Secondary Logo

Journal Logo

Original article

Long-term results and radiographic findings of percutanous vertebroplasties with polymethylmethacrylate for vertebral osteoporotic fractures

XU, Bao-shan; HU, Yong-cheng; YANG, Qiang; XIA, Qun; MA, Xin-long; JI, Ning

Editor(s): HAO, Xiu-yuan

Author Information
doi: 10.3760/cma.j.issn.0366-6999.2012.16.006
  • Free


Vertebral osteoporotic fractures (VOFs) are very common in the elderly with osteoporosis.1 Since percutaneous vertebroplasty (PVP) was first reported for the successful treatment of vertebral angiomas in France in 1987, this procedure has been widely used in the treatment of VOFs, with immediate pain relief and improved mobility. PVP is a minimally invasive procedure involving percutaneous injection of acrylic cement (polymethyl methacrylate (PMMA)) into a weakened vertebral body under fluoroscopic guidance; however, injection of PMMA into the vertebral body is not absorbed, and little information is available concerning the long-term results after this procedure.2–7 In this retrospective study with 7-11 years of follow-up, we evaluated the long-term clinical results and radiological changes after PVPs.



Between February 2000 and March 2004, 51 patients with VOFs were treated by PVPs with PMMA in our department. These sequential patients included 35 women and 16 men, with an average age of 70.6 years (range, 59-83 years). The VOFs occurred within 3 weeks of treatment in 26 cases, within 3 months in 15 cases, and after 3 months in 10 cases. All of the patients had back pain around the fracture site, and 10 of the patients complained of radicular pain in the thoracic wall or the flank. The pain was aggravated by mobility, and the symptoms responded poorly to conservative therapy. The patients had limited mobility, and 35 patients were bedridden due to VOFs. The presence of symptomatic nerve root or cord compression at the intervention level was deemed a contraindication to PVP.

All patients underwent plain radiographs, bone mineral density measurements, and CT and MRI examinations. The VOFs involved the T6-L4 segments. The posterior walls of the vertebral bodies were involved by fractures in 10 patients, but without neurologic compromise. The VOFs were present in multiple levels in seven patients. Physical examination and MRI was used to determine the symptomatic segment. The presence of a low signal on the T1-weighted image, with an elevated signal on the T2-weighted and fat-suppressed images indicated new fractures or non-union of fractures.8


The PVPs were performed under fluoroscopic guidance. The transpedicle approach was generally used; however, an approach between the pedicle and the rib head was adopted for 4 patients with thin pedicles. Lidocaine (1%) was injected to provide local anesthesia. A 9-gauge bone marrow biopsy needle was inserted into the anterior part of the vertebral body under fluoroscopic guidance. The powder of PMMA (Palacos, Schering-Plough, Beigium) was mixed with sterile barium sulfate at a ratio of 20% to increase radiopacity, and the thoroughly mixed cement was filled into 2.5 ml syringes. When the cement had the consistency of a paste, the cement was slowly injected into the vertebral body through the needle under strict fluoroscopic guidance. The cement was injected until the cortex was reached or resistance was met. The injection was immediately stopped when epidural, foraminal, or venous leakage was detected. Extreme caution was used when the posterior vertebral body cortex was not intact, as was the case in 10 patients. Complete filling of the vertebrae with cement was specifically not attempted.

The injection was performed at 61 vertebral levels (29 thoracic and 32 lumbar). Because cross-filling occurred in most cases, only 5 vertebrae had contralateral punctures and injections. The total volume of cement injected per vertebrae averaged 4.4 ml (range, 2-6 ml). After the procedure, the patients were asked to remain supine for 4 hours to allowing complete curing of the PMMA. Plain radiographs and CT scans were performed to evaluate cement filling and leakage. Intravenous antibiotics were administered for 1 day, and anti-osteoporotic medications were used routinely for at least 3 months.


Patients were requested to make return visits at intervals of 3 months, 1 year, and every 2 years post-operatively. The pain was assessed with a visual analog scale (VAS) ranging from 0 to 10, with 0 representing “no pain” and 10 representing “the most severe pain.” Marked pain relief was defined as a reduction in the pain score of >5, and moderate pain relief was defined as a reduction in the pain score between 3 and 5. Mobility was graded as walking without difficulty (grade 1), walking with assistance (grade 2), and bedridden (grade 3). The patients were asked whether they were satisfied with the procedure, and if needed, whether they would undergo a PVP again.

