Percutaneous Vertebral Augmentation in Metastatic Disease: State of the Art

Indications

PV is an interventional procedure in which bone cement, usually polymethyl methacrylate (PMMA), is injected under radiologic guidance into the collapsed vertebral body. This procedure was first reported by Galibert in 1987⁵ for the treatment of painful aggressive vertebral hemangiomas and subsequently used in the United States in 1993.⁶ Since then, the technique has been widely applied and has evolved as the treatment of choice for painful osteoporotic fractures. The second most common use of PV and PMMA is for the treatment of painful vertebral fractures caused by metastatic disease or multiple myeloma.

PV also works well in those metastatic cases in which intractable pain is not associated with vertebral body collapse.⁷

Further reported indications are spinal pseudoarthrosis, intravertebral vacuum phenomenon, Langerhans cell histiocytosis, osteogenesis imperfecta, and Paget disease.⁸

Recent studies show that PV should be performed before radiotherapy for the treatment of those spinal metastases that have led to pathologic compression without neurological deficits.

In selected cases, PV is a minimally invasive alternative to open surgery.

PK is an evolution/modification of PV, combining the analgesic effect of PV and the possibility of restoring normal height of the collapsed vertebral body. This improves the kyphotic deformity frequently associated with vertebral fractures of the thoracic segment. Introduced in 2001, it employs a height-restoration device, known as “balloon tamp,” that when inflated restores the original shape of the vertebral body while creating a cavity in which to inject the cement.

The ideal candidates for this augmentation procedure are those patients who complain of midline nonradiating back pain that increases with weight bearing and is exacerbated by manual palpation of the spinous process of the involved vertebrae. This typical symptomatology subsides with recumbency and/or sitting. Exceptions are patients with thoracic spine fractures in whom pain radiates to the ribs or those with fractures at the level of the conus medullaris, where pain may radiate to the hips without evidence of cord compression.⁹

Detailed anamnesis, accurate neurologic examination, and recent radiographic imaging are mandatory to exclude spinal cord compromise and/or retropulsed bony fragments in the canal, which obviously represent a contraindication to the procedure.

Indications of both procedures have been defined in the guidelines published by the Society of Interventional Radiology in 2003 and recently updated by the Cardiovascular and Interventional Radiological Society of Europe.¹⁰ Indications include painful osteoporotic vertebral fractures after 3 weeks of analgesic therapy, painful vertebrae due to benign or malignant primary or secondary bone tumors, painful vertebral fractures with osteonecrosis (Kümmell disease), reinforcement of the vertebral body prior to a surgical procedure, and chronic traumatic vertebral fractures with non-union. Absolute contraindications include asymptomatic vertebral fractures, pain improving with medical therapy, ongoing infection, osteoporotic patient prophylaxis, uncorrectable coagulopathy, myelopathy due to retropulsion of bone (canal compromise), and an allergic reaction to PMMA or the opacification agent. Relative contraindications are radicular pain, vertebral fractures >70% of height loss, severe spinal stenosis, asymptomatic retropulsion of bony fragment, tumor extension into canal/epidural space, and lack of surgical backup.

Although disruption of the posterior cortex of the vertebral body has been considered as a relative contraindication, new techniques allow efficacious PV in these circumstances.

Multilevel procedures (3–4 levels) should be avoided in patients with low cardiopulmonary reserve (ie, chronic obstructive airway disease or congestive cardiac failure) as these patients may be at high risk for symptomatic pulmonary fat embolism.¹¹

Informed consent should include discussion regarding failure to obtain pain relief as well as complications. The eventuality of open stabilization or urgent decompressive surgery should also be discussed with the patient.

Radiology

Traditional anteroposterior and laterolateral X-rays show the degree of vertebral compression, eventual osteolysis, extent of pedicle involvement, and fracture or cortical destruction.

Computed tomography (CT) may be useful to further define the extent of vertebral collapse, the location and extent of any lytic process, the visibility and degree of involvement of the pedicles, the presence of cortical destruction and epidural or foraminal stenosis caused by tumor extension or bone fragment displacement, and to estimate the needle path and size.

Magnetic resonance imaging (MRI) is pivotal in triaging patients before PV or PK. In fact, signal changes within the vertebral body marrow suggest edematous changes in a healing fracture, which are a necessary condition for the procedure. Sequences that are particularly sensitive to the presence of edema are fat-suppressed T2-weighted and fat-suppression inversion recovery images, which may also be useful in planning the vertebroplastic procedure.¹¹

Concomitant limitations of these specific MRI sequences are that they may reveal increased activity up to 2 years after fracture, whereas it is widely accepted that patients with fractures older than 6 months do not benefit from vertebroplasty.¹²

Bone nuclear scanning has also been used to identify recent fractures in patients with multiple involved levels or in patients in whom MRI is not possible because of a pacemaker or stent. The recent fracture is typically heralded by an intense radioisotope uptake.¹³

Radiological investigation is a fundamentally delicate tool focusing on the most recent symptomatic fractured level. It helps solve those unclear cases in which physical examination is not sufficient to determine which of several adjacent fractures is symptomatic or which fracture has no pain on palpation over the involved vertebra and/or there is no correlation between the affected site and the pain localization.

Techniques

PV and PK require a detailed knowledge of spine anatomy and an intensive training in fluoroscopic imaging interventional procedures. The procedure should be performed in an appropriately sterile area. Broad-spectrum antibiotics can be administered just before the treatment. An anesthesiologist or other physician able to undertake rate and pulse oximetry must be present continuously. The typology of anesthesia should be selected on an individual basis. A generous amount of local anesthetic, especially into the periosteum, is suggested to maintain communication with the patient. Emergency measures should be available in the operating theater. In selected cases, general anesthesia is used.

Good-quality imaging biplanar or C-arm fluoroscopy with a radiolucent table is mandatory for maximal procedural safety, to correctly identify the anatomical structures (eg, pedicle, posterior wall).

There are various types of cement (methyl methacrylate powder) currently in use. They basically differ in their polymerization times, and the practitioner performing the procedure should be very skilled in this specific aspect. Cement opacification with barium sulfate specifically designed for use in vertebroplasty is required.

