Vertebral Augmentation

Kyphoplasty and vertebroplasty currently are used to treat neurologically intact patients suffering from vertebral compression fractures resulting from osteoporosis and certain neoplasms. Percutaneous vertebroplasty (PV) is an imaging-guided procedure that reinforces a compromised vertebra with polymethylmethacrylate (PMMA), alleviating pain and improving the patient’s mobility. Initially described in 1987 as a treatment for painful hemangiomas,¹ the procedure now is used most widely to treat fractures and destructive lesions that cause vertebral collapse and pain. The goal of vertebral augmentation procedures is to reduce fracture pain by stabilizing the vertebral fracture, allowing patients to return to their daily living activities and exercise.

The primary indications for PV are persistently painful compression fractures, most commonly related to osteoporosis, that are unresponsive to correct medical treatment, and benign or malignant osteolytic neoplasms (e.g., hemangioma, metastasis, and myeloma). In the setting of osteoporotic fracture, PV is done primarily for pain management and secondarily to prevent further collapse.² For neoplasm, the indications are somewhat broader, and the procedure may be done for pain management and/or stabilization. PV is not an ablative procedure, but reinforcement may reduce pain while facilitating other therapies such as radiation therapy or surgical resection and fixation and minimizing the risk of further collapse or fracture.

The specific indications for PV are as follows:

The indications have been expanded to include nonunions of chronic traumatic fractures, which typically show a pseudarthrosis with a fluid-filled cleft or cystic findings on MRI evaluation.

As with any spine operation, precise patient selection is essential to the therapeutic success of PV. Patients with prior unsuccessful spinal surgery generally do not respond as well to PV.

The absolute contraindications to PV for either osteoporotic or neoplastic conditions are as follows:

Relative contraindications include the following:

Other, less common, relative contraindications are (1) retropulsion of fracture fragment(s) causing substantial spinal canal compromise (e.g., a burst-type fracture); (2) neoplasm extending into the epidural space with substantial spinal canal compromise; (3) severe vertebral collapse such that it would be technically challenging to place a needle; (4) chronic stable fracture without pain; and (5) treatment of more than three levels at any one time. Acute traumatic fracture of a nonosteoporotic vertebra also is considered a contraindication.

Patient Selection

Preoperative evaluation entails imaging to demonstrate the level of fracture (by plain radiography), to confirm an “active” fracture by MRI (demonstrating bone marrow edema) or nuclear medicine scintigram (“bone scan”), and excluding unsuspected malignancy or degenerative disease (e.g., herniated discs). This usually entails obtaining recent MRI examinations, within 7 to 10 days of consultation.

Plain radiography is useful for showing the vertebral level involved and to determine the degree of collapse. With complete or near complete vertebral collapse, PV usually is unlikely to be successful.

A specific protocol is employed with sagittal T1-weighted conventional spin echo (CSE), sagittal T2-weighted fast spin echo (FSE), or turbo spin echo (TSE) with fat saturation and axial T2-weighted FSE through areas of abnormality. If fat suppression is not available for the T2-weighted images, then a short tau inversion recovery (STIR) sequence should be substituted. The fluid-sensitive sequences (T2 with fat suppression or STIR images) are necessary to identify an “active” compression fracture. While the causes of bone marrow edema are legion, in the context of compression fractures and appropriate clinical symptoms, it often indicates a persistent pain generator. The age of the compression fracture is less important than the presence of bone marrow edema on MRI. This becomes relevant for people who have had symptoms for months or years due to unhealed fractures.

Concordant pain evaluation also is a critical aspect for determining whether patients with benign compression fractures will respond to PV. Under fluoroscopy, the back may be percussed and palpated over the spinous processes, marking the areas of tenderness relative to the recognized compression fracture. The patient also may assume any position that triggers the pain while the physician marks the painful site. If the pain corresponds to the fracture level (usually within one level above or below), then the symptoms are likely to be amenable to PV at that vertebral level. If pain is not concordant (for instance, lumbosacral pain associated with a thoracolumbar fracture), PV is much less likely to help.

If the patient is a candidate for PV, a further workup is required. Patients taking warfarin must have it discontinued prior to the procedure and potentially be switched to heparin for the periprocedural period. Patients with a fever, leukocytosis, or an elevated erythrocyte sedimentation rate must wait until the cause is diagnosed and treated (presumably an infection). For patients taking steroids, it is necessary to work closely with the primary care provider to manage the dose during the periprocedural period.

