+Model

DIII-714; No. of Pages 5 ARTICLE IN PRESS

Diagnostic and Interventional Imaging (2015) xxx, xxx—xxx

ORIGINAL ARTICLE /Musculoskeletal imaging

Accuracy of core needle biopsy for the

diagnosis of osteosarcoma: A retrospective
analysis of 73 patients
T. Taupin a,∗, A.-V. Decouvelaere b, G. Vaz c, P. Thiesse a

a
Centre de lutte contre le cancer Léon-Bérard, service de radiologie, 28,
promenade Léa-et-Napoléon-Bullukian, 69008 Lyon, France
b
Centre de lutte contre le cancer Léon-Bérard, service d’anatomie et de cytologie
pathologiques, 28, promenade Léa-et-Napoléon-Bullukian, 69008 Lyon, France
c
Hôpital Édouard-Herriot, service de chirurgie orthopédique, place d’Arsonval,
69008 Lyon, France

KEYWORDS Abstract
Core needle biopsy; Purpose: The goals of this retrospective study were to evaluate the accuracy of core needle
Surgical biopsy; biopsy (CNB) for the diagnosis of osteosarcoma and to identify criteria that may predict failed
Osteosarcoma; CNB.
Sclerotic Materials and methods: From 2002 to 2012, 73 patients with a total of 73 osteosarcomas under-
osteosarcoma went CNB. Patients demographics and procedure details were recorded, including tumor size,
tumor characteristics (hemorrhagic or not, lytic, sclerotic [> 50% bone condensation], or mixed),
the type of anesthesia, the number of tissue samples, the size of the biopsy needle and pathol-
ogy report. Procedures were analyzed in terms of sensitivity, specificity, positive predictive
value (PPV) and negative predictive value (NPV).
Results: A diagnosis was not made in 5/73 patients (6.8%) with an overall sensitivity of 93.1%,
a specificity of 100%, a PPV of 100% and a NPV of 99.9%. No complications due to CNB were
observed. No criteria were identified as predictors of CNB failure.
Conclusion: Even in the presence of sclerotic tumors, CNB should be the first line diagnostic test
for suspected osteosarcomas, pending performance by a well-trained radiologist and reading
by a specialized pathologist.
Level of evidence: IV.
© 2015 Éditions françaises de radiologie. Published by Elsevier Masson SAS. All rights reserved.

Abbreviations: MRI, magnetic resonance imaging; CNB, core needle biopsy; PPV, positive predictive value; NPV, negative predictive
value; G, Gauge; CI, confidence interval; C, contributive; NC, non-contributive.
∗ Corresponding author.

E-mail address: Tatiana taupin@hotmail.com (T. Taupin).

http://dx.doi.org/10.1016/j.diii.2015.09.013
2211-5684/© 2015 Éditions françaises de radiologie. Published by Elsevier Masson SAS. All rights reserved.