Most operators aim to obtain sufficient cement placement into the vertebral body using a monopedicular approach, thus reducing the procedure time.

PMMA injection into the vertebral body is performed when the trochar (uni- or bipedicular approach) has been deepened into the ventral portion of the vertebral body.

The cement should harden to “toothpaste” consistency before injecting. The injection should be stopped once the cement spreads to the posterior third of the vertebral body.

The cement column should ideally spread among the superior and inferior endplates and between the two pedicles. Such cement filling should prevent the collapse of the treated vertebral body.

If PMMA diffuses into a blood vessel or toward the posterior cortical margin, injection must be immediately stopped. Use of high-viscosity cement and small-volume injection is recommended in order to minimize the risk of PMMA leakage.

As Belkoff et al¹⁴ reported, maximal filling of the compressed vertebral body is not necessary; 2 mL of cement are sufficient for restoring vertebral body strength. Inadequate cement in the unstable fractured area may be responsible for unrelieved pain.

PK (Figure 1) differs from PV (Figure 2) in that it involves the percutaneous placement of balloons (called “tamps”) into the vertebral body with an inflation/deflation sequence that creates a cavity before the cement injection. PK may restore the vertebral body height and reduce the kyphotic angulation of the compression fracture before PMMA injection.

Proposed Mechanisms of Pain Relief

The most accepted theory indicates augmentation with cement increases the fracture mechanical load threshold, stabilizing the vertebra.¹⁵ In tumor fractures, pain relief is related primarily to vertebral body stabilization and secondarily to the induction of tumor necrosis and the destruction of sensitive nerve endings.¹⁶ The last two effects seem to be directly connected to the local heat produced by the highly exothermic reaction of the PMMA polymerization. The antitumoral effect is supported by the local cytotoxic effect of PMMA on rapidly proliferating cells.¹⁷

The heat effect is related to the degree and duration of the heat-exposure period. However, this does not explain why the analgesic effect is not proportionally related to the volume of PMMA injected. Low-volume injection has been found to be as effective as high-volume injection in relieving pain. On these bases, although a unilateral approach may allow a satisfying filling of the vertebra in terms of stabilization, it might not be sufficient when an additional antitumoral effect is desired.

Interestingly, despite the unequivocal effect on pain, the literature remains unclear about the reliability of vertebroplasty in achieving bone stabilization and preventing future vertebral fracturing.²

Results

Pain relief and increased mobility are expected within 24 hours following PV.¹⁸ Significant pain relief is expected in >70% of patients with vertebral malignancies, in >90% of patients with osteoporotic fractures, and in about 80% of patients with hemangioma. Pain relief usually persists over months to years.¹⁹

The significant decrease in back pain resulting from PV and PK in patients with pathologic fractures dramatically improves their quality of life.²⁰ In particular, the procedure-related benefits include reduction/withdrawal of analgesic drugs and improvement in physical mobility. (It is well known that the lower the intake of analgesics, the higher the perception of quality of life, thanks to the disappearance of drowsiness and nausea resulting in an improved appetite.) PK has not been shown to be better than PV in terms of pain relief or quality of life. Eck and colleagues²¹ found in a recent meta-analysis that although both methods provide significant improvement in visual analogue scale scores, there is a statistically greater improvement in pain relief with PV than with PK. In their literature review, Cloft and Jensen²² concluded that there was no proven advantage of PK over PV with regard to pain relief, vertebral height restoration, and complication rate. Moreover, Mathis²³ found that the claims of superior height restoration by PK are insufficiently documented.

PK is estimated to be 2.5–7.0 times more expensive than PV due to additional equipment, general anesthesia, and hospital costs.²² Based on 2006 available data, Mathis²³ found no substantial scientific, procedural, or economic advantages supporting PK's superiority over PV.

Complications

Complications are rare but can be dramatic. Minor complications such as PMMA extension into the disc require no therapy and have no clinical consequence. Major complications may consist of PMMA invasion of the spinal canal with related neurological deficit. The latter may require urgent laminectomy and evacuation of the extruded cement to prevent permanent sequelae.

According to the Society of Interventional Radiology (SIR), the major complication rate is <1% and reaches about 5% in tumor cases.²⁴ Cement leakage is documented in 30%–72.5% of cases on radiographs and in 87.9%–93% of cases on CT.²⁵ PMMA can flow outside the vertebral body posteriorly into the spinal canal and neural foramina. This is usually due to destruction of the vertebral body posterior cortex and of the medial/inferior cortex of the pedicle, but it rarely necessitates urgent surgical decompression. Leakage can occur into the paravertebral veins, mostly without clinical consequences.²⁶ Asymptomatic extension into the inferior vena cava may also occur²⁷ but involves possible systemic complications such as pulmonary embolism and paradoxical cerebral arterial PMMA emboli.¹⁸ Asymptomatic pulmonary embolization may occur in up to 4%–6.8% of patients.²⁸

Local metastasis occurring in the needle track after PV has also been reported.²⁹ An increase in local pain and fever may occur following the procedure but usually resolves within 72 hours.¹⁸

PK complications are mostly related to incorrect placement of hardware or cement extravasation and may result in neurologic insult.³⁰ The incidence of cement leakage during PK is in the range 8.6%–33%. The reasons for a lower cement extravasation percentage with PK include (1) balloon tamping of a cavity that is surrounded by a shell of impacted cancellous bone, (2) the ability to determine the amount of cement to be injected thanks to the knowledge of the volume of fluid used to inflate the balloon, and (3) lower injection pressure during cement injection.

Infective complications are overall rare. However, meticulous attention to a sterile technique is warranted and should include preoperative intravenous antibiotic administration.

Adjacent-Level Fractures

An increased risk of adjacent-level fracture following PV has been noticed with long-term follow-up. The reported rate of new fractures varies from 7% to 20% within 1 year of follow-up and affects mainly the immediately adjacent levels.³¹ In a recent meta-analysis of the literature, the authors found that the risk of developing a new fracture after kyphoplasty is 14.1%, which is similar to the 17.9% estimated following vertebroplasty.

A local, unfavorable biomechanical situation seems to be included in those patients who suffer adjacent-level fracturing, while the nonadjacent fracture group is more likely to be related to an ongoing disease process such as osteoporosis. Thus, risk factors include osteoporosis, previous vertebral fracture, and organ transplant, which expose the patient to a high osteoporosis risk.

