Eur Spine J DOI 10.1007/s00586-014-3244-8
ORIGINAL ARTICLE
Magnetic resonance imaging changes of intervertebral discs after kyphoplasty M. A. Ko¨nig • S. Panzer • J. Schulz M. Bierschneider • B. M. Boszczyk
•
Received: 3 December 2013 / Revised: 14 February 2014 / Accepted: 16 February 2014 Ó Springer-Verlag Berlin Heidelberg 2014
Abstract Background Minimal-invasive cement augmentation techniques gained popularity recently. Long-term studies, however, are still not available focusing on the effect of possible acceleration of intervertebral disc degeneration. Materials and methods Fifteen patients (average age 67.1 ± 6.9 years, range 58–77; 10 female, 5 male) with acute or osteoporotic fractures were included in this study and MRI scans were performed before surgery and after a mean follow-up period of 15.2 months (range 8–27 months). Out of these patients, seven were available for a long-term MRI scan after a mean of 94.3 months (range 84–96 months). Disc degeneration and injuries were graded according to published Pfirrmann and Oner scales. Results A total of 43 intervertebral discs with moderate initial degeneration were examined pre-operatively and at the first follow-up. Twenty were available for the long-termfollow-up. At the first follow-up, 3 (1.3 %) discs showed a degenerative progression of 1 grade compared to the preoperative MRI. Only one injured and one uninjured disc (0.4 %) showed progressive degeneration of 1 grade in the
M. A. Ko¨nig B. M. Boszczyk (&) The Centre for Spinal Studies and Surgery, Queens Medical Centre Nottingham, Nottingham, UK e-mail: [email protected] M. A. Ko¨nig e-mail: [email protected] S. Panzer Department of Radiology, Berufsgenossenschaftliche Unfallklinik Murnau, Murnau, Germany J. Schulz M. Bierschneider Department of Neurosurgery, Berufsgenossenschaftliche Unfallklinik Murnau, Murnau, Germany
long-term follow up. No intervertebral disc in-between bisegmental cement augmentation showed acceleration of degenerative changes. Conclusion Despite several limitations regarding patients’ age and lack of performed perfusion MRI scans, this study suggests that vertebral cement augmentation through kyphoplasty has no significant influence on disc degeneration even after a long period. The absence of severe disc degeneration after vertebral augmentation supports further clinical trials, which should incorporate endplate perfusion studies for detailed information regarding disc perfusion. Keywords Kyphoplasty Intervertebral disc degeneration Magnetic resonance imaging Vertebral compression fracture
Introduction In the course of the last two decades, encouraging clinical results have been achieved in the management of osteoporotic vertebral fractures through the techniques of vertebroand kyphoplasty [1, 5, 9–12, 15, 17, 34]. The minimally invasive nature of these methods and the low overall complication rate have consequently prompted the expansion of the indications for vertebral augmentation to traumatic fracture types or fractures due to early onset of secondary osteoporosis [6, 29]. However, no long-term adverse effects of vertebral augmentation with polymethylmethacrylate (PMMA) have been published and very little is known of the biological effects upon the endplate and intervertebral disc. A 30 % decrease of the perfusion surface seems to be sufficient to attain a 50 % perfusion deficit to the nucleus [32] and as the supply of nutrients to the intervertebral disc is provided to a significant degree through the vascularised
123
Eur Spine J
vertebral endplate [25, 27], compromise of blood flow to and from the endplate may be expected to have a detrimental effect. Diffusion in degenerative discs also becomes, compared to healthy uninjured discs, more and more irregular and patchy [26]. This in turn could, in theory, lead to accelerated intervertebral disc degeneration with the possibility of a delayed onset of symptoms associated therewith. With the expansion of the indications for vertebral augmentation, procedure-related disc degeneration could reach clinical significance. Despite the clinical importance of this topic, only very little attention has been paid to the actual extent of disc changes after vertebral cement augmentation [24]. The purpose of this investigation is, therefore, twofold: (a)
(b)
To analyse the influence of primary endplate and disc injury severity on subsequent disc degeneration after vertebral augmentation. To analyse the influence of vertebral augmentation on the degeneration of adjacent discs in general.
Materials and methods Hospital records were reviewed for patients treated with kyphoplasty for osteoporotic or traumatic vertebral fractures of the thoracolumbar and lumbar spine whereas cases with neoplastic disease of the spine or infections were excluded. Patients with osteoporotic fractures were selected if the preoperative MRI scan was available at our institution and patients with traumatic fractures were selected if these were isolated injuries, regardless of the availability of a preoperative MRI scan. In total, 24 patients were identified, approached and counselled regarding the proposed followup investigation. Fifteen patients (average age 67.1 ± 6.9 years, range 58–77; 10 female, 5 male) agreed to MRI follow-up (Tables 1, 2, 3) and informed consent was obtained from all patients for the extent of the MRI investigation. Of these, seven patients were available for long-term follow-up MRI scan. As far as could be determined, refusal of participation was not related to the initial procedure or the outcome thereof in any of the declining patients. One of the recruited patients (CC) had suffered traumatic paraplegia following an L1 fracture more than 20 years previously, with resulting inactivity-induced osteoporosis. Five patients (SK, WM, EV, BH, MM) presented with primary osteoporosis and two patients with cortisone induced secondary osteoporosis in the setting of rheumatoid arthritis (BI, PH). In the remaining two patients (PA, BF), vertebral injuries following minor trauma had occurred, without obvious or known osteoporosis. In addition to the osteoporosis patients, five further patients (MK, AE, AR, BF, PH) with single genuine traumatic fractures were included. During the follow-up period, three patients (CC, PH, WM)
123
suffered additional fractures, three of which were treated with kyphoplasty in the first 8 weeks after the first cement augmentation and one who was treated conservatively. These vertebrae were added to the investigation as they were adjacent to treated vertebrae and, therefore, could potentially influence the outcome of discs already included in the study (Patient CC and PH see Tables 1 and 2). The adjacent disc levels with untreated vertebral bodies were identified and included in order to observe the natural history of disc degeneration. In total, 51 discs were observed with 23 in the thoracic spine and 28 in the lumbar spine at the first follow-up. At the long-term follow-up, nine thoracic and 11 lumbar discs were included. Surgical technique Twenty fractures were treated with a biportal access kyphoplasty, three via a unilateral perpendicular approach and two with semi-open kyphoplasty. All cement augmentations had fluoroscopically controlled needle placement to achieve an optimal result of balloon fracture reduction and cement application [33]. The average amount of PMMA cement used in the lumbar spine was 7.6 ± 1.9 ml (range 4–10.5 ml) and 5.7 ± 1.3 ml (range 4–8 ml) in the thoracic vertebrae and, therefore, all fractures were treated adequately [7]. Seven discs were in-between bi-segmental cement augmentations. MRI evaluation The MRI scan prior to surgery was available for osteoporotic fracture patients and in 3 trauma patients. In addition to the pre-operative fracture classification according to the scale proposed by Magerl et al. [18], the presentation of the fracture on the initial MRI was also reviewed. The affected vertebral body and intervertebral discs were graded according to the systems proposed by Pfirrmann et al. [23] (Table 4) and by Oner et al. [19, 20] (Table 5). The grading system by Oner et al. [19, 20] assesses the injury severity of the vertebral body, endplates, adjacent discs and spinal ligaments. The grading system by Pfirrmann et al. [23] grades degeneration according to the state of hydration in uninjured discs adjacent to fractured vertebrae. At both follow-up assessments, the degree of degeneration of all discs adjacent to augmented vertebral bodies was graded according to Pfirrmann et al. [23]. The MRI investigation pre- and post-operatively was performed with the same MRI machine (1.0 Tesla, Magnetom Expert, Siemens, Erlangen, Germany). The longterm follow-up MRI scans were all done with 1.5T (Achieva, Philips Medical Solutions, Best, The Netherlands). MRI settings were TSE; T2; TE 120 ms; TR 4,000 and TIR, TE 60 ms; TR3366.
