Introduction

Kyphoplasty, along with vertebroplasty and newer, related procedures, are forms of vertebral body augmentation (VBA). The procedures employ percutaneous injection of polymethylmethacrylate (PMMA) acrylic cement into a fractured vertebral body to restore strength and stiffness. Other indications, such as pathologic fractures from metastasis, are becoming more common. High energy, bursting, and extension fracture patterns should be avoided because of the increased PMMA extravasation risk.

In appropriately indicated patients, kyphoplasty yields excellent early pain relief and return to activity. Potential disadvantages of kyphoplasty include procedural risks, such as cement leakage and possible fracture of adjacent segment.

Brief Description

Spinal osteoporosis alone is asymptomatic. If allowed to progress, however, it confers increasing risk of fragility fracture. The principal manifestation of osteoporotic vertebral compression fractures (VCFs) is back pain. Some minimally symptomatic patients do not present for medical evaluation.¹ Others require hospital admission for unrelenting pain. Typically, over 3 months, the fracture heals and the back pain subsides.² Although the nonunion rate is low, not all VCFs heal.

Back pain can persist after fracture healing. From 33% to 75% of fractures precipitate chronic back pain.³ The chronic pain has been attributed to hyperkyphosis, leading to excessive muscular strain. Excessive anterior vertebral body loading engendered by this malalignment may propagate stress fractures in the surrounding endplates.⁴ Late kyphosis is occasionally associated with myelopathy.⁵

Indications and Contraindications

The goal of kyphoplasty is to interrupt the cycle of pain and functional decline associated with VCFs. Given the limited data comparing long-term impacts of kyphoplasty relative to nonoperative management, injecting all fractures cannot be justified. Because many patients improve quickly, most patients should try nonoperative management before considering kyphoplasty.

The duration of this nonoperative trial is inversely related to the patient’s pain level and functional limitations. Consider early intervention in patients unable to return to ambulation after a few days. Protracted bed rest may be riskier than procedural risks. At least 150,000 VCFs per year are refractory to nonoperative measures and require hospitalization, with bed rest and IV narcotics. Ambulatory patients should undergo 4 to 8 weeks of nonoperative care. In this group, treatment often includes limited contact thoracolumbar bracing, activity limitations, and sparing use of pain medications. For fractures of L2 and above, a CASH or Jewett brace is recommended. Low lumbar fractures may respond to a chairback brace. Fractures above T6 are more frequently related to metastasis than osteoporosis. Fractures less likely to improve with standard medical management include those with the following:

Over time, kyphoplasty indications have gradually been expanded to include conditions such as multiple myeloma and osteolytic metastases. Moreover, kyphoplasty has been added to open decompression and internal fixation procedures. Hybrid procedures may be indicated for more complex fracture patterns, significant compression of the neural elements, and neoplastic lesions with cortical destruction.⁷ Another hybrid option combines radiosurgery and kyphoplasty. Conventional radiotherapy remains the index treatment in many patients with vertebral body metastasis.⁷ Used alone, radiation is associated with delayed pain relief and further vertebral collapse due to both the previous bone erosion and the radiation itself. Newer radiation therapy techniques allow more focused radiation to be applied via intense treatments over a shorter time course.

Absolute contraindications to kyphoplasty include the following:

In this setting, “young” suggests patients younger than 65 years. The stronger the host bone, the less effectively polymethylmethacrylate restores stiffness. Patients with good bone stock fracture only after high energy loading. In this setting, PMMA leakage is more common. Calcium phosphate kyphoplasty (and other resorbable materials) is under study for this indication. Though isolated reports suggest pain improvement with kyphoplasty for sacral fractures, this indication is not widely accepted.

While less common than VCF, osteoporotic burst fractures (senile burst fractures) are not rare. Any fracture precipitating more than 50% height loss will have associated posterior cortical compromise. In many cases, this compromise takes the form of cortical buckling. When the canal occlusion is less than 33%, kyphoplasty can be considered. On the other hand, in the face of cortical comminution, avoid percutaneous kyphoplasty because of the increased risk of cement extravasation. In patients with neurologic injury, open surgery may be required.