Both Grados et al¹⁹ and Lin et al³² postulated that intradiscal leakage of PMMA may correlate with mechanical consequences on adjacent vertebrae, particularly those with osteoporosis, with a consequent increase in fracture risk. Recent papers,¹⁹ however, reported an incidence of adjacent vertebral fracture (1 year or less) after PV equivalent to the one expected in untreated osteoporotic vertebral fractures.

Neither volume of cement injected nor extravasation of cement into the intravertebral disc seems to increase the likelihood of subsequent adjacent vertebral fracture. A retrospective study showed that a targeted exercise program (Rehabilitation of Osteoporosis Program–Exercise [ROPE] incorporating isometric back-extensor muscle strengthening and proprioceptive postural retraining) after PV significantly decreases fracture recurrence; refracture rates were also lower in the rehabilitation group compared to the vertebroplasty-only group.³³

Newer biomaterials may help to prevent adjacent-level fracturing, thus diminishing the overall complication rates.³⁴ The development of new fractures subsequent to PK has also been reported.³⁵

Discussion

The treatment of painful vertebral metastases remains a major challenge in patients under palliative care. While radiotherapy has represented the gold standard for many years, there is now increasing evidence supporting the combination of surgery with radiation as more efficacious.³⁶ Numerous retrospective studies validate the efficacy and safety of PV. However, no study has investigated the therapeutic effect and safety of this technique in metastatic fractures. The only phase I/II clinical study of PV as a palliative for painful malignant vertebral compression fractures³⁷ confirms the procedure efficacy without reporting any severe complication.

The “painkiller” effect is clearly superior and more rapid when compared to radiotherapy. This becomes particularly useful in those patients with poor prognosis. Since this therapy is not designed to exert an antitumor effect but rather to provide pain relief by strengthening weakened vertebrae, pain recurrence is unavoidable if the metastatic foci expand.

An evidence-based assessment of PV has shown that the costs of this procedure are relatively low if compared with open surgical interventions for vertebral compression fractures and still inferior to the conservative treatment, which may consist of prolonged bed rest, analgesic drugs, orthotic devices, and eventual complications due to immobilization.

Maximal attention to indications and contraindications is pivotal in this procedure in which the percutaneous approach would otherwise represent a great disadvantage in case of major complications ([Table 1] and [Table 2]).

Recent reviews and editorials have called for a more critical evaluation of these procedures.³⁸ Controlled, multicentered trials are needed to prove its short- and long-term safety, efficacy, and cost–effectiveness in the treatment of metastatic vertebral fractures.

In addition, the procedures have reached widespread popularity, making it difficult to gain consent from patients to be included in the conservative treatment group of a randomized study.

Studies also need to be done to compare PV and PK in various disease states in a randomized fashion. Although PV does not restore vertebral body height, the procedure may be more appropriate than PK in high-risk patients or when the indication is solely for pain control.

The superiority of one technique over the other remains controversial. In spite of the necessity for more vigorous research, percutaneous augmentation with both procedures is promising in the treatment of painful vertebral fractures due to malignant infiltration.

With careful selection, adequate training, and meticulous attention to detail during the procedure, devastating complications can be dramatically reduced. More recently, the combination of PK with radiosurgery has proven to be effective in treating pathological fractures and in reducing the risk of fracture progression related to radiosurgery.³⁹

There is growing interest in the use of less invasive techniques for the treatment of spinal metastases. One example is percutaneous radiofrequency ablation coupled with PMMA injection for debulking tumor and vertebral stabilization.⁴⁰

Another novel technique involves the combination of percutaneous tumor debulking, PK, and intravertebral administration of a mixture of radiolabeled ¹⁵³Sm-EDTPM and PMMA.⁴¹

Conclusions

PV and PK are minimally invasive techniques used to treat painful vertebral compression fractures. There is a growing body of evidence indicating that these procedures are efficacious in alleviating pain associated with vertebral compression fractures. However, they need to be performed by a very skilled operator to prevent those rare but eventually serious complications. Because of this, recent reviews and editorials have called for a more critical evaluation of these procedures.

Indications

PV is an interventional procedure in which bone cement, usually polymethyl methacrylate (PMMA), is injected under radiologic guidance into the collapsed vertebral body. This procedure was first reported by Galibert in 1987⁵ for the treatment of painful aggressive vertebral hemangiomas and subsequently used in the United States in 1993.⁶ Since then, the technique has been widely applied and has evolved as the treatment of choice for painful osteoporotic fractures. The second most common use of PV and PMMA is for the treatment of painful vertebral fractures caused by metastatic disease or multiple myeloma.

PV also works well in those metastatic cases in which intractable pain is not associated with vertebral body collapse.⁷

Further reported indications are spinal pseudoarthrosis, intravertebral vacuum phenomenon, Langerhans cell histiocytosis, osteogenesis imperfecta, and Paget disease.⁸

Recent studies show that PV should be performed before radiotherapy for the treatment of those spinal metastases that have led to pathologic compression without neurological deficits.

In selected cases, PV is a minimally invasive alternative to open surgery.

PK is an evolution/modification of PV, combining the analgesic effect of PV and the possibility of restoring normal height of the collapsed vertebral body. This improves the kyphotic deformity frequently associated with vertebral fractures of the thoracic segment. Introduced in 2001, it employs a height-restoration device, known as “balloon tamp,” that when inflated restores the original shape of the vertebral body while creating a cavity in which to inject the cement.

The ideal candidates for this augmentation procedure are those patients who complain of midline nonradiating back pain that increases with weight bearing and is exacerbated by manual palpation of the spinous process of the involved vertebrae. This typical symptomatology subsides with recumbency and/or sitting. Exceptions are patients with thoracic spine fractures in whom pain radiates to the ribs or those with fractures at the level of the conus medullaris, where pain may radiate to the hips without evidence of cord compression.⁹

Detailed anamnesis, accurate neurologic examination, and recent radiographic imaging are mandatory to exclude spinal cord compromise and/or retropulsed bony fragments in the canal, which obviously represent a contraindication to the procedure.