Please cite this article in press as: Taupin T, et al. Accuracy of core needle biopsy for the diagno-
sis of osteosarcoma: A retrospective analysis of 73 patients. Diagnostic and Interventional Imaging (2015),
http://dx.doi.org/10.1016/j.diii.2015.09.013
+Model
DIII-714; No. of Pages 5 ARTICLE IN PRESS
2 T. Taupin et al.
Osteosarcoma is the most frequent malignant bone tumor
in adolescents and young adults [1]. It represents 15% of all
primary bone tumors confirmed by biopsy. There are differ-
ent types of primary osteosarcoma and the most frequent
one is intramedullary high-grade osteosarcoma, which rep-
resents 75% of all osteosarcomas [2]. When the diagnosis of
osteosarcoma is suggested by imaging findings, a biopsy must
be performed to obtain a tissue sample of the tumor. For
many years surgical biopsy was the first line technique for
musculoskeletal tumors, because it provided an adequate
tissue specimen for further histopathological assessment
[3,4].
Some authors have suggested that core needle biopsy
(CNB) should be the first line diagnostic technique for
osteosarcoma [5—8]. On the other hand the value of CNB
as the first line diagnostic test in osteosclerotic tumors has
not been confirmed in the literature, compared to surgical
biopsies, which are highly sensitive and specific, but at the
penalty of invasiveness and extra costs [9—11].
Some studies have not identified any factors of failure for
CNB [12]. In others the failure rate of CNB was greater for
bone tumors compared to soft tissue tumors [13]. Sclerotic
or necrotic elements can be a risk factor of diagnostic fail-
ure of CNB [14,15]. The portion of lytic tumor and the size of
the tumor were also shown to influence the diagnostic accu-
racy of CNB in a study in 88 bone tumors [16]. The accuracy
of percutaneous biopsy has been confirmed for mixed bone
tumors [5,17].
The goals of this retrospective study were to evaluate
the accuracy of core needle biopsy (CNB) for the diagnosis
of osteosarcoma and to identify criteria that may predict
failed CNB.
Materials and methods
Patients
The study population consisted of 73 patients with a total
of 73 osteosarcomas. They were 48 boys (65.7%) and 25 girls
(34.2%) with a mean age of 25 years ± 18.45 (SD) (range:
2—88 years). All these patients underwent primary or sec-
ondary CNB after an unsuccessful surgical biopsy (n = 3).
They were investigated in our institution from the year 2002
to the year 2012. Imaging examinations and imaging reports
were available for all patients. Thirty-three patients (33/73;
45%) have died at the time of the study. The study was con-
ducted following the guidelines of our institutional review
board.
Procedure
CNB was performed by different radiologists (fellows or staff
physicians throughout the 10 years of the study). All were
senior radiologists with at least one year of experience in
CNB, without a surgeon by their side. The biopsy site was
sterilized and anesthetized, then an incision was made using
the surgical approach that would be taken in case of future
surgery and after consultation with the surgeon to avoid any
seeding during CNB [18]. Tissue samples (from 1 to 12) were
obtained using an 8- to 18-G (Gauge) automatic or semi-
automatic needle with a coaxial system. When possible,
Please cite this article in press as: Taupin T, et al. Accuracy of core needle biopsy for the diagno-
sis of osteosarcoma: A retrospective analysis of 73 patients. Diagnostic and Interventional Imaging (2015),
http://dx.doi.org/10.1016/j.diii.2015.09.013
the different passes through the tumor targeted the non-
necrotic tissues that surrounded these osteosarcomas as
showed by pre-biopsy magnetic resonance imaging examina-
tion or the soft tissue when it was invaded and more easily
accessible.
If necessary a trephine needle was used in the sclerotic
portions, after placement of a bone trocar (Osteo-Site® 11 G
10 cm beveled bone biopsy needle, Cook Medical, Blooming-
ton USA; Monopty® 14 G 10 cm Bard Biopsy Systems, Tempe,
USA). The material was used according to the manufacturer
instructions. The site of entry of the skin was tattooed with
India ink. The patients were then monitored for 30 minutes
before discharge accompanied by a third party. In case of
general anesthesia, the patient spent the night in the hos-
pital.
In the pathology department a portion of the biopsy sam-
ple was fixed in 10%, buffered formol and the other portion
was either kept in a cool place, or in a liquid preservative
such as RPMI medium and decalcified before analysis [19,20].
Methods of evaluation
The following data were obtained from the files of each
individual patient:
• the smallest and largest tumor dimensions (taking into
account extraosseous extension), hemorrhagic or not, the
location and whether the tumor was lytic, sclerotic (at
least 50% bone condensation) or mixed;
• the type of anesthesia (oral premedication 1 h before the
procedure, then local anesthesia ± inhaled nitrous oxide);
• the number of tissue specimens and the size of the needle
used;
• the pathology reports used as judgment criteria.
Statistical analysis
The diagnostic accuracy of CNB for the diagnosis of
osteosarcoma (pathological diagnosis) as well as sensitiv-
ity, specificity, positive predictive value (PPV) and negative
predictive value (NPV) were calculated. The odds ratio (OR)
were used to identify criteria that may be associated with
failed CNB. Statistical analysis was performed using S.A.S
9.3 software. Differences were searched for using the Fisher
exact test for qualitative variables and the Student t test for
quantitative variables. The normality of the distribution of
the quantitative variables was confirmed.
Results
Three of the 73 lesions (3/73, 4.1%) were hemorrhagic.
The mean largest tumor size was 105 ± 51.2 mm (SD)
(range: 35—260) and of the second largest tumor size was
59 ± 27.2 mm (SD) (range: 17—160).
Seven tumors were biopsied with an 8-G system (7/73,
9.5%), 3 with 11-G (3/73, 4.1%), 34 with 14-G (34/73, 46.5%),
1 with 16-G (1/73, 1.4%), and 28 with an 18-G (28/73,
38.3%), after using a manual bone marrow biopsy needle
if necessary.
Besides premedication and local anesthesia, 25/73
patients (38.3%) received an inhalation of nitrous oxide,
4/73 (5.5%) received an Emlapatch 5% (AstraZeneca,
+Model
DIII-714; No. of Pages 5 ARTICLE IN PRESS
Accuracy of core needle biopsy for the diagnosis of osteosarcoma 3