Open surgery is indicated in patients with osteoporotic bones who also have progressive neurologic deficit. Unfortunately, in this frail population, operative intervention confers high risk. Similarly, spinal instrumentation systems often fail in osteoporotic bone. PMMA augmentation increases screw pull-out strength. Combination of kyphoplasty with open decompression restores anterior column load bearing and limits the scope of the reconstruction necessary.

Description of the Device

Kyphoplasty requires one or two high quality fluoroscopes and a kyphoplasty kit. The traditional set begins with a modified Jamshidi needle and a guide wire. Other systems remove the guide wire step (“express” and “one step”). Ultimately, each system is used to safely place two working cannulae through which KyphX balloon tamps can be inserted into the vertebral body. Smaller cannulae are available for upper thoracic vertebrae. The balloon tamps are modified angioplasty balloons. Currently, three sizes are available and are selected based on the size of the fractured vertebral body: 10, 15, and 20 mm. The balloons attach to a syringe with an integral pressure gauge. In the operating room, the balloons are prepped at the back table by instilling 10 ml of radiopaque contrast media. As volume is added to the balloon, the balloon pressure (measured in psi) increases. As the tamp displaces bone, the pressure gradually decays.

Kyphon (Sunnyvale, CA) manufactures several specific balloon products thought to assist in challenging clinical scenarios. For example, a bidirectional balloon (KyphX Elevate) emphasizes craniocaudal expansion and limits mediolateral enlargement. Another single-direction balloon (KyphX Exact) is deployed through a metal housing, which is thought to control balloon direction. These tamps confer additional cost to the procedure. There are no data demonstrating improved outcomes or decreased risk with these devices.

The system also includes bone void fillers, each of which holds 1.5 ml of PMMA. The bone void fillers are cannulae with plungers that allow gradual backfilling of the void created by the tamp. A modified bone void filler, the biopsy device, has sharper tips and can be deployed through the working cannula. PMMA with added barium to enhance fluoroscopic visibility and a mixing system are also available in a separate kit.

Background of Scientific Testing and Clinical Outcomes

The source of pain relief after kyphoplasty remains unclear. Currently, most authors suggest that restoration of strength and stiffness to the fractured vertebral body relieves pain. Both cement volume and percentage of the vertebral body filled can predict postaugmentation bone strength and stiffness. Overall, the more PMMA inserted, the higher the postinsertion vertebral strength and stiffness.

Outcomes data include a number of retrospective studies. Very recently prospective data have been reported from the FREE trial.¹⁰ This trial included 21 sites in 8 countries that enrolled 300 patients with acute VCF and randomized them to either kyphoplasty (149) or nonoperative care (151). As of this writing, the complete paper has not been published, but early pain relief seems to be a clear advantage of kyphoplasty. Whether that advantage persists is more difficult. The primary outcome was the difference in the Short Form (SF-36) physical component summary at 1 month. Quality of life measurements and spine radiographs were assessed through 12 months. Kyphoplasty subjects reported greater improvement than controls in their SF-36 physical component (5.2 point difference; p < 0001) at one month). By 12 months, the difference declined to 1.5 points and was no longer significant (p = .2). Kyphoplasty improved quality of life by the 1-point EuroQol questionnaire at 1 (0.18 points; 95% CI, 0.08–0.28; p < .001) and 12 (0.12; 95% CI, 0.01–0.22; p = .025) months. Back function, as measured by the 24-point Roland-Morris scale, was improved by 4.0 points by kyphoplasty at 1 month (p < .001) and 2.6 points at 12 months (p = .001). Kyphoplasty patients reported fewer days with limited activity, less back pain, and less use of analgesics and walking aids.

Of note, the FREE study was funded by the manufacturer and many of its authors are Kyphon consultants. On the other hand, three other small studies comparing kyphoplasty with conventional medical treatment also found that kyphoplasty consistently improved pain and physical function, with results sustained at 6 months.^11–13

In 2005, Hadjipavlou et al¹⁴ combined the available vertebroplasty and kyphoplasty outcome reports in an effort to compare the procedures. Using meta-regression techniques, the authors found that individual study design had a considerable impact on subsequent analysis. For prospective studies, the rates of success with vertebroplasty and kyphoplasty were not significantly different at 92% and 93% respectively. However, in retrospective studies, kyphoplasty was more successful (95% vs. 86%; p = .019)