Indications of both procedures have been defined in the guidelines published by the Society of Interventional Radiology in 2003 and recently updated by the Cardiovascular and Interventional Radiological Society of Europe.¹⁰ Indications include painful osteoporotic vertebral fractures after 3 weeks of analgesic therapy, painful vertebrae due to benign or malignant primary or secondary bone tumors, painful vertebral fractures with osteonecrosis (Kümmell disease), reinforcement of the vertebral body prior to a surgical procedure, and chronic traumatic vertebral fractures with non-union. Absolute contraindications include asymptomatic vertebral fractures, pain improving with medical therapy, ongoing infection, osteoporotic patient prophylaxis, uncorrectable coagulopathy, myelopathy due to retropulsion of bone (canal compromise), and an allergic reaction to PMMA or the opacification agent. Relative contraindications are radicular pain, vertebral fractures >70% of height loss, severe spinal stenosis, asymptomatic retropulsion of bony fragment, tumor extension into canal/epidural space, and lack of surgical backup.

Although disruption of the posterior cortex of the vertebral body has been considered as a relative contraindication, new techniques allow efficacious PV in these circumstances.

Multilevel procedures (3–4 levels) should be avoided in patients with low cardiopulmonary reserve (ie, chronic obstructive airway disease or congestive cardiac failure) as these patients may be at high risk for symptomatic pulmonary fat embolism.¹¹

Informed consent should include discussion regarding failure to obtain pain relief as well as complications. The eventuality of open stabilization or urgent decompressive surgery should also be discussed with the patient.

Radiology

Traditional anteroposterior and laterolateral X-rays show the degree of vertebral compression, eventual osteolysis, extent of pedicle involvement, and fracture or cortical destruction.

Computed tomography (CT) may be useful to further define the extent of vertebral collapse, the location and extent of any lytic process, the visibility and degree of involvement of the pedicles, the presence of cortical destruction and epidural or foraminal stenosis caused by tumor extension or bone fragment displacement, and to estimate the needle path and size.

Magnetic resonance imaging (MRI) is pivotal in triaging patients before PV or PK. In fact, signal changes within the vertebral body marrow suggest edematous changes in a healing fracture, which are a necessary condition for the procedure. Sequences that are particularly sensitive to the presence of edema are fat-suppressed T2-weighted and fat-suppression inversion recovery images, which may also be useful in planning the vertebroplastic procedure.¹¹

Concomitant limitations of these specific MRI sequences are that they may reveal increased activity up to 2 years after fracture, whereas it is widely accepted that patients with fractures older than 6 months do not benefit from vertebroplasty.¹²

Bone nuclear scanning has also been used to identify recent fractures in patients with multiple involved levels or in patients in whom MRI is not possible because of a pacemaker or stent. The recent fracture is typically heralded by an intense radioisotope uptake.¹³

Radiological investigation is a fundamentally delicate tool focusing on the most recent symptomatic fractured level. It helps solve those unclear cases in which physical examination is not sufficient to determine which of several adjacent fractures is symptomatic or which fracture has no pain on palpation over the involved vertebra and/or there is no correlation between the affected site and the pain localization.

Techniques

PV and PK require a detailed knowledge of spine anatomy and an intensive training in fluoroscopic imaging interventional procedures. The procedure should be performed in an appropriately sterile area. Broad-spectrum antibiotics can be administered just before the treatment. An anesthesiologist or other physician able to undertake rate and pulse oximetry must be present continuously. The typology of anesthesia should be selected on an individual basis. A generous amount of local anesthetic, especially into the periosteum, is suggested to maintain communication with the patient. Emergency measures should be available in the operating theater. In selected cases, general anesthesia is used.

Good-quality imaging biplanar or C-arm fluoroscopy with a radiolucent table is mandatory for maximal procedural safety, to correctly identify the anatomical structures (eg, pedicle, posterior wall).

There are various types of cement (methyl methacrylate powder) currently in use. They basically differ in their polymerization times, and the practitioner performing the procedure should be very skilled in this specific aspect. Cement opacification with barium sulfate specifically designed for use in vertebroplasty is required.

Most operators aim to obtain sufficient cement placement into the vertebral body using a monopedicular approach, thus reducing the procedure time.

PMMA injection into the vertebral body is performed when the trochar (uni- or bipedicular approach) has been deepened into the ventral portion of the vertebral body.

The cement should harden to “toothpaste” consistency before injecting. The injection should be stopped once the cement spreads to the posterior third of the vertebral body.

The cement column should ideally spread among the superior and inferior endplates and between the two pedicles. Such cement filling should prevent the collapse of the treated vertebral body.

If PMMA diffuses into a blood vessel or toward the posterior cortical margin, injection must be immediately stopped. Use of high-viscosity cement and small-volume injection is recommended in order to minimize the risk of PMMA leakage.

As Belkoff et al¹⁴ reported, maximal filling of the compressed vertebral body is not necessary; 2 mL of cement are sufficient for restoring vertebral body strength. Inadequate cement in the unstable fractured area may be responsible for unrelieved pain.

PK (Figure 1) differs from PV (Figure 2) in that it involves the percutaneous placement of balloons (called “tamps”) into the vertebral body with an inflation/deflation sequence that creates a cavity before the cement injection. PK may restore the vertebral body height and reduce the kyphotic angulation of the compression fracture before PMMA injection.

Proposed Mechanisms of Pain Relief

The most accepted theory indicates augmentation with cement increases the fracture mechanical load threshold, stabilizing the vertebra.¹⁵ In tumor fractures, pain relief is related primarily to vertebral body stabilization and secondarily to the induction of tumor necrosis and the destruction of sensitive nerve endings.¹⁶ The last two effects seem to be directly connected to the local heat produced by the highly exothermic reaction of the PMMA polymerization. The antitumoral effect is supported by the local cytotoxic effect of PMMA on rapidly proliferating cells.¹⁷

The heat effect is related to the degree and duration of the heat-exposure period. However, this does not explain why the analgesic effect is not proportionally related to the volume of PMMA injected. Low-volume injection has been found to be as effective as high-volume injection in relieving pain. On these bases, although a unilateral approach may allow a satisfying filling of the vertebra in terms of stabilization, it might not be sufficient when an additional antitumoral effect is desired.

