Introduction

The first percutaneous vertebroplasty was performed by a French group in 1984, and reported in the literature in 1987, for the treatment of painful cervical haemangioma [1]. This technique has rapidly become standard care for treatment in medically refractory painful vertebral compression fractures. Vertebral fractures usually become evident due to pain which can have variable intensities and may affect a patient’s quality of life, including back pain, functional limitations, depression, disability, height loss caused by vertebral collapse, spinal instability and, in many cases, kyphotic deformity that could compromise lung capacity [2, 3].

There are many causes of vertebral compression fractures such as osteoporosis, trauma, neoplasm and haemangioma. Some patients do not respond to conservative treatments including analgesics, braces, immobilization and physical therapy [4–6], and some develop negative effects following heavy and prolonged analgesic treatment. Furthermore many patients with severe pain require assistance and hospitalisation for continuous care.

Several authors claim that vertebroplasty is a successful technique, with a 70–90% of higher effectiveness for short-term pain reduction and return to normal activity, but long-term effectiveness is still debated [7–9]. We analysed a large cohort of patients who did not respond to conservative management with analgesics and who underwent vertebroplasty for vertebral fracture caused by different pathological processes. The aim of this study is to present new data about this procedure through the study of a large number of vertebroplasties with short- and long-term follow-up. The effectiveness on pain relief in the short term is again proven.

Materials and methods

We performed a retrospective, nonrandomized review of 624 patients who underwent vertebroplasty between May 2001 and May 2006, at our centre. Institutional review board approval was obtained for this study.

Before the procedure, patients were informed of the benefits and risks of the treatments being proposed, and all patients signed a written consent form to the study in accordance with the national legislation and the Declaration of Helsinki.

Patient selection and evaluation

Between May 2001 and May 2006, 624 patients (202 male, 422 female) underwent vertebroplasty to treat 1,253 vertebral compression fractures: 483/624 (77.4%) vertebroplasties were performed for compression fractures related to osteoporosis; 96/624 (15.4%) for neoplastic involvement or bone metastasis, 23 (3.7%) for myeloma, 3 (0.5%) for haemangiomas, 4 (0.6%) for trauma and 15 (2.4%) for other causes.

Osteoporosis was the most common cause of vertebral fractures, especially in women above 50 years old. The average age of our population was 72 years (range 55–89). The vertebrae to be treated were located in the thoracic segment (504 vertebrae), lumbar segment (737) and sacral segment (12) (Fig. 1).

Fig. 1
figure 1

Distribution of vertebral fractures

Full size image

Patients included in this study were not limited to those with vertebral fractures due to osteoporosis, but different causes of somatic fractures were included (metastasis, myeloma, haemangioma and trauma) (Table 1).

Table 1 Different causes of vertebral fractures
Full size table

Patients were evaluated by three different radiologists with extensive experience in spinal imaging through a multidisciplinary approach that includes complete history, physical examination and radiological imaging. Vertebroplasty was typically offered to patients with radiological evidence of acute or subacute vertebral fractures. The inclusion criteria were: vertebral fracture with height loss of the somatic body (minimal 15%) on imaging, invalidating back pain refractive to medical therapy (defined by minimal or no pain relief with the administration of analgesics or presence of negative effects) for at least 4 weeks, focal tenderness on physical examination related to the level of the vertebral fracture.

For each patient the imaging assessment was obtained in the period preceding percutaneous vertebroplasty in our department. Magnetic resonance imaging (MRI) was performed in addition to routine conventional X-ray radiography and computed tomography (CT) in all patients to evaluate the presence of oedema on T2-weighted sequences, index of acute compression fractures, that could be a predictive factor of a favourable response to the procedure [10].

CT, before vertebroplasty, evaluates the integrity of the posterior somatic wall and helps to assess the eventual posterior displacement of fragments. Furthermore, CT allows measurement of the pedicular diameter, which may influence the size of the needle chosen for puncture, especially in the more slender thoracic vertebral pedicles.

The pain and functional degree were evaluated in all patients using a self-assessment questionnaire submitted before the procedure and in follow-up. This questionnaire included a 0–10 points visual analogue scale (VAS) for the evaluation of pain.

