outcome possible,ccentuating the value of CT-guided PCNB procedures.

Our current retrospective audit described the diagnostic per-ormance of the initial CT-guided PCNB of extremities’ long boneesions in our center over a three-year period. It showed an over-ll diagnostic yield of 87.75%. However; exclusion of non-surgicallyroven cases resulted in a diagnostic accuracy of 82.85%.

This compares well with the published diagnostic yield of per-utaneous musculoskeletal biopsies falling in the range of 69–98%rom different musculoskeletal regions with most of these studiesncluding repeat biopsy data in their results [4,6–11,14–17,19–21].

Our results are exclusively limited to the data of initial biopsyampling of the extremity long bone lesions.

While performing this audit, we recognized the femur as theost common involved extremity long bone followed by the

umerus and tibia. This corresponds to previous literature reports6,8,14,19,20] and affirms the known distribution of primary bonearcomas around the metabolically active growth plates of theseones.

The highest diagnostic yield for metastatic bone disease andrimary bone malignancy categories in our series is in keepingith a number of previous reports [6,7,10,15,22,23]. This could be

ttributed to the compact nature of malignant neoplasm’s cells andtroma so that a true-cut biopsy core will be mostly comprehensiveor their tissue diagnosis [20,24].

In the current study, the non-diagnostic cases varied betweenenign lesions (osteoid osteoma and non-ossifying fibroma), low-rade malignancies (adamantinoma and giant cell tumor), systemicematologic malignancy (large-cell lymphoma) and the indeter-inate entity of Langerhans cell histiocytosis. Their pathologic

eports described the presence of fibrous tissue, necrosis, or blood.ikewise, similar tissue compositions were associated with falseegative results in previous works [7,9,14,20,22].

These results could be attributed to multiple factors: first, inase of benign mesenchymal lesions the dominant tissue archi-

ecture is few cells imbedded in an abundant loose stroma,n arrangement that necessitates a substantial tissue sam-le to conclude precise histological typing by the pathologist19,25,26]. Similarly, low-grade tumors like giant-cell tumor and

nal of Radiology 83 (2014) 360– 365

admantinoma show heterogenic cells with non-specific morphol-ogy which require the assessment of tissue structure rather thancellular typing. Second, hematologic malignancies are formed ofloose tissues; some of them may be necrotic as in large-celllymphomas, or more likely prone to damage during patho-logic specimen preparation interfering with their phenotypingby immune-histochemistry [2,6,23]. Thirdly, both lymphoma andhistiocytosis lack specific imaging features “great mimickers”along with the previous contemplations represent diagnosticchallenges for both the radiologist and pathologist [23]. Lastly,all non-diagnostic samples in our series were for completelyintra-osseous lesions. This may represent a potential limitation,hampering adequate tissue sampling of different parts of thelesions [19,20,24].

The matrix of a lesion, as suggested by its imaging features ondifferent imaging modalities, is a major determinant for biopsyinstrumentation choice and percutaneous biopsy decision.

Our work revealed higher diagnostic yield for lesions of mixedmatrices followed by sclerotic lesions and lytic lesions. Our resultsagree with the work of Jelinek et al. [9] who found high diagnosticyield for the sclerotic lesions. Contrarily, previous works [6,19,20]reported higher diagnostic results in case of lytic and mixed lesionscompared to sclerotic lesions. While Omura et al. [24] found no sig-nificant difference in the diagnostic yield of these different lesions;the difference was significant in a recent [26] retrospective evalua-tion of CT features associated with indeterminate musculoskeletalbiopsy results.

We believe that good diagnostic yield for sclerotic lesions inour study is explained by orthopedic-radiologist collaborative com-munication to decide which part to be biopsied, providing thepathologists with relevant clinical and imaging data as well asimprovement in biopsy instrumentation; such as cutting needlesand orthopedic hammers that made biopsying these lesions easier.However; the small size of our study population is a notable bias.

The presence of extra-osseous soft-tissue component orextremely attenuated cortex around the lesion facilitates biopsy,makes the lesion amenable to biopsy and is associated with higherdiagnostic yield [8,10,24,26,27].

Our results agree with those data where all the six non-diagnostic cases were associated with lack of extra-osseous softtissue component.