Eur Spine J Table 1 Patients with single-level kyphoplasty: initial data and first follow-up Patient
Age
Sex
AE
72
F
Level
Volume (ml)
Access
4.5 ? 4.5
bilat
T12/L1 L1 L1/L2
AR
66
F
L2/L3 L3
4?4
64
M
T12/L1 L1
5.5 ? 5
64
F
T10/T11 T11 T11/T12
MK
69
M
76
M
59
F
F
67
75
IN
3
1
IN
4
IN
4 4
3
3
3
4
5
5
2
1
3
4
1
M
T11/T12
3
3
1
M
T12/L1 T10/T11
4 3
4 3
4
4
3
3
IN
4
3?1
L4/L5 L5
5
L5/S1
3
3
3
open uni 3
2
2
1
2
A3.1
2
1
1
3
A3.1
1
2
1
3
A3.1
2
2
1
3
A1.1
1
2
1
3
A1.2
1
1
1
2
A1.2
1
2
2
1
3
A3.1
3
2
2
2
3
A3.3
1
2
2
2
2
A1.2
2
1
1
3
A1.2
2
2
1
3
A3.1
1 2 1 1 1 4
3
A3.1
1
2 3
3
1 1 1
bilat
L5/S1 F
1
1
L4/L5 L5
1
1 3
T11/T12
F
2
1 1
CONS
2
1 1
4
bilat
A3.1
1
1
3?3
3
1 1
4
bilat
1
1
4
3?3
2
4
1
bilat
2
2
T11/T12
2?2
AO
4 1
1
T10/T11
COR
1
1
bilat
PLC
1 2
3
4?4
PLL
1
4 4
ALL
1
2
bilat
T12/L1
75
2
DI
3
T12a
BI
EPi
L1/L2
T11
76
4
3
T11/T12
EV
4
3
3?3
EPs
1
3
T12 WM
3
3
T11 SK
3
T12/L1
T12 61
3
bilat
T12/L1 PA
3
T12/L1 L1 PH
3?3
3
bilat
T11/T12 T12
MM
2.5 ? 3
3
bilat
L1/L2 LM
Pfirr FU 1
bilat
L3/L4 BF
Pfirr pre
1 1
Patient data and MRI grading results for injured vertebral bodies M male, F female. Levels vertebral body and adjacent intervertebral discs. Volume applied cement volume. Access: unilat unilateral, bilat bilateral, open uni open unilateral. Pfirr pre pre-operative Pfirrmann grade, PfirrFU 1 Pfirrmann grade at first follow up [23] (see Table 5 for grading). AO fracture classification according to Magerl et al. [18]. DI; EPs; EPi; COR; ALL; PLL; PLC spinal column injury classification according to Oner et al. [19, 20] (see Table 4 for abbreviations and grades). ‘‘IN’’ in DI columns indicate injured discs. ‘‘CONS’’ conservative fracture treatment. ‘‘OLD’’ old traumatic fracture a
Fractured vertebra after pre-operative MRI scan
All authors of this investigation graded all MR images separately and blindly according to the Pfirrmann and Oner classifications. Wherever there was a grading difference, a consensus of all three was made and the result obtained. This
investigation was focused specifically on disc degeneration after cement augmentation and, therefore, no attempt was made to correlate clinical outcome, grade of height restoration or kyphosis angle with MRI findings.
123
Eur Spine J Table 2 Patients with multilevel kyphoplasties: initial data and first follow-up Patient
Age
Sex
CC
57
F
Level
Volume (ml)
Access
4
unilat
T8/T9 T9a T9/T10 T10
3?4 3.5 ? 3.5 6
4?4 4?4
56
F
T8/T9 T9
4
4
IN
4
3
4
4 4
4 4
EPi
1
1
DI
ALL
PLL
PLC
COR
AO
1 A1.1. 1 1
2
2
1
1
3
A1.2
1 1
1
A1.1 4
2
1
2
1
1
3
A1.1
1
2
1
1
A3.1
1
2
1
2
A1.1
1
1
1
1
A3.1
1
1
1
2
A3.1
2
3
1
3
A3.3
1
4
IN
3
5
5
5
5 4
3
3
1 1 1
1 3
2 2
1
unilat
T9/T10
1 3
bilat
L5/S1 PH
4
bilat
L4/L5 L5
4
EPs
OLD
L1/L2 L3/L4 L4
4
unilat
T12/L1 L1
4
bilat
T11/T12 T12
Pfirr FU 1
bilat
T10/T11 T11a
Pfirr pre
1 1
T10 T10/T11 T11
3?3
bilat
T11/T12 T12
3
L1a
4?4
L2/L3
BH
69
F
4.5 ? 4.5
4
3
3
3
3
IN
3
2 1
1
bilat
1 1
2
bilat
L1/L2 L2 L3
4 open uni
T12/L1
1 1
1
A3.1 1 2
2
1
L3/L4
3
3
1
L3/L4
4
4 PMMA
1
4
4 PMMA
L4
4?4
bilat
L4/L5 L5
4?4
L5/S1
1
bilat 4
1
1
2
A1.1.
2
1
1
1
A1.1
2
2
1
3
A3.1
1 2
4
4
2
1 1
Patient data and MRI grading results for injured vertebral bodies M male, F female. Levels vertebral body and adjacent intervertebral discs. Volume applied cement volume. Access: unilat unilateral; bilat bilateral, open uni open unilateral. Pfirr pre pre-operative Pfirrmann grade; PfirrFU 1 Pfirrmann grade at first follow up [23] (see Table 5 for grading). AO fracture classification according to Magerl et al. [18]. DI; EPs; EPi; COR; ALL; PLL; PLC spinal column injury classification according to Oner et al. [19, 20] (see Table 4 for abbreviations and grades). ‘‘IN’’ in DI columns indicate injured discs. ‘‘CONS’’ conservative fracture treatment. ‘‘OLD’’ old traumatic fracture a
Fractured vertebra after pre-operative MRI scan
Results Patient data with the first follow-up are summarised in Tables 1 and 2. Patient data of the long-term follow-up are shown in Table 3. For didactic reasons, the tables show
123
only the affected vertebrae and adjacent discs. A total of 15 patients with a sum of 25 fractures ranging from T9 to L5 were included. Seven fractures were classified as A1.1 (28 %), 4 as A1.2 (16 %), 12 as A3.1 (48 %) and 2 as A3.3 (8 %) fractures according to Magerl [18].
Eur Spine J Table 3 Single level and multilevel kyphoplasties at long-term follow-up Patient
Age
Sex
Level
AR
66
F
L2/L3 L3
Volume (ml)
4?4
Access
64
M
T12/L1 L1
5.5 ? 5
MK
64
69
F
M
T11/12 T12
59
F
76
F
57
F
3
4
4
4
IN
3
3 4
IN
4
5
bilat
2
T11/ T12
3
3
4
1
T12/L1
3
3
4
1
T11/ T12
4
4
4
1
4
4
4
IN
4
4
L4/L5 3?1
1
bilat
2
bilat
1
1
1
3
3
3
1
4
4
4
1
4
4
4
4
unilat
T10
3?4
1
bilat
T10/ T11
4
4
IN
4
4
3
4
4
L1/L2
4
4
4
L3/L4
4
4
4
T11
3.5 ? 3.5
bilat
T11/ T12 T12
6
L1
L4
L5/S1
A3.1
2
2
1
2
A3.1
2
1
1
3
A3.1
1
2
1
3
A3.1
1
2
1
3
A1.2
2
1
1
3
A1.2
1
A1.1.
2
2
1
1
3
A1.2
1
A1.1 4
2
1
2
1
1
3
A1.1
1
2
1
1
A3.1
1
2
1
2
A1.1
1
OLD
4?4
bilat
L4/L5 L5
3
1 1
unilat
T12/L1
1
1 1
4
2
2 2
L5/S1
T9/T10
2
1
T8/T9 T9
AO
4 3 1
4?4
COR
2 4
bilat
PLC
4 1
4
3 ?3
PLL
1 2
4
ALL
1 1
IN
DI
4
L5 CC
3
EPi
4
T12 T12/L1 EV
3
EPs
T12/L1
T12 PH
2.5 ? 3
PfirrFU 2
bilat
L1/L2 LM
Pfirr FU 1
bilat
L3/L4 BF
Pfirr pre
IN 4?4
1 3
3
3
bilat
2 1
5
5
2
5
1 1
Patient data and MRI grading results at long-term follow-up M male, F female, Levels vertebral body and adjacent intervertebral discs, Volume applied cement volume. Access: unilat unilateral, bilat bilateral, open uni open unilateral, Pfirr pre pre-operative Pfirrmann grade, PfirrFU 1 Pfirrmann grade at first follow up, PfirrFU 2 Pfirrmann grade long-term follow-up [23] (see Table 5 for grading). AO fracture classification according to Magerl et al. [18]. DI; EPs; EPi; COR; ALL; PLL; PLC spinal column injury classification according to Oner et al. [19, 20] (see Table 4 for abbreviations and grades). ‘‘IN’’ in DI columns indicates injured discs. ‘‘CONS’’ conservative fracture treatment. ‘‘OLD’’ old traumatic fracture
The initial follow-up includes 21 fractures and 43 surveyed discs of which seven were adjacent to bisegmental augmentation. Fifty-one discs distant to vertebral augmentation were observed to determine the natural disc degeneration.
