Joint Bone Spine 73 (2006) 532–537

http://france.elsevier.com/direct/BONSOI/

Original article

Percutaneous computed-tomography-guided
biopsy of the spine: 229 procedures
Hasan Kamil Sucu a,*, Canan Çiçek a, Türkan Rezanko b, Hamdi Bezircioğlu a, Yusuf Erşahin c,
Mine Tunakan b, Mustafa Minoğlu a
a
Department of Neurosurgery, Atatürk Training and Research Hospital, Izmir, Turkey
b
Department of Pathology, Atatürk Training and Research Hospital, Izmir, Turkey
c
Division of Pediatric Neurosurgery, Ege University Faculty of Medicine, Izmir, Turkey
Received 7 September 2005; accepted 18 January 2006
Available online 29 March 2006

Abstract
Objectives: Percutaneous biopsy of the spine is an effective and well-evaluated procedure. Only very few series containing more than a
hundred patients have been reported so far and there is no agreement about the factors affecting the diagnostic rate. We aimed to find out if
there is any factor influencing the success rate of the spinal biopsy using our biopsy series.
Methods: Two hundred and twenty-nine procedures were performed in 201 patients between November 2001 and August 2005. All proce-
dures were performed under computed tomography guidance. The side was chosen according to the extension of the lesion. When the lesion was
in the midline or extended to both sides, we preferred to obtain biopsy from the right side. The puncture point and the needle trajectory were
planned on both lateral computed tomography scout scan and axial scans.
Results: We found that the diagnostic rate was not affected by the variables such as age, gender, type and diameter of the biopsy needle,
diagnosis as well as lesion localization and level. The success rate of the repeat biopsies was considerably lower than the first procedures.
Conclusions: The diagnostic rate is not affected by any of the variables but the approach, chosen can vary with the level, localization, and
lesion characteristics.
© 2006 Elsevier SAS. All rights reserved.

Keywords: Computerized tomography; Percutaneous spinal biopsy; Spine

1. Introduction published the transpedicular biopsy technique. We also de-

The percutaneous spinal biopsy can provide an accurate di-
agnosis and facilitate treatment in the various spinal pathologi-
cal entities. Its efficacy in the management of spinal lesions has
been extensively evaluated in the literature [1–9].
The most common approach to biopsy of the vertebral
bodies is posterolateral [2–4,7,8]. However, different ap-
proaches have been described as an alternative to the classic
posterolateral approach because of the technique-related diffi-
culties and the risk of complications. Brugieres et al. [10] de-
scribed the transcostovertebral approach for performing biopsy
in the thoracic region in 1990. One year later Renfrew et al. [7]
scribed the transforaminodiskal approach in 2003 [11].
It is generally accepted that the larger needle diameter [3,12,
13] and the osteolytic lesions [8,10,14] are the factors increas-
ing the diagnostic rate in the percutaneous CT-guided spinal
biopsies. However dubious results regarding the gender [5,12,
15] and spinal level [5,8,12,14–17] have been shown, and there
is no consensus in these matters. There has been no study com-
paring the diagnostic rates of different biopsy techniques.
We aimed to find out if there is any factor influencing the
success rate of the spinal biopsy using our biopsy series.
2. Methods

Between November 2001 and August 2005, 229 percuta-

* Correspondingauthor. neous spinal biopsy procedures were performed in 201 patients
E-mail address: hsucu@yahoo.com (H.K. Sucu). (Table 1). There were 115 male and 86 female patients, ran-
1297-319X/$ - see front matter © 2006 Elsevier SAS. All rights reserved.
doi:10.1016/j.jbspin.2006.01.013
 H.K. Sucu et al. / Joint Bone Spine 73 (2006) 532–537 533