Interestingly, despite the unequivocal effect on pain, the literature remains unclear about the reliability of vertebroplasty in achieving bone stabilization and preventing future vertebral fracturing.²

Results

Pain relief and increased mobility are expected within 24 hours following PV.¹⁸ Significant pain relief is expected in >70% of patients with vertebral malignancies, in >90% of patients with osteoporotic fractures, and in about 80% of patients with hemangioma. Pain relief usually persists over months to years.¹⁹

The significant decrease in back pain resulting from PV and PK in patients with pathologic fractures dramatically improves their quality of life.²⁰ In particular, the procedure-related benefits include reduction/withdrawal of analgesic drugs and improvement in physical mobility. (It is well known that the lower the intake of analgesics, the higher the perception of quality of life, thanks to the disappearance of drowsiness and nausea resulting in an improved appetite.) PK has not been shown to be better than PV in terms of pain relief or quality of life. Eck and colleagues²¹ found in a recent meta-analysis that although both methods provide significant improvement in visual analogue scale scores, there is a statistically greater improvement in pain relief with PV than with PK. In their literature review, Cloft and Jensen²² concluded that there was no proven advantage of PK over PV with regard to pain relief, vertebral height restoration, and complication rate. Moreover, Mathis²³ found that the claims of superior height restoration by PK are insufficiently documented.

PK is estimated to be 2.5–7.0 times more expensive than PV due to additional equipment, general anesthesia, and hospital costs.²² Based on 2006 available data, Mathis²³ found no substantial scientific, procedural, or economic advantages supporting PK's superiority over PV.

Complications

Complications are rare but can be dramatic. Minor complications such as PMMA extension into the disc require no therapy and have no clinical consequence. Major complications may consist of PMMA invasion of the spinal canal with related neurological deficit. The latter may require urgent laminectomy and evacuation of the extruded cement to prevent permanent sequelae.

According to the Society of Interventional Radiology (SIR), the major complication rate is <1% and reaches about 5% in tumor cases.²⁴ Cement leakage is documented in 30%–72.5% of cases on radiographs and in 87.9%–93% of cases on CT.²⁵ PMMA can flow outside the vertebral body posteriorly into the spinal canal and neural foramina. This is usually due to destruction of the vertebral body posterior cortex and of the medial/inferior cortex of the pedicle, but it rarely necessitates urgent surgical decompression. Leakage can occur into the paravertebral veins, mostly without clinical consequences.²⁶ Asymptomatic extension into the inferior vena cava may also occur²⁷ but involves possible systemic complications such as pulmonary embolism and paradoxical cerebral arterial PMMA emboli.¹⁸ Asymptomatic pulmonary embolization may occur in up to 4%–6.8% of patients.²⁸

Local metastasis occurring in the needle track after PV has also been reported.²⁹ An increase in local pain and fever may occur following the procedure but usually resolves within 72 hours.¹⁸

PK complications are mostly related to incorrect placement of hardware or cement extravasation and may result in neurologic insult.³⁰ The incidence of cement leakage during PK is in the range 8.6%–33%. The reasons for a lower cement extravasation percentage with PK include (1) balloon tamping of a cavity that is surrounded by a shell of impacted cancellous bone, (2) the ability to determine the amount of cement to be injected thanks to the knowledge of the volume of fluid used to inflate the balloon, and (3) lower injection pressure during cement injection.

Infective complications are overall rare. However, meticulous attention to a sterile technique is warranted and should include preoperative intravenous antibiotic administration.

Adjacent-Level Fractures

An increased risk of adjacent-level fracture following PV has been noticed with long-term follow-up. The reported rate of new fractures varies from 7% to 20% within 1 year of follow-up and affects mainly the immediately adjacent levels.³¹ In a recent meta-analysis of the literature, the authors found that the risk of developing a new fracture after kyphoplasty is 14.1%, which is similar to the 17.9% estimated following vertebroplasty.

A local, unfavorable biomechanical situation seems to be included in those patients who suffer adjacent-level fracturing, while the nonadjacent fracture group is more likely to be related to an ongoing disease process such as osteoporosis. Thus, risk factors include osteoporosis, previous vertebral fracture, and organ transplant, which expose the patient to a high osteoporosis risk.

Both Grados et al¹⁹ and Lin et al³² postulated that intradiscal leakage of PMMA may correlate with mechanical consequences on adjacent vertebrae, particularly those with osteoporosis, with a consequent increase in fracture risk. Recent papers,¹⁹ however, reported an incidence of adjacent vertebral fracture (1 year or less) after PV equivalent to the one expected in untreated osteoporotic vertebral fractures.

Neither volume of cement injected nor extravasation of cement into the intravertebral disc seems to increase the likelihood of subsequent adjacent vertebral fracture. A retrospective study showed that a targeted exercise program (Rehabilitation of Osteoporosis Program–Exercise [ROPE] incorporating isometric back-extensor muscle strengthening and proprioceptive postural retraining) after PV significantly decreases fracture recurrence; refracture rates were also lower in the rehabilitation group compared to the vertebroplasty-only group.³³

Newer biomaterials may help to prevent adjacent-level fracturing, thus diminishing the overall complication rates.³⁴ The development of new fractures subsequent to PK has also been reported.³⁵

Discussion

The treatment of painful vertebral metastases remains a major challenge in patients under palliative care. While radiotherapy has represented the gold standard for many years, there is now increasing evidence supporting the combination of surgery with radiation as more efficacious.³⁶ Numerous retrospective studies validate the efficacy and safety of PV. However, no study has investigated the therapeutic effect and safety of this technique in metastatic fractures. The only phase I/II clinical study of PV as a palliative for painful malignant vertebral compression fractures³⁷ confirms the procedure efficacy without reporting any severe complication.

The “painkiller” effect is clearly superior and more rapid when compared to radiotherapy. This becomes particularly useful in those patients with poor prognosis. Since this therapy is not designed to exert an antitumor effect but rather to provide pain relief by strengthening weakened vertebrae, pain recurrence is unavoidable if the metastatic foci expand.