The long-term follow-up included five lumbar and seven thoracic vertebral fractures. Twenty adjacent discs of which four were between augmented vertebrae were evaluated. Twenty discs remote from augmented vertebrae
123
Eur Spine J Table 4 MRI classification of spinal column injury according to Oner et al. [20] Grade 1
Grade 2
Grade 3
Grade 4
Disc (DI)
No evidence of injury in the disc space
Rupture and/or debris in the anterior half of the disc space
Rupture and/or debris in the posterior half of the disc space
Involvement of the whole disc. Either the disc is completely herniated into the vertebral body and/or there is rupture and debris in the whole disc space
Endplate (EP)
Only plastic deformity of the endplate without disruption
Disruption in the posterior half of the endplate
Disruption of the whole endplate
Vertebral body (COR)
Less than one-third of the volume of the vertebral body is involved (bone marrow oedema)
Disruption in the anterior half of the endplate(evident discontinuity of low signal line) One to two-thirds of the vertebral body is involved
Anterior longitudinal ligament (ALL)
No evidence of injury
The ligament is slackened but continuous
The ligament is ruptured
Posterior longitudinal ligament (PLL)
No evidence of injury
The ligament is attached to the extruding bone fragment from the posterior cortex and continuous
The ligament is ruptured
Posterior ligamentary complex (PLC)
No evidence of injury
Oedema in the interspinous space without evident discontinuity or elongation
Elongation of the interspinous space without evident discontinuity
More than twothirds of the vertebral body is involved
Clear disruption
Endplate injury is differentiated between the superior (EPs) and inferior endplate (EPi)
were once again observed for disc degeneration without
Pre-operative MRI
Table 5 MRI classification of intervertebral disc degeneration according to Pfirrmann et al. [23] Grade 1
Grade 2
Grade 3
Grade 4
Grade 5
The structure of the disc is homogenous, with a bright hyperintense white signal and a normal disc height
The structure of the disc is in homogenous, with a hyperintense white signal. The distinction between nucleus and anulus is clear, and the disc height is normal, with or without horizontal grey bands
The structure of the disc is in homogenous with an intermediate grey signal intensity. The distinction between nucleus and anulus is unclear, and the disc height is normal or slightly decreased
The structure of the disc is in homogenous, with a hypointense dark grey signal intensity. The distinction between nucleus and anulus is lost, and the disc height is normal or moderately decreased
The structure of the disc is in homogenous, with a hypointense black signal intensity. The distinction between nucleus and anulus is lost, and the disc space is collapsed
cement augmentation. MRI evaluation A first follow-up MRI scan was obtained after a mean follow-up time of 15.2 months (range 8–27 months).The second MRI scan, defined as long-term follow-up, included seven patients after a mean follow-up time of 94.3 months (range 84–96 months). An exemplary case of a L1 fracture is shown in Fig. 1, with pre-operative findings followed by the first follow-up and long-term results.
123
Disruption of spinal ligaments could be excluded in all patients, but the extent of vertebral body injury and endplate disruption varied considerably. Only three treated vertebrae had limited injury of the vertebral body (grade 1). The remaining treated vertebrae were found to have a grade 2 (5 vertebrae) or grade 3 (14 vertebrae) injury. Since three fractures occurred after the baseline MRI scan, grading was not possible in these cases. Grade 1 was present in 30 (57.7 %) of the total of 52 endplates (superior and inferior) of the injured vertebrae. Grade 2 and 3 anterior or posterior disruption was present in 19 endplates (36.5 %). Severe injury was identified in 3
Eur Spine J
Fig. 1 Patient BF: L1 fracture. EPs 2, EPi 1, DI 4 (T12/L1) and DI 2 (L1/L2) pre-operatively according to Oner [19, 20] (Table 4). Disc degeneration grade 3 at follow-up according to Pfirrmann [23] (Table 5) and no changes in long-term follow-up
(5.8 %) endplates of three vertebrae in three patients. Uninjured discs adjacent to augmented vertebral bodies mostly showed mild to moderate degeneration preoperatively. Nineteen discs showed a Pfirrmann grade 3 degeneration (44.2 %), 15 (35 %) had grade 4 and 2 (3.8 %) had grade 5 degeneration pre-operatively. Seven injured discs were identified, four had rupture or debris in the anterior half of the disc space and three had involvement of the whole disc (Table 1). The group without cement augmentation showed similar degeneration: five had grade 2 (3.9 %), 32 grade 3 (62.7 %), 11 grade 4 (21.6 %) and 3 (5.8 %) grade 5 degeneration according to the Pfirrmann classification.
severe endplate injuries showed no progressive degeneration similar to the discs in between bilateral cement augmentations. PMMA disc leaks in both adjacent disc levels after kyphoplasty occurred in one patient, with an inferior endplate injury grade 4 and a superior endplate injury grade 1 (Patient BH Table 2), but no further disc degeneration was seen at follow-up. Degeneration in discs apart from the cement augmentation appeared in 14 cases in the thoracic and lumbar spine. The majority of the thoracic discs (4) degenerated from Pfirrmann grade 3 to 4 and one disc degenerated from grade 2 to 3. In the lumbar spine, six discs showed degeneration of grade 3–4, two from grade 2 to 3 and one from grade 4 to 5.
MRI follow-up
MRI long-term follow-up
At the first follow-up, the majority of uninjured discs did not show any evidence of progressive degeneration. Only three discs showed degeneration of one Pfirrmann grade from 3 to 4. Two discs were below the treated vertebra with grade 1 or 2 inferior endplate injuries, while one disc was above the cement augmentation with a vertebral superior endplate injury of grade 1. One instance of disc degeneration occurred in between a fresh osteoporotic fracture T12 (Magerl A1.1) and the old traumatic fracture L1 (Patient CC, Table 1), while the two remaining disc degenerations appeared in the lower thoracic area T9 and T11 (Magerl A1.2 and A3.1) of two patients (PA and PH Table 1). The seven injured discs presented with grade 3 or 4 degenerative changes. Interestingly, discs with associated
Table 3 summarises the data of seven patients from preoperative findings and both follow-up measurements to point out possible disc degeneration. 19 (95 %) discs under surveillance showed no degeneration compared to preoperative findings or the first follow-up. The intervertebral disc of T12/L1 in one patient had a progressive degeneration from 3 to 4 according to Pfirrmann with no initial superior and inferior endplate injury after a traumatic fracture of T12 (Patient MK Table 3). One initial injury with total disc involvement showed degeneration of 1 grade from 4 to 5 compared to the first follow up with grade 4 superior and grade 3 inferior endplate injury (Patient LM Table 3). Only one of the three early disc degenerations at the first follow-up was available for long-
123
Eur Spine J
term results but this disc showed no progressive degeneration (Patient CC Table 3). Again, no disc degeneration was seen in discs in between bisegmental cement augmentations. In the surveyed group apart from cement augmentation, four new degenerations of 1 grade occurred in the lumbar spine and four in the thoracic spine. One thoracic disc showed even a degeneration of two grades from 2 to 4 (Patient PH Table 3). Two discs degenerated even more from the first follow-up to the long-term follow-up from 3 to 4.
Discussion Cement augmentation with PMMA is established as a treatment option for osteoporotic fractures and lytic tumour lesions [1, 9–11]. Advantages are of course the less invasive procedure itself and good mechanical support [4, 34], making vertebro- and kyphoplasties suitable for a large group of patients with proven effectiveness of these treatment options [12, 15, 17, 22]. Adverse effects were only reported during or immediately after cement application ranging from asymptomatic appearances to patients’ death [3, 8, 13, 14, 16, 21, 28, 30, 31] or focusing on the adjacent vertebral bodies [2]. No long-term studies have been published and no information focusing on the effect of cement augmentation on the intervertebral discs is currently available. This restricts its use with younger patients due to concerns related to the effect of PMMA upon adjacent intervertebral disc vitality and long-term sequelae. The intervertebral disc as largest inert structure in the human body is highly dependent on endplate diffusion for nutritive agents and diffusion surface reduction causes deficits in the nucleus [25, 26, 32]. Qian et al. [24] interestingly showed in their prospective study of 97 patients an accelerated disc degeneration after a short-term follow-up of 2 years. Up to 52.6 % of the intervertebral discs next to the cement augmented vertebral body showed degenerative changes. Unfortunately, no information is given regarding existing disc degeneration, so no conclusion could be made if cement augmentation accelerates disc degeneration. It is of note that 29 % of the control group were found to have disc degeneration. At least one of the case examples shows degeneration of multiple discs. The subjects seem to also have had a less degenerative overall outset than the patients presented in this observation. Other factors include a different ethnic population and a single observer in the Qian study. In the present investigation, 15 patients were followed up with an MRI (at an average of 15.2 months after being treated) with PMMA kyphoplasty for one or more vertebral fractures. Out of these 15 patients, seven were applicable
123
for a long-term follow-up after a mean of 94.3 months, to measure possible long-term adverse effects of cement augmentation to the intervertebral discs. Uninjured intervertebral discs adjacent to treated vertebral bodies did not show a rapid progression of degeneration according to the Pfirrmann grading system in majority of the cases. Three discs had a progression of 1 grade according to Pfirrmann [23] at the first follow-up and injured discs had moderate disc degeneration at further follow-up. In the long-term follow-up of the seven patients, no progression of degeneration was seen in 19 out of 20 adjacent discs indicating that cement augmentation procedures with PMMA do not significantly affect the endplate diffusion in vertebral fractures. Even in previously injured discs, PMMA cement seems not to accelerate disc degeneration significantly since only one disc with severe endplate injury had deteriorated by 1 grade. Interestingly, no accelerated disc degeneration was found in intervertebral discs with bisegmental cement augmentation. Compared with the natural history of disc degeneration in these patients, PMMA cement augmentation does not seem to accelerate disc degeneration dramatically indicating safe usage even if bisegmental augmentation is needed. Despite the fact that this is the first study focusing on disc degeneration after cement augmentation in a long-term follow-up, it has several limitations. First of all, patients with osteoporotic fractures usually present with disc degeneration due to their age which limits the comparison of pre- and post-operative intradiscal changes. This is because results may vary if the disc is less degenerated and, therefore, needs a higher diffusive threshold for greater supply with nutritive substances. Unfortunately, out of this small cohort of patients, only seven patients could be investigated in the long-term follow-up, which reduces the obtained information significantly regarding the results. Possibly more accelerated disc degeneration could have been found if more patients had been available for the last follow-up. In addition, this study lacks a proper control group as presented in the study by Qian et al. [24]. However, intervertebral discs without influence of cement augmentation were included for each patient in terms of observing the natural history of disc degeneration. As a last limitation, the endplate diffusion of the affected disc was not controlled with contrast studies according to Rajasekaran to determine the diffusion surface after kyphoplasty [25, 26].
