Downloaded from www.ajronline.org by 95.90.213.214 on 03/16/17 from IP address 95.90.213.214. Copyright ARRS.

For personal use only; all rights reserved

Acute Osteoporotic Vertebral

Collapse: Open Study on Percutaneous
Injection of Acrylic Surgical Cement in
20 Patients

Catherine Cyteval1 OBJECTIVE. The aim of this study was to determine the efficacy of percutaneous vertebro-
Marie P. Baron Sarrab#{232}re1 plasty in treating painful spinal osteoporotic collapse.

Jean 0. Roux1 SUBJECTS AND METHODS. Twenty-three cases of vertebral collapse were evaluated
with CT and MR imaging to determine osteoporotic origin and recent evolution. Percutaneous
Eric Thomas2
vertebroplasties were performed using CT guidance. The 20 patients included in the study (17
Christian Jorgensen2
women. 3 men: 62-92 years old) had acute pain of less than 1 months duration that hindered am-
Francis Blotman2
bulation and required treatment with narcotic drugs. They underwent this procedure for analgesic
Jacques Sany2 purposes. The analogic visual scale of Huskisson was used for pain when scoring assessment.
Patrice 1 RESULTS. In 15 patients (75%), pain relief was complete within 24 hr after injection. Anal-
gesic administration was stopped in 14 patients. Mild pain persisted in three (15%) ofthe remain-
ing five patients. In one other patient (5%), crural pain was observed with cement leakage in the
psoas muscle. In the fifth patient (5%). pain recurred after the patient was lifted. The pain was re-
lated to a new acute collapse of an adjacent vertebrae.
CONCLUSION. Vertebroplasty for the treatment of osteoporotic vertebral collapse is a
minimally invasive procedure that provides immediate pain relief and enables the patient to be-
come quickly mobile.

more collapsed vertebrae on radiographs of the tho-

I broplasty
veloped techniques
and
n recent years. arepercutaneous
now havebeingbeenverte-
used
de-
racic or lumbar
age and porosis: and benignancy
of the collapse as determined
spine with no risk
and acute character
by CT and MR imag-
factor other than

extensively in some clinics for pain relief and

ing according to recently described patterns 17, 8).
strengthening of weakened vertebral bodies
Moreover. in all cases. the extent of vertebral-body
I l-6]. Because the worlds population is ag- height collapse was not an inclusion criterion. Unen-
ing. increasing numbers of elderly persons are hanced CT was performed on a HiSpeed scanner
sustaining vertebral compression fractures due (General Electric Medical Systems. Milwaukee. WI)
to osteopenia. The aim of this study was to in- with 3-mm-thick contiguous sections over the entire
vestigate, using clinical scoring, the results of height of the collapsed vertebra. MR imaging was
vertebroplasty under CT guidance in patients performed with a l.5-T unit (Magnetom Vision: Sie-
with painful osteoporotic vertebral collapse. mens, Erlangen. Germany) using spin-echo TI-
weighted sequences (TRJFE and range. 5(X)/l2-30)
short inversion time inversion recovery (TRiTE.
4000/60: inversion time, 150 msec) in the sagittal
Subjects and Methods
plane (5-mm-thick sections). These examinations
Populotion
also revealed multiple vertebral collapses. MR imag-
Received December 17, 1998; accepted after revision
June 3, i999. Twenty patients ( I 7 women. three men: 62-92 ing was interpreted by two radiologists to exclude
1 Department of Radiology, Hopital Lapeyronie, 371 av du years old) with osteop()rotic vertebral collapse un- other disorders that could have resulted in patient
Doyen Gaston Giraud, 34295 Montpellier Cedex 5, France. derwent vertebroplasty at our institution. Patients pain. For clinical assessment. the analogic visual