Wu et al. [19] recommended obtaining three cores from bonylesions and four cores from soft-tissue lesions for optimized tis-sue sampling. Although we did not record the number of harvestedbiopsies, our routine is to get a minimum of three biopsies in everycase and reached six in some cases. This, we postulate, was a con-tributor to the higher diagnostic yield in our series and is supportedby results of previous works [19].

The current work showed 82.85% diagnostic accuracy for theinitial CT-guided PCNB in extremity long bone lesions. The highaccuracy in our series could be attributed to (a) appropriateselection of the procedure technique because of effective communi-cation between the requesting orthopedican and interventionalist,(b) state of conscious sedation allowing adequate number of biop-sies with acceptable cores, and (c) evolution in knowledge aboutbiopsy techniques in recent years with its tremendous effect onproper patient selection as well as marked improvement in biopsyinstrument technology.

Our study is inherently limited by its retrospective design. Thesmall sample size and lack of confirmatory pathologic tissues in asubset of 14 patients are added limitations to the current work. Weacknowledge that as we are a tertiary referral center, our results are

skewed by selection bias. Also our study is biased by the fact thatmultiple pathologists with different experiences were involved ininterpretation of the biopsies. They also, were not blinded to theimaging features of the biopsied tissues with its well established

n Jour

cdp

5

tlrtaeema

C

e

A

Am

R

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

M.R. Nouh, H.M. Abu Shady / Europea

riteria in bony lesions. This helped them to narrow the differentialiagnosis yet our results emphasize the success of this multidisci-linary practice.

. Conclusion

Despite these limitations, our study concludes a high diagnos-ic performance of initial CT-guided PCNB in extremity long boneesions especially in the settings of malignant, mixed and/or scle-otic lesions as well as extra-osseous exophytic tissue growth. Onhe other hand, there were some inherent limitations in benignnd/or cystic lesions where assessment of tissue architecture isssential for proper pathologic diagnosis. Additionally, our studymphasizes the importance of managing these patients in a tertiaryusculoskeletal center with an effective interdisciplinary team

pproach for best results.

onflict of interest

No sources of financial support and/or possible conflicts of inter-st to declare.

cknowledgement

The authors thank Dr. Deena I. Al-Refai, radiology consultant atl-Razi Orthopedic Hospital, Kuwait, for linguistic revision of theanuscript.

eferences

[1] Ilaslan H, Schils J, Nageotte W, Lietman SA, Sundaram M:. Clinical pre-sentation and imaging of bone and soft-tissue sarcomas. Cleve Clin J Med2010;77(Suppl):S2–7.

[2] Kant S, Lakhoo K. Primary bone tumours. In: Ameh EA, Bickler SW, editors.Paediatric surgery: a comprehensive text for Africa, vol. II. Seattle, WA, USA:Global HELP Organization; 2011. p. 633–8.

[3] Subhawong TK, Wang X, Durand DJ, et al. Proton MR spectroscopy inmetabolic assessment of musculoskeletal lesions. AJR Am J Roentgenol2012;198(1):162–72.

[4] Welker JA, Henshaw RM, Jelinek J, Shmookler BM, Malawer MM. The per-cutaneous needle biopsy is safe and recommended in the diagnosis ofmusculoskeletal masses. Cancer 2000;89:2677–86.

[5] Bickels J, Jelinek J, Shmookler B, Neff RS, Malawer MM. Biopsy of muscolu-loskeletal tumors. Current concepts. Clin Orthop Relat Res 1999;368:212–9.

[6] Vieillard MH, Boutry N, Chastanet P, Duquesnoy B, Cotten A, Cortet B. Contribu-tion of percutaneous biopsy to the definite diagnosis in patients with suspectedbone tumor. Joint Bone Spine 2005;72(1):53–60.

[7] Mitsuyoshi G, Naito N, Kawai A, et al. Accurate diagnosis of musculoskeletallesions by core needle biopsy. J Surg Oncol 2006;94:21–7.

[

nal of Radiology 83 (2014) 360– 365 365

[8] Datir A, Pechon P, Saifuddin A. Imaging guided percutaneous biopsy of patho-logic fractures: a retrospective analysis of 129 cases. AJR Am J Roentgenol2009;193:504–8.