high-grade highly undifferentiated osteosarcoma. The fail-

Table 1 Different tumor locations in 73 patients with
ure in patient 2 was due to the lack of cartilage cap in
osteosarcoma.
the biopsy specimens, which was successfully obtained with
Left calcaneum 1 (1.3%) a surgical biopsy for a diagnosis of a low-grade parosteal
Left acetabular cup 1 (1.3%) osteosarcoma. In case 3 an external pathologist suggested
Right ethmoid bone 1 (1.3%) an osteosarcoma or a lesion secondary to sarcomatoid
Right femur 17 (23.3%) carcinoma. Patients 4 and 5 correspond to chondromyx-
Left femur 24 (32.9%) oid fibroma-like osteosarcoma and highly undifferentiated
Right humerus 3 (4.1%) fibroblastic osteosarcoma, which are rare entities. The final
Left humerus 5 (6.8%) diagnosis of patient 5 was obtained by surgical biopsy,
Right ilium 2 (2.7%) requested by the pathologist.
Left ilium 1 (1.3%)
Right mastoid 1 (1.3%)
Mandible 1 (1.3%)
Right fibula 2 (2.7%) Discussion
Right radius 1 (1.3%)
Sacrum 1 (1.3%) In our study, the rate of failure of CNB was 6.8%, with a
Left scapula 2 (2.7%) sensitivity of 93.1% for the diagnosis of osteosarcomas. The
Right tibia 6 (8.2%) sensitivity was 89.6% in non-lytic osteosarcomas (mixed and
Left tibia 4 (5.4%) sclerotic). Our sensitivity was high compared to the study by
Wu et al., which had a sensitivity of 87% for lytic lesions and
only 57% for sclerotic lesions [16]. The difference in these
Rueil-Malmaison, France), 1/73 (1.4%) was hypnotized and results might be explained by the fact that our study was
1/73 patient (1,4%) received an injection of morphine. Four performed in a reference cancer center. The sensitivity was
CNBs (4/73, 5.5%) were performed under general anesthe- very high (93.3%) in tumors presenting with at least 50% of
sia. Forty-one tumors were femoral (41/73, 56.1%). The bone condensation. This may be because the radiologist pays
different locations are listed in Table 1. There were no more attention to the quality of the biopsy specimens in this
biopsy-related complications. type of tumor.
A diagnosis was not obtained with CNB in 5/73 patients The PPV of CNB is very high in musculoskeletal tumors,
(5/73, 6.8%). No false-positive findings were observed. CNB with 74% [21], 93% [15] and even 100% [5]. The efficacy drops
had an overall sensitivity of 93.1% (95% CI: 0.85—0.97), a to 57% [16] and 66% [22] in sclerotic tumors. In two studies,
specificity of 100% (95% CI: 0.99—1.00), a PPV of 100% (95% the PPV of surgical biopsies was 96% in 50 musculoskeletal
CI: 0.95—1.00), and a NPV of 99.9% (95% CI: 0.99—1.00). tumors [23], and 82% in 110 osteosarcomas [3]. Data from the
The percentage of purely sclerotic tumors (> 50% of bone literature suggest that the accuracy of surgical biopsy and
condensation) was the same between non-contributive and CNB is similar. However such a direct comparison is difficult
contributive CNB: (1/5 non-contributive (20%); and 14/68 because of differences in study populations.
contributive (20.6%)). The difference was not significant Ultrasound and computed tomography (CT) can be used