Exclusion criteria for the procedure included response to medical treatment, untreatable coagulopathy, systemic or local infections (osteomyelitis, spondylodiscitis), presence of radicular symptoms or neurologic deficits, spinal stenosis and noncorrelating pain.

Vertebroplasty procedure

The procedure was performed under local anaesthesia (injecting a mixture of lidocaine 1% with bupivacaine 0.25%) in the prone position on a fluoroscopy table (90% of patients) or in a CT unit (10% of patients) through a 22-G spinal needle both at the skin level and deep to include the periostium of the pedicle (Fig. 2).

Fig. 2
figure 2

A 63-year-old female patient with back pain due postmenopausal osteoporosis. A CT volume rendering postprocessing. Parasagittal view of the thoracolumbar spine with unipedicular approach in different vertebral levels. B CT of PMMA injection during vertebroplasty

Full size image

Biplanar fluoroscopy was mandatory for this procedure to be performed under fluoroscopic guidance. Intraosseous venography is not performed in our institution. Needles with connection tubing and cement injection syringes (11−13 G) were used. In 88% of cases transpedicular access was unilateral; a bilateral approach was performed only in a few cases (12%) when the diffusion of cement was not satisfactory with a unilateral approach (especially at the beginning of our study).

Trochar/s advanced through the pedicle, sloping anteriorly, medially and caudally. A poly(methyl methacrylate) (PMMA) mixture was injected into the vertebral body (mean 2.5 ml) after careful imaging to confirm location of the trochar/s in the anteromedial portion of vertebral body; cement injection was executed on lateral view with continuous fluoroscopic monitoring, with close attention at the posterior margin of the vertebral body and at the epidural space. During cement deposition, frequent fluoroscopic controls in both planes were required to ensure that the material remained within the vertebral body without migrating into the surrounding venous plexus. Injection was terminated when adequate filling of the vertebral body was obtained or when the cement reached the posterior quarter of the vertebral body.

A maximum of six vertebral levels were treated in a single session based on clinician comfort levels. Additional levels were treated in repeat sessions as required. Up to three vertebrae were treated in the same section in 20 patients (16 with osteoporosis, two with metastasis and two with myeloma), and we also performed prophylactic vertebroplasty in vertebrae positioned between two vertebral fractures (Fig. 3).

Fig. 3
figure 3

A 70-year-old female patient with osteoporosis: treatment of six vertebral levels. A Unipedicular trochars approach. Lateral radiograph. B Lateral view of PMMA injection. C CT volume rendering postprocessing

Full size image

Sixty one patients underwent a second percutaneous vertebroplasty for new vertebral fractures.

Outcome measures

At the end of the procedure, the patient remained supine in bed. After 4 h patients were examined by radiologists and VAS was obtained [11]. VAS was indicated on a ten-point scale ranging from 0 (no pain) to 10 (worst pain). The analgesic use was classified as no medication (0), use of paracetamol (1), use of nonsteroidal anti-inflammatory drug (NSAIDs) (2), and use of opiate derivatives (3). At 4 h postprocedure, patients were asked to report their pain as being improved, unchanged, or worse than before the procedure.

Follow-up was assessed at 1 week and at 3, 6 and 12 months by VAS; patients were seen in our department or contacted by phone thereafter for our data set. Patients also underwent spine radiography at 3 months after the vertebroplasty procedure.

In follow-up period the radiologist asked whether or not the procedure had calmed the pain for which they were treated and if they still took analgesic compounds. If patients were referred with sudden pain, they underwent magnetic resonance imaging of the region involved.

Statistical analysis

Statistical analysis was performed to assess the mean difference of visual analogue pain score (VAS). A paired two-tailed t-test was applied to VAS values pre- and post-vertebroplasty and to each follow-up time point to assess the significance of pain response and change with time. P values less than 0.05 were considered significant. In addition, because the nature of the lesions varied among patients, the univariate analysis of variance (ANOVA) test was applied to analyse VAS score changes in different types of lesions.