Address correspondence to C. Cyteval. were selected on the basis of four clinical criteria scale of Huskisson [91 was used for pain scoring.
and two imaging criteria. The clinical criteria were and the scores ranged from 7.5 to 9.5 (on a scale of
2Department of Rheumatology, Hopital Lapeyronie,
acute pain of less than I months duration requiring 1-10. with 10 being most painful)
in all patients.
34295 Montpellier Cedex 5, France.
narcotic drugs. an inability to stand because of pain. Only recent collapses with pain of less than one
AJR i999;i73:i685-i690
no clinical findings of neurologic complications. no months duration and abnomial MR vertebral body
0361-803X199/1136-1685 recent history of trauma. and no malignant disease signal image were treated (Figs. IA and IB). Three
American Roentgen Ray Society history. Imaging criteria were the presence of one or of the 20 patients had pain. with two acute vertebral

AJR:173, December 1999 1685

 f: Fig. 1.-Acute and chronic osteoporotic vertebral collapse in 77-year-old
.- - woman with acute pain for 3 weeks.
A,T1-weighted MR image (TRITE, 500/12) shows acute collapse of 13 and L4
vertebrae, with preserved areas of normal intensity. Chronic vertebral col-
lapse of L5 shows normal fat intensity signal.
B, Short inversion time inversion recovery image (TRITE, 4000/60: flip angle,
1800: inversion time, 150 msec) shows band of high signal intensity in supe-
nor part of 13 and L4 vertebral bodies.
C and 0, CT scans obtained before (C) and after (Dl vertebroplasty show
posterolateral approach and good needle position in L3.
Downloaded from www.ajronline.org by 95.90.213.214 on 03/16/17 from IP address 95.90.213.214. Copyright ARRS. For personal use only; all rights reserved

E-H, Serial CT scans obtained during injection in L4 vertebral body.

er4: k 4-

E F

1686 AJR:173, December 1999

 Cement Injections for Osteoporotic Vertebral Collapse

collapses. and underwent vertebroplasty at these two Patients were permitted to stand the next day. The Complications
levels during the same procedure. Twenty-three ce- procedure was assessed on the basis of standing abil- No general complications occurred.
ment injections were performed in 20 procedures: 10 ity and was scored using the analogic visual scale of The patient presenting with a new vertebral
at the thoracic level (two on Tb. two on TI I. and Huskisson. Scores before and after the procedure
collapse within the month after vertebroplasty
six on Tl2) and 13 at the lumbar level (eight on LI. were compared using the Students t test. After 6
was treated with traditional therapy (no narcotic
two on L2. two on L3. and one on L4). months. all patients were reevaluated clinically and
Downloaded from www.ajronline.org by 95.90.213.214 on 03/16/17 from IP address 95.90.213.214. Copyright ARRS. For personal use only; all rights reserved

