Infections of the Thoracic Spine

Published on 11/04/2015 by admin

Filed under Orthopaedics

Last modified 22/04/2025

Print this page

rate 1 star rate 2 star rate 3 star rate 4 star rate 5 star
Your rating: none, Average: 0 (0 votes)

This article have been viewed 2340 times

47 Infections of the Thoracic Spine

KEY POINTS

Introduction

The last several decades have witnessed a rise in reported spinal infections. This increase has largely been attributed to factors associated with a growing elderly population. Improvements in medical care have directly resulted in prolonged life expectancy with more individuals living longer with chronic diseases. As a result, various medical conditions associated with advanced age, such as diabetes or illnesses that lead to immunocompromise, predispose patients to developing spinal infections. Additionally, as individuals are living longer, more elderly patients are seeking to undergo spinal procedures for degenerative conditions that otherwise, left untreated, result in debilitating pain. Both minor procedures such as discography and epidural injections and extensive spinal fusion surgeries pose the risk of direct bacterial inoculation of the spine.

Infections of the spine are characterized either by their microbiology or by the location of pathology. From a microbiology standpoint, spinal infections are differentiated by pyogenic or granulomatous etiologies. Pyogenic infections are generally of bacterial origin. Granulomatous spinal infections encompass fungal etiologies, but include some bacterial sources, and refer primarily to the histologic course of the infection. Spinal tuberculosis is by far the most common of the granulomatous spinal infections worldwide.

Spinal infections are also classified by the primary location of pathogenesis. Sole involvement of the disc space is referred to as discitis. Osteomyelitis is an infection of the bony spine (Figure 47-1). Osteodiscitis or spondylodiscitis is combined involvement of the intervertebral disc and the vertebra. Abscess or granulation formation can occur in a subdural, epidural, or paravertebral location (Figure 47-2). Frequently, spinal infections invade all compartments of the spinal column, including the soft tissues, bony spine, and within the spinal canal.

Spinal osteomyelitis is estimated to occur in 1 in 100,000 to 250,000, and accounts for 2% to 7% of all cases of osteomyelitis. Spinal osteomyelitis occurs more commonly among older individuals, with approximately one half of all patients being over 50 years of age. Similarly, epidural abscesses occur in adults and are estimated to occur in 0.2 to 1.2 per 10,000 hospital admissions annually. When bacterial spinal infections occur in younger individuals, they are more commonly seen in intravenous drug users. Both osteomyelitis and epidural abscesses generally occur in the thoracic and lumbar spine, with thoracic infections representing over a third to half of all cases, and lumbar infections accounting for a majority of the remainder. Cervical spine infections are estimated to account for only 5% to 14% of all cases.

Outside the United States, spinal tuberculosis still represents a considerable health care problem. Tuberculosis is relatively common in underdeveloped countries where malnutrition and overcrowding are present. It is estimated that 2 billion people have tuberculosis worldwide, with 9 million new cases each year. Approximately 5% of these patients have spinal involvement. Spinal tuberculosis is a major source of morbidity, representing the most common cause of nontraumatic paraplegia in underdeveloped countries.

While the incidence of spinal infections is increasing, management of these conditions is also dramatically evolving. Earlier detection, better screening and surveillance, and advanced imaging modalities have improved diagnosis of spinal infections and identification of pathogenic organisms. More effective antimicrobial pharmacotherapy has led to better medical treatment with clearance of infection and less recurrence. Surgical treatment options have incorporated advances in surgical technique, instrumentation, and biomedical technology to increase eradication of infection, preservation of neurological function, restoration of spinal alignment, and prevention of deformity and chronic pain.

Pathophysiology

The pathophysiology of spinal infection ultimately begins with the individual’s underlying predisposing risk factors. Advanced age, diabetes, and multiple medical comorbidities are associated with increased risk for spinal infection. Additionally, spinal surgery, intravenous drug use, and immunocompromise contribute to further risk. Infection generally metastasizes hematogenously to the spine from extraspinal sources such as the urinary tract, respiratory system, skin or soft tissue infections, or cardiac vegetations. Direct inoculation from surgery, percutaneous procedures, or penetrating trauma is an additional modality for bacterial seeding. Local invasion to the spine also occurs from infected adjacent or contiguous sources such as the retroperitoneal, abdominopelvic, pleural, or retropharyngeal spaces. Spread of infection can also occur within the spinal column by direct extension from the bony or soft tissue elements to the epidural space.