Conclusion PMMA cement augmentation appears not to have a significant effect on accelerated disc degeneration in osteoporotic fractures. In this study, only minor progression of
Eur Spine J
degeneration was found in 5 out of 25 discs under surveillance in previously injured and uninjured discs. Bisegmental kyphoplasty did not accelerate disc degeneration. Future studies should include perfusion investigation of the endplates. If vertebral augmentation techniques are deemed useful in young patients with traumatic fractures, more focus should be laid on the intervertebral discs, since PMMA cement might influence endplate diffusion more in less degenerated discs. Conflict of interest No benefits or funds in any form have been received or will be received from a commercial party related directly or indirectly to the subject of this article.
References 1. Berlemann U, Heini PF (2002) Percutaneous cementing techniques in treatment of osteoporotic spinal sintering. Unfallchirurg 105:2–8 2. Berlemann U, Ferguson SJ, Nolte LP, Heini PF (2002) Adjacent vertebral failure after vertebroplasty. J Bone Jt Surg Br 84:748–752 3. Bernhard J, Heini PF, Villiger PM (2003) Asymptomatic diffuse pulmonary embolism caused by acrylic cement: an unusual complication of percutaneous vertebroplasty. Ann Rheum Dis 62:85–86 4. Belkoff SM, Mathis JM, Fenton DC, Scribner RM, Reiley ME, Talmadge K (2001) An ex vivo biomechanical evaluation of an inflatable bone tamp used in the treatment of compression fracture. Spine 26:151–156 5. Boszczyk BM, Bierschneider M, Robert B, Jaksche H (2003) Treatment of severe osteoporotic fractures through a microsurgical interlaminary approach. In: Szpalski M, Gunzburg R (eds) Vertebral osteoporotic compression fractures. Lippincott WW, Philadelphia, pp 179–188 6. Boszczyk B, Bierschneider M, Potulski M, Robert B, Vastmans J, Jaksche H (2002) Extended kyphoplasty indications for stabilization of osteoporotic vertebral compression fractures. Unfallchirurg 105:952–957 7. Boszczyk B (2010) Volume matters: a review of procedural details of two randomised controlled vertebroplasty trials of 2009. Eur Spine J 19(11):1837–1840 8. Chen HL, Wong CS, Ho ST, Chang FL, Hsu CH, Wu CT (2002) A lethal pulmonary embolism during percutaneous vertebroplasty. Anesth Analg 95:1060–1062 9. Dudeney S, Lieberman IH, Reinhardt MK, Hussein M (2002) Kyphoplasty in the treatment of osteolytic vertebral compression fractures as a result of multiple myeloma. J Clin Oncol 20:2382–2387 10. Fourney DR, Schomer DF, Nader R et al (2003) Percutaneous vertebroplasty and kyphoplasty for painful vertebral body fractures in cancer patients. J Neurosurg 98:21–30 11. Garfin SR, Hansen AY, Reiley MA (2001) Kyphoplasty and vertebroplasty for the treatment of painful osteoporotic compression fractures. Spine 26:1511–1515 12. Garfin SR, Lin G, Lieberman I, Phillips F, Truumees E (2001) Retrospective analysis of the outcomes of balloon kyphoplasty to treat vertebral body compression fracture refractory to medical management. Eur Spine J 10:S7–S8 (Abstract) 13. Harrington KD (2001) Major neurological complications following percutaneous vertebroplasty with polymethylmethacrylate. J Bone Jt Surg Am 84:1070–1073
14. Jang JS, Lee SH, Jung SK (2002) Pulmonary embolism of polymethylmethacrylate after percutaneous vertebroplasty. Spine 27:E416–E418 15. Ledlie JT, Renfro M (2003) Balloon kyphoplasty: One-year outcomes in vertebral body height restoration, chronic pain and activity levels. J Neurosurg 98:36–42 16. Lee BJ, Lee SR, Yoo TY (2002) Paraplegia as a complication of percutaneous vertebroplasty with polymethylmethacrylate. Spine 27:E419–E422 17. Lieberman IH, Dudeney S, Reinhardt MK, Bell G (2001) Initial outcome and efficacy of ‘‘kyphoplasty’’ in the treatment of painful osteoporotic vertebral compression fractures. Spine 26:1631–1638 18. Magerl F, Aebi M, Gertzbein SD, Harms J, Nazarian S (1994) A comprehensive classification of thoracic and lumbar injuries. Eur Spine J 3:184–201 19. Oner FC, van der Rijt RR, Ramos LMP, Dhert WJA, Verbout AJ (1998) Changes in the disc space after fractures of the thoracolumbar spine. J Bone Jt Surg Br 80:833–839 20. Oner FC, van Gils APG, Faber AJ, Dhert WJA, Verbout AJ (2002) Some complications of common treatment schemes of thoracolumbar spine fractures can be predicted with magnetic resonance imaging. Spine 27:629–636 21. Padovani B, Kasriel O, Brunner P, Peretti-Viton P (1999) Pulmonary embolism caused by acrylic cement: a rare complication of percutaneous vertebroplasty. Am J Neuroradiol 20:375–377 22. Perez-Higueras A, Alvarez L, Rossi RE, Quinones D, Al-Assir I (2002) Percutaneous vertebroplasty: long-term clinical and radiological outcome. Neuroradiology 44:950–954 23. Pfirrmann CW, Metzdorf A, Zanetti M, Hodler J, Boos N (2001) Magnetic resonance classification of lumbar intervertebral disc degeneration. Spine 26(17):1873–1878 24. Qian J, Yang H, Jing J, Zhao H, Ni L, Tian D, Wang Z (2012) The early stage adjacent disc degeneration after percutaneous vertebroplasty and kyphoplasty in the treatment of osteoporotic VCFs. PLoS One 7(10):e46323. doi:10.1371/journal.pone. 0046323 25. Rajasekaran S, Naresh-Babu J, Murugan S (2007) Review of postcontrast MRI studies on diffusion of human lumbar discs. J Magn Reson Imaging 25:410–418 26. Rajasekaran S, Babu JN, Arun R, Armstrong BR, Shetty AP, Murugan S (2004) ISSLS prize winner: a study of diffusion in human lumbar discs: a serial magnetic resonance imaging study documenting the influence of the endplate on diffusion in normal and degenerate discs. Spine 29(23):2654–2667 27. Rajasekaran S, Venkatadass K, Naresh Babu J, Ganesh K, Shetty AP (2008) Pharmacological enhancement of disc diffusion and differentiation of healthy, ageing and degenerated discs: results from in vivo serial post-contrast MRI studies in 365 human lumbar discs. Eur Spine J 17(5):626–643 28. Ratliff J, Nguyen T, Heiss J (2001) Root and spinal cord compression from methylmethacrylate vertebroplasty. Spine 26(13):E300–E302 29. Ross PD, Genant HK, Davis JW, Miller PD, Wasnich RD (1993) Predicting vertebral fracture incidence from prevalent fractures and bone density among none-black, osteoporotic women. Osteoporos Int 3:120–126 30. Scroop R, Eskridge J, Britz GW (2002) Paradoxical cerebral embolization of cement during intraoperative vertebroplasty: case report. Am J Neuroradiol 23:868–870 31. Tozzi P, Abdelmoumene Y, Corno AF, Gersbach PA, Hoogewoud HM, von Segesser LK (2002) Management of pulmonary embolism during acrylic vertebroplasty. Ann Thorac Surg 74:1706–1708 32. van der Werf M, Lezuo P, van Donkelaar CC, Ito K (2007) Inhibition of vertebral enplate perfusion results in decreased
123
Eur Spine J intervertebral disc intranuclear diffusive transport. J Anat 211:769–774 33. Wenger W, Markwalder TM (1999) Surgically controlled, transpedicular methyl methacrylate vertebroplasty with flouroscopic guidance. Acta Neurochir 141:625–631
123
34. Wilson DR, Myers ER, Mathis JM et al (2000) Effect of augmentation on the mechanics of vertebral wedge fractures. Spine 25:158–165
ORIGINAL ARTICLE
Magnetic resonance imaging changes of intervertebral discs after kyphoplasty M. A. Ko¨nig • S. Panzer • J. Schulz M. Bierschneider • B. M. Boszczyk
•
Received: 3 December 2013 / Revised: 14 February 2014 / Accepted: 16 February 2014 Ó Springer-Verlag Berlin Heidelberg 2014
Abstract Background Minimal-invasive cement augmentation techniques gained popularity recently. Long-term studies, however, are still not available focusing on the effect of possible acceleration of intervertebral disc degeneration. Materials and methods Fifteen patients (average age 67.1 ± 6.9 years, range 58–77; 10 female, 5 male) with acute or osteoporotic fractures were included in this study and MRI scans were performed before surgery and after a mean follow-up period of 15.2 months (range 8–27 months). Out of these patients, seven were available for a long-term MRI scan after a mean of 94.3 months (range 84–96 months). Disc degeneration and injuries were graded according to published Pfirrmann and Oner scales. Results A total of 43 intervertebral discs with moderate initial degeneration were examined pre-operatively and at the first follow-up. Twenty were available for the long-termfollow-up. At the first follow-up, 3 (1.3 %) discs showed a degenerative progression of 1 grade compared to the preoperative MRI. Only one injured and one uninjured disc (0.4 %) showed progressive degeneration of 1 grade in the