by standard radiography to assess the analgesic status

drugs, bed rest for 2 weeks. and a lumbar cast).
and morphology of treated and adjacent vertebrae. Methyl methacrylate leaks were detected
Technique during the procedure in eight patients. Leaks
Results
Percutaneous vertebroplasties were perfomied us- were located in adjacent disks in five patients
Analgesic Effect
ing Cl guidance as recommended by two groups of (Fig. 4). in the lumbar venous plexus once (Fig.
All patients received narcotic drugs after hos-
authors [4. 5. 101. Sedatives (diazepam. 20 mg: pethi- 5), and in the foramina twice (Fig. 6). No clini-
pital admission. In 14 patients. analgesic treat-
dine. 100 mg: and au-opine. 0.25 mg) were given I hr cal complications occurred in patients with disk
ments were stopped the day after the procedure.
before the patient was psitioned on the table. and the or IV leaks. Among the two patients with Ibra-
In six other patients. nonnarcotic drugs were
sedative effect ended after 4 hr. Lidocaine hydrochlo- minal leaks. one patient experienced crural pain
ride was injected as a local anesthetic. A single-side enough to allow lifting. All patients were able to
when standing up 24 hr after the procedure:
transpedicular approach was used at the thoracic level stand within 24 hr after treatment. Pain decreased
however, the pain was partly related to forami-
to avoid pleural tears. whereas a paravertebral ap- in all patients.
nal cement leakage and partly to the static lum-
proach was preferentially used at the L2-L4 levels be- The pain decrease. as scored by the Huskisson
cause it facilitated reaching the center of the vertebral
bar abnormality. This complication was treated
scale. was significant (p < .001) (Fig. 3). Within
body (Fig. 2). For both techniques. patients were in the with steroid therapy for 2 days. and no further
the first week after vertebroplasty, 15 patients had
prone position. A 13-gauge I I-cm-long biopsy needle treatment was required. The pain decreased but
minimal pain (a score of 1.5-3.5 on the analogic
(Biomid: Galhini. Mirandola, Italy) was carefully ad- persisted somewhat when walking. The second
visual scale), five patients had moderate pain (a
vanced into the anterior third of the vertebral body. patient. despite a cement leakage reaching the
score of4 or 5). and one patient ofthose five con-
The surgical cement (Palacos R Gentamicine: Scher- psoas muscle, showed clear improvement (from
tinued to have cniral pain (score. 6). Improve-
ing-Plough. Levallois-Perret. France) was composed 9.5 to 3.5 on the analogic visual scale) and expe-
of polymethylmethacrylate with additional antibiotics ment was observed within the first 24 hr in all
rienced no crural or sciatic pain.
(gentamicine) and radiopaque chemical compounds cases. and all patients were discharged from the
(zirconium dioxide). Helical CT was performed of the hospital within the week.
whole vertebral body after acrylic cement injection ( I At the I-month follow-up examination, pain Discussion
ml) to monitor the cement progression: no CT recon- had recurred in one patient at the same lumbar Approximately 30% of postmenopausal
struction was used. In the absence of leaks or pain. an level. This patient presented with a new po- white women have osteoporosis. and most of
additional 2 ml of cement was injected. followed by
rotic vertebral collapse of an underlying verte- these women have vertebral fractures I 1 11.
another helical acquisition. and the same CT scan was
bra. At the 6-month follow-up. four patients These fractures have a substantial negative im-
repeated until the cement reached the posterior verte-
presented with a new vertebral collapse with pact on function and quality of life. The cost to
bral wall or until paravertebral or epidural cement
pain at other vertebral levels. but without the society was nearly $250 million in the United
leaks occurred (Figs. IC-I H). A total of 4-8 ml
(mean. 5 ml for the thoracic spine. 6 ml for the lumbar local recurrence of pain. No new collapse of States in 1995 1 12). Although these fractures
spine) was injected. and all treated vertebrae were injected vertebral bodies was noted during the rarely require hospitalization. some of them
filled more than 50%. The duration of the procedure follow-up period. nor were any morphologic cause incapacitating pain requiring long periods
was approximately 30 mm. vertebra modifications noted. of strict immobilization, and aged patients risk

10

07
U

. 5
>
. 4
Do
0
( 3
C

0 hr 24hr
Day of Procedure

Fig. 2.-82-year-old woman who underwent vertebro- Fig. 3-Graph of analogic visual scores before and 24 hr after procedure shows decreased pain in all patients.
plasty of L2 vertebral body. CT scan oftranspedicular ap-
proach shows needle and start of cement injection.

AJR:173, December 1999 1687

 Cyteval et al.

that the risk (pneumothorax if a posterolateral ap-

proach is used, neurologic complication if a
transpedicular approach is used) is not warranted
for the treatment of a nonmalignant disorder. In
our experience. patients with these collapses
should not be bedridden.
Downloaded from www.ajronline.org by 95.90.213.214 on 03/16/17 from IP address 95.90.213.214. Copyright ARRS. For personal use only; all rights reserved