Bacterial Pathogenesis

Hematogenous seeding of the spine may occur via either arterial or venous pathways. The venous plexi that drain from within the spinal canal communicate with plexi that form a venous ring around each vertebral body. This venous system communicates with the venous drainage of the pelvis. Batson demonstrated that this venous pathway is a valveless system in which microorganisms may circulate and lodge in the low-flow end-organ vasculature surrounding the vertebral body. Alternatively, direct bacterial seeding of the vertebral body may occur from ascending and descending arterial branches that send penetrating vessels to the vertebral body.

Pathologic sequelae of spinal infection include loss of spinal alignment with progressive deformity, and risk of neurological compromise. Bacterial involvement of the spinal column with subsequent inflammatory infiltration causes bony destruction and eventually erodes the subchondral plate to involve the relatively avascular disc space (Figure 47-3). Advanced bone loss, particularly across multiple adjacent segments, combined with disc space narrowing, leads to progressive kyphotic deformity. neurological compromise may result from severe bony destruction, resulting in pathologic fracture with retropulsed bony fragments into the canal. Epidural abscess formation or extension of inflammatory granulation tissue into the canal can cause direct compression of the spinal cord or nerve roots. Additionally, septic thrombosis of veins within the epidural space or the arteriolar supply can cause ischemic injury. Particularly, in a spinal cord already compromised by mechanical compression from either an abscess or fracture, hypoperfusion from thrombosed feeding arteries or draining veins may lead to rapid neurological deterioration.

Gram-positive cocci are the most prevalent inciting organism, representing 50% to 67% of all causative organisms. Staphylococcus aureus is the most prevalent bacteria identified, accounting for 80% of all gram-positive infections, and 55% of all spinal infections. In a meta-analysis of 915 patients with epidural abscess, S. aureus was identified as the causative organism in 73.2% of cases. Gram-negative bacteria, particularly Escherichia coli and Proteus, are more frequently identified in patients with preexisting urinary tract infections. Pseudomonas aeruginosa is most common among immunocompromised patients or intravenous drug users. Indolent infections are more likely to occur with low-virulence organisms such as Streptococcus viridans or Staphylococcus epidermisdis.

Clinical Presentation

The clinical presentation of bacterial spinal infections often depends on the virulence of the organism, the duration of infection, and the overall integrity of the patient’s immune system. Improved diagnostic modalities have led to earlier detection of disease, with initiation of appropriate medical therapy often before patients develop systemic illness or potentially irreversible neurological compromise. Over 90% of patients with pyogenic osteomyelitis present with axial neck or back pain as the primary complaint. The pain is generally characterized as insidious and nonmechanical in nature, and unrelieved by recumbency. Patients frequently note local spine tenderness with limited range of motion. Constitutional symptoms associated with infection such as fevers, chills, and malaise may also be present: however, an elevated temperature is only found in 52% of patients at time of presentation.

Neurological findings are less common with pyogenic osteomyelitis. A review of the literature reveals that only 17% of patients with bacterial osteomyelitis have neurological signs or symptoms on initial presentation. Alternatively, neurological complaints are frequently associated with acute bacterial epidural abscess, with 56% of patients presenting with motor deficits, and 36% with radicular pain. The clinical triad of localized spine pain, fever, and progressive neurological deficit is seen, however, in only 36% of patients with epidural abscess.

Spinal tuberculosis has a similar presentation to bacterial osteomyelitis, with most presenting with spine pain and localized tenderness. Unlike bacterial osteomyelitis, however, patients with tuberculosis present with a more insidious course. A mean duration of symptoms prior to diagnosis is 6.1 months. neurological deficits at the time of presentation are also more prevalent with tuberculosis, with 44.9% of patients having neurological findings. Motor function abnormalities are present in 34.6% of patients with spinal tuberculosis, with 6.4% being paraplegic at time of presentation.

Diagnostic Evaluation

The initial evaluation of a patient suspected of spinal infection includes standard serologic markers for infection or inflammation. A basic panel includes peripheral white blood cell count (WBC), erythrocyte sedimentation rate (ESR), and C-reactive protein (CRP). WBC is elevated at time of presentation, however, in only 42% of cases, and often normalizes in patients with chronic infection. ESR and CRP are markers of inflammation and demonstrate high sensitivity for spinal infection. CRP, an acute phase protein, increases within 4 to 6 hours of infection. ESR begins to increase only several days after the onset of infection and peaks at 7 to 8 days. ESR is elevated in over 90% of patients with spinal infection; however, ESR and CRP lack specificity, and may be increased in patients either with infection or with other inflammatory disorders. Individuals suspected of tuberculosis are assessed with a PPD and subsequently sputum staining for acid-fast bacilli.

Definitive diagnosis of spinal infection is made upon identifying the causative organism from positive culture. Prompt blood and urine cultures are obtained immediately on presentation, as infection commonly spreads to the spine either from the genitourinary tract or hematogenously. Positive blood cultures identify the inciting organism in 25% to 59% of cases. Ideally, cultures are obtained prior to initiating antimicrobial therapy to obviate the potential of a sterile nondiagnostic culture.

