Trauma

Lumbar spine trauma often requires spinal stabilization, particularly when two- or three-column injury is present. Multiple classification systems have been developed for spinal fractures, including those of Denis,¹ McAfee and colleagues,² Gertzbein and associates,³ and Magerl and coworkers.⁴ Most available systems suffer from one or more limitations, including a degree of complexity that limits routine clinical use, failure to include clinically relevant information such as the neurological status of the patient, and lack of guidance with regard to clinical management.⁵ Vaccaro and colleagues recently proposed a new classification of thoracolumbar injuries, the Thoracolumbar Injury Classification and Severity Score (TLICS), which attempts to overcome these limitations. TLICS focuses on three parameters, with points given for the various subcategories.⁵ First, the injury morphology is assessed as compression without burst (1 point), compression with burst (1 point), translational-rotational (3 points), or distraction (4 points). Second, the integrity of the posterior ligamentous complex (PLC) is assessed as intact (0 points), suspected or indeterminate injury (2 points), or clearly injured (3 points). Third, the neurological status of the patient is assessed as intact (0 points), nerve root injury (2 points), complete cord or conus injury (2 points), incomplete cord or conus injury (3 points), or cauda equina injury (3 points). A TLICS of 3 or less suggests a nonoperative injury, and a score of 5 or more suggests that an operative procedure should be considered. A TLICS of 4 suggests a situation that might be managed conservatively or operatively. In addition, based on neurological status and the integrity of the PLC, Vaccaro and colleagues provided suggested surgical approaches.⁵

Degenerative Disease

Indications for fusion in the setting of lumbar degenerative spine disease remain matters of ongoing debate and study. Resnick and coworkers recently provided a series of recommendations regarding this population based on the best available medical literature (Table 303-1).^6–22 In addition, lumbar instrumentation is typically necessary to maintain correction and stability in patients with degenerative scoliosis or iatrogenic flat back syndrome who fail conservative management and desire surgical treatment.^23–26

TABLE 303-1 Summary of Guidelines for the Performance of Fusion Procedures for Degenerative Disease of the Lumbar Spine

Data from references 6 to 22.

Dynamic stabilization devices are designed to maintain or restore normal spinal motion. Although the indications for these devices are still being clarified, many of the applications are related to degenerative disease.^27–32 Khoueir and associates recently provided a salient overview of the disorders in which these devices are expected to play a role: (1) controlling motion in the iatrogenically destabilized spine such as may occur with decompression procedures, (2) augmentation of interbody fusion with promotion of increased anterior load sharing, (3) protection or restoration of degenerated facet joints and intervertebral disks, (4) as part of a 360-degree circumferential segment reconstruction with motion preservation, (5) reduction of destructive forces at instrumentation-bone interfaces in patients with osteopenia or osteoporosis, and (6) reduction of fusion-related sequelae such as adjacent-level degeneration and pseudarthosis.³³

Infection

Vertebral osteomyelitis can frequently be treated with parenteral antibiotics. Failure of medical therapy may necessitate surgical débridement. Although no rigorously defined criteria exist, spinal stabilization may be necessary when more than 50% of the vertebral body has been eroded or if aggressive débridement would produce an unstable spine. Historically, débridement and stabilization were performed as two separate procedures to optimize infection clearance. However, several reports have demonstrated successful treatment with combined débridement and stabilization, with instrumentation including titanium cages and pedicle screw and rod systems, performed either through a combined anterior and posterior approach^34–37 or entirely through a posterior approach.^36,38 This combined approach has produced favorable results with regard to clearing infection, achieving ultimate spinal stability, and permitting earlier ambulation and rehabilitation. This approach has been similarly favorable for spinal tuberculosis.^39–43

Tumor

Surgical treatment of primary and metastatic spine tumors may be indicated to alleviate pain, provide stability, and decompress the neural elements.⁴⁴ The type, location, and degree of destruction of the tumor, as well as the overall treatment goals of surgery, determine whether the optimal approach is anterior, posterior, or combined anterior and posterior.⁴⁵ For metastatic disease, it must be remembered that the primary goal of surgery is typically palliation, and the goal should be maximization of quality of life with minimization of surgical complications.⁴⁶