The informed consent process usually is a comprehensive discussion of the risks, benefits, and alternatives to PV. Risks include bleeding, infection, allergic reaction to the PMMA, fracture (pedicle or rib), and PMMA extravasation. Pneumothorax also is a potential risk for thoracic levels. Informed consent should also explain the goal of treatment (e.g., pain management, prevention of further deformity, stabilization for radiation therapy), theoretical background (e.g., stabilization by internal “casting”), technical aspects (e.g., needle procedure versus open surgery), and clinical data regarding PV. This often includes answering questions regarding the nature of osteoporosis and the status of PV as “standard of care” versus an “experimental procedure,” including the off-label use of the cement (PMMA) and the modification of PMMA by additional sterile barium to increase opacification (thus allowing early detection of extravasation). The possible need for emergent spinal decompression should be addressed with the patient.

Anesthesia for PV may be accomplished with neuroleptic anesthesia (i.e., conscious sedation), with the advantage that the patient is able to report any symptoms of a complication immediately, allowing the operator to halt the injection and minimize further injury. It also facilitates PV as a same-day outpatient procedure with a quicker recovery period. For these reasons, intravenous CS (IVCS) is considered the preferred mode of sedation for PV. Intravenous antibiotic coverage is recommended using cefazolin, or vancomycin if the patient has penicillin allergy, as infection prophylaxis.

Technical Performance of Vertebroplasty

The traditional mode for PV has been to perform bilateral injections via a transpedicular or dorsolateral (parapedicular) approach using a two-needle technique (Figs. 220-1 to 220-5). Unipedicular injections also can provide substantial vertebral reinforcement when enough cement is injected to cross the midline of the vertebral body and provide a similar distribution to bipedicular injections.³

FIGURE 220-1 Parasagittal approach. Anteroposterior projection fluoroscopic images show the typical position for the parasagittal transpedicular puncture. A, Note that the initial puncture site is with the needle tip located in the superolateral corner of the ring shadow of the pedicle (arrowhead). B, The needle has been advanced to the ventral one third of the vertebral body, and the tip is somewhat centrally located but still ipsilateral. Note the intradiscal cement from injection of the level above (arrow).

FIGURE 220-2 “Scotty dog” approach. Oblique projection fluoroscopic image shows the typical position for transpedicular puncture when using the “Scotty dog” approach. Note that the target is the ring shadow of the pedicle in an approximately 45-degree obliquity (arrowhead points to the pedicle below the punctured level for reference). This trajectory places the needle tip centrally and facilitates filling of both sides from a unilateral injection.

FIGURE 220-3 Needle-tip position. Sagittal projection fluoroscopic image shows appropriate needle-tip position within the ventral one quarter to one third of the vertebral body. Note the small amount of cement injected that has begun to fill the fracture line (arrow).

FIGURE 220-4 Cement extravasation. Sagittal projection fluoroscopic image shows a small amount of cement that has extravasated ventrally (arrow) outside of the vertebral body into the retroperitoneum. Extravasation in this location usually is not clinically significant.

FIGURE 220-5 Cement distribution. The axial CT image shows hyperattenuating cement in the ventral and mid-portions of the vertebral body. Note that the dorsal vertebral margin is concave (arrowhead). This forms the dorsal vertebral line on lateral radiography or fluoroscopy and also is the location where the basivertebral venous plexus perforates the cortex (not shown). Therefore, cement should not be injected in an attempt to fill the entire dorsal aspect of the vertebral body under fluoroscopy.

The modalities used for imaging are either fluoroscopy or CT. PV can be accomplished with either single-plane or biplane fluoroscopy. Biplane fluoroscopy is considered ideal because one can view orthogonal projections without changing tube position, greatly reducing procedure time. CT requires repositioning the gantry between needle manipulations, imaging, and injection, consuming useful injection. CT fluoroscopy allows the user to generate real-time CT sections while performing procedures, but concerns with CT fluoroscopy revolve around the increased radiation dose delivered to personnel and the operator in the room during the procedure. For most purposes, C-arm fluoroscopy provides adequate imaging, is readily available, and is less time-consuming.

The needles used are standard bone marrow biopsy-type needles that have a Luer-lock connection. Either beveled or pointed tips can be used effectively, and which to use is based largely on operator preference and experience. The goal is to fill the focus of the fracture, which typically is ventral and rostral, and to avoid injection into the larger vascular sinusoids of the basivertebral plexus, which could allow extravasation into the epidural space. Initial cortical puncture and purchase should be obtained by hand, and then needle advancement may be by hand or with the assistance of a mallet. For lumbar levels, 10- to 11-gauge needles are used, but for the thoracic levels, some authors advocate using 13- to 14-gauge needles. The needle tips should be advanced to the ventral one fourth to one third of the vertebral body, approaching the midline.