M. A. Ko¨nig B. M. Boszczyk (&) The Centre for Spinal Studies and Surgery, Queens Medical Centre Nottingham, Nottingham, UK e-mail: [email protected] M. A. Ko¨nig e-mail: [email protected] S. Panzer Department of Radiology, Berufsgenossenschaftliche Unfallklinik Murnau, Murnau, Germany J. Schulz M. Bierschneider Department of Neurosurgery, Berufsgenossenschaftliche Unfallklinik Murnau, Murnau, Germany
long-term follow up. No intervertebral disc in-between bisegmental cement augmentation showed acceleration of degenerative changes. Conclusion Despite several limitations regarding patients’ age and lack of performed perfusion MRI scans, this study suggests that vertebral cement augmentation through kyphoplasty has no significant influence on disc degeneration even after a long period. The absence of severe disc degeneration after vertebral augmentation supports further clinical trials, which should incorporate endplate perfusion studies for detailed information regarding disc perfusion. Keywords Kyphoplasty Intervertebral disc degeneration Magnetic resonance imaging Vertebral compression fracture
Introduction In the course of the last two decades, encouraging clinical results have been achieved in the management of osteoporotic vertebral fractures through the techniques of vertebroand kyphoplasty [1, 5, 9–12, 15, 17, 34]. The minimally invasive nature of these methods and the low overall complication rate have consequently prompted the expansion of the indications for vertebral augmentation to traumatic fracture types or fractures due to early onset of secondary osteoporosis [6, 29]. However, no long-term adverse effects of vertebral augmentation with polymethylmethacrylate (PMMA) have been published and very little is known of the biological effects upon the endplate and intervertebral disc. A 30 % decrease of the perfusion surface seems to be sufficient to attain a 50 % perfusion deficit to the nucleus [32] and as the supply of nutrients to the intervertebral disc is provided to a significant degree through the vascularised
123
Eur Spine J
vertebral endplate [25, 27], compromise of blood flow to and from the endplate may be expected to have a detrimental effect. Diffusion in degenerative discs also becomes, compared to healthy uninjured discs, more and more irregular and patchy [26]. This in turn could, in theory, lead to accelerated intervertebral disc degeneration with the possibility of a delayed onset of symptoms associated therewith. With the expansion of the indications for vertebral augmentation, procedure-related disc degeneration could reach clinical significance. Despite the clinical importance of this topic, only very little attention has been paid to the actual extent of disc changes after vertebral cement augmentation [24]. The purpose of this investigation is, therefore, twofold: (a)
(b)
To analyse the influence of primary endplate and disc injury severity on subsequent disc degeneration after vertebral augmentation. To analyse the influence of vertebral augmentation on the degeneration of adjacent discs in general.
Materials and methods Hospital records were reviewed for patients treated with kyphoplasty for osteoporotic or traumatic vertebral fractures of the thoracolumbar and lumbar spine whereas cases with neoplastic disease of the spine or infections were excluded. Patients with osteoporotic fractures were selected if the preoperative MRI scan was available at our institution and patients with traumatic fractures were selected if these were isolated injuries, regardless of the availability of a preoperative MRI scan. In total, 24 patients were identified, approached and counselled regarding the proposed followup investigation. Fifteen patients (average age 67.1 ± 6.9 years, range 58–77; 10 female, 5 male) agreed to MRI follow-up (Tables 1, 2, 3) and informed consent was obtained from all patients for the extent of the MRI investigation. Of these, seven patients were available for long-term follow-up MRI scan. As far as could be determined, refusal of participation was not related to the initial procedure or the outcome thereof in any of the declining patients. One of the recruited patients (CC) had suffered traumatic paraplegia following an L1 fracture more than 20 years previously, with resulting inactivity-induced osteoporosis. Five patients (SK, WM, EV, BH, MM) presented with primary osteoporosis and two patients with cortisone induced secondary osteoporosis in the setting of rheumatoid arthritis (BI, PH). In the remaining two patients (PA, BF), vertebral injuries following minor trauma had occurred, without obvious or known osteoporosis. In addition to the osteoporosis patients, five further patients (MK, AE, AR, BF, PH) with single genuine traumatic fractures were included. During the follow-up period, three patients (CC, PH, WM)
123
suffered additional fractures, three of which were treated with kyphoplasty in the first 8 weeks after the first cement augmentation and one who was treated conservatively. These vertebrae were added to the investigation as they were adjacent to treated vertebrae and, therefore, could potentially influence the outcome of discs already included in the study (Patient CC and PH see Tables 1 and 2). The adjacent disc levels with untreated vertebral bodies were identified and included in order to observe the natural history of disc degeneration. In total, 51 discs were observed with 23 in the thoracic spine and 28 in the lumbar spine at the first follow-up. At the long-term follow-up, nine thoracic and 11 lumbar discs were included. Surgical technique Twenty fractures were treated with a biportal access kyphoplasty, three via a unilateral perpendicular approach and two with semi-open kyphoplasty. All cement augmentations had fluoroscopically controlled needle placement to achieve an optimal result of balloon fracture reduction and cement application [33]. The average amount of PMMA cement used in the lumbar spine was 7.6 ± 1.9 ml (range 4–10.5 ml) and 5.7 ± 1.3 ml (range 4–8 ml) in the thoracic vertebrae and, therefore, all fractures were treated adequately [7]. Seven discs were in-between bi-segmental cement augmentations. MRI evaluation The MRI scan prior to surgery was available for osteoporotic fracture patients and in 3 trauma patients. In addition to the pre-operative fracture classification according to the scale proposed by Magerl et al. [18], the presentation of the fracture on the initial MRI was also reviewed. The affected vertebral body and intervertebral discs were graded according to the systems proposed by Pfirrmann et al. [23] (Table 4) and by Oner et al. [19, 20] (Table 5). The grading system by Oner et al. [19, 20] assesses the injury severity of the vertebral body, endplates, adjacent discs and spinal ligaments. The grading system by Pfirrmann et al. [23] grades degeneration according to the state of hydration in uninjured discs adjacent to fractured vertebrae. At both follow-up assessments, the degree of degeneration of all discs adjacent to augmented vertebral bodies was graded according to Pfirrmann et al. [23]. The MRI investigation pre- and post-operatively was performed with the same MRI machine (1.0 Tesla, Magnetom Expert, Siemens, Erlangen, Germany). The longterm follow-up MRI scans were all done with 1.5T (Achieva, Philips Medical Solutions, Best, The Netherlands). MRI settings were TSE; T2; TE 120 ms; TR 4,000 and TIR, TE 60 ms; TR3366.