Brief Overview of Historical Instrumentation

Modern attempts at spine instrumentation began in the 1940s with the Harrington system, which used a distraction-compression rod.⁴⁷ Although Harrington rods represented a significant advancement, their limitations soon became evident. Overdistraction frequently produced flat back syndrome, and neural compression by the laminar hooks and hook failure were not uncommon. In addition, the Harrington system lacked the ability to apply segmental corrective forces, although the addition of segmental wiring provided limited segmental correction.⁴⁷

The Luque rod system was developed as an alternative to the Harrington system.⁴⁸ Luque rods require fixation points above and below the affected segment and employ sublaminar wires to provide segmental fixation. The limitations of this system included neurological deficits associated with sublaminar wire passage or migration of the rods through a laminectomy defect.⁴⁹ In addition, this system provided very limited resistance to axial loads, limiting its utility in cases in which the anterior column was compromised.

The Cotrel-Dubousset (CD) system, developed in the late 1980s, consisted of multiple laminar hooks that were connected by a rod.⁵⁰ The CD system improved on the Harrington and Luque systems by requiring fixation of fewer segments above and below the diseased level. Pedicle screws were used in place of hooks at levels requiring laminectomy.

Metallic Pedicle Screw-Rod Systems

Boucher⁵¹ initially reported the use of pedicle screws for spinal fixation in the early 1950s, and Roy-Camille⁵² later popularized their use for lumbar fracture, pseudarthrosis, metastases, primary spine tumor, lumbosacral fusion, and spondylolisthesis. Pedicle screws offer considerable segmental control and enable fusion of fewer levels than was required with Harrington or Luque rods. Polyaxial screw heads have been introduced to facilitate connection of the screws to the rod or plate system.⁵³ Care must be used when placing pedicle screws in order to prevent injury to the neural structures and dura.

A variety of materials are used to construct pedicle screw-rod systems. Until the late 1990s, pedicle screw-rod systems were primarily composed of stainless steel. Given the considerable interference of stainless steel on computed tomography (CT) and magnetic resonance imaging (MRI) studies, titanium-alloy systems were developed. Titanium-alloy implants have been shown to be compatible with high-quality MR imaging and to result in a significant reduction of artifacts on CT imaging.^54,55 In addition, titanium-alloy implants have been shown to produce greater bone ongrowth compared with stainless steel implants, resulting in greater bone-pedicle screw fixation.⁵⁶ However, studies have suggested that the fatigue life of titanium spinal implants is inferior to that of steel, especially at notch sites resulting from rod contouring.^57,58 Cobalt-chromium rods were designed to offer improved mechanical strength compared with titanium-alloy, while still maintaining improved imaging characteristics relative to stainless steel. Significantly elevated levels of cobalt and chromium ions have been identified in the serum of patients following implantation of metal-on-metal Maverick-type artificial lumbar disks.⁵⁹ Whether elevated levels of these ions are encountered following implantation of cobalt-chromium rods and whether there are any long-term implications remain to be studied.

Polyetheretherketone Rods

Polyetheretherketone (PEEK), a synthetic semicrystalline thermoplastic polymer, has been used in cervical and lumbar interbody cages since the 1980s.^32,60 The modulus of elasticity of PEEK is between that of cortical and cancellous bone, which simulates the load characteristics of the spine. PEEK rods have recently been developed and provide a semirigid alternative to metallic-based rods. PEEK rods are used with a modified top-loading metallic screw (CD Horizon Legacy; Medtronic Sofamor Danek, Inc., Memphis, TN.) and allow limited motion while still providing resistance to marked flexion, extension, axial loading, and lateral rotation.³²

There are several potential applications for PEEK rods.³² First, they may be used in the treatment of spinal instability, such as spondylolisthesis or degenerative disk disease, to subject adjacent levels to less stress while still facilitating osseous fusion. Second, a PEEK rod and a domino connector may be used to treat adjacent-level disease following prior instrumented fusion. Third, a PEEK rod may be used to provide a tension band following decompression in a patient with spondylolisthesis in which arthrodesis is not the objective. Although there are several potential theoretical benefits of semirigid fixation with PEEK rods, supporting literature demonstrating these benefits remains very limited at this time.³²