An evidence-based assessment of PV has shown that the costs of this procedure are relatively low if compared with open surgical interventions for vertebral compression fractures and still inferior to the conservative treatment, which may consist of prolonged bed rest, analgesic drugs, orthotic devices, and eventual complications due to immobilization.

Maximal attention to indications and contraindications is pivotal in this procedure in which the percutaneous approach would otherwise represent a great disadvantage in case of major complications ([Table 1] and [Table 2]).

Recent reviews and editorials have called for a more critical evaluation of these procedures.³⁸ Controlled, multicentered trials are needed to prove its short- and long-term safety, efficacy, and cost–effectiveness in the treatment of metastatic vertebral fractures.

In addition, the procedures have reached widespread popularity, making it difficult to gain consent from patients to be included in the conservative treatment group of a randomized study.

Studies also need to be done to compare PV and PK in various disease states in a randomized fashion. Although PV does not restore vertebral body height, the procedure may be more appropriate than PK in high-risk patients or when the indication is solely for pain control.

The superiority of one technique over the other remains controversial. In spite of the necessity for more vigorous research, percutaneous augmentation with both procedures is promising in the treatment of painful vertebral fractures due to malignant infiltration.

With careful selection, adequate training, and meticulous attention to detail during the procedure, devastating complications can be dramatically reduced. More recently, the combination of PK with radiosurgery has proven to be effective in treating pathological fractures and in reducing the risk of fracture progression related to radiosurgery.³⁹

There is growing interest in the use of less invasive techniques for the treatment of spinal metastases. One example is percutaneous radiofrequency ablation coupled with PMMA injection for debulking tumor and vertebral stabilization.⁴⁰

Another novel technique involves the combination of percutaneous tumor debulking, PK, and intravertebral administration of a mixture of radiolabeled ¹⁵³Sm-EDTPM and PMMA.⁴¹

Conclusions

PV and PK are minimally invasive techniques used to treat painful vertebral compression fractures. There is a growing body of evidence indicating that these procedures are efficacious in alleviating pain associated with vertebral compression fractures. However, they need to be performed by a very skilled operator to prevent those rare but eventually serious complications. Because of this, recent reviews and editorials have called for a more critical evaluation of these procedures.

Indications

PV is an interventional procedure in which bone cement, usually polymethyl methacrylate (PMMA), is injected under radiologic guidance into the collapsed vertebral body. This procedure was first reported by Galibert in 1987⁵ for the treatment of painful aggressive vertebral hemangiomas and subsequently used in the United States in 1993.⁶ Since then, the technique has been widely applied and has evolved as the treatment of choice for painful osteoporotic fractures. The second most common use of PV and PMMA is for the treatment of painful vertebral fractures caused by metastatic disease or multiple myeloma.

PV also works well in those metastatic cases in which intractable pain is not associated with vertebral body collapse.⁷

Further reported indications are spinal pseudoarthrosis, intravertebral vacuum phenomenon, Langerhans cell histiocytosis, osteogenesis imperfecta, and Paget disease.⁸

Recent studies show that PV should be performed before radiotherapy for the treatment of those spinal metastases that have led to pathologic compression without neurological deficits.

In selected cases, PV is a minimally invasive alternative to open surgery.

PK is an evolution/modification of PV, combining the analgesic effect of PV and the possibility of restoring normal height of the collapsed vertebral body. This improves the kyphotic deformity frequently associated with vertebral fractures of the thoracic segment. Introduced in 2001, it employs a height-restoration device, known as “balloon tamp,” that when inflated restores the original shape of the vertebral body while creating a cavity in which to inject the cement.

The ideal candidates for this augmentation procedure are those patients who complain of midline nonradiating back pain that increases with weight bearing and is exacerbated by manual palpation of the spinous process of the involved vertebrae. This typical symptomatology subsides with recumbency and/or sitting. Exceptions are patients with thoracic spine fractures in whom pain radiates to the ribs or those with fractures at the level of the conus medullaris, where pain may radiate to the hips without evidence of cord compression.⁹

Detailed anamnesis, accurate neurologic examination, and recent radiographic imaging are mandatory to exclude spinal cord compromise and/or retropulsed bony fragments in the canal, which obviously represent a contraindication to the procedure.

Indications of both procedures have been defined in the guidelines published by the Society of Interventional Radiology in 2003 and recently updated by the Cardiovascular and Interventional Radiological Society of Europe.¹⁰ Indications include painful osteoporotic vertebral fractures after 3 weeks of analgesic therapy, painful vertebrae due to benign or malignant primary or secondary bone tumors, painful vertebral fractures with osteonecrosis (Kümmell disease), reinforcement of the vertebral body prior to a surgical procedure, and chronic traumatic vertebral fractures with non-union. Absolute contraindications include asymptomatic vertebral fractures, pain improving with medical therapy, ongoing infection, osteoporotic patient prophylaxis, uncorrectable coagulopathy, myelopathy due to retropulsion of bone (canal compromise), and an allergic reaction to PMMA or the opacification agent. Relative contraindications are radicular pain, vertebral fractures >70% of height loss, severe spinal stenosis, asymptomatic retropulsion of bony fragment, tumor extension into canal/epidural space, and lack of surgical backup.

Although disruption of the posterior cortex of the vertebral body has been considered as a relative contraindication, new techniques allow efficacious PV in these circumstances.

Multilevel procedures (3–4 levels) should be avoided in patients with low cardiopulmonary reserve (ie, chronic obstructive airway disease or congestive cardiac failure) as these patients may be at high risk for symptomatic pulmonary fat embolism.¹¹

Informed consent should include discussion regarding failure to obtain pain relief as well as complications. The eventuality of open stabilization or urgent decompressive surgery should also be discussed with the patient.

Radiology

Traditional anteroposterior and laterolateral X-rays show the degree of vertebral compression, eventual osteolysis, extent of pedicle involvement, and fracture or cortical destruction.

Computed tomography (CT) may be useful to further define the extent of vertebral collapse, the location and extent of any lytic process, the visibility and degree of involvement of the pedicles, the presence of cortical destruction and epidural or foraminal stenosis caused by tumor extension or bone fragment displacement, and to estimate the needle path and size.