ging in age from 11 to 81 years (mean 57 years). All biopsy

procedures were performed by the two neurosurgeons (HKS,
CÇ) and the data regarding the patients and lesion, details of
the procedures, type and size of the needles, pathological ex-
aminations and results of the cultures were prospectively re-
corded.
The advances in the magnetic resonance imaging allow de-
tection of vertebral lesions at an earlier stage and there is an
increase in proportion of vertebral lesions with intensity
changes on MRI scans only when the patients are referred for
biopsy. Because of this we did not divided the lesions into two
groups as osteolytic and osteosclerotic prospectively.
Needle biopsy was performed in all spinal column seg-
ments, only except for C1 (20 cervical, 66 thoracic, 117 lum-
bar, 23 sacral, two lumbo-sacral and one coccygeal lesion). In
five patients we were able to obtain biopsy samples from two
vertebrae at the same puncture (Fig. 1). The target was the
posterior part of vertebra in 15 and the anterior part of vertebra
(vertebral body) in 214 biopsy procedures.
We used the anterolateral approach for uppermost thoracic
spine in four cases for the first time to our knowledge (Fig. 2).
Anterolateral approach is easier and more reliable than poster-
ior approaches particularly for T1 and some T2 lesions (Ta-
ble 3). The anterolateral approach to the vertebral body was
used in 11 and four times in the cervical region (Fig. S1; see
the Supplementary Material available with this article online)
and in the uppermost thoracic vertebra, respectively.
The subgroups of posterior approach to the vertebral body
are as follows: the classical posterolateral (paravertebral) ap-
proach in 18, intertransversocostal approach in 20, transpedicu-
lar approach in 54 and transforaminodiskal approach in 107
procedures.
All procedures were performed under CT guidance. All pa-
tients were injected with diazepam and of diclofenac sodium
Fig. 1. (a) 69-year-old male patient had been suspected L5-S1 spondylodiscitis
intramuscularly 30 min before the procedure as premedication. and the diagnosis confirmed by percutaneous spinal biopsy. Jhamshidi biopsy
The skin of the biopsy site was infiltrated with the local anes- needle is advance perpendicular to the patient’s vertical axis. (b) Entering the
thetic (prilocaine). We initially used 14 Gauge (G) Tru-Cut pedicle, the needle first passes the S1 corpus then L5-S1 intervertebral disc and
needle (23 patients) and had a difficulty in penetrating the cor- L5 corpus lastly. (This procedure could obtain biopsy specimen shaped as a
cylinder consisting bone at the both end and intervertebral disc tissue between
tical bone (P < 0.0001). Then we started using the 10-cm-long them.).
Jhamshidi needles. We switched to the 11G- Jhamshidi needles
in order to obtain greater specimen after using the 13G- Jham-
shidi needles at first.
The side of the approach was determined on the CT scans, ime and rifampicin were administered in patients with non-spe-
obtained just before the procedure. The side that the lesion had cific infections.
been predominantly located was chosen for biopsy. When the Generally, chi-square tests were used in statistical analyses.
lesion was in the midline or extended to both sides, we pre- Only for comparing the mean ages of diagnostic and non-diag-
ferred to obtain biopsy from the right side just because we nostic groups t-test was used. P < 0.05 was considered statisti-
put the biopsy table on the right sight of the patient. The punc- cally significant.
ture point and the needle trajectory were planned on both lat-
eral CT scout scan and axial scans. 3. Results
Every biopsy specimen was processed for pathological ex-
aminations and culture. In addition, biopsy samples were also 3.1. Diagnostic rates
sent for acid-fast staining and microscopic examination for
searching pathogen microorganism. The inconclusive patholo- 3.1.1. Overall
gical examinations that had no effect on the management of the In 197 of 229 procedures we were able to obtain the biopsy
patients were considered unsuccessful. Vancomycin, cefotax- samples from the desired places of the spine. In 15 biopsy sam-
 534
Table 1
The summary of our 229 biopsy procedures (201 patients)
Sample obtained, 197 biopsies No sample, 32 biopsies
Follow up, 182 biopsies No follow up, Biopsy sample couldn’t obtained from the right place of the spine: 31 cases The biopsy
Accurate diagnosis, 142 procedures Inconclusive results, 15 biopsies Surgeon related reasons: 6 cases Needle related reasons: 25 cases procedure was
No of No of Type of Pathological diagnosis 40 procedures The needle The needle The needle was The needle abandoned