Dynamic Stabilization

Posterior dynamic stabilization is one of the most rapidly evolving fields in spinal surgery.³³ As an alternative to fusion, dynamic stabilization devices are intended to maintain or restore intervertebral motion and simulate the behavior of the normal spine. Khoueir and colleagues recently described a classification system for posterior dynamic stabilization devices.³³ This system divides these devices into three categories: interspinous spacer devices, pedicle screw-rod–based devices, and total facet replacement systems.

Open Technique

The patient is first placed under general anesthesia, intubated, then positioned prone on a radiolucent surgical table, such as a Jackson table. Pressure points are appropriately padded, and the surgical field is prepared and draped in a sterile fashion. Unless contraindicated, use of a blood product recycling unit should be considered, although a recent cost-benefit analysis questioned its cost-effectiveness.⁶³ Fluoroscopy is used to mark an incision that ends about 5 to 7 cm above and below the vertebral levels to be instrumented. A midline skin incision is made, and subperiosteal muscle dissection is performed to expose the segments to be instrumented. The dissection is extended to expose the lateral tips of the transverse processes.

Once adequate bony exposure is achieved, the external landmarks for pedicle screw placement are identified. In the lumbar spine, the starting point for pedicle cannulation is typically defined as the intersection of the axial plane through the middle of the transverse process and the sagittal plane through the superior facet (Fig. 303-1A). The entry site for the first sacral pedicle is at the inferolateral portion of the superior S1 facet (Fig. 303-1D). Fluoroscopy is used to confirm each entry site, and a high-speed drill is then used to create a pilot hole just through the cortex. A pedicle finder is then gently advanced through the pilot hole to cannulate the pedicle and into the vertebral body (see Fig. 303-1A). Fluoroscopy can be used to determine the sagittal and axial trajectories. A ball-tip feeler is then used to palpate the trajectory created by the pedicle finder to assess for breaches. Recannulation of the pedicle using a modified trajectory may be necessary if a breach is identified. The trajectory is then prepared with a tap (Fig. 303-1B), followed by reassessment for evidence of breaches and introduction of the screw (Fig. 303-1C). Preoperative imaging can be used to preselect screw sizes, with the desired depth being about 70% to 80% of the vertebral body. To avoid injury to vascular and visceral structures, screws should not breach the anterior aspect of the vertebral body. Once all screws have been placed, the desired rod length is measured, cut, and contoured. The rod is secured into place with locking nuts, and levels may be distracted or compressed as indicated (Fig. 303-1D to F). Pedicle screws may be either monaxial or polyaxial, with the latter deigned to facilitate rod placement. One or more cross-linking devices may be used to link the rods horizontally. Triggered electromyographic stimulation has been suggested as an adjunct to optimize safe pedicle screw placement.^64,65

FIGURE 303-1 Diagrammatic depiction of lumbar pedicle screw and rod placement. A, A high-speed drill is used to score the cortical surface at the pedicle entry site. B, A pedicle finder probe is advanced to cannulate the pedicle. C, After the pedicle has been cannulated with a pedicle finder, a ball-tip probe is used to assess for breaches. A tap may be used to prepare the pedicle pathway. Shown is a pedicle screw being placed in the cannulated pedicle trajectory. D, The screw heads are aligned to receive the rod that has been cut to fit and contoured. E, Locking nuts are used to secure the rod to the screw heads. Shown is a tool to enable distraction of the segmental level. F, A compressor tool is used to apply segmental compressive forces.

(Images courtesy of DePuy Spine, Inc., Raynham, Massachusetts.)

Minimally Invasive Technique

Minimally invasive surgical techniques have been developed for posterior instrumentation of the lumbar spine to reduce the morbidity associated with the traditional open surgical approach.⁶² Potential advantages of using a minimally invasive approach for instrumentation include reduced blood loss, decreased infection risk, and less soft tissue and muscle trauma. Minimally invasive techniques for the placement of pedicle screw-rod systems in the lumbar spine have been previously described.^66–69 The best documented approach uses the bull’s-eye technique for placement of pedicle screws.