Aside from pain relief, a major benefit of VBA lies in the restoration of mobility. In one series of 11 wheelchair-bound cancer patients, 73% were able to walk shortly after vertebroplasty.¹⁵ Other studies reported restoration of mobility after kyphoplasty in 84% to 100%.^8,16 In terms of other types of physical functioning, a number of different outcome measures have been used. In a retrospective analysis of patients with painful osteoporotic VCF, 49 patients who were available for follow-up at a mean 9-month interval had an improvement in visual analogue pain scale score of seven points (p < .05), and an improvement in Roland-Morris Disability Survey of 11 points (p < .05).¹⁷

In a retrospective analysis of 52 patients with 82 painful osteoporotic VCFs, kyphoplasty restored 4.6 mm and 3.9 mm to the heights of the anterior and medial columns, respectively.¹⁷ The mean Cobb angle increased by 14%. In a meta-analysis, Hadjipavlou et al concluded that, although postural reduction can improve vertebral height following a compression fracture, better reductions are obtained with kyphoplasty than with vertebroplasty.¹⁴ Better reductions may be achieved with earlier treatment.

Clinical Presentation and Evaluation

Successful kyphoplasty hinges on distinction of compression fracture pain from other etiologies. Clinical assessment involves an evaluation of the patient’s spinal alignment and gait, followed by palpation of the spine, ilium, sacrum, and paravertebral tissues. The importance of local tenderness over the involved spinous process as a principal sign of a painful VCF has been analyzed in two studies. In the first, which comprised 10 patients, Gaughen et al¹⁸ noted that local tenderness was not present despite imaging findings suggestive of an acute fracture. Recently, Gaitanis and coworkers¹⁶ found that spinous process tenderness corresponded to the level of pathology in 100% of osteolytic tumors and in 96% of VCFs when correlated with magnetic resonance imaging (MRI) findings of an acute fracture.

Several imaging techniques are employed in the evaluation of a painful VCF. Recently, flexion and extension or standing and supine lateral radiographs have been used to assess fracture mobility. A number of studies have examined the presence of intravertebral clefts. Although the exact cause of these intraosseous nitrogen pockets has been debated, the so-called Kummel sign may characterize pseudarthrosis. A cone-down lateral view directly perpendicular to the involved level is required in the assessment, because these clefts can easily be missed with standing lateral radiographs alone.¹⁹

Magnetic resonance imaging is an important technique for detection of osteoporotic compression fractures (Figure 34-1). It is more sensitive than plain radiography, with a reported accuracy of 96%.¹⁹ Fracture acuity (or failure of healing) is also best observed as intense signal on sagittal MRI with short tau inversion recovery (STIR) sequences (Figures 34-2 and 34-3).²⁰ For patients unable to undergo MRI, the combination of a technetium bone scan with computed tomography (CT) of the scintigraphically active levels can provide useful information on relatively fresh vertebral fractures (Figure 34-4).²¹

FIGURE 34-1 In this sagittal MRI, a patient has gradually increased collapse of superior endplate with stress injury to the pars and progressive kyphosis and translation. This patient complained of both a chin on chest deformity and progressive myelopathy. Kyphoplasty is not indicated in this case.

FIGURE 34-2 Sagittal T2-weighted or STIR MRIs are critical images in the evaluation of a patient with suspected painful osteoporotic vertebral compression fractures. In this STIR image of a patient with a lumbar transitional vertebral, multiple injuries are seen, especially acute L1 and L2 superior endplate injuries. The acute injuries demonstrate marked marrow edema diffusely. In the L1 lesion, a band of edema is seen from anterior to posterior along the fracture line. These bands often reflect “reducible” fractures.

FIGURE 34-3 The T1-weighted MRI gives better anatomic information than the STIR. Look for evidence of metastatic change such as soft tissue extension or extension of the marrow signal through the pedicle.

FIGURE 34-4 For patients unable to have an MRI, a bone scan can be helpful in identifying acute or subacute fractures. In this case, note the marked uptake at the T12 level.