H.K. Sucu et al. / Joint Bone Spine 73 (2006) 532–537

biopsies patients lesion for biopsies (False negative: 22 Couldn’t reach the punctured the able to puncture the couldn’t puncture because of the
73 73 Infection 51 non-specific Misdiagnosis: 4; vertebrae in 4 cases wrong place in vertebrae but the vertebrae in pain: 1 case
infections Necrosis or chronic (4 Jhamshidi) 2 cases couldn’t get 6 cases (6 Tru-cut)
22 tuberculoses inflammatory granulation (1 Jhamshidi, enough specimen
22 21 Primary 8 lymphoma tissue, but the etiology 1 Tru-cut) in 19 cases
neoplasm 7 plasmocytoma is not certain: 9 (18 Jhamshidi,
1 chronic myelocytic Etiology is infection, 1 Jhamshidi +
leukemia but it is not certain Tru-Cut)
1 Ewing’s sarcoma whether it is NSE or
2 chordoma TBC: 3
1 osteoblastoma Etiology is malignancy,
1 chondrosarkoma but the type of it could
1 osteosarkoma not been defined: 2
40 39 Metastasis 6 lung cancer
4 kidney cancer
2 breast cancer
2 thyroid cancer
2 prostat cancer
1 carcinoma of ovary
1 colon cancer
1 malignant melanoma
21 unknown
7 6 Normal 6 normal bone
 H.K. Sucu et al. / Joint Bone Spine 73 (2006) 532–537 535

Table 2
Diagnostic rates for procedures and patients
Total Excluded Diagnosed Diagnostic rate
Number Number Cases
Procedures 229 15 142 66% (142/214)
Patients 201 15 139 75% (139/186)
Biopsy 197 15 142 78% (142/182)
samples

3.1.4. First vs. Repeat biopsies

Because we had got inconclusive results, in 26 patients
biopsy procedures were performed two times and in one pa-
tient biopsy procedures were performed three times. The suc-
cess rates of the first (68.8%) and repeated (44%) biopsies var-
ied significantly (P = 0.014).

3.1.5. Anterior vs. posterior sampling

When we sampled the biopsy specimen from the posterior
part of vertebrae diagnostic rate was 80.0%. In contrary, diag-
nostic rate was 65.3% for vertebral body. However this differ-
ence was not significant.

3.1.6. Diagnostic rates in terms of level

Our diagnostic rate for cervical biopsies (65.0) was not dif-
fering from other parts of spine (66.5%). We didn’t find sig-
nificant difference between the diagnostic rates of the lumbar
(65.1%) and thoracic biopsies (68.3%). In our series, the diag-
nostic rate for sacral biopsies was 68.0% (17/25) and was not
different from the other parts of the spine.

3.1.7. Approaches
We did not find a difference between the diagnostic rates for
different approaches for thoracolumbar region.