Results

Data were collected from 624 patients who underwent vertebroplasty at 1,253 vertebral levels. In no cases did PMMA flow out into the spinal vertebral canal. We reported 72 (11.5%) cases of cement venous leakage and 63 (10.1%) cases of diskal leakage (Fig. 4).

Fig. 4
figure 4

Distribution of complications

Full size image

Two patients (0.32%) developed a clinically silent cement pulmonary embolism. In both cases, embolism was immediately recognized during the procedure and patients subsequently underwent unenhanced chest CT which confirmed this complication. Only one patient (0.16%) presented radiculopathy, which was successfully treated with oral drugs. No patients presented with complications such as haematoma, infections, paraplegia or other major complications; no deaths were attributable to the vertebroplasty procedure.

A statistically significant decrease of pain was observed after the vertebroplasty procedure: a significant reduction of VAS scores at 1 week and at 3, 6 and 12 months after treatment was seen in patients that had completed the entire follow-up of 12 months (609 patients, 97.5%). Most patients reported an improvement in their mobility and quality of life; only 29 patients (4.65%) did not report any back pain improvement.

Whatever the cause for vertebral collapse, the VAS score evaluation pretreatment and post-treatment proved an important decrease in pain (Fig. 5).

Fig. 5
figure 5

Comparison of the average VAS pre and postvertebroplasty in different pathologies

Full size image

Table 2 shows VAS values obtained for the different pathological processes: before vertebroplasty, there was a mean VAS score of 8.2 ± 2.6 for myeloma, 8.2 ± 2.5 for osteoporosis, 8.6 ± 2.7 for metastasis, 8.0 ± 2.5 for haemangioma, 7.1 ± 2.1 in the case of trauma. The overall results show a reduction of the median pain score from 8 ± 2.5 to 1.5 ± 0.4 post-PVT; the average VAS score after the procedure was 2.1 ± 0.7 for myeloma, 1.5 ± 0.5 for osteoporosis, 2.1 ± 0.6 for metastasis, 0.4 ± 0.1 for haemangioma and 1.4 ± 0.4 for trauma.

Table 2 VAS results prevertebroplasty (VTP) and postvertebroplasty (VTP)
Full size table

Some patients treated, underwent a second percutaneous vertebroplasty for new vertebral fractures, in other localizations of disease: 52 (10.8%) new fractures occurred due to osteoporosis, seven to metastasis and two to myeloma. The patients treated a second time showed results comparable to those of patients who had already undergone PVT.

Thirty one (6.4%) osteoporotic fractures involved the contiguous vertebrae: 12 at the level above and 19 at the level below. The same treatment protocol previously described was applied at the second level.

Outcome measures

According to the visual analogue pain score, a significant decrease in pain was seen between pre- and post-procedure values. The pain response was similar at 4 h postprocedure evaluation and 1 year (Fig. 6).

Fig. 6
figure 6

Comparison of VAS score before and after vertebrosplasty

Full size image

VAS response was statistically significant at every evaluation with a P value of < 0.001. By comparing VAS scores obtained from the different patient groups (osteoporosis, metastasis, myeloma, haemangioma and trauma cases) we did not find statistically significative differences in decrease in pain (P > 0.05).

Discussion

Although vertebroplasty is rapidly spreading as a treatment for vertebral compression fractures, the evidence for efficacy is weighed largely on symptomatic improvement in perceived pain from case series. The procedure is also designed to improve ambulation and ability to perform daily life activities and to reduce drug treatment.

In our study, an important number of patients developed vertebral fracture due to osteoporosis. Osteoporosis affects women four times more than men and is more frequently found in Caucasians. There are many factors that can put an individual at risk: estrogen deficiency or menopause, senile osteoporosis, excessive physical exercise or eating disorders, chronic diseases and related medications, lifestyle choices including smoking, excessive alcohol, poor diet. It is estimated that approximately 20–25% of individuals have severe symptoms from osteoporosis [12]. Patient presentation can vary, but the most common complaint is pain which can be debilitating.

Our results showed that vertebroplasty was very efficient in early drastic reduction of pain in 94.4% of patients: these results are substantially in accordance with recent studies that reported success rates above 85% [13–16]. Our success might be attributed to the following factors: strict evaluation of clinical indications, preprocedural evaluation with MRI, good quality fluoroscopic equipment and ultraspecialised operators’ expertise. Furthermore the bilateral approach was occasionally employed and this is a possible factor influencing the incidence of adverse events.