Because pain improvement is the key as-

sessment goal of vertebroplasty in vertebral
porotic collapses. we used a quantitative score
to grade pain [9]. Our study design differs
from that used by Jensen et al. [19]. We con-
sider that MR imaging or scintigraphy are es-
sential for detecting all vertebral collapses and
for allowing treatment of multiple recent le-
sions in the same procedure. The only patient
Fig. 4.-CT scan shows cement leak in disk at end of Fig. 5.-CT scan shows cement leak in lumbar venous
procedure (arrow) in 74-year-old woman. plexus (arrow) in 82-year-old woman. with early pain recurrence had a new adjacent
vertebral collapse. However, retrospective
analysis of MR data showed that the new col-
lapsed vertebra already presented an abnormal
major complications with long-term bed rest. Gangi et al. [17] reported complete relief of back signal of his upper end plate without any col-
Minimally invasive techniques are thus of some pain using this technique. with no further col- lapse at the time of vertebroplasty. This sug-
interest, particularly in patients whose on-going lapse of vertebral bodies and no complications. gests that the treatment of all lesions at the
medical treatment is often inefficient [13]. In 1997, Jensen et al. [19] reported stabilization same time could have avoided this collapse.
The percutaneous injection of cement into with polymethylmethacrylate of age-related or which we related to the procedure (Fig. 7).
vertebral bodies for the treatment ofvertebral he- steroid-induced osteoporotic vertebrae with par- The guidance method is a controversial point
mangioma was pioneered in France [14-16]. tial compression fractures. In that study, 90% of concerning the vertebroplasty procedures. One
This technique has now been expanded to pallia- patients showed marked pain relief within 1 group of authors has argued that fluoroscopic
five stabilization of metastatic diseases of the week oftreatment, along with increased mobility guidance with the transpedicular approach is
spine [I I, 12]. Although relatively few cases and a lower narcotics requirement. safe. even without using venograms [ 1 1. An-
have been reported in the literature, the pain relief Our study was designed to evaluate the effi- other group recommends the use of venograms
offered by this technique was judged significant cacy of percutaneous vertebroplasty in acute to prevent early cement leaks in the vena cava or
in all reports [I, 17]. In addition to its use for pri- vertebral collapses. We did not include chronic in the perivertebral veins or both [19]. We and
mary metastatic disease in vertebral bodies, per- pain because it is more difficult to correlate other groups of authors [4. 5] recommend CT to
cutaneous injection of cement into severely that pain with the collapse of one vertebra. In- facilitate accurate needle placement. particu-
osteoporotic vertebral bodies has been per- deed, chronic pain could be the consequence larly in the thoracic spine using the transpedicu-
formed. Mathis et al. [18] describe a woman with of the collapse on vertebral statics. lar approach. CT also allows accurate
multiple compression fractures of the lower tho- In the inclusion time ofour study, two patients evaluation of cement leakage. but close correla-
racic and lumbar spine after steroid therapy that with painful osteoporotic collapses did not un- tion between CT findings and clinical complica-
were greatly improved by multiple percutaneous dergo vertebroplasty because of the location of tions is essential to confirm the validity of
vertebroplasties. In a series of four patients, the vertebral collapses (T5 and T6). We believe postprocedural CT evaluation. The potential

Fig. 6-74-year-old woman who presented after procedure with left cruralgia that decreased rapidly in 1 week.
A and B, CT scans obtained before (A) and during (B) vertebroplasty show left foramina (arrow, A) before injection and suitable position of needle in middle of vertebral
body. Note absence of contrast material and calcification in region of neural foramen (arrow, A).
C, CT scan obtained after vertebroplasty shows leak of methyl methacrylate through left foramina (arrow).

1688 AJR:173, December 1999

 Cement Injections for Osteoporotic Vertebral Collapse

We noted no general complications, particu-

larly no pulmonary emboli caused by acrylic
cement as described by Padovani et al. [22].
These authors considered that a greater risk of
venous leak occurs with highly vascular lesions
Downloaded from www.ajronline.org by 95.90.213.214 on 03/16/17 from IP address 95.90.213.214. Copyright ARRS. For personal use only; all rights reserved

such as metasta.ses or vertebral angiomas.