Fig. 2. (a) The anterolateral approach for the uppermost thoracic region. (b) The
needle is starting to enter the anterior-inferior part of T2 vertebrae. T3 vertebrae
can be seen posterior to the T2 vertebrae.
ples of 15 patients, pathological examination turned out to be a
normal bone matrix and had been lost to follow-up. Therefore,
the data of these 15 procedures were not included in the statis-
tical analyses. (Table 1). The overall success rates per proce-
dure and per patient were 66% and 75%, respectively (Table 2).
3.1.2. Age–gender
We did not find any significant difference between the
male’s diagnostic rate (74.8%) and female’s diagnostic rate
(74.7%). The mean age of the diagnostic group was 56.3 years
and for the non-diagnostic group was 54.1 years. This differ-
ence is not statistically significant.
3.1.3. Needle type and diameter
The needle types (Jhamshidi vs. tru-cut) did not have a sig-
nificant difference in terms of diagnostic rate. Also, we did not
find any significant difference between the large (8 and 11
gauge) and small (13 gauge) diameter Jhamshidi needles.
3.2. Diagnosis
We were able to make the diagnosis in 142 biopsy proce-
dures (139 patients) (Table 1). In four patients the pathological
diagnosis of needle biopsy was different from that of the open
surgery. The different pathological diagnoses of needle biopsy
and open surgery were as follows: lymphoma and Ewing sar-
coma in S2, non-specific infection and plasmocytoma in T2
and T4, indifferent malignant tumor and myxopapillary epen-
dymoma in S5, respectively. The culture from the biopsy ma-
terial was positive in only 10 patients (9 Mycobacterium tuber-
culosis and one Kytococcus sedentarius).
3.3. Complications
We had one major complication. In a 69-year-old male pa-
tient who had undergone a spinal biopsy using the classic pos-
terolateral approach, a large retroperitoneal hematoma devel-
oped (Fig. S2). The patient was operated on for the
hematoma and survived. No vessel injury was detected at sur-
gery and bleeding was presumably from the psoas muscle.
Only one patient could not tolerate the pain and we had to
abandon the procedure. In all other patients no back or radicu-
lar pain was observed, either immediately after biopsy or dur-
536 H.K. Sucu et al. / Joint Bone Spine 73 (2006) 532–537
Table 3
Summary of the recommended biopsy approaches according to the level of the lesions
Condition
Lesion in both anterior (vertebral body) and posterior (neural arch) of the
vertebra
Lesion in several vertebrae
Lesion in or near the pedicle
Lesion in cervical region
Lesion in T1 vertebral body
Lesion in T2 vertebral body patient’s neck is long and thin.
Lesion in the disc space of T2–L5
a large paravertebral mass
Lesion in T3–T6 vertebral body
Lesion in T7–L5 vertebral body and near upper end plate
Lesion in T7–L5 vertebral body and near lower end plate
Lesion in L5–S1 disc space with neighboring end plates
Lesion in the sacrum
ing the days that followed, and no complications were ob-
served.
4. Discussion
Many authors have suggested that some factors may influ-
ence the success rate of spinal biopsies. Kattapuram et al. [12]
found the mean age of the diagnostic group (56 years) older
than the non-diagnostic group (54 years) and proposed that
there was a trend toward better results in female patients
(97%) than in male (86%).
Fyfe et al. [3], Kattapuram et al. [12], and Ward et al. [13]
reported that larger needles in diameter had a higher diagnostic
rate than the smaller ones.
Diagnostic rates were compared mainly between thoracic
biopsies and lumbar biopsies in the reported series. Generally
lower diagnostic rates were found in thoracic level biopsies in
many previous studies [8,14–16]. Brugieres et al. [16] and Lis
et al. [5] found high accuracy rates (90% and 100% respec-
tively) for cervical biopsies. Kattapuram et al. [12], Kornblum
et al. [15] and Lis et al. [5] reported similar high diagnostic
rates for sacral biopsies such as 86.7%, 92%, and 96% respec-
tively. On contrary, Ozerdemoglu et al. [17] found very low
diagnostic rate for sacral needle biopsies (12%).
In spite of these reports we did not find any relation be-
tween the success rate of spinal biopsies and the factors above
mentioned. Similarly, we did not find a difference between the
diagnostic rates for different approaches but we have learnt
from our experiences that different approaches can be appro-
priate for different conditions (Table 3).
Posterolateral (paravertebral) approach should not be the
first choice because of its high complication risk [3,8,10,12],
except for the disc space biopsies [6]. It can also be preferred
in large paravertebral masses or abscesses. The transcostover-
tebral approach can be performed in thoracic region only. A
narrow space between the elongated transverse process and
neighboring rib may hinder the penetration of the vertebral
body. Although the transpedicular technique seems to be reli-
able [7,13], it is not easy to penetrate the cortical shell of the