The cement flow must be monitored carefully, and the injection should cease if there is resistance or if cement approaches the dorsal vertebral margin. If any extraosseous cement is identified, the injection should be terminated immediately. Up to three levels can be done in a single operation. The risk of doing more numerous levels simultaneously is the possibility that PMMA could cause fat emboli syndrome and respiratory compromise through marrow displacement.

Postprocedure, the patient remains flat in bed for at least 1 hour, followed by bedrest for an additional 1 to 3 hours with the head of bed elevated less than 45 degrees. If any new, severe pain or new neurologic symptoms develop, a CT scan should be done immediately to evaluate the distribution of cement.

Treating Tumors

Vertebroplasty may play an important role in palliation for the patient with vertebral metastases and in improving independence and function during more definitive systemic therapy. Tumors that are particularly radiosensitive are most appropriate to PV, because local control can be obtained as easily after stabilization as before. Tumors that require resection with a surgical margin, such as primary malignancies or locally aggressive benign tumors, are not appropriate for PV. Although there is some theoretical margin of tumor-kill associated with the thermal effect of the PMMA mass,⁸ the application of PV in the setting of neoplasm is not intended as an ablative procedure, and the goal is to provide structural support primarily, with the secondary goal of pain relief. The amount of PMMA used may be greater in an osteolytic spinal lesion than in a compression fracture, because the trend to minimize injectant volume in osteoporosis (i.e., filling the fracture line) does not apply in an erosive or destructive lesion.

Metastases are the lesions most commonly treated by PV, but large, symptomatic hemangiomas also may require ablation and stabilization with PMMA injection. Hemangiomas have a benign histology, but may grow aggressively and cause pain through either tissue distortion or pathologic fracture. PV stabilizes the vertebral body and obliterates the vascular sinusoids that make up the mass of the hemangioma. Subsequent surgery may then be focused on decompression, if needed.⁹ In this fashion, PV is another adjuvant treatment akin to intralesional sclerosis or embolotherapy preoperatively. PV also is effective in treating vertebral metastases that result in pain or instability, providing immediate and long-term pain relief.¹⁰ MRI is crucial for preoperative planning to evaluate the soft tissue extent of the tumor (e.g., spinal canal involvement, neuroforaminal encroachment, and status of the posterior longitudinal ligament).

The goal in PMMA application during tumor treatment is to obtain craniocaudal filling (attempting to get superior-to-inferior end-plate filling) that crosses the midline transversely, providing a vertical strut. The cement may fill part of the tumor, but this is not the goal. Metastatic lesions and plasma cell tumors are not cystic structures—any space filled with cement likely represents tumor tissue that has been displaced somewhere. There is a small risk that the neoplastic tissue may be displaced into the canal or neural foramen, causing root or cord compression. Adjuvant therapy (surgery, chemotherapy, or radiation therapy) will be required to control the tumor locally. For cases involving malignant neoplasm, there often is concern regarding vascular seeding of the displaced tumor as the PMMA infiltrates the tumor bed, displacing tumor into the vascular system. Although there is no literature to support or refute this concern, it is believed to be a nontrivial risk.

Complications

Fortunately, major complications are uncommon. The most common complication is radicular pain caused by migration of cement into the epidural venous plexus. In most patients, intradiscal and paravertebral leaks of cement have no clinical importance.¹¹ However, there have been case reports of severe neurologic complications, underscoring the need for appropriate safeguards as outlined previously.¹² Permanent paralysis has been reported, but is exceptionally uncommon if the procedure is performed in a controlled, image-guided fashion (by using a biplane real-time fluoroscopy suite). Rib, pedicle, or transverse process fractures also have been noted. One case has been reported of a pulmonary embolism caused by acrylic cement. This rare complication was believed to have occurred because perivertebral venous migration was not recognized.¹³ The anticipated complication rate is higher when treating neoplasms (10%) than for osteoporotic compression fractures (1% to 3%).

Outcomes

Taylor et al. performed a systematic review and meta-regression to compare the efficacy and safety of balloon kyphoplasty and vertebroplasty for the treatment of vertebral compression fractures, and to examine the prognostic factors that predict outcome. They found level III evidence to support balloon kyphoplasty and vertebroplasty as effective therapies in the management of patients with symptomatic osteoporotic vertebral compression fractures refractory to conventional medical therapy. Although there was a good ratio of benefit to harm for both procedures, balloon kyphoplasty appeared to offer the better adverse event profile.¹⁴