Similar to the open technique, the patient is first placed under general anesthesia, intubated, then positioned prone on a radiolucent surgical table, such as a Jackson table. Pressure points are appropriately padded, and the surgical field is prepared and draped in a sterile fashion. Anteroposterior fluoroscopy is then aligned to provide an en face view of the pedicles at the first desired level of instrumentation (Fig. 303-2A). Care must be taken to ensure that the pedicles are well aligned. At the working vertebral level, both the superior and inferior end plates should be aligned, and the spinous process should be in the midline. In addition, the pedicle should be visualized in the upper half of the vertebral body.

FIGURE 303-2 Percutaneous pedicle screw placement using the bulls-eye technique. A, Anteroposterior fluoroscopy is used to align the pedicles en face at the level to be drilled. A Jamshidi needle is aligned en face at the left L5 pedicle. B, The K wire is driven into the Jamshidi needle that is en face with the targeted pedicle. C, K wires have been placed into three pedicles, and a Jamshidi needle is aligned with the right L5 pedicle. D, Lateral view of the lumbar spine after K wires have been placed at L4 and L5 pedicles. E, A cannulated tap is used to prepare the pedicle trajectory. F, Pedicle screws have been placed into the left L4 and L5 pedicles. K wires remain on the right side at the same level.

Using fluoroscopic imaging, the tip of a Jamshidi needle is placed on the skin overlying the center of the pedicle, and a scalpel is used to make approximately a 2-cm vertical skin incision, centered at the tip of the needle. The Jamshidi needle is then carefully advanced through the incision, directed toward the underlying pedicle. Fluoroscopy and tactile feedback are used to place the tip of the Jamshidi needle in the center of the pedicle. The needle is then aligned to provide an en face view of both the needle and the pedicle (see Fig. 303-2A). A K wire is then driven about 2 cm beyond the tip of the Jamshidi needle (Fig. 303-2B). Next, the Jamshidi needle is removed, leaving the K wire in place (Fig. 303-2A to C). A K wire is similarly placed in each pedicle planned for instrumentation. On lateral-view fluoroscopy, the K wires are then driven to a depth of about two thirds of the vertebral body (Fig. 303-2D). A cannulated tap is passed over each K wire and used to prepare the trajectory (Fig. 303-2E), followed by placement of cannulated pedicle screws (Fig. 303-2F). Efforts have been made to develop accurate navigation systems for minimally invasive pedicle screw placement to reduce radiation exposure for both the surgeon and patient.^70–73

A variety of systems have been developed for minimally invasive rod passage following pedicle screw placement. Typically, minimally invasive rod passage is based on specific screw extenders that are attached to the screws, extend out of the wound, facilitate rod passage, and are then removed after the rod is secured in place (Fig. 303-3A and B). One of the first systems developed was the Sextant (Medtronic Sofamor Danek, Inc.), which can be used to instrument either one or two levels (see Fig. 303-3). After the pedicle screws have been placed, the extenders lock together, and a Sextant arm, with the rod attached at the end, is then connected to the extenders (see Fig. 303-3B). The Sextant arm then permits introduction of the rod through the screw heads (Fig. 303-3C and D). Locking caps are then placed through the extenders, and the extenders are removed.

FIGURE 303-3 Rod placement using the Sextant System (Medtronic Sofamor Danek, Inc.). A, The assembled pedicle screw, screw driver, and screw extender is shown on the left, and the individual components are shown on the right. Also shown is an example of a rod used with the Sextant system (bottom right). B, Screw extenders attached to screws in the left L4 and L5 pedicles have been connected, and the Sextant arm, with rod connected at the end, has been attached. C, The Sextant arm is rotated. D, The rod is delivered into the screw heads.

Because the rod passage technique employed by Sextant is limited to two levels, alternative techniques have been developed to pass rods minimally invasively for longer constructs. For example, the Viper2 (DePuy Spine, Inc., Raynham, Massachusetts) (Fig. 303-4) and the Longitude (Medtronic Sofamor Danek, Inc.) (Fig. 303-5) systems both enable minimally invasive rod passage following percutaneous placement of pedicle screws and can be used to instrument a single level or for substantially longer constructs.