Eur Spine J Table 1 Patients with single-level kyphoplasty: initial data and first follow-up Patient
Age
Sex
AE
72
F
Level
Volume (ml)
Access
4.5 ? 4.5
bilat
T12/L1 L1 L1/L2
AR
66
F
L2/L3 L3
4?4
64
M
T12/L1 L1
5.5 ? 5
64
F
T10/T11 T11 T11/T12
MK
69
M
76
M
59
F
F
67
75
IN
3
1
IN
4
IN
4 4
3
3
3
4
5
5
2
1
3
4
1
M
T11/T12
3
3
1
M
T12/L1 T10/T11
4 3
4 3
4
4
3
3
IN
4
3?1
L4/L5 L5
5
L5/S1
3
3
3
open uni 3
2
2
1
2
A3.1
2
1
1
3
A3.1
1
2
1
3
A3.1
2
2
1
3
A1.1
1
2
1
3
A1.2
1
1
1
2
A1.2
1
2
2
1
3
A3.1
3
2
2
2
3
A3.3
1
2
2
2
2
A1.2
2
1
1
3
A1.2
2
2
1
3
A3.1
1 2 1 1 1 4
3
A3.1
1
2 3
3
1 1 1
bilat
L5/S1 F
1
1
L4/L5 L5
1
1 3
T11/T12
F
2
1 1
CONS
2
1 1
4
bilat
A3.1
1
1
3?3
3
1 1
4
bilat
1
1
4
3?3
2
4
1
bilat
2
2
T11/T12
2?2
AO
4 1
1
T10/T11
COR
1
1
bilat
PLC
1 2
3
4?4
PLL
1
4 4
ALL
1
2
bilat
T12/L1
75
2
DI
3
T12a
BI
EPi
L1/L2
T11
76
4
3
T11/T12
EV
4
3
3?3
EPs
1
3
T12 WM
3
3
T11 SK
3
T12/L1
T12 61
3
bilat
T12/L1 PA
3
T12/L1 L1 PH
3?3
3
bilat
T11/T12 T12
MM
2.5 ? 3
3
bilat
L1/L2 LM
Pfirr FU 1
bilat
L3/L4 BF
Pfirr pre
1 1
Patient data and MRI grading results for injured vertebral bodies M male, F female. Levels vertebral body and adjacent intervertebral discs. Volume applied cement volume. Access: unilat unilateral, bilat bilateral, open uni open unilateral. Pfirr pre pre-operative Pfirrmann grade, PfirrFU 1 Pfirrmann grade at first follow up [23] (see Table 5 for grading). AO fracture classification according to Magerl et al. [18]. DI; EPs; EPi; COR; ALL; PLL; PLC spinal column injury classification according to Oner et al. [19, 20] (see Table 4 for abbreviations and grades). ‘‘IN’’ in DI columns indicate injured discs. ‘‘CONS’’ conservative fracture treatment. ‘‘OLD’’ old traumatic fracture a
Fractured vertebra after pre-operative MRI scan
All authors of this investigation graded all MR images separately and blindly according to the Pfirrmann and Oner classifications. Wherever there was a grading difference, a consensus of all three was made and the result obtained. This
investigation was focused specifically on disc degeneration after cement augmentation and, therefore, no attempt was made to correlate clinical outcome, grade of height restoration or kyphosis angle with MRI findings.
123
Eur Spine J Table 2 Patients with multilevel kyphoplasties: initial data and first follow-up Patient
Age
Sex
CC
57
F
Level
Volume (ml)
Access
4
unilat
T8/T9 T9a T9/T10 T10
3?4 3.5 ? 3.5 6
4?4 4?4
56
F
T8/T9 T9
4
4
IN
4
3
4
4 4
4 4
EPi
1
1
DI
ALL
PLL
PLC
COR
AO
1 A1.1. 1 1
2
2
1
1
3
A1.2
1 1
1
A1.1 4
2
1
2
1
1
3
A1.1
1
2
1
1
A3.1
1
2
1
2
A1.1
1
1
1
1
A3.1
1
1
1
2
A3.1
2
3
1
3
A3.3
1
4
IN
3
5
5
5
5 4
3
3
1 1 1
1 3
2 2
1
unilat
T9/T10
1 3
bilat
L5/S1 PH
4
bilat
L4/L5 L5
4
EPs
OLD
L1/L2 L3/L4 L4
4
unilat
T12/L1 L1
4
bilat
T11/T12 T12
Pfirr FU 1
bilat
T10/T11 T11a
Pfirr pre
1 1
T10 T10/T11 T11
3?3
bilat
T11/T12 T12
3
L1a
4?4
L2/L3
BH
69
F
4.5 ? 4.5
4
3
3
3
3
IN
3
2 1
1
bilat
1 1
2
bilat
L1/L2 L2 L3
4 open uni
T12/L1
1 1
1
A3.1 1 2
2
1
L3/L4
3
3
1
L3/L4
4
4 PMMA
1
4
4 PMMA
L4
4?4
bilat
L4/L5 L5
4?4
L5/S1
1
bilat 4
1
1
2
A1.1.
2
1
1
1
A1.1
2
2
1
3
A3.1
1 2
4
4
2
1 1
Patient data and MRI grading results for injured vertebral bodies M male, F female. Levels vertebral body and adjacent intervertebral discs. Volume applied cement volume. Access: unilat unilateral; bilat bilateral, open uni open unilateral. Pfirr pre pre-operative Pfirrmann grade; PfirrFU 1 Pfirrmann grade at first follow up [23] (see Table 5 for grading). AO fracture classification according to Magerl et al. [18]. DI; EPs; EPi; COR; ALL; PLL; PLC spinal column injury classification according to Oner et al. [19, 20] (see Table 4 for abbreviations and grades). ‘‘IN’’ in DI columns indicate injured discs. ‘‘CONS’’ conservative fracture treatment. ‘‘OLD’’ old traumatic fracture a
Fractured vertebra after pre-operative MRI scan
Results Patient data with the first follow-up are summarised in Tables 1 and 2. Patient data of the long-term follow-up are shown in Table 3. For didactic reasons, the tables show
123
only the affected vertebrae and adjacent discs. A total of 15 patients with a sum of 25 fractures ranging from T9 to L5 were included. Seven fractures were classified as A1.1 (28 %), 4 as A1.2 (16 %), 12 as A3.1 (48 %) and 2 as A3.3 (8 %) fractures according to Magerl [18].
Eur Spine J Table 3 Single level and multilevel kyphoplasties at long-term follow-up Patient
Age
Sex
Level
AR
66
F
L2/L3 L3
Volume (ml)
4?4
Access
64
M
T12/L1 L1
5.5 ? 5
MK
64
69
F
M
T11/12 T12
59
F
76
F
57
F
3
4
4
4
IN
3
3 4
IN
4
5
bilat
2
T11/ T12
3
3
4
1
T12/L1
3
3
4
1
T11/ T12
4
4
4
1
4
4
4
IN
4
4
L4/L5 3?1
1
bilat
2
bilat
1
1
1
3
3
3
1
4
4
4
1
4
4
4
4
unilat
T10
3?4
1
bilat
T10/ T11
4
4
IN
4
4
3
4
4
L1/L2
4
4
4
L3/L4
4
4
4
T11
3.5 ? 3.5
bilat
T11/ T12 T12
6
L1
L4
L5/S1
A3.1
2
2
1
2
A3.1
2
1
1
3
A3.1
1
2
1
3
A3.1
1
2
1
3
A1.2
2
1
1
3
A1.2
1
A1.1.
2
2
1
1
3
A1.2
1
A1.1 4
2
1
2
1
1
3
A1.1
1
2
1
1
A3.1
1
2
1
2
A1.1
1
OLD
4?4
bilat
L4/L5 L5
3
1 1
unilat
T12/L1
1
1 1
4
2
2 2
L5/S1
T9/T10
2
1
T8/T9 T9
AO
4 3 1
4?4
COR
2 4
bilat
PLC
4 1
4
3 ?3
PLL
1 2
4
ALL
1 1
IN
DI
4
L5 CC
3
EPi
4
T12 T12/L1 EV
3
EPs
T12/L1
T12 PH
2.5 ? 3
PfirrFU 2
bilat
L1/L2 LM
Pfirr FU 1
bilat
L3/L4 BF
Pfirr pre
IN 4?4
1 3
3
3
bilat
2 1
5
5
2
5
1 1
Patient data and MRI grading results at long-term follow-up M male, F female, Levels vertebral body and adjacent intervertebral discs, Volume applied cement volume. Access: unilat unilateral, bilat bilateral, open uni open unilateral, Pfirr pre pre-operative Pfirrmann grade, PfirrFU 1 Pfirrmann grade at first follow up, PfirrFU 2 Pfirrmann grade long-term follow-up [23] (see Table 5 for grading). AO fracture classification according to Magerl et al. [18]. DI; EPs; EPi; COR; ALL; PLL; PLC spinal column injury classification according to Oner et al. [19, 20] (see Table 4 for abbreviations and grades). ‘‘IN’’ in DI columns indicates injured discs. ‘‘CONS’’ conservative fracture treatment. ‘‘OLD’’ old traumatic fracture
The initial follow-up includes 21 fractures and 43 surveyed discs of which seven were adjacent to bisegmental augmentation. Fifty-one discs distant to vertebral augmentation were observed to determine the natural disc degeneration.