Even though percutaneous vertebroplasty
does not modify the long-term outcome as
compared with conservative therapy. it facili-
tates analgesia and early mobility. which may
be crucial in older patients. In addition. verte-
broplasty is obviously a cost-effective proce-
dure (mean hospitalization. 5 days) when
compared with the cost of conservative ther-
apy (which in our center includes 2 weeks of
hospitalization and 3 weeks in a rehabilitation
center). Accurate cost-effective analysis
should also include additional costs incurred
by bed rest in older patients.
We performed percutaneous vertebroplasty
in 20 patients with acute porotic vertebral col-
lapse. The procedure was fast and produced no
major complications. All patients showed im-
mediate objective improvement (graded by an-
Fig. 7-77-year-old woman with new porotic vertebral collapse of 13 1 month after vertebroplasty of Li and L4.
A, Ti-weighted MR image (TRITE, 500/12) shows acute collapse of Li and L4, with preserved areas of normal intensity. alogic visual scale), and only one vertebral
Chronic vertebral collapse of L2 and L5 shows normal fat intensity signal. MR image shows abnormal signal of upper collapse of an adjacent vertebra was seen at the
end plate of 13 without any collapse at time of vertebroplasty. 6-month follow-up. Although our study had no
B, Short inversion time inversion recovery image (TRT1E, 4000/60; flip angle, 1 800:inversion time, i 50 msec) shows band
of high signal intensity in superior part of Li, L3, and L4 vertebral bodies.
control group, our results are encouraging. and
we recommend this treatment for patients with
uncontrollable pain that prevents ambulation.
Our treatment facilitates analgesia and allows
disadvantage of CT guidance alone is the lack shown to decrease local infection risk in or- early standing even if it does not modify the
of real-time visualization of cement leaks. Ac- thopedic prostheses. long-term outcome when compared with con-
cording to Barr et al. [10], this drawback may The leakage rate must be compared only in servative therapy.
be avoided by smaller and more closely moni- studies that use fluoroscopy to perform the
tored injections. We observed a relatively high procedure and CT to visualize eventual leaks,
incidence of leakage. Although these clinical because CT is more accurate for detecting sub- References
consequences were observed in only one pa- tle leaks, as described by Cotten et al. [20]. I . Cotten A, Boutry N, Cortet B. et al. Percutaneous
tient, smaller and more closely monitored bolus Our leakage rate was better than the rate of vertebroplasty: state of the art. RadioGraphics 1998:
injections could be recommended. Ideally. ac- those researchers (40% versus 72%, respec- 18:311-320
cording to Barr et al.. the safest procedure tively). However, randomized prospective tn- 2. Galibert P. Deramond H. Rosat P. Le Gars D. Preim-

would involve a combination of CT and fluoro- mary note on the treatment of vertebral hemangioma
als are essential for accurate comparisons.
by percutaneous acrylic vertebroplasty. Neurvehirur-
scopic guidance. Only prospective randomized A major problem of local treatment is the
gie 1987:33:66-68
studies could compare guidance methods in risk of collapse in vertebrae adjacent to the 3. Deramond H, Despriester C. Galibert P. Le Gars D.
terms of safety. duration, and cost. one treated that may be osteoporotic but still Percutaneous vertebroplasty with polymethyl-
Our single-side technique has the advan- not collapsed [3]. Gardos et al. [21] reported methacrylate: technique, indications. and results.
tage of being easy and fast. However. because a high relative risk of collapse of adjacent Radio! Clin North Am 1998:36:533-546
4. Kaemmerlen P. Thiesse P. Bouvard H, Biron P.
no randomized study has been conducted to treated vertebrae. We observed only one early
Momex F, Jonas P. Venebroplastie percutan#{233}edans
compare single and double approaches, defin- adjacent vertebral collapse that could already
le traitement des m#{233}tastases:technique et r#{233}sultats.
J
itive conclusions as to the benefits and draw- have been suspected on the preprocedural
Radio! 1989:70:557-562
backs of these techniques cannot yet be put MR imaging. In the clinical and radiologic 5. Kaemmerlen P. Thiesse P. Jonas P. et al. Percutane-
forward. The double approach should be bet- follow-up 6 months after the procedure, we ous injection of orthopedic cement in metastatic r-
ter for filling of the vertebrae. However. a did not note any new collapse of adjacent ver- tebral lesions. New EngI J Med 1989:321:121