FIGURE 303-4 Rod placement using the Viper2 System (DePuy Spine, Inc.). A, Screw extender with attached pedicle screw. Note the large opening in the extender available to accommodate the rod. B to D, The rod is sequentially passed through the openings in the extenders. Note the tool that can be used to align the extenders to facilitate rod passage (D). Also note that initial rod introduction does not require a separate incision. E, Once the rod is in place, locking nuts are placed to secure the rod to the screw heads.

(Images courtesy of DePuy Spine, Inc., Raynham, Massachusetts.)

FIGURE 303-5 Rod placement using the Longitude System (Medtronic Sofamor Danek, Inc.). A, Screw extender with attached pedicle screw. B, Screw extenders are aligned to receive the rod. C, The rod is attached to the end of a handheld device, a separate skin incision entry point is made, and the rod sequentially passed through the openings in the extenders. Intraoperative images show the contoured rod before passage (D) and a locking nut being placed after the rod has been placed (E).

(Images in A, B, and C courtesy of Medtronic Sofamor Danek, Inc., Memphis, TN.)

Achieving Arthrodesis

With open instrumentation, arthrodesis is promoted by clearing the bony surfaces, including laminae, facet joints, and transverse processes, of soft tissue and decorticating the surface with a high-speed drill. Decortication is best performed before rod placement to maximize the bony surface that can be accessed with the high-speed drill. Cancellous bone graft is then packed into the joints and placed along the decorticated surfaces. Autogenous iliac crest bone graft may be used depending on surgeon preference, the amount of graft needed, and the level of risk for pseudarthrosis. Investigations are ongoing regarding the use of agents such as bone morphogenetic protein (BMP) to assist in the fusion process.^35,74–76

Minimally invasive approaches for lumbar spine instrumentation do not provide adequate exposure to enable generous decortication and placement of bone graft. Instead, minimally invasive posterior lumbar instrumentation is often placed to supplement an interbody fusion, such as with a direct lateral diskectomy and graft placement, transforaminal lumbar interbody fusion (TLIF), or posterior lumbar interbody fusion (PLIF).

Degenerative Spondylolisthesis

Spondylolisthesis, the forward slippage of one vertebra relative to the adjacent vertebra, can result from degenerative disease. Degenerative spondylolisthesis may warrant surgical treatment in the setting of significant back or leg pain that is unresponsive to nonsurgical measures, if neurological deficits develop, or if the listhesis is shown to be progressive. Several studies have been designed to compare treatment approaches.^77–91 Martin and associates conducted a systematic review of the surgical management of degenerative lumbar spondylolisthesis.⁷⁸ Based on 13 studies, that were generally of low methodologic quality, satisfactory outcomes were significantly more likely with fusion than with decompression alone (relative risk, 1.40; 95% confidence interval, 1.04 to 1.89; P < .05). In addition, there was a significantly greater probability of achieving fusion with the use of instrumentation (relative risk, 1.37; 95% confidence interval, 1.07 to 1.75; P < .05). However, use of instrumentation was not associated with a significant improvement in clinical outcome (relative risk, 1.19; 95% confidence interval, 0.92 to 1.54; P > .05). Nevertheless, the optimal treatment approach for degenerative spondylolisthesis continues to be controversial and remains an area of active study.