There are patients in whom both MRI and CT imaging is useful. For those with questionable endplate erosion, the greater bone–soft tissue contrast of the CT scan often demonstrates erosions more clearly (Figure 34-5). Similarly, a fine-cut (2 mm) CT scan with sagittal reconstructions may demonstrate small lytic lesions not otherwise seen in the fractured vertebral body on MRI (Figure 34-6). Most commonly, however, the CT is ordered as an adjunct to MRI in patients with canal compromise from their fracture.

FIGURE 34-5 CT scans are useful for anatomic detail in patients unable to have an MRI. In patients with unusual fracture patterns or in those in whom cortical compromise or metastasis is suspected, order a CT scan for its excellent bone–soft tissue contrast. In this case of a prostate cancer metastasis, note the lytic lesion in the posterior aspect of the vertebral body with the destruction of the posterior cortex. This patient would not be a good candidate for percutaneous kyphoplasty, but mini-open or hybrid procedures could be considered if needed.

FIGURE 34-6 This axial CT image obtained in a patient noted to have a compression fracture without trauma was found to have both a hemangioma (on the right) and a lytic metastasis (on the left).

Operative Technique

Kyphoplasty procedures may be performed in the operating room or in the angiography suite. These procedures can be done under local anesthesia with intravenous sedation or under general anesthesia. There are advantages to both approaches. General anesthesia is associated with more comfortable prone positioning and less involuntary motion. On the other hand, rib fractures during positioning can occur.

Kyphoplasty patients are positioned prone on a radiolucent operating table or surgical frame. Lordotic positioning is maintained with bolsters. Lordosis allows a positional reduction. Later, when the balloons are removed, the lordotically positioned patient will be less likely to lose the reduction achieved. With this in mind, the radiolucent Wilson frame often makes lordosis difficult to achieve. A Jackson frame may allow better lordotic placement, but may be less comfortable for awake patients.

Kyphoplasty begins with true anteroposterior (AP) and lateral fluoroscopic images (Figures 34-7 and 34-8). Ensure a true AP with the spinous process in the midline between the pedicles. On the lateral view, the pedicles should line up and yield a clear view of the foramen and the posterior vertebral cortex. Both images should show the endplates of the level selected as a single line, not an oval.

FIGURE 34-7 These images exhibit craniocaudal and lateral intraoperative views in the operating room during a two-level kyphoplasty. In this case, a single fluoroscopy unit was used and positioning assessed in the AP (A) and lateral (B) planes. Four working cannulae have been placed. Through the cannulae are seen the inflatable balloon tamps attached to pressure syringes containing contrast medium. Serial inflation is undertaken gradually to effect reduction.

FIGURE 34-8 This series of fluoroscopic images demonstrates the kyphoplasty procedure beginning with a lateral scout image (A). Note that the pedicles line up so that the foramen can be seen clearly. In this case, a biopsy was obtained through the cannula (B). The bone void filler or the special biopsy needle can be used for this purpose. A syringe is attached to the needle and mild suction applied. An 8-gauge core is obtained. These cores may obviate open biopsy in cases in which Tru-Cut and Jamshidi biopsies were not diagnostic. In C and D, balloons have been deployed and an excellent reduction of the superior endplate is noted. On the AP view (D), note the medialization of the balloons and the alignment of the spinous process equidistant between the pedicles. In E, from another patient, the “air vertebrogram” left when the balloons have been removed is noted. Note the backfilling of the void with PMMA. In the final lateral view (F), excellent fill of the void is noted. Additional PMMA has been injected to fill the interstices around the void. Some of the reduction achieved with the balloons was lost, however.

When possible, biplanar fluoroscopy should be employed. This saves considerable time when switching from AP to lateral. If only one machine is available, mark the fluoroscope positions achieved, so they are easily re-achieved. Most typically, a transpedicular route to the vertebra is selected. In some thoracic cases, the narrow and straight pedicle precludes appropriate medialization and an extrapedicular approach is required. Most authors recommend a bilateral approach.

Beginning with AP fluoroscopy, an 11-gauge Jamshidi needle is placed at the 10 o’clock or 2 o’clock position on the pedicular ring. Unlike pedicle screws, the goal is not to proceed “straight down the barrel,” but rather to medialize through the cylinder of the pedicle. Therefore start at the lateral border and aim medially. Once in bone, verify your trajectory on the lateral image. If the AP and lateral images do not demonstrate a clearly intrapedicular position, an en face or oblique view is useful.