Magnetic resonance imaging (MRI) is pivotal in triaging patients before PV or PK. In fact, signal changes within the vertebral body marrow suggest edematous changes in a healing fracture, which are a necessary condition for the procedure. Sequences that are particularly sensitive to the presence of edema are fat-suppressed T2-weighted and fat-suppression inversion recovery images, which may also be useful in planning the vertebroplastic procedure.¹¹

Concomitant limitations of these specific MRI sequences are that they may reveal increased activity up to 2 years after fracture, whereas it is widely accepted that patients with fractures older than 6 months do not benefit from vertebroplasty.¹²

Bone nuclear scanning has also been used to identify recent fractures in patients with multiple involved levels or in patients in whom MRI is not possible because of a pacemaker or stent. The recent fracture is typically heralded by an intense radioisotope uptake.¹³

Radiological investigation is a fundamentally delicate tool focusing on the most recent symptomatic fractured level. It helps solve those unclear cases in which physical examination is not sufficient to determine which of several adjacent fractures is symptomatic or which fracture has no pain on palpation over the involved vertebra and/or there is no correlation between the affected site and the pain localization.

Techniques

PV and PK require a detailed knowledge of spine anatomy and an intensive training in fluoroscopic imaging interventional procedures. The procedure should be performed in an appropriately sterile area. Broad-spectrum antibiotics can be administered just before the treatment. An anesthesiologist or other physician able to undertake rate and pulse oximetry must be present continuously. The typology of anesthesia should be selected on an individual basis. A generous amount of local anesthetic, especially into the periosteum, is suggested to maintain communication with the patient. Emergency measures should be available in the operating theater. In selected cases, general anesthesia is used.

Good-quality imaging biplanar or C-arm fluoroscopy with a radiolucent table is mandatory for maximal procedural safety, to correctly identify the anatomical structures (eg, pedicle, posterior wall).

There are various types of cement (methyl methacrylate powder) currently in use. They basically differ in their polymerization times, and the practitioner performing the procedure should be very skilled in this specific aspect. Cement opacification with barium sulfate specifically designed for use in vertebroplasty is required.

Most operators aim to obtain sufficient cement placement into the vertebral body using a monopedicular approach, thus reducing the procedure time.

PMMA injection into the vertebral body is performed when the trochar (uni- or bipedicular approach) has been deepened into the ventral portion of the vertebral body.

The cement should harden to “toothpaste” consistency before injecting. The injection should be stopped once the cement spreads to the posterior third of the vertebral body.

The cement column should ideally spread among the superior and inferior endplates and between the two pedicles. Such cement filling should prevent the collapse of the treated vertebral body.

If PMMA diffuses into a blood vessel or toward the posterior cortical margin, injection must be immediately stopped. Use of high-viscosity cement and small-volume injection is recommended in order to minimize the risk of PMMA leakage.

As Belkoff et al¹⁴ reported, maximal filling of the compressed vertebral body is not necessary; 2 mL of cement are sufficient for restoring vertebral body strength. Inadequate cement in the unstable fractured area may be responsible for unrelieved pain.

PK (Figure 1) differs from PV (Figure 2) in that it involves the percutaneous placement of balloons (called “tamps”) into the vertebral body with an inflation/deflation sequence that creates a cavity before the cement injection. PK may restore the vertebral body height and reduce the kyphotic angulation of the compression fracture before PMMA injection.

Proposed Mechanisms of Pain Relief

The most accepted theory indicates augmentation with cement increases the fracture mechanical load threshold, stabilizing the vertebra.¹⁵ In tumor fractures, pain relief is related primarily to vertebral body stabilization and secondarily to the induction of tumor necrosis and the destruction of sensitive nerve endings.¹⁶ The last two effects seem to be directly connected to the local heat produced by the highly exothermic reaction of the PMMA polymerization. The antitumoral effect is supported by the local cytotoxic effect of PMMA on rapidly proliferating cells.¹⁷

The heat effect is related to the degree and duration of the heat-exposure period. However, this does not explain why the analgesic effect is not proportionally related to the volume of PMMA injected. Low-volume injection has been found to be as effective as high-volume injection in relieving pain. On these bases, although a unilateral approach may allow a satisfying filling of the vertebra in terms of stabilization, it might not be sufficient when an additional antitumoral effect is desired.

Interestingly, despite the unequivocal effect on pain, the literature remains unclear about the reliability of vertebroplasty in achieving bone stabilization and preventing future vertebral fracturing.²

Results

Pain relief and increased mobility are expected within 24 hours following PV.¹⁸ Significant pain relief is expected in >70% of patients with vertebral malignancies, in >90% of patients with osteoporotic fractures, and in about 80% of patients with hemangioma. Pain relief usually persists over months to years.¹⁹

The significant decrease in back pain resulting from PV and PK in patients with pathologic fractures dramatically improves their quality of life.²⁰ In particular, the procedure-related benefits include reduction/withdrawal of analgesic drugs and improvement in physical mobility. (It is well known that the lower the intake of analgesics, the higher the perception of quality of life, thanks to the disappearance of drowsiness and nausea resulting in an improved appetite.) PK has not been shown to be better than PV in terms of pain relief or quality of life. Eck and colleagues²¹ found in a recent meta-analysis that although both methods provide significant improvement in visual analogue scale scores, there is a statistically greater improvement in pain relief with PV than with PK. In their literature review, Cloft and Jensen²² concluded that there was no proven advantage of PK over PV with regard to pain relief, vertebral height restoration, and complication rate. Moreover, Mathis²³ found that the claims of superior height restoration by PK are insufficiently documented.

PK is estimated to be 2.5–7.0 times more expensive than PV due to additional equipment, general anesthesia, and hospital costs.²² Based on 2006 available data, Mathis²³ found no substantial scientific, procedural, or economic advantages supporting PK's superiority over PV.

Complications

Complications are rare but can be dramatic. Minor complications such as PMMA extension into the disc require no therapy and have no clinical consequence. Major complications may consist of PMMA invasion of the spinal canal with related neurological deficit. The latter may require urgent laminectomy and evacuation of the extruded cement to prevent permanent sequelae.