Degenerative Disk Disease

There are few rigorously designed, randomized prospective studies to address the effectiveness of fusion for the treatment of degenerative disk disease.^16,92,93 Fritzell and colleagues reported a multicenter randomized controlled trial from the Swedish Lumbar Spine Study Group in 2001.⁹² This study included 294 patients who met the following criteria: (1) disabling low back pain for at least 2 years, (2) required to have radiographic and clinical evidence of spondylosis at either or both L4-5 and L5-S1, and (3) deemed to be surgical candidates. Patients were randomized to one of two arms, consisting of nonoperative or operative treatment. The nonoperative arm included physical therapy, education, and other pain-relieving modalities based on the discretion of the treating physician. Twenty-six surgeons provided operative treatment, which consisted of posterolateral fusion, posterolateral fusion and pedicle screw fixation, or interbody fusion supplemented with posterolateral fusion and pedicle screw fixation. The only significant demographic or clinical difference between the nonoperative and operative patient groups was that the latter had a greater degree of comorbidity. Two-year follow-up was achieved in 98% of patients and included interim evaluations at 6 months and 1 year. A battery of standardized outcomes measures were used, including the pain Visual Analog Scale (VAS) score, Oswestry Disability Index (ODI) for low back pain, the Million VAS, the General Function Scale (GFS), Work Status, a patient satisfaction survey, and an independent evaluation by a different spine surgeon. Rigorous statistical analyses were performed, including intention to treat and worst-case scenarios.

The data from Fritzell and colleagues clearly demonstrated that the surgical group had significantly improved outcome based on pain relief, disability, and degree of satisfaction compared with the nonsurgical group. All outcome measures assessed were significantly better in the surgically treated patients compared with the patients treated nonoperatively. The study by Fritzell and colleagues has been criticized as being biased against nonoperative therapy because patients had already failed a trial of nonoperative treatment before entering the study.⁹⁴

Brox and associates conducted a randomized prospective study comparing operative and nonoperative treatment for low back pain. Inclusion criteria included the following: (1) low back pain for at least 1 year, (2) ODI score of 30 to 100, and (3) radiographic evidence of degeneration at either or both L4-5 and L5-S1.⁹³ Patients randomized to operative treatment underwent instrumented posterolateral fusion. Patients randomized to nonoperative treatment underwent a very rigorous program designed for patients with low back pain. This included cognitive therapy, supervised physical therapy (average of 25 hours per week for 8 weeks), and a home program of exercises. Outcome measures included a modified ODI, pain VAS score, medication use, GFS, the Waddel Fear Avoidance Belief Questionnaire, and a patient satisfaction score. Of the 121 patients evaluated for inclusion, only 64 were ultimately randomized, with 37 to surgical treatment and 27 to nonsurgical treatment. Except for a greater number of men in the surgical group, the patients in the two treatment arms did not differ significantly.

At the 1-year end point, there was a 97% follow-up rate, and both groups improved significantly from baseline on all outcome measures. The primary outcome measure was the modified ODI, which improved by 15.6 points in the surgical group and 13.3 points in the nonsurgical group. The mean difference of 2.3 points was adjusted to 2.7 points, after adjusting for differences in sex distribution, but remained statistically nonsignificant. The results of the secondary outcome measures primarily favored surgical treatment, with this arm having significantly greater improvement in pain (both back and lower extremity), less emotional distress, and higher success ratings (based on both patient and independent physician assessments). Fingertip-to-floor distance showed significantly greater improvement in the nonsurgical group compared with the surgical group. The data from the study of Brox and associates must be interpreted carefully owing to very large confidence intervals and the small patient population.

Trauma

The most common traumatic lumbar spine fractures are compression and burst fractures. Although most compression fractures and stable burst fractures may be treated nonsurgically,⁹⁵ for those cases that do warrant surgical treatment, instrumentation offers several advantages.⁹⁶ These advantages include the ability to correct kyphotic deformity, to help protect neural elements from injury, to facilitate fusion through rigid immobilization, to reduce pain, and to facilitate earlier patient mobilization. Whether surgical treatment is best achieved through posterior, anterior, or combined approaches depends on the integrity of the anterior and posterior columns and remains a controversial topic.^97–101 The superiority of pedicle screw systems over other forms of posterior instrumentation for unstable lumbar spine fractures has been demonstrated.^102–105 Several authors have reported favorable outcomes using minimally invasive techniques in the management of thoracolumbar trauma.^61,106,107

Although it is generally accepted that posterior instrumentation with pedicle screws can be an effective means of treating unstable lumbar fractures,¹⁰⁸ there continues to be considerable controversy about whether short-segment pedicle instrumentation is adequate treatment.^{100,109–114} McLain and coworkers provided a preliminary report on instrumentation failure in 10 of 19 patients treated with short-segment pedicle instrumentation for thoracolumbar fractures.¹⁰⁹ These failures included progressive kyphosis related to bending of the screws in 6 patients, kyphosis due to bony collapse or vertebral translation in 3 patients, and segmental kyphosis following fracture of a caudad screw in 1 patient. The authors suggested that posterior screw fixation alone may be inadequate.