(OR: 0.96; 95% CI: 0.1; 10.8) (P > 0.999). for CNB guidance depending on tumor characteristics, size,
Non-contributive CNB included more mixed (scle- accessibility and the choice of the radiologist. Usually three
rotic + lytic) and sclerotic tumors: 3/5: 60%. Of the 68 high quality tissue specimens are sufficient to obtain a
contributive biopsies 26/68 (38%) included this type of pathological diagnosis [16].
tumor. However, this difference was not significant (OR: Our study has limitations. Of note, the size of the biopsy
2.39; 95% CI: 0.2562; 30.439) (P = 0.37). needles varied from 8 to 18 G and the number of biopsy spec-
Contributive and non-contributive CNBs for the differ- imens varied from 1 to 12. The radiologist chose the most
ent types of tumors are listed in Table 2. The sensitivity of appropriate guidance method (ultrasound or CT) depending
CNB was 89.6% for non-lytic osteosarcomas (mixed or purely specific tumor characteristics [7]. Moreover, the patholo-
sclerotic) and 93.3% in purely sclerotic tumors. gist knew the results of imaging examinations. It must be
No variables were identified as predictors of failed CNB. acknowledged that MR imaging is helpful to suggest a specific
The 5 non-contributive CNB are described in Table 3. Patient diagnosis in bone tumors.
1 included a single tissue sample with a 14-G needle, of a There were no false-positive findings because the pathol-
ogists were experienced in sarcomatous diseases. If the size
of the tissue specimen was insufficient for a formal patho-
Table 2 Number of contributive core needle biopsies logical diagnosis, surgical biopsy was requested as for the
(C CNB) or non-contributive (NC CNB), in 73 patients with 5 failed CNB. This is a frequent problem when obtaining a
osteosarcoma. diagnosis by CNB in other centers.
Like in the literature and in our study, if there are more
NC CNB C CNB than three tissue specimens, the number of specimens is
Osteosclerotic (+ 50% of 1/5 (20%) 14/68 (20.6%) not a risk factor of failure [16]. The diameter of the needle,
osteosclerosis) larger than 18 G, is not a risk factor of failure either. None of
Mixed (osteosclerotic + lytic) 2/5 (40%) 12/68 (17.6%) the factors studied were found to be significant risk factors
Lytic 2/5 (40%) 42/68 (61.8%) of failed CNB in our 5 patients. The small number of failed
CNB probably did not allow to identify predictors of failure.

Please cite this article in press as: Taupin T, et al. Accuracy of core needle biopsy for the diagno-
sis of osteosarcoma: A retrospective analysis of 73 patients. Diagnostic and Interventional Imaging (2015),
http://dx.doi.org/10.1016/j.diii.2015.09.013
 4

DIII-714; No. of Pages 5

+Model
http://dx.doi.org/10.1016/j.diii.2015.09.013
sis of osteosarcoma: A retrospective analysis of 73 patients. Diagnostic and Interventional Imaging (2015),
Please cite this article in press as: Taupin T, et al. Accuracy of core needle biopsy for the diagno-

Table 3 Description of the 5 core needle biopsy failure.