Plain radiographs and CT scans were performed after >7 years of follow-up. The vertebral body anterior heights and Cobb angles were measured on plain radiographs, and the augmented vertebral bodies were observed on radiographs and CT scans.

Statistical methods

Clinical symptoms and radiographs were assessed and compared pre-operatively, immediately post-operatively, and at the final follow-up evaluation. Statistical data were analyzed with paired t-test and chi squared test to ascertain statistical significance. P<0.05 was considered statistically significant.


Post-operative clinical results

The PVP procedure was technically successful in all of the treated vertebrae. All of the patients experienced pain relief within 4-72 hours (mean, 12 hours) with an improved level of mobility. Marked-to-complete pain relief was achieved in 37 patients, and moderate pain relief was achieved in 14 patients. The VAS score decreased from 8.3±2.9 pre-operatively to 2.1±1.9 immediately post-operatively. All the 35 patients who were bedridden before surgery had improved mobility, and could walk with assistance or without difficulty in 1-2 days.

Complications and cement leakage

There were no major adverse events or complication in our patients. Cement leaks were detected in 10 vertebrae of nine patients on post-operative CT scans, including minimal epidural leaks through basivertebral vascular foramina in three levels, paravertebral leaks through cortical fractures in four levels, and adjacent intradisc leaks through endplate fractures in four levels; however, there were no neurological complications or clinical symptoms resulting from cement leaks.

Follow-up results

During the follow-up period, the clinical results were stable in 46 patients with sustained pain relief and improved activities, and five patients had recurrent painful episodes related to the procedure. Two patients had mild pain 2 and 3 weeks post-operatively, and edema signals of bone marrow around the cement were noted on MRI. Three patients had new collapse of vertebrae adjacent to the augmented vertebrae 3 weeks, 1 month, and 1 year post-operatively. The recurrent symptoms resolved with conservative treatment in all five patients.

After >7 years of follow-up, eight patients had died from causes unrelated to the intervention, 12 patients were lost to follow-up, thus leaving 31 patients available for evaluation (follow-up rate, 72.1%). Among these 31 patients, the average length of follow-up was 9.2 years (range, 7-11 years), and the PMMA was injected at 43 levels with a mean volume of 4.3 ml per level (range, 2-6 ml).

For the 31 patients followed up >7 years, the VAS decreased from 8.3±2.6 pre-operatively, to 2.1±1.6 immediately post-operatively, and 1.0±0.9 at the final follow-up evaluation, with a significant improvement in ambulatory ability (Table 1). Three patients had new collapse of vertebrae at adjacent levels that resolved with conservative treatment. All of the patients were satisfied with the PVP and reported that they would undergo the procedure again if needed.

Table 1
Table 1:
Mobility of 31 patients followed over 7 years

The mean anterior vertebral heights were 18.7 mm pre-operatively, 19.6 mm immediately post-operatively, and 19.3 mm at the final follow-up evaluation. The mean Cobb angle was 17.4 degrees pre-operatively, 16.6 degrees post-operatively, and 16.9 degrees at the final follow-up evaluation. Indeed, the changes in vertebral heights and Cobb angles were not significant. Based on the radiographs, neither loose nor displaced cement was detected, and there were no new fractures or collapse at the augmented vertebrae. On CT scans, the cement closely contacted or infiltrated the trabecular bone. The boundary between the cement and trabecular bone was indistinct and there was no evident radiolucent gap between the cement and trabecular bone (Figure 1).

Figure 1.
Figure 1.:
An 80-year-old woman with osteoporotic fracture of T11-12, after intervention of PVP, she had stable results of pain relief and improved mobility during 10 years of follow-up. Plain radiograph (A), CT of T12 (B), MRI of T12 (C), CT of L1 (D) and MRI of L1 (E) at 10 years after PVP showing the infiltration of PMMA into trabecular bone without evident radiolucent gap between them.