study on a population with vertebral collapses tebrae. These data are in agreement with the 6. Weill A, Chiras J, Simon JM, Sola-Martinez T, En-
kaoua E. Spinal meta.stases: indications for and re-
caused by metastatic disease showed that results of Deramond et al. [31, who concluded
suits of percutaneous injection of acrylic suiical
pain relief was not proportional to the per- that little risk is specifically due to the proce-
cement. Radiology 1996:199:241-247
centage of lesion filling [20]. We agree with dure and that collapses may occur in any pa- 7. Laredo JD, Lakhdari K. Bellaiche L Hamze B, Jan-
Jensen et al. [19] on the importance of sup- tient at all vertebral levels during osteoporotic klewicz P. Tubiana JM. Acute vertebral collapse: CT
plementary antibiotics. which have been disease evolution. findings in benign and malignant nontraumatic

AJR:173, December 1999 1689

 Cyteval et al.

cases. Radiology 1995:194:41-48 broplasty in symptomatic cervical vertebral hae- 19. Jensen ME. Evans Al, Mathis JM. Kallmes DF.
8. Cuenod CA, Laredo JD, Chevret S. et al. Acute ver- mangiomas: report of 2 cases. Neumradiologv Cloft HJ. Dion JE. Percutaneous polymethyl-
tebral collapse due to osteoporosis or malignancy: 1996:38:389-391 methacrylate vertebroplasty in the treannent of os-
appearance on unenhanced and gadolinium-en- 15. DoussetV, Mousselard H, de Monck dUser L et al. teoporotic vertebral body compression fractures:
hanced MR images. Radiology 1996;199:541-549 Asymptomatic cervical haemangioma treated by technical aspect. AJNR 1997:18:1897-1904
9. Huskisson EC. Measurement of pain. Lancer 1974; percutaneous vertebroplasty. Neumradiology 1996: 20. Cotten A, Dewatre F, Cortet B, et al. Percutaneous
vertebroplasty for osteolytic metastases and rny-
Downloaded from www.ajronline.org by 95.90.213.214 on 03/16/17 from IP address 95.90.213.214. Copyright ARRS. For personal use only; all rights reserved

2:1127-1131 38:392-394
10. Barr MS. Barr JD, Milton S. Percutaneous svrtebro- 16. Ide C. Gangi A. RimmelinA, et al. Vertebral haeman- eloma: effect of the percentage of lesion filling and
plasty (invited commentary). RadioGraphics 1998; giomas with spinal cord compression: the place of the leakage of methyl methacrylate at clinical fol-
18:320-322 preoperative percutaneous vertebroplasty with methyl low-up. Radiology 1996:200:525-530
11. Melton LI. Epidemiology of spinal osteoporosis. methacrylate. Neumradiology 1996:38:585-589 21. Ganios F. Hardy N, Caymile G, et al. Traitement des
Spine 1997;22:2S-l IS 17. Gangi A, Kastler BA, Dietemann JL. Percutaneous fractures vert#{233}bralesost#{233}oponxiques par vert#{233}bro-
12. Gunnar B, Andersson J, Weinstein iN. Focus issue vertebroplasty guided by a combination of Cl and plasliepercutan#{233}e(abstr).RevRhwn EdFrl997;1 1:757
on osteoporosis. Spine 1997;22:IS fluomscopy.AJNR 1994:15:83-86 22. Padovani B. Kasriel 0, Brunner P. Pereui-Viton P.
13. BrostromPG,LaneJ.Augmentationofosteopoiiic wr- 18. Mathis MJ, Petit M, NaffN. Percutaneous vertebro- Pulmonary embolism caused by acrylic cement: a
tebralbodies: 199722:38S-42S
11JtUrediIeCfiCnS.frW plasty treatmentofstemid-inducedosteoporotic corn- rare complication of percutaneous vertebroplasty.
14. Feydy A, Cognard C. MiauxY. et al. Acrylic verte- pression fractures.Arthritis Rheum 1998:41:171-175 AJNR 1999:20:375-377

1690 AJR:173, December 1999