According to the Society of Interventional Radiology (SIR), the major complication rate is <1% and reaches about 5% in tumor cases.²⁴ Cement leakage is documented in 30%–72.5% of cases on radiographs and in 87.9%–93% of cases on CT.²⁵ PMMA can flow outside the vertebral body posteriorly into the spinal canal and neural foramina. This is usually due to destruction of the vertebral body posterior cortex and of the medial/inferior cortex of the pedicle, but it rarely necessitates urgent surgical decompression. Leakage can occur into the paravertebral veins, mostly without clinical consequences.²⁶ Asymptomatic extension into the inferior vena cava may also occur²⁷ but involves possible systemic complications such as pulmonary embolism and paradoxical cerebral arterial PMMA emboli.¹⁸ Asymptomatic pulmonary embolization may occur in up to 4%–6.8% of patients.²⁸

Local metastasis occurring in the needle track after PV has also been reported.²⁹ An increase in local pain and fever may occur following the procedure but usually resolves within 72 hours.¹⁸

PK complications are mostly related to incorrect placement of hardware or cement extravasation and may result in neurologic insult.³⁰ The incidence of cement leakage during PK is in the range 8.6%–33%. The reasons for a lower cement extravasation percentage with PK include (1) balloon tamping of a cavity that is surrounded by a shell of impacted cancellous bone, (2) the ability to determine the amount of cement to be injected thanks to the knowledge of the volume of fluid used to inflate the balloon, and (3) lower injection pressure during cement injection.

Infective complications are overall rare. However, meticulous attention to a sterile technique is warranted and should include preoperative intravenous antibiotic administration.

Adjacent-Level Fractures

An increased risk of adjacent-level fracture following PV has been noticed with long-term follow-up. The reported rate of new fractures varies from 7% to 20% within 1 year of follow-up and affects mainly the immediately adjacent levels.³¹ In a recent meta-analysis of the literature, the authors found that the risk of developing a new fracture after kyphoplasty is 14.1%, which is similar to the 17.9% estimated following vertebroplasty.

A local, unfavorable biomechanical situation seems to be included in those patients who suffer adjacent-level fracturing, while the nonadjacent fracture group is more likely to be related to an ongoing disease process such as osteoporosis. Thus, risk factors include osteoporosis, previous vertebral fracture, and organ transplant, which expose the patient to a high osteoporosis risk.

Both Grados et al¹⁹ and Lin et al³² postulated that intradiscal leakage of PMMA may correlate with mechanical consequences on adjacent vertebrae, particularly those with osteoporosis, with a consequent increase in fracture risk. Recent papers,¹⁹ however, reported an incidence of adjacent vertebral fracture (1 year or less) after PV equivalent to the one expected in untreated osteoporotic vertebral fractures.

Neither volume of cement injected nor extravasation of cement into the intravertebral disc seems to increase the likelihood of subsequent adjacent vertebral fracture. A retrospective study showed that a targeted exercise program (Rehabilitation of Osteoporosis Program–Exercise [ROPE] incorporating isometric back-extensor muscle strengthening and proprioceptive postural retraining) after PV significantly decreases fracture recurrence; refracture rates were also lower in the rehabilitation group compared to the vertebroplasty-only group.³³

Newer biomaterials may help to prevent adjacent-level fracturing, thus diminishing the overall complication rates.³⁴ The development of new fractures subsequent to PK has also been reported.³⁵

Discussion

The treatment of painful vertebral metastases remains a major challenge in patients under palliative care. While radiotherapy has represented the gold standard for many years, there is now increasing evidence supporting the combination of surgery with radiation as more efficacious.³⁶ Numerous retrospective studies validate the efficacy and safety of PV. However, no study has investigated the therapeutic effect and safety of this technique in metastatic fractures. The only phase I/II clinical study of PV as a palliative for painful malignant vertebral compression fractures³⁷ confirms the procedure efficacy without reporting any severe complication.

The “painkiller” effect is clearly superior and more rapid when compared to radiotherapy. This becomes particularly useful in those patients with poor prognosis. Since this therapy is not designed to exert an antitumor effect but rather to provide pain relief by strengthening weakened vertebrae, pain recurrence is unavoidable if the metastatic foci expand.

An evidence-based assessment of PV has shown that the costs of this procedure are relatively low if compared with open surgical interventions for vertebral compression fractures and still inferior to the conservative treatment, which may consist of prolonged bed rest, analgesic drugs, orthotic devices, and eventual complications due to immobilization.

Maximal attention to indications and contraindications is pivotal in this procedure in which the percutaneous approach would otherwise represent a great disadvantage in case of major complications ([Table 1] and [Table 2]).

Recent reviews and editorials have called for a more critical evaluation of these procedures.³⁸ Controlled, multicentered trials are needed to prove its short- and long-term safety, efficacy, and cost–effectiveness in the treatment of metastatic vertebral fractures.

In addition, the procedures have reached widespread popularity, making it difficult to gain consent from patients to be included in the conservative treatment group of a randomized study.

Studies also need to be done to compare PV and PK in various disease states in a randomized fashion. Although PV does not restore vertebral body height, the procedure may be more appropriate than PK in high-risk patients or when the indication is solely for pain control.

The superiority of one technique over the other remains controversial. In spite of the necessity for more vigorous research, percutaneous augmentation with both procedures is promising in the treatment of painful vertebral fractures due to malignant infiltration.

With careful selection, adequate training, and meticulous attention to detail during the procedure, devastating complications can be dramatically reduced. More recently, the combination of PK with radiosurgery has proven to be effective in treating pathological fractures and in reducing the risk of fracture progression related to radiosurgery.³⁹

There is growing interest in the use of less invasive techniques for the treatment of spinal metastases. One example is percutaneous radiofrequency ablation coupled with PMMA injection for debulking tumor and vertebral stabilization.⁴⁰

Another novel technique involves the combination of percutaneous tumor debulking, PK, and intravertebral administration of a mixture of radiolabeled ¹⁵³Sm-EDTPM and PMMA.⁴¹

Conclusions

PV and PK are minimally invasive techniques used to treat painful vertebral compression fractures. There is a growing body of evidence indicating that these procedures are efficacious in alleviating pain associated with vertebral compression fractures. However, they need to be performed by a very skilled operator to prevent those rare but eventually serious complications. Because of this, recent reviews and editorials have called for a more critical evaluation of these procedures.