VOF is a common clinical complication of osteoporosis, and a major health care problem. VOF may result in persistent severe pain and limited mobility, and significantly impact the quality of life. Most patients with VOFs recover with rest or analgesics; however, prolonged bed rest may promote bone loss in the vertebral bodies, making patients more susceptible to future fractures. A PVP is a minimally invasive procedure to consolidate the vertebral bodies that have been structurally weakened by osteoporosis or osteolytic lesions. Since PVP was first reported in 1987 for the management of a painful vertebral hemangioma, PVPs have been widely accepted for the treatment of VOFs without neurologic damage, and excellent results of immediate pain relief and improved mobility have been reported in >90% of the patients.2–6 Although cement leakage during PVPs occurs at a relatively high percentage (20%-73%; approximately 30% in benign lesions and 65% in malignant tumors), most of the leakage does not result in clinical complications. The posterior cortex was involved occasionally in VOFs, and with the increased risk of epidural leaks, necessitating a very cautious injection. In our series, complete or partial pain relief was obtained in all 51 patients within 4-72 hours, and all 35 patients who were bedridden pre-operatively could walk with aid or without difficulty in 1-2 days. Cement leakage was detected in 9 of 51 patients on post-operative CT scans, but without neurologic compromise.

The injection of PMMA into vertebral bodies is not absorbed, and little information is available concerning the long-term results after this procedure. Sporadic reports relating the long-term clinical results of PVPs for VOFs have demonstrated durable pain relief post-operatively.2–7 In our 31 patients, who were followed up for 7-11 years, sustained pain relief was obtained in most patients, except for three who experienced new collapse that resolved after conservative treatment. The VAS decreased dramatically at the final follow-up evaluation, with significant improvement in ambulatory ability. Thus, our results showed that PVPs are effective in relieving the pain associated with VOFs, and the analgesic effect was long-lasting when no new VOFs occurred.

Despite some literature suggesting obvious positional reduction during PVPs, the reduction in vertebral heights and degrees of kyphosis in our patients was not significant immediately after the PVPs, and the changes in vertebral heights and Cobb angles were not significant at the final follow-up evaluation based on the radiographs.9,10 Unilateral punctures and injections were adopted in most of the patients because cross-filling occurred in most cases and a mean volume of 4.3 ml cement was injected per vertebra. Unilateral injections are comparable to bilateral injections in restoration of vertebral body strength and height, and the clinical results were not proportional to the volume of injection.11,12 As reported in the literature, partial vertebral height restoration achieved at PVP did not result in additional pain relief or improved quality of life, but the long-term biomechanical influence of kyphosis remained uncertain.10

The mechanism underlying pain relief following PVPs for painful VOFs remains unclear. Stabilization of microfractures and reduction of mechanical forces may alleviate pain. Biomechanical studies have shown that the strength or load-bearing capacity of a single vertebra is significantly increased following augmentation when compared to the intact strength. Destruction of sensitive nerve endings in surrounding tissue probably occurs in response to mechanical, vascular, chemical, and thermal forces; however, PMMA injections in rabbit vertebral bodies cause a minimal necrotic exothermal effect, indicating that pain relief may be produced by the stabilizing effects of the cement.13 Otherwise, the PMMA and barium sulfate injected into the vertebral bodies may induce foreign body reactions, but the effects of PMMA on the vertebral cancellous bone have not been well-documented.14

Several reports have demonstrated histologic changes in human vertebral bodies after augmentation with PMMA. Togawa et al15 reported foreign body giant cells and macrophages in the fibrous membranes around the PMMA, and a few spicules of necrotic bone associated with creeping substitution. Braunstein et al16 reported large amounts of newly-formed callus surrounding the injected PMMA, filling the spaces between the vertebral endplates, cancellous bone, and the injected PMMA. Some authors have reported that the radiologic characteristics of PMMA-augmented vertebrae in the long term; specifically, the cement remained unchanged and there was no modification of anatomic structures in contact with the cement.7 Lin et al17 reported an increased risk of re-fractures of cemented vertebrae in osteoporotic patients with larger height restoration. Experiments with augmented osteoporotic vertebrae showed fissures along the bone-cement interface at the end of fatigue loading, indicating that the fractures of the augmented vertebrae may be due to uneven deformation of the vertebrae.18,19

In our 31 patients followed up >7 years, only two patients had mild pain and bone marrow edema around the cement 2 and 3 weeks post-operatively, neither loose nor displaced cement was detected on the radiographs, and there was no new fracture or collapse at the augmented vertebrae. On CT scans, the cement closely contacted or infiltrated the trabecular bone. The boundary between the cement and trabecular bone was indistinct and there was no evident radiolucent gap between the cement and trabecular bone. This finding indicates the long-term safety of PVP with PMMA, and in particular, the safety of cement in contact with bone and discs.