The long-term follow-up included five lumbar and seven thoracic vertebral fractures. Twenty adjacent discs of which four were between augmented vertebrae were evaluated. Twenty discs remote from augmented vertebrae
123
Eur Spine J Table 4 MRI classification of spinal column injury according to Oner et al. [20] Grade 1
Grade 2
Grade 3
Grade 4
Disc (DI)
No evidence of injury in the disc space
Rupture and/or debris in the anterior half of the disc space
Rupture and/or debris in the posterior half of the disc space
Involvement of the whole disc. Either the disc is completely herniated into the vertebral body and/or there is rupture and debris in the whole disc space
Endplate (EP)
Only plastic deformity of the endplate without disruption
Disruption in the posterior half of the endplate
Disruption of the whole endplate
Vertebral body (COR)
Less than one-third of the volume of the vertebral body is involved (bone marrow oedema)
Disruption in the anterior half of the endplate(evident discontinuity of low signal line) One to two-thirds of the vertebral body is involved
Anterior longitudinal ligament (ALL)
No evidence of injury
The ligament is slackened but continuous
The ligament is ruptured
Posterior longitudinal ligament (PLL)
No evidence of injury
The ligament is attached to the extruding bone fragment from the posterior cortex and continuous
The ligament is ruptured
Posterior ligamentary complex (PLC)
No evidence of injury
Oedema in the interspinous space without evident discontinuity or elongation
Elongation of the interspinous space without evident discontinuity
More than twothirds of the vertebral body is involved
Clear disruption
Endplate injury is differentiated between the superior (EPs) and inferior endplate (EPi)
were once again observed for disc degeneration without
Pre-operative MRI
Table 5 MRI classification of intervertebral disc degeneration according to Pfirrmann et al. [23] Grade 1
Grade 2
Grade 3
Grade 4
Grade 5
The structure of the disc is homogenous, with a bright hyperintense white signal and a normal disc height
The structure of the disc is in homogenous, with a hyperintense white signal. The distinction between nucleus and anulus is clear, and the disc height is normal, with or without horizontal grey bands
The structure of the disc is in homogenous with an intermediate grey signal intensity. The distinction between nucleus and anulus is unclear, and the disc height is normal or slightly decreased
The structure of the disc is in homogenous, with a hypointense dark grey signal intensity. The distinction between nucleus and anulus is lost, and the disc height is normal or moderately decreased
The structure of the disc is in homogenous, with a hypointense black signal intensity. The distinction between nucleus and anulus is lost, and the disc space is collapsed
cement augmentation. MRI evaluation A first follow-up MRI scan was obtained after a mean follow-up time of 15.2 months (range 8–27 months).The second MRI scan, defined as long-term follow-up, included seven patients after a mean follow-up time of 94.3 months (range 84–96 months). An exemplary case of a L1 fracture is shown in Fig. 1, with pre-operative findings followed by the first follow-up and long-term results.
123
Disruption of spinal ligaments could be excluded in all patients, but the extent of vertebral body injury and endplate disruption varied considerably. Only three treated vertebrae had limited injury of the vertebral body (grade 1). The remaining treated vertebrae were found to have a grade 2 (5 vertebrae) or grade 3 (14 vertebrae) injury. Since three fractures occurred after the baseline MRI scan, grading was not possible in these cases. Grade 1 was present in 30 (57.7 %) of the total of 52 endplates (superior and inferior) of the injured vertebrae. Grade 2 and 3 anterior or posterior disruption was present in 19 endplates (36.5 %). Severe injury was identified in 3
Eur Spine J
Fig. 1 Patient BF: L1 fracture. EPs 2, EPi 1, DI 4 (T12/L1) and DI 2 (L1/L2) pre-operatively according to Oner [19, 20] (Table 4). Disc degeneration grade 3 at follow-up according to Pfirrmann [23] (Table 5) and no changes in long-term follow-up
(5.8 %) endplates of three vertebrae in three patients. Uninjured discs adjacent to augmented vertebral bodies mostly showed mild to moderate degeneration preoperatively. Nineteen discs showed a Pfirrmann grade 3 degeneration (44.2 %), 15 (35 %) had grade 4 and 2 (3.8 %) had grade 5 degeneration pre-operatively. Seven injured discs were identified, four had rupture or debris in the anterior half of the disc space and three had involvement of the whole disc (Table 1). The group without cement augmentation showed similar degeneration: five had grade 2 (3.9 %), 32 grade 3 (62.7 %), 11 grade 4 (21.6 %) and 3 (5.8 %) grade 5 degeneration according to the Pfirrmann classification.
severe endplate injuries showed no progressive degeneration similar to the discs in between bilateral cement augmentations. PMMA disc leaks in both adjacent disc levels after kyphoplasty occurred in one patient, with an inferior endplate injury grade 4 and a superior endplate injury grade 1 (Patient BH Table 2), but no further disc degeneration was seen at follow-up. Degeneration in discs apart from the cement augmentation appeared in 14 cases in the thoracic and lumbar spine. The majority of the thoracic discs (4) degenerated from Pfirrmann grade 3 to 4 and one disc degenerated from grade 2 to 3. In the lumbar spine, six discs showed degeneration of grade 3–4, two from grade 2 to 3 and one from grade 4 to 5.
MRI follow-up
MRI long-term follow-up
At the first follow-up, the majority of uninjured discs did not show any evidence of progressive degeneration. Only three discs showed degeneration of one Pfirrmann grade from 3 to 4. Two discs were below the treated vertebra with grade 1 or 2 inferior endplate injuries, while one disc was above the cement augmentation with a vertebral superior endplate injury of grade 1. One instance of disc degeneration occurred in between a fresh osteoporotic fracture T12 (Magerl A1.1) and the old traumatic fracture L1 (Patient CC, Table 1), while the two remaining disc degenerations appeared in the lower thoracic area T9 and T11 (Magerl A1.2 and A3.1) of two patients (PA and PH Table 1). The seven injured discs presented with grade 3 or 4 degenerative changes. Interestingly, discs with associated
Table 3 summarises the data of seven patients from preoperative findings and both follow-up measurements to point out possible disc degeneration. 19 (95 %) discs under surveillance showed no degeneration compared to preoperative findings or the first follow-up. The intervertebral disc of T12/L1 in one patient had a progressive degeneration from 3 to 4 according to Pfirrmann with no initial superior and inferior endplate injury after a traumatic fracture of T12 (Patient MK Table 3). One initial injury with total disc involvement showed degeneration of 1 grade from 4 to 5 compared to the first follow up with grade 4 superior and grade 3 inferior endplate injury (Patient LM Table 3). Only one of the three early disc degenerations at the first follow-up was available for long-
123
Eur Spine J
term results but this disc showed no progressive degeneration (Patient CC Table 3). Again, no disc degeneration was seen in discs in between bisegmental cement augmentations. In the surveyed group apart from cement augmentation, four new degenerations of 1 grade occurred in the lumbar spine and four in the thoracic spine. One thoracic disc showed even a degeneration of two grades from 2 to 4 (Patient PH Table 3). Two discs degenerated even more from the first follow-up to the long-term follow-up from 3 to 4.
Discussion Cement augmentation with PMMA is established as a treatment option for osteoporotic fractures and lytic tumour lesions [1, 9–11]. Advantages are of course the less invasive procedure itself and good mechanical support [4, 34], making vertebro- and kyphoplasties suitable for a large group of patients with proven effectiveness of these treatment options [12, 15, 17, 22]. Adverse effects were only reported during or immediately after cement application ranging from asymptomatic appearances to patients’ death [3, 8, 13, 14, 16, 21, 28, 30, 31] or focusing on the adjacent vertebral bodies [2]. No long-term studies have been published and no information focusing on the effect of cement augmentation on the intervertebral discs is currently available. This restricts its use with younger patients due to concerns related to the effect of PMMA upon adjacent intervertebral disc vitality and long-term sequelae. The intervertebral disc as largest inert structure in the human body is highly dependent on endplate diffusion for nutritive agents and diffusion surface reduction causes deficits in the nucleus [25, 26, 32]. Qian et al. [24] interestingly showed in their prospective study of 97 patients an accelerated disc degeneration after a short-term follow-up of 2 years. Up to 52.6 % of the intervertebral discs next to the cement augmented vertebral body showed degenerative changes. Unfortunately, no information is given regarding existing disc degeneration, so no conclusion could be made if cement augmentation accelerates disc degeneration. It is of note that 29 % of the control group were found to have disc degeneration. At least one of the case examples shows degeneration of multiple discs. The subjects seem to also have had a less degenerative overall outset than the patients presented in this observation. Other factors include a different ethnic population and a single observer in the Qian study. In the present investigation, 15 patients were followed up with an MRI (at an average of 15.2 months after being treated) with PMMA kyphoplasty for one or more vertebral fractures. Out of these 15 patients, seven were applicable
123
for a long-term follow-up after a mean of 94.3 months, to measure possible long-term adverse effects of cement augmentation to the intervertebral discs. Uninjured intervertebral discs adjacent to treated vertebral bodies did not show a rapid progression of degeneration according to the Pfirrmann grading system in majority of the cases. Three discs had a progression of 1 grade according to Pfirrmann [23] at the first follow-up and injured discs had moderate disc degeneration at further follow-up. In the long-term follow-up of the seven patients, no progression of degeneration was seen in 19 out of 20 adjacent discs indicating that cement augmentation procedures with PMMA do not significantly affect the endplate diffusion in vertebral fractures. Even in previously injured discs, PMMA cement seems not to accelerate disc degeneration significantly since only one disc with severe endplate injury had deteriorated by 1 grade. Interestingly, no accelerated disc degeneration was found in intervertebral discs with bisegmental cement augmentation. Compared with the natural history of disc degeneration in these patients, PMMA cement augmentation does not seem to accelerate disc degeneration dramatically indicating safe usage even if bisegmental augmentation is needed. Despite the fact that this is the first study focusing on disc degeneration after cement augmentation in a long-term follow-up, it has several limitations. First of all, patients with osteoporotic fractures usually present with disc degeneration due to their age which limits the comparison of pre- and post-operative intradiscal changes. This is because results may vary if the disc is less degenerated and, therefore, needs a higher diffusive threshold for greater supply with nutritive substances. Unfortunately, out of this small cohort of patients, only seven patients could be investigated in the long-term follow-up, which reduces the obtained information significantly regarding the results. Possibly more accelerated disc degeneration could have been found if more patients had been available for the last follow-up. In addition, this study lacks a proper control group as presented in the study by Qian et al. [24]. However, intervertebral discs without influence of cement augmentation were included for each patient in terms of observing the natural history of disc degeneration. As a last limitation, the endplate diffusion of the affected disc was not controlled with contrast studies according to Rajasekaran to determine the diffusion surface after kyphoplasty [25, 26].