Biopsy of an abnormal spinal lesion can confirm the diagnosis of infection as well as isolate the inciting organism. Percutaneous closed biopsy is performed using computed tomography (CT) or fluoroscopic guidance. Closed biopsy demonstrates a reported accuracy of 70% to 100% in identifying the causative organism. Open surgical biopsy is indicated in the setting of a nondiagnostic closed biopsy in a patient with persistent clinical infection or deterioration despite broad-spectrum medical therapy, or for lesions inaccessible percutaneously. Open biopsy is diagnostic in over 80% of patients, likely due to a larger bony sample. A high concordance rate is observed in patients with both positive blood and biopsy specimens, reinforcing the importance of early blood culture sampling prior to initiating antimicrobial pharmacotherapy.

Imaging

Plain spine x-rays may demonstrate characteristic findings associated with osteomyelitis or osteodiscitis, and often serve as a rapid method for surveying the full spinal axis for potential infection. Disc space narrowing is the earliest and most consistent radiographic finding, occurring in 74% of cases, generally after approximately 2 to 4 weeks. Enlargement of the paravertebral shadow may indirectly suggest a thoracic paravertebral abscess. After 3 to 6 weeks, leukocyte infiltration into the subchondral bone and vertebral body leads to bony destructive changes, appearing as a lytic area in the anterior aspect of the vertebral body adjacent to the disc, or blurring of the endplates. With advanced bone loss, the vertebral body collapses. Thirty-six inch standing x-rays are essential for assessing progression of sagittal and coronal plane deformity in severe cases. With chronic disease (after 8 to 12 weeks), reactive bone formation and endplate sclerosis occurs. Ultimately, the reparative process results in new bone formation and hypertrophic changes. Eventually, 50% of cases lead to spontaneous fusion; however, it may require several years for this to take place. The remaining cases likely form a fibrous ankylosis which may similarly effectively immobilize the involved segment.

Radionuclide studies are capable of detecting and localizing infection before abnormal findings are observed on plain radiographs. Gallium scanning demonstrates 89% sensitivity, 85% specificity, and 86% accuracy for diagnosing disc space infections. Technetium scanning is 90% sensitive, 78% specific, and 94% accurate. Combined gallium and technetium scanning is reported to have 94% accuracy. SPECT is a sensitive bone scintigraphic modality for early detection of osteomyelitis and is often performed in conjunction with technetium and gallium scanning.

CT imaging is beneficial for evaluating the extent of bony destruction. Axial CT imaging demonstrates the presence of retropulsed fragments and the degree of canal compromise in the setting of pathologic fracture. Sagittal reconstructed CT imaging may reveal endplate osteopenia as an early finding of infection. Superb detailing of bony anatomy may be useful for preoperative planning in cases necessitating surgical intervention. Also, CT imaging can delineate adjacent soft tissue abscess or granulation tissue that may require operative debridement.

Magnetic resonance imaging (MRI) is the gold standard for radiologic evaluation of spinal infection. MRI demonstrates high sensitivity (96%), specificity (92%), and accuracy (94%). Intravenous gadolinium further delineates areas of abnormal enhancement and facilitates localization of infection to the vertebral body, intervertebral disc, or epidural space. Optimal visualization of the neural elements allows for evaluation of canal compromise or spinal cord compression. MRI is readily capable of delineating paravertebral abscesses. Multiplanar imaging allows for full evaluation of the complete spinal column in sagittal and axial planes to assess for the extent of involvement.

Management

Management of spinal infections has dramatically evolved over the last several decades. Advances in imaging allow for prompt diagnosis with initiation of appropriate antimicrobial pharmacotherapy, often early in the clinical course. Improved surgical technique combined with developments in spinal instrumentation has resulted in decreased surgical morbidity and better long-term clinical outcomes. The general principles of treatment for spinal infections, regardless of medical or surgical intervention, are fundamentally the same. The primary objectives are to eradicate infection, preserve neurological function, maintain spinal alignment, and prevent pain.

Medical Therapy

Medical therapy for spinal infection consists primarily of antimicrobial pharmacotherapy. Most patients with vertebral osteomyelitis respond successfully to nonsurgical treatment. The main tenet of medical therapy is identification of the inciting organism with either a positive blood or biopsy specimen, and initiation of an appropriate antimicrobial agent. The selection of either a single or multi-drug regimen is dictated by the virulence and resistance of the causative organism. Therefore, optimal treatment is entirely dependent on isolating an organism. As a result, antimicrobial treatment is withheld in patients that are neurologicalally and clinical stable until definitive cultures are obtained. Patients presenting with sepsis or progressive deterioration may necessitate empirical broad-spectrum coverage until an organism is identified.