A high risk of new VOFs after PVP has been frequently reported, especially at adjacent levels. The risk factors affecting new VOFs include lower bone mineral density, older patient age, intradiscal cement leakage, thoracolumbar location of the initial VOF, a greater restoration rate of vertebral height, more pre-existing vertebral fractures, and high levels of bone resorption markers.20–23 The levels of re-fracture after PVPs often involve the thoracolumbar junction, thus the degree of osteoporosis and altered biomechanics due to resistant kyphosis may be the most important factors for new VOFs after PVPs. The increased loading in structures adjacent to the augmented vertebra may be responsible for adjacent VOFs because the cement inhibits the normal endplate bulge in the augmented vertebra, and thus pressurizes the adjacent disc, which subsequently increases the loading of the untreated vertebra. Some reports have shown that PVPs do not increase the incidence of new VOFs compared with conservative treatment, indicating an elevated risk of new adjacent VOFs unrelated to the PVP.18,24

In our series, anti-osteoporotic medications were used routinely for at least 3 months. Only three new collapses of vertebrae adjacent to the augmented vertebrae were noted 3 weeks, 1 month, and 1 years post-operatively, and the recurrent symptoms resolved with conservative treatment. Whether or not increased physical activity results in an elevated risk of new VOFs is uncertain, therefore patients should be carefully counseled after a PVP to optimize medical therapy for osteoporosis and to use extreme care when engaging in physical activity.25

Based on this 7-11 year follow-up study, PVPs restored patient mobility and provided immediate and durable pain relief of symptomatic VOFs, and the PMMA injected into the vertebral body combined closely with host trabecular bone without adverse reactions, indicating that PVPs are safe and useful in the treatment of VOFs.