These results were obtained at very little discomfort in terms of hospital stay and adverse effects: all of patients were discharged the same day as the procedure. This is of particular interest because most patients treated with PVT are elderly and frail.

We performed PVT in all patients, regardless of the underlying lesion: it is interesting to note that the outcome was similar. No significant difference was observed in terms of pain reduction between patients with osteoporosis (improvement of 6.9 points on VAS), metastatic involvement (6.6), myeloma (6.2), haemangioma (7.6) and traumatic lesions (5.9). Therefore, there was no significant difference in VAS score after vertebroplasty during follow-up in accordance with Layton et al. [17]. Nor did age seem to influence the outcome.

The average number of treated vertebrae for our patients was higher comparison with other studies [18]. Therefore we performed PVT on unfractured vertebrae located between two collapsed bodies to reduce the risk of secondary fracture.

There was a low complication rate in our study, and all of these have been described in previous vertebroplasty case reports [5, 19–21].

Paraspinal, intradiskal, and venous cement extravasation rates for PVT are relatively high and were documented to occur in up to 88% of patients [22–26]. In our opinion, in accordance with Hulme et al. [23], the best method to monitor cement leakage is postoperative CT. This should identify all leaks, but this needs to be balanced against increased patient radiation exposure and cost. Most cement leakages are asymptomatic but the long-term effect of these “benign” leaks is unknown; in our experience we had only two patients that developed asymptomatic pulmonary embolism and during our follow-up they did not present chest pain or cardiovascular distress. In our study, 135/624 of patients had some form of cement leakage documented by fluoroscopy as intraprocedural presence of cement outside the vertebral body. Small cement leakage alone should not be considered a complication, but a stopping point for injection. It is important to recognize cement leakage during the procedure so that it can be terminated or the contralateral pedicle entered. Only three patients had complications related to cement leakage: two patients were found to have cement pulmonary embolism (clinically silent) and a single patient developed radiculopathy. Furthermore no procedure-related infection, haematoma or needle rupture was encountered.

In our case series some patients developed new clinically significant contiguous fractures during follow-up (31/1,253 levels in patients who developed new fractures contiguous to treated vertebrae), particularly at the beginning of our study. Most of these patients that we re-treated were affected by osteoporosis. Because of this, we decided to perform prophylactic vertebroplasty in vertebral segment between two collapsed vertebrae. Our results show lower percentages of new fractures compared with other recent studies [27–28].

There are many limitations to our study. First, although demographic and technical information were collected prospectively, the patients had to be recalled for pain evaluation. Retrospective study is a potential for bias, as patients might not recall pain before treatment or, if treatment has been very successful, might enhance the memory of pain. Nevertheless, the reported benefits of vertebroplasty in terms of pain reduction were constant even as time from vertebroplasty to phone interview increased: this suggests that the examination was homogenous because the outcome opinion is not influenced by time.

Second, without a comparison group it is not possible to consider whether vertebroplasty is the treatment of choice rather than conservative treatment of painful vertebral fractures. Vertebroplasty has yet to be validated by a prospective randomized study. Nowadays randomized prospective studies are currently being performed internationally, as the investigational vertebroplasty efficacy and safety trial (INVEST): after the results of these studies, we can explicitly declare that vertebroplasty has been validated as a treatment of choice for patients with painful vertebral fractures. Our study shows that vertebroplasty is a reliable procedure in the treatment of vertebral fractures caused by a wide range of different diseases.

Third, it is possible that during the follow-up questions we could inadvertently influence the responses of patients, depending on how the questions were asked. As many patients were older, it is possible that responses to the follow-up questions were occasionally provided by family members.

In conclusion, vertebroplasty should be considered a safe and reliable treatment in symptomatic acute vertebral fractures, in patients with severe pain refractory to analgesic drug therapy. Vertebroplasty requires hospitalization and must be performed in accredited centres. Furthermore, the multilevel treatment might be effective in preventing new fractures in high-risk patients.