Patient 1 Patient 2 Patient 3 Patient 4 Patient 5
Location Left tibia Left femur Sacrum Left femur Left humerus

ARTICLE IN PRESS
1st dimension 48 mm 43 mm 60 mm 173 mm 45 mm
2nd dimension 28 mm 33 mm 60 mm 88 mm 30 mm
Hemorrhagic No No No No No
Predominant development Osseous Osseous Osseous Osseous and soft Osseous and soft
tissue tissue
Rather lytic or sclerotic Lytic Sclerotic Lytic Lytic and sclerotic Lytic and
sclerotic
1st passage needle diameter 14 14 14 14 14
(Gauge)
Total number of samples 1 3 3 5 3
Type of sedation Local anesthesia Local Local anesthesia Local General
anesthesia + nitrous anesthesia + nitrous anesthesia
oxide gas oxide gas
Histological type Uncertainty in the histopathologic diagnostic Diagnosis confirmed by an open biopsy
High-grade very Parosteal low-grade Osteosarcoma Chondromyxoïd Very
undifferentiated osteosarcoma versus fibroma like undifferentiated
osteosarcoma sarcomatoid osteosarcoma osteosarcoma
carcinoma
metastasis

T. Taupin et al.
+Model
DIII-714; No. of Pages 5 ARTICLE IN PRESS
Accuracy of core needle biopsy for the diagnosis of osteosarcoma
Our study confirms the value of CNB as the first line diag-
nostic technique for sclerotic osteosarcoma. Our results are
similar to those reported in the literature for post-biopsy
complications (none in our series). Complications of surgi-
cal biopsies occur between 4 and 19% of procedures, and
may include seromas, hematomas, infections, bone frac-
tures, fistula up to the skin surface, with tumor seeding
along the biopsy path. The risk of a technical error in sur-
gical biopsies performed by a non-specialized surgeon is 3
to 8 times greater, in particular for the choice of the inci-
sion site, which sometimes negatively influences patient
outcome [3—5,8,24]. CNB complications are the same, but
occur in less than 10% of procedures [16]. CNBs are also less
expensive, with a smaller incision, can be scheduled during
the week, have a shorter healing time, and thus reduce the
time necessary to initiate neoadjuvant therapy [5,23].
In conclusion, the overall sensitivity of percutaneous
CNB is high, between 93.1%, and 89.6% in a subpopula-
tion of mixed and sclerotic lesions. Although the extent of
bone condensation may affect the efficacy of results, bone
condensation is not a significant risk factor of failure. We did
not identify any variable that may predict CNB failure. We
suggest performing CNB as the first line diagnostic technique
when osteosarcoma is suspected even if sclerotic lesions
with at least 50% condensation are identified. The patient
should be referred to a reference center, to undergo CNB
and a specialized pathological evaluation. In tumors that
are technically difficult to reach, or require general anes-
thesia, a multidisciplinary consensus should be obtained to
determine the most appropriate technique.
Disclosure of interest
The authors declare that they have no competing interest.
References
[1] Kaste SC. Imaging pediatric bone sarcomas. Radiol Clin North
Am 2011;49:749—65.
[2] Bonar SF. Central low-grade osteosarcoma: a diagnostic chal-
lenge. Skeletal Radiol 2012;41:365—7.
[3] Mankin HJ, Lange TA, Spanier SS. The hazards of biopsy in
patients with malignant primary bone and soft-tissue tumors.
J Bone Joint Surg Am 1982;64:1121—7.
[4] Mankin HJ, Mankin CJ, Simon MA. The hazards of the biopsy,
revisited. Members of the Musculoskeletal Tumor Society. J
Bone Joint Surg Am 1996;78:656—63.
[5] Ray-Coquard I, Ranchère-Vince D, Thiesse P, et al. Evalua-
tion of core needle biopsy as a substitute to open biopsy
in the diagnosis of soft-tissue masses. Eur J Cancer 2003;39:
2021—5.