Under lateral fluoroscopic view, advance the Jamshidi to the midway point of the pedicle. Return to the AP view and verify tip position. Until the Jamshidi has passed through the posterior cortical margin of the vertebral body, it must be lateral to the medial pedicle wall on the AP image. If the needle has been medialized appropriately, return to lateral, and advance to 1 to 2 mm past the posterior vertebral body margin. Now the needle should be just barely across the medial pedicle border on the AP. Remove the Jamshidi stylet and place a guide pin.

The osteointroducer instruments are passed over the guide pin. The blunt dissector of the osteointroducer and guide pin are removed, leaving the working cannula in place just anterior to the posterior cortical margin of the vertebral body. Better medialization allows for more aggressive anterior placement. For harder bone, use the provided drill to prepare the path for the bone void filler. Live or pulsed fluoroscopy is recommended when approaching the anterior cortex.

Insert IBT to within 4 mm of the anterior cortex. Inflate the balloon to 50 psi (pounds per square inch) pressure to maintain its position and tamponade the bone. Place instruments through the opposite pedicle in similar fashion. Once the contralateral balloon has been placed, inflate both IBTs in 0.5-ml increments. Once inserted into the vertebral body, the balloons are gradually inflated using visual (radiographic), and volume and pressure controls (via a digital manometer), to reduce the fracture deformity.

Monitor AP, lateral, and oblique images for IBT position in relation to cortices. Sequentially inflate until the following inflation endpoint is reached:

A number of acrylic cements are available. Though the PMMA kits used with total joint arthroplasty can be employed, cement formulations specifically designed for vertebral augmentation may have better handling and setting characteristics. VBA cements also have extra sterile barium added to the polymer powder to increase its radiopacity.

With the balloons removed, bone filler devices are advanced into the distal portion of the cavity. Retrograde fill with PMMA is then undertaken using fluoroscopic monitoring. For kyphoplasty, the PMMA is placed into bone filler devices (BFDs). Then it is left in the device until it reaches a toothpaste consistency. Early implantation with runny PMMA increases leak risk. Operating room temperatures may affect PMMA polymerization times. Occasionally, warm saline solution is useful to accelerate setting of the PMMA.

Using the plunger, apply the PMMA under continuous fluoroscopy. Inject slightly more PMMA than final IBT inflation volume to allow intercalation of the material into surrounding trabeculae. The wound may be closed with a suture or Steri-Strip.

Postoperative Care

No braces or particular postoperative precautions are needed after kyphoplasty. That said, osteoporotic patients should be restricted in terms of heavy lifting and the carrying of heavy weight away from the body or above shoulder level. Osteoporotic patients should avoid concurrent bending and lifting. Ensure that the patient has been evaluated for and treated for their underlying osteoporosis.

Other postoperative care is fairly straightforward. Many patients will have remaining axial weakness. Consider physical therapy for patients who are weak or have ongoing muscular pain. Wound issues are typically minimal except for those patients taking blood thinners. Address nutritional issues when needed.

Complications and Avoidance

Kyphoplasty complications can be categorized: medical, anesthesia related, instrument placement, and PMMA problems. In most cases, failure to improve is due to inappropriate patient selection. The more diffuse the patient’s pain, the less likely they are to benefit from VBA. Placement of PMMA into the spine may increase the risk of adjacent segment fracture.

Kyphoplasty patients are, by definition, frail. Medical and anesthesia issues are not unusual in this elderly patient population. On the other hand, VBA procedures are not significantly physiologically taxing. When medical problems occur, they can be ascribed to the procedure itself or to preexisting cardiac and pulmonary problems. In markedly functionally limited patients, the risks of activity restriction in terms of deep vein thrombosis, pulmonary embolus (PE), and opiate-related complications are likely underreported and could be riskier than operative treatment.

Many patients in this age group take anticoagulant medications. When possible, reverse these agents before kyphoplasty. In particular, patients with multiple fractures, concomitant rib fractures, and osteoporotic bursting patterns are at higher risk for procedural and medical complications. Biopsies should be performed with kyphoplasty in patients with a history of cancer or an absence of concomitant trauma.