[9] Jelinek JS, Murphey MD, Welker JA, et al. Diagnosis of primary bone tumors withimage-guided percutaneous biopsy: experience with 110 tumors. Radiology2002;223:731–7.

10] Hau A, Kim I, Kattapuram S, et al. Accuracy of CT-guided biopsies in 359 patientswith musculoskeletal lesions. Skeletal Radiol 2002;31:349–53.

11] Torriani M, Etchebehere M, Amstalden E. Sonographically guided core nee-dle biopsy of bone and soft tissue tumors. J Ultrasound Med 2002;21(3):275–81.

12] Liu PT, Valadez SD, Chivers FS, Roberts CC, Beauchamp CP. Anatomically basedguidelines for core needle biopsy of bone tumors: implications for limb sparingsurgery. Radiographics 2007;27:189–205.

13] Espinosa LA, Jamadar DA, Jacobson JA, et al. CT-guided biopsy of bone: a radi-ologist’s perspective. AJR Am J Roentgenol 2008;190(5):W283–9.

14] Altuntas AO, Slavin J, Smith PJ, et al. Accuracy of computed tomographyguided core needle biopsy of musculoskeletal tumours. ANZ J Surg 2005;75(4):187–91.

15] Puri A, Shingade VU, Agarwal MG, et al. CT-guided percutaneous core needlebiopsy in deep seated musculoskeletal lesions: a prospective study of 128 cases.Skeletal Radiol 2006;35(3):138–43.

16] Fraser-Hill MA, Renfrew DL. Percutaneous needle biopsy of musculoskele-tal lesions. 1. Effective accuracy and diagnostic utility. AJR Am J Roentgenol1992;158:809–12.

17] Dupuy DE, Rosenberg AE, Punyaratabandhu T, Tan MH, Mankin HJ. Accuracy ofCT-guided needle biopsy of musculoskeletal neoplasms. AJR Am J Roentgenol1998;171:759–62.

18] Saghieh S, Masrouha KZ, Musallam KM, et al. The risk of local recurrence alongthe core-needle biopsy tract in patients with bone sarcomas. Iowa Orthop J2010;30:80–3.

19] Wu JS, Goldsmith JD, Horwich PJ, Shetty SK, Hochman MG. Bone and soft-tissue lesions: what factors affect diagnostic yield of image-guided core-needlebiopsy? Radiology 2008;248:962–70.

20] Rimondi E, Rossi G, Bartalena T, et al. Percutaneous CT-guided biopsy ofthe musculoskeletal system: results of 2027 cases. Eur J Radiol 2011;77:34–42.

21] Pohlig F, Kirchhoff C, Lenze U, et al. Percutaneous core needle biopsy versusopen biopsy in diagnostics of bone and soft tissue sarcoma: a retrospectivestudy. Eur J Med Res 2012;17:29.

22] Logan PM, Connell DG, O’Connell JX, Munk PL, Janzen DL. Percutaneous biopsyof musculoskeletal tumors: an algorithm for selection of specific biopsy tech-niques. AJR Am J Roentgenol 1996;166:137–41.

23] Yang J, Frassica FJ, Fayad L, Clark DP, Weber KL. Analysis of nondiagnostic resultsafter image-guided needle biopsies of musculoskeletal lesions. Clin OrthopRelat Res 2010;468(11):3103–11.

24] Omura MC, Motamedi K, UyBico SD, Seeger LL. Revisiting CT-guided percu-taneous core-needle biopsy of muscloskeletal lesions: contributors to biopsysuccess. AJR Am J Roentgenol 2011;197(2):457–61.

25] McCarthy EF. CT-guided needle biopsies of bone and soft tissue tumors: apathologist’s perspective. Skeletal Radiol 2007;36:181–2.

26] Hwang S, Lefkowitz RA, Landa J, et al. Percutaneous CT-guided bone biopsy:diagnosis of malignancy in lesions with initially indeterminate biopsy results

and CT features associated with diagnostic or indeterminate results. AJR Am JRoentgenol 2011;197:1417–25.

27] Jakanani GC, Saifuddin A. Percutaneous image-guided needle biopsy of riblesions: a retrospective study of diagnostic outcome in 51 cases. Skeletal Radiol2013;42(1):85–90.