Conclusion PMMA cement augmentation appears not to have a significant effect on accelerated disc degeneration in osteoporotic fractures. In this study, only minor progression of
Eur Spine J
degeneration was found in 5 out of 25 discs under surveillance in previously injured and uninjured discs. Bisegmental kyphoplasty did not accelerate disc degeneration. Future studies should include perfusion investigation of the endplates. If vertebral augmentation techniques are deemed useful in young patients with traumatic fractures, more focus should be laid on the intervertebral discs, since PMMA cement might influence endplate diffusion more in less degenerated discs. Conflict of interest No benefits or funds in any form have been received or will be received from a commercial party related directly or indirectly to the subject of this article.
References 1. Berlemann U, Heini PF (2002) Percutaneous cementing techniques in treatment of osteoporotic spinal sintering. Unfallchirurg 105:2–8 2. Berlemann U, Ferguson SJ, Nolte LP, Heini PF (2002) Adjacent vertebral failure after vertebroplasty. J Bone Jt Surg Br 84:748–752 3. Bernhard J, Heini PF, Villiger PM (2003) Asymptomatic diffuse pulmonary embolism caused by acrylic cement: an unusual complication of percutaneous vertebroplasty. Ann Rheum Dis 62:85–86 4. Belkoff SM, Mathis JM, Fenton DC, Scribner RM, Reiley ME, Talmadge K (2001) An ex vivo biomechanical evaluation of an inflatable bone tamp used in the treatment of compression fracture. Spine 26:151–156 5. Boszczyk BM, Bierschneider M, Robert B, Jaksche H (2003) Treatment of severe osteoporotic fractures through a microsurgical interlaminary approach. In: Szpalski M, Gunzburg R (eds) Vertebral osteoporotic compression fractures. Lippincott WW, Philadelphia, pp 179–188 6. Boszczyk B, Bierschneider M, Potulski M, Robert B, Vastmans J, Jaksche H (2002) Extended kyphoplasty indications for stabilization of osteoporotic vertebral compression fractures. Unfallchirurg 105:952–957 7. Boszczyk B (2010) Volume matters: a review of procedural details of two randomised controlled vertebroplasty trials of 2009. Eur Spine J 19(11):1837–1840 8. Chen HL, Wong CS, Ho ST, Chang FL, Hsu CH, Wu CT (2002) A lethal pulmonary embolism during percutaneous vertebroplasty. Anesth Analg 95:1060–1062 9. Dudeney S, Lieberman IH, Reinhardt MK, Hussein M (2002) Kyphoplasty in the treatment of osteolytic vertebral compression fractures as a result of multiple myeloma. J Clin Oncol 20:2382–2387 10. Fourney DR, Schomer DF, Nader R et al (2003) Percutaneous vertebroplasty and kyphoplasty for painful vertebral body fractures in cancer patients. J Neurosurg 98:21–30 11. Garfin SR, Hansen AY, Reiley MA (2001) Kyphoplasty and vertebroplasty for the treatment of painful osteoporotic compression fractures. Spine 26:1511–1515 12. Garfin SR, Lin G, Lieberman I, Phillips F, Truumees E (2001) Retrospective analysis of the outcomes of balloon kyphoplasty to treat vertebral body compression fracture refractory to medical management. Eur Spine J 10:S7–S8 (Abstract) 13. Harrington KD (2001) Major neurological complications following percutaneous vertebroplasty with polymethylmethacrylate. J Bone Jt Surg Am 84:1070–1073
14. Jang JS, Lee SH, Jung SK (2002) Pulmonary embolism of polymethylmethacrylate after percutaneous vertebroplasty. Spine 27:E416–E418 15. Ledlie JT, Renfro M (2003) Balloon kyphoplasty: One-year outcomes in vertebral body height restoration, chronic pain and activity levels. J Neurosurg 98:36–42 16. Lee BJ, Lee SR, Yoo TY (2002) Paraplegia as a complication of percutaneous vertebroplasty with polymethylmethacrylate. Spine 27:E419–E422 17. Lieberman IH, Dudeney S, Reinhardt MK, Bell G (2001) Initial outcome and efficacy of ‘‘kyphoplasty’’ in the treatment of painful osteoporotic vertebral compression fractures. Spine 26:1631–1638 18. Magerl F, Aebi M, Gertzbein SD, Harms J, Nazarian S (1994) A comprehensive classification of thoracic and lumbar injuries. Eur Spine J 3:184–201 19. Oner FC, van der Rijt RR, Ramos LMP, Dhert WJA, Verbout AJ (1998) Changes in the disc space after fractures of the thoracolumbar spine. J Bone Jt Surg Br 80:833–839 20. Oner FC, van Gils APG, Faber AJ, Dhert WJA, Verbout AJ (2002) Some complications of common treatment schemes of thoracolumbar spine fractures can be predicted with magnetic resonance imaging. Spine 27:629–636 21. Padovani B, Kasriel O, Brunner P, Peretti-Viton P (1999) Pulmonary embolism caused by acrylic cement: a rare complication of percutaneous vertebroplasty. Am J Neuroradiol 20:375–377 22. Perez-Higueras A, Alvarez L, Rossi RE, Quinones D, Al-Assir I (2002) Percutaneous vertebroplasty: long-term clinical and radiological outcome. Neuroradiology 44:950–954 23. Pfirrmann CW, Metzdorf A, Zanetti M, Hodler J, Boos N (2001) Magnetic resonance classification of lumbar intervertebral disc degeneration. Spine 26(17):1873–1878 24. Qian J, Yang H, Jing J, Zhao H, Ni L, Tian D, Wang Z (2012) The early stage adjacent disc degeneration after percutaneous vertebroplasty and kyphoplasty in the treatment of osteoporotic VCFs. PLoS One 7(10):e46323. doi:10.1371/journal.pone. 0046323 25. Rajasekaran S, Naresh-Babu J, Murugan S (2007) Review of postcontrast MRI studies on diffusion of human lumbar discs. J Magn Reson Imaging 25:410–418 26. Rajasekaran S, Babu JN, Arun R, Armstrong BR, Shetty AP, Murugan S (2004) ISSLS prize winner: a study of diffusion in human lumbar discs: a serial magnetic resonance imaging study documenting the influence of the endplate on diffusion in normal and degenerate discs. Spine 29(23):2654–2667 27. Rajasekaran S, Venkatadass K, Naresh Babu J, Ganesh K, Shetty AP (2008) Pharmacological enhancement of disc diffusion and differentiation of healthy, ageing and degenerated discs: results from in vivo serial post-contrast MRI studies in 365 human lumbar discs. Eur Spine J 17(5):626–643 28. Ratliff J, Nguyen T, Heiss J (2001) Root and spinal cord compression from methylmethacrylate vertebroplasty. Spine 26(13):E300–E302 29. Ross PD, Genant HK, Davis JW, Miller PD, Wasnich RD (1993) Predicting vertebral fracture incidence from prevalent fractures and bone density among none-black, osteoporotic women. Osteoporos Int 3:120–126 30. Scroop R, Eskridge J, Britz GW (2002) Paradoxical cerebral embolization of cement during intraoperative vertebroplasty: case report. Am J Neuroradiol 23:868–870 31. Tozzi P, Abdelmoumene Y, Corno AF, Gersbach PA, Hoogewoud HM, von Segesser LK (2002) Management of pulmonary embolism during acrylic vertebroplasty. Ann Thorac Surg 74:1706–1708 32. van der Werf M, Lezuo P, van Donkelaar CC, Ito K (2007) Inhibition of vertebral enplate perfusion results in decreased
123
Eur Spine J intervertebral disc intranuclear diffusive transport. J Anat 211:769–774 33. Wenger W, Markwalder TM (1999) Surgically controlled, transpedicular methyl methacrylate vertebroplasty with flouroscopic guidance. Acta Neurochir 141:625–631
123
34. Wilson DR, Myers ER, Mathis JM et al (2000) Effect of augmentation on the mechanics of vertebral wedge fractures. Spine 25:158–165