bone [11]. Another limiting factor of the transpedicular ap-
Recommendation
Obtain the biopsy specimen from the posterior part of vertebrae
Select the lumbar region first because it is easy to perform the biopsy procedure and
it has lower complication risk, higher diagnostic rate and then select the sacral region
and lower complication risk
Transpedicular
Anterolateral approach
Anterolateral approach
Anterolateral or transpedicular approach
Posterolateral approach
Posterolateral approach
Transcostovertebral approach
Transforaminodiscal approach
Transpedicular approach
Perform transpediculo-disco-vertebral approach
Transpedicular approach
proach can be the pedicle width. [18,19]. This technique is
particularly useful when the lesion is located in or near the
pedicle. The approach we used mostly was the transforamino-
discal one, because of its easy application and a low complica-
tion risk [11]. We prefer this approach especially to obtain
biopsy sample from the cranial part of vertebra.
The biopsy procedure can be repeated easily if the first one
fails. However the success rate in the repeat biopsies was con-
siderably lower than that of the first biopsies. Performing a
repeat biopsy may lead to a delay in the appropriate treatment.
Instead of the needle biopsy, open biopsy can be considered in
those cases. The technically difficult location and the patient’s
incompatibility may presumably lead to the lower diagnostic
rate of repeat biopsies.
In conclusion, percutaneous spinal biopsy is a simple and
repeatable procedure. Its complication rate is acceptable. Pa-
tients usually tolerate spinal biopsy well with local anesthesia.
The diagnostic rate does not seem to vary with any of the vari-
ables evaluated in this study.
Supplementary Material
Supplementary data (Fig. S1, S2) associated with this article
can be found, in the online version, at doi: 10.1016/j.
jbspin.2006.01.013.
References
[1] Babu NV, Titus VT, Chittaranjan S, Abraham G, Prem H, Korula RJ.
Computed tomographically guided biopsy of the spine. Spine 1994;19:
2436–42.
[2] Bender CE, Berquist TH, Wold LE. Imaging-assisted percutaneous
biopsy of the thoracic spine. Mayo Clin Proc 1986;61:942–50.
[3] Fyfe IS, Henry AP, Mulholland RC. Closed vertebral biopsy. J Bone
Joint Surg Br 1983;65:140–3.
[4] Laredo JD, Bard M. Thoracic spine: percutaneous trephine biopsy. Radi-
ology 1986;160:485–9.
[5] Lis E, Bilsky MH, Pisinski L, Boland P, Healey JH, O’Malley B, et al.
Percutaneous CT-guided biopsy of osseous lesion of the spine in patients
with known or suspected malignancy. AJNR 2004;25:1583–8.
 H.K. Sucu et al. / Joint Bone Spine 73 (2006) 532–537
[6] Pierot L, Boulin A. Percutaneous biopsy of the thoracic and lumbar
spine: transpedicular approach under fluoroscopic guidance. Am J Neu-
roradiol 1999;20:23–5.
[7] Renfrew DL, Whitten CG, Wiese JA, el-Khoury GY, Harris KG. CT-
guided percutaneous transpedicular biopsy of the spine. Radiology
1991;180:574–6.
[8] Stoker DJ, Kissin CM. Percutaneous vertebral biopsy: a review of 135
cases. Clin Radiol 1985;36:569–77.
[9] Legrand E, Flipo RM, Guggenbuhl P, Masson C, Maillefert JF, Soubrier
M, et al. Management of nontuberculous infectious discitis. Treatments
used in 110 patients admitted to 12 teaching hospitals in France. Joint
Bone Spine 2001;68:504–9.
[10] Brugieres P, Gaston A, Heran F, Voisin MC, Marsault C. Percutaneous
biopsies of the thoracic spine under CT guidance: transcostovertebral ap-
proach. J Comput Assist Tomogr 1990;14:446–8.
[11] Sucu HK, Bezircioglu H, Çiçek C, Ersahin Y. CT-guided percutaneous
transforamino-diskal biopsy of vertebral body lesions. J Neurosurg 2003;
99(Suppl):51–5.
[12] Kattapuram SV, Khurana JS, Rosenthal DI. Percutaneous needle biopsy
of the spine. Spine 1992;17:561–4.
[13] Ward JC, Jeanneret B, Oehlschlegel C, Magerl F. The value of percuta-
neous transpedicular vertebral bone biopsies for histologic examination.
537
Results of an experimental histopathologic study comparing two biopsy
needles. Spine 1996;21:2484–90.
[14] Ghelman B, Lospinuso MF, Levine DB, O’Leary PF, Burke SW. Percu-
taneous computed-tomography-guided biopsy of the thoracic and lumbar
spine. Spine 1991;16:736–9.
[15] Kornblum MB, Wesolowski DP, Fischgrund JS, Herkowitz HN. Com-
puted tomography-guided biopsy of the spine. A review of 103 patients.
Spine 1998;23:81–5.
[16] Brugieres P, Revel MP, Dumas JL, Heran F, Voisin MC, Gaston A. CT-
guided vertebral biopsy. A report of 89 cases. J Neuroradiol 1991;18:
351–9.
[17] Ozerdemoglu RA, Thompson RC, Transfeldt EE, Cheng EY. Diagnostic
value of open and needle biopsies in tumors of the sacrum. Spine 2003;
28:909–15.
[18] Misenhimer GR, Peek RD, Wiltse LL, Rothman SL, Widell EH. Ana-
tomic analysis of pedicle cortical and cancellous diameter as related to
screw size. Spine 1989;14:367–72.
[19] Zindrick MR, Wiltse LL, Doornik A, Widell EH, Knight GW, Patward-
han AG, et al. Analysis of the morphometric characteristics of the thor-
acic and lumbar pedicles. Spine 1987;12:160–6.