In contrast, Parker and colleagues reported very favorable results for 46 patients who were treated with short-segment pedicle screw fixation (one level above the fracture to one level below the fracture) for thoracolumbar spine fractures.¹¹⁰ One patient died of unrelated causes. All the remaining 45 patients were followed at least 40 months and had successful fracture healing in virtual anatomic alignment. In deciding which patients should be treated with short-segment posterior fixation, the authors used the Load-Sharing Classification.¹¹⁵

Afzal and associates evaluated the use of short-segment pedicle screw instrumentation and augmentation vertebroplasty in 16 patients with lumbar burst fractures.¹¹¹ The central and anterior height of the vertebral body were restored to 72% and 82% of the estimated intact height on average, respectively. They noted four complications, including three cases of cement leakage without clinical consequence and one superficial wound infection. All patients demonstrated good fracture reduction. The authors concluded that posterior short-segment pedicle fixation in conjunction with balloon vertebroplasty is a feasible option to address all three vertebral columns through a single approach.

Infection

Operative stabilization may be indicated for lumbar vertebral osteomyelitis and diskitis. However, there are several areas of controversy and active research regarding the optimal timing and approach for these surgical procedures. Traditional teaching suggests that, if spinal reconstruction is indicated, clearing the infection, which may include an open débridement, should precede spinal instrumentation. Multiple recent studies have suggested that débridement and instrumented stabilization of the infected and unstable spine may be performed in a single stage. Carragee reported a retrospective review of 17 cases of spinal instrumentation for active pyogenic vertebral osteomyelitis with greater than 2-years of follow-up.³⁷ Twelve of the 17 patients were immunocompromised. Eight patients had significant postoperative complications, which include two instrumentation failures and one wound dehiscence. All the patients were mobilized directly after instrumentation, and none of the patients had evidence of persistent infection at last follow-up. The author concluded that instrumentation did not appear to compromise the ability to clear infection. Other reports have demonstrated similar success of instrumentation in the setting of active infection, including both pyogenic^36,38 and tuberculous.^{38–40,42,43} The use of titanium implants, including pedicle screws, laminar hooks, rods, and cages, has been specifically investigated and does not appear to negatively affect infection clearance, although this remains controversial.^34–3641

In cases of spinal osteomyelitis that warrant operative stabilization, both anterior and posterior approaches may be required. With the use of modern posterior instrumentation, several authors have suggested that a posterior-only approach may be adequate. Güzey and coworkers reported on 19 patients with thoracic or lumbar tuberculous spondylitis who were treated with a single-stage posterior decompression and débridement, followed by placement of interbody grafts if necessary, instrumentation, and posterior or posterolateral grafts.⁴² The only instrumentation failure was fracture of a single pedicle screw at 3 months of follow-up, and all patients demonstrated at least some improvement of neurological status at a mean follow-up of 53 months. Thirteen patients had a kyphotic deformity, with a mean angulation of 18.2 degrees (range, 0 to 42 degrees). Although this did not significantly change following surgery (mean, 17.3 degrees; range, 0 to 42 degrees), there was only a 2.8-degree loss of correction (range, 2 to 5 degrees) at a mean follow-up of 44 months, and there was no progression of kyphosis beyond 15 months in any patient.

Lee and associates reported 16 patients with lumbar tuberculous spondylitis who were treated by posterior lumbar interbody fusion with autogenous iliac bone grafting and pedicle screw instrumentation.⁴³ Bony fusion was achieved within 1 year in 15 patients. They noted a preoperative lordotic angle of 27.8 degrees (range 9 to 45 degrees) that improved to 35.8 degrees (range 28 to 48 degrees) at a mean follow-up of 33 months.