[6] Rimondi E, Rossi G, Bartalena T, et al. Percutaneous CT-guided
biopsy of the musculoskeletal system: results of 2027 cases.
Eur J Radiol 2011;77:34—42.
[7] De Bazelaire C, Coffin A, Cohen S, Scemama A, de Kerviler
E. Biopsies in oncology. Diagn Interv Imaging 2014;95:
647—57.
Please cite this article in press as: Taupin T, et al. Accuracy of core needle biopsy for the diagno-
sis of osteosarcoma: A retrospective analysis of 73 patients. Diagnostic and Interventional Imaging (2015),
http://dx.doi.org/10.1016/j.diii.2015.09.013
5
[8] Welker JA, Henshaw RM, Jelinek J, Shmookler BM, Malawer
MM. The percutaneous needle biopsy is safe and recom-
mended in the diagnosis of musculoskeletal masses. Cancer
2000;89:2677—86.
[9] Leffler SG, Chew FS. CT-guided percutaneous biopsy of scle-
rotic bone lesions: diagnostic yield and accuracy. AJR Am J
Roentgenol 1999;172:1389—92.
[10] Li Y, Du Y, Luo TY, Yang HF, Yu JH, Xu XX, et al. Factors influ-
encing diagnostic yield of CT-guided percutaneous core needle
biopsy for bone lesions. Clin Radiol 2014;69:e43—7.
[11] Perrier L, Buja A, Mastrangelo G, et al. Clinicians’ adher-
ence versus non adherence to practice guidelines in the
management of patients with sarcoma: a cost-effectiveness
assessment in two European regions. BMC Health Serv Res
2012;12:82.
[12] Altuntas AO, Slavin J, Smith PJ, et al. Accuracy of computed
tomography guided core needle biopsy of musculoskeletal
tumours. ANZ J Surg 2005;75:187—91.
[13] Yang J, Frassica FJ, Fayad L, Clark DP, Weber KL. Anal-
ysis of nondiagnostic results after image-guided needle
biopsies of musculoskeletal lesions. Clin Orthop Relat Res
2010;468:3103—11.
[14] Ayala AG, Zornosa J. Primary bone tumors: percutaneous
needle biopsy. Radiologic-pathologic study of 222 biopsies.
Radiology 1983;149:675—9.
[15] Mitsuyoshi G, Naito N, Kawai A, Kunisada T, Yoshida A, Yanai
H, et al. Accurate diagnosis of musculoskeletal lesions by core
needle biopsy. J Surg Oncol 2006;94:21—7.
[16] Wu JS, Goldsmith JD, Horwich PJ, Shetty SK, Hochman
MG. Bone and soft-tissue lesions: what factors affect diag-
nostic yield of image-guided core-needle biopsy? Radiology
2008;248:962—70.
[17] Jelinek JS, Murphey MD, Welker JA, et al. Diagnosis of primary
bone tumors with image-guided percutaneous biopsy: experi-
ence with 110 tumors. Radiology 2002;223:731—7.
[18] UyBico SJ, Motamedi K, Omura MC, Nelson SD, Eilber FC,
Eckardt J, et al. Relevance of compartmental anatomic guide-
lines for biopsy of musculoskeletal tumors: retrospective
review of 363 biopsies over a 6-year period. J Vasc Interv Radiol
2012;23:511—8.
[19] Santini-Araujo E, Olvi LG, Muscolo DL, Velan O, Gonzalez ML,
Cabrini RL. Technical aspects of core needle biopsy and fine
needle aspiration in the diagnosis of bone lesions. Acta Cytol
2011;55:100—5.
[20] Gomez-Brouchet A, Bouvier C, Decouvelaere A-V, Larousserie
F, Aubert S, Leroy X, et al. Place of the pathologist in
the management of primary bone tumors (osteosarcoma and
Ewing’s family tumors after neoadjuvant treatment). Ann
Pathol 2011;31:455—65.
[21] Yao L, Nelson SD, Seeger LL, Eckardt JJ, Eilber FR. Primary mus-
culoskeletal neoplasms: effectiveness of core-needle biopsy.
Radiology 1999;212:682—6.
[22] Moore TM, Meyers MH, Patzakis MJ, Terry R, Harvey JP.
Closed biopsy of musculoskeletal lesions. J Bone Joint Surg Am
1979;61:375—80.
[23] 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;78:644—9.
[24] Adams SC, Potter BK, Pitcher DJ, Temple HT. Office-based core
needle biopsy of bone and soft tissue malignancies: an accurate
alternative to open biopsy with infrequent complications. Clin
Orthop Relat Res 2010;468:2774—80.