1. Chinese Orthopaedic Association. Diagnosis and treatment of osteoporotic fractures. Orthop Surg 2009; 1: 251-257.
2. Tanigawa N, Kariya S, Komemushi A, Nakatani M, Yagi R, Kohzai M, et al. Percutaneous vertebroplasty for osteoporotic compression fractures: long-term evaluation of the technical and clinical outcomes. Am J Roentgenol 2011; 196: 1415-1418.
3. Anselmetti GC, Manca A, Hirsch J, Montemurro F, Isaia G, Osella G, et al. Percutaneous vertebroplasty in osteoporotic patients: an institutional experience of 1634 patients with long-term follow-up. J Vasc Interv Radiol 2011; 22: 1714-1720.
4. Thillainadesan J, Schlaphoff G, Gibson KA, Hassett GM, McNeil HP. Long-term outcomes of vertebroplasty for osteoporotic compression fractures. J Med Imaging Radiat Oncol 2010; 54: 307-314.
5. Mazzantini M, Carpeggiani P, d'Ascanio A, Bombardieri S, Di Munno O. Long-term prospective study of osteoporotic patients treated with percutaneous vertebroplasty after fragility fractures. Osteoporos Int 2011; 22: 1599-1607.
6. Kim JH, Yoo SH, Kim JH. Long-term follow-up of percutaneous vertebroplasty in osteoporotic compression fracture: minimum of 5 years follow-up. Asian Spine J 2012; 6: 6-14.
7. Franc J, Lehmann P, Saliou G, Monet P, Kocheida EM, Daguet E, et al. Vertebroplasty: 10 years clinical and radiological follow-up. J Neuroradiol 2010; 37: 211-219.
8. Ma X, Wang LX, Wang HL, Jiang L, Lu FZ, Jiang JY. Value of preoperative magnetic resonance imaging measurements in thoracic percutaneous vertebroplasty using unilateral puncture. Chin Med J 2010; 123: 2983-2988.
9. Zhang JD, Poffyn B, Sys G, Uyttendaele D. Comparison of vertebroplasty and kyphoplasty for complications. Orthop Surg 2011; 3: 158-160.
10. McKiernan F, Faciszewski T, Jensen R. Does vertebral height restoration achieved at vertebroplasty matter? J Vasc Interv Radiol 2005; 16: 973-979.
11. Steinmann J, Tingey CT, Cruz G, Dai Q. Biomechanical comparison of unipedicular versus bipedicular kyphoplasty. Spine 2005; 30: 201-205.
12. Jin YJ, Yoon SH, Park KW, Chung SK, Kim KJ, Yeom JS, et al. The volumetric analysis of cement in vertebroplasty: relationship with clinical outcome and complications. Spine 2011; 36: E761-E772.
13. Urrutia J, Bono CM, Mery P, Rojas C. Early histologic changes following polymethylmethacrylate injection (vertebroplasty) in rabbit lumbar vertebrae. Spine 2008; 33: 877-882.
14. Togawa D, Kovacic JJ, Bauer TW, Reinhardt MK, Brodke DS, Lieberman IH. Radiographic and histologic findings of vertebral augmentation using polymethylmethacrylate in the primate spine percutaneous vertebroplasty versus kyphoplasty. Spine 2006; 31: E4-E10.
15. Togawa D, Bauer TW, Lieberman IH, Takikawa S. Histologic evaluation of human vertebral bodies after vertebral augmentation with polymethyl methacrylate. Spine 2003; 28: 1521-1527.
16. Braunstein V, Sprecher CM, Gisep A, Benneker L, Yen K, Schneider E, et al. Long-term reaction to bone cement in osteoporotic bone: new bone formation in vertebral bodies after vertebroplasty. J Anat 2008; 212: 697-701.
17. Lin WC, Lee YC, Lee CH, Kuo YL, Cheng YF, Lui CC, et al. Refractures in cemented vertebrae after percutaneous vertebroplasty: a retrospective analysis. Eur Spine J 2008; 17: 592-599.
18. Kamano H, Hiwatashi A, Kobayashi N, Fuwa S, Takahashi O, Saida Y, et al. New vertebral compression fractures after prophylactic vertebroplasty in osteoporotic patients. Am J Roentgenol 2011; 197: 451-456.
19. Tseng YY, Yang TC, Tu PH, Lo YL, Yang ST. Repeated and multiple new vertebral compression fractures after percutaneous transpedicular vertebroplasty. Spine 2009; 34: 1917-1922.
20. Lu K, Liang CL, Hsieh CH, Tsai YD, Chen HJ, Liliang PC. Risk factors of subsequent vertebral compression fractures after vertebroplasty. Pain Med 2012; 13: 376-382.
21. Lee KA, Hong SJ, Lee S, Cha IH, Kim BH, Kang EY. Analysis of adjacent fracture after percutaneous vertebroplasty: does intradiscal cement leakage really increase the risk of adjacent vertebral fracture? Skeletal Radiol 2011; 40: 1537-1542.
22. Chen WJ, Kao YH, Yang SC, Yu SW, Tu YK, Chung KC. Impact of cement leakage into disks on the development of adjacent vertebral compression fractures. J Spinal Disord Tech 2010; 23: 35-39.
23. Komemushi A, Tanigawa N, Kariya S, Kojima H, Shomura Y, Tokuda T, et al. Biochemical markers of bone turnover in percutaneous vertebroplasty for osteoporotic compression fracture. Cardiovasc Intervent Radiol 2008; 31: 332-335.
24. Lin H, Bao LH, Zhu XF, Qian C, Chen X, Han ZB. Analysis of recurrent fracture of a new vertebral body after percutaneous vertebroplasty in patients with osteoporosis. Orthop Surg 2010; 2: 119-123.
25. Rad AE, Gray LA, Sinaki M, Kallmes DF. Role of physical activity in new onset fractures after percutaneous vertebroplasty. Acta Radiol 2011; 52: 1020-1023.

osteoporotic fractures; vertebroplasty; polymethylmethacrylate; treatment outcome

© 2012 Chinese Medical Association