The most devastating technical complication of kyphoplasty arises from PMMA extravasation. Leakage is clinically silent in the vast majority of cases, with symptomatic leaks representing only a small portion of the total.²² PMMA may extravasate into the vascular tree, disc space, anterior and lateral soft tissues, and spinal canal. Extravasation is most common in metastatic osteolytic tumors or myeloma.¹⁵

Interestingly, leakage into the central canal is better tolerated in most cases than intraforaminal leak; however, when symptomatic, central canal extravasation leads to more devastating neurological symptoms, such as paraplegia. In most cases, symptoms are transient and respond well to nerve root blocks or oral medication; rarely do they require surgical decompression.²³

Along with the more viscous cement applied, void creation and bone compacting effects may decrease extravasation rates compared with vertebroplasty. A cadaveric study by Belkoff et al²⁴ reported reduced rates of PMMA extravasation after kyphoplasty compared with vertebroplasty. In a series of patients with metastatic disease, Fourney et al²⁵ reported a 9% extravasation rate after vertebroplasty, but no cases of extravasation following kyphoplasty.

Another serious complication of VBA procedures is postoperative infection. Simple wound infections can be identified and treated easily, but deep space infections including those of the cement mantle are serious and difficult to fully eradicate without removal of the cement bolus. Concurrent infection, even in distant organ systems, is a contraindication to kyphoplasty

Improper instrumentation placement most frequently stems from difficulty delineating the bony anatomy in patients in whom poor bone quality coexists with spinal deformity, such as degenerative scoliosis or marked spondylosis. Once the instruments are in place, care must be taken not to apply too much force, because leverage may lead to fractures. Pedicle and transverse process fractures may lead to postoperative pain, irritate local nerve roots, or destabilize the spine. Finally, these breaches create a path for inadvertent leakage of cement into the spinal canal. In one multicenter study, instrument placement problems led to postoperative hematoma in two patients, and a direct injury to the spinal cord when an extrapedicular approach was used on a vertebra with a fractured pedicle.²⁶

Methacrylate monomer is toxic. Some recommend that more than 30 ml PMMA be injected per session.²⁷ The more viscous the cement, the less likely it is that untoward blood pressure or blood gas effects will occur.¹⁴

Several VBA reports suggest an increased risk of secondary fractures adjacent to the augmented vertebra.^28,29 Two small studies suggest that kyphoplasty decreases adjacent fracture risk. Kasperk and colleagues^11,30 found that at 6-month follow-up, 30% (6 of 20) nonoperatively treated patients developed secondary fractures, whereas only 12.5% of 40 kyphoplasty patients had secondary fractures. Similarly, Komp et al¹² reported that 65% of 17 nonoperatively treated patients had new fractures, whereas only 37% of 19 kyphoplasty patients had additional fractures. In the FREE study, on the other hand, at 12 months, new vertebral fractures were slightly higher but not statistically significantly different between the kyphoplasty (41.8%) and nonsurgical (37.8%) groups (p = .5).¹⁰ The exact effects of VBA on adjacent levels probably vary with steroid exposure, spinal level, local spondylosis, and muscular factors; these require further study.

Conclusions and Discussion

Kyphoplasty has been widely available for roughly 10 years. During that time, its popularity as a percutaneous means of stabilizing osteoporotic compression fractures has skyrocketed. Vertebroplasty remains widely popular and highly successful as well. Use of balloon tamps appears to improve fracture reduction and decreases cement leakage rates. But leakage remains problematic in higher energy fractures and those associated with retropulsion. A number of newer procedures are evolving to compete with vertebroplasty and kyphoplasty.

Kyphoplasty should be considered in patients with persistent pain and functional limitation after fracture despite a trial of nonoperative management. Markedly limited and bedridden patients should be offered earlier treatment. In all cases, maximal management of the underlying osteoporosis should be pursued.

A number of controversies remain. Reports conflict as to the degree of reduction achieved and its impact on clinical outcomes. Kyphoplasty adds significant cost, radiation exposure, and operative time over vertebroplasty. So far, clear benefit has not been confirmed.

Though current data suggest that kyphoplasty patients experience marked early pain reduction and return to activity, its impact on long-term outcomes is unclear. As we study these fractures more closely, our ability to predict which fractures are likely to collapse and which may heal uneventfully with observation should improve.