Antimicrobial therapy is generally delivered parenterally for a minimum of 6 weeks. A 25% failure rate is observed in patients treated with antibiotics for less than 4 weeks. Serial serologic evaluation of ESR is an effective measure of therapeutic response. After 6 weeks of intravenous antibiotics, some advocate continuing oral therapy until the ESR has diminished by a minimum of one half the pretreatment level to prevent relapse. A two-thirds reduction in ESR from pretreatment levels is an indication of complete eradication of infection. In addition to antimicrobial pharmacotherapy, immobilization with an external orthosis is recommended for patients with severe pain, greater than 50% vertebral height loss, or involvement of the thoracolumbar junction.

Medical treatment for spinal tuberculosis is primarily reserved for patients without any neurological involvement. The Medical Research Council Committee for Research on Tuberculosis in the Tropics concluded that treatment for spinal tuberculosis in developing countries consists of ambulatory pharmacotherapy with 6- or 9-month regimens of isoniazid or rifampin. In Western countries, drug therapy for spinal tuberculosis is 6 months of isoniazid, rifampin, and pyrazinamide. Others advocate a more aggressive approach to spinal tuberculosis with 12 months of treatment, beginning with isoniazid, ethambutol, rifampin, and pyrazinamide for the first 2 months, followed by tailoring of the therapy based on sensitivities. Multimodal therapy is often necessary due to potential drug resistance, as well as the decreased accessibility of certain agents to different involved organ systems. Unfortunately, many of these agents have potential side effects, with the risk of liver failure among the more clinically significant.

Indications for Surgical Intervention

There are several indications for surgical intervention for spinal infection. Open surgical biopsy to determine the bacteriologic diagnosis is recommended in patients with nondiagnostic cultures or closed biopsy. Patients in sepsis refractory to medical treatment may require abscess drainage or debridement of necrotic tissue to facilitate penetration of antimicrobial therapy to sites of active infection. Individuals presenting with acute neurological deficit resulting from spinal cord compression require emergent decompression. Delaying surgical intervention in neurologicalally compromised patients may be cautiously reserved in those who are too significantly medically compromised to undergo surgery, and those who present with over 72 hours of neurological deficit. Chronic pain and significant deformity are relative indications for surgical intervention.

Patients with spinal tuberculosis and neurological deficit generally are require radical debridement with bone grafting and stabilization. There are data to suggest that patients with tuberculosis and mild neurological deficits may respond to medical therapy alone. In a study of 200 cases of patients with spinal tuberculosis and neurological impairment, 38% of patients recovered with only medical therapy. Sixty-two percent, however, ultimately required surgery, with 69% of surgically treated patients having a complete neurological recovery. A direct correlation between duration of neurological symptoms prior to surgery and time for recovery from paraplegia supports early operative intervention in patients with neurological impairment. With prompt surgical treatment, better neurological outcomes and prevention of deformity can be expected.

Surgical Management

Several important issues require consideration once it is determined that a patient requires surgical intervention. The primary issue is deciding upon an appropriate surgical approach and fusion technique, from a broad spectrum of operative modalities previously described. Anterior approaches include anterior debridement and fusion with or without instrumentation. Posterior approaches involve a posterior decompression, debridement, and instrumented fusion. Circumferential approaches include anterior debridement with strut grafting and instrumentation with posterior supplemental fixation in a single-stage or delayed fashion. Ultimately, surgical decision-making is dependent upon whether the primary pathology is ventral, dorsal, or circumferential, and whether the infected tissue requires complete or partial debridement. Additional factors include the degree of preexisting deformity, determining the optimal technique for restoring spinal alignment, and whether spinal reconstruction and stabilization are necessary. Last, given the propensity for significant medical comorbidities in this patient population, serious consideration must be given toward selecting a surgical approach that the patient can tolerate with minimized morbidity.

Timing of surgical intervention is also a critical factor. Patients with acute neurologicalal deficits secondary to spinal cord compression require emergent decompression to prevent irreversible injury. Persistent sepsis despite medical therapy, with significant abscesses and infected or necrotic tissue, may necessitate urgent drainage or debridement to decrease the overall infectious burden and facilitate antimicrobial penetration. Acute instability that threatens neurological structures demands immediate immobilization and may require urgent operative stabilization. Delayed surgical intervention is indicated for patients that are stable neurologically and clinically, but have disabling pain or evidence of chronic progressive deformity. Generally, in these instances, surgical instrumented stabilization and arthrodesis is performed after the acute infection is cleared.

The use of instrumentation and certain grafting techniques in an acutely infected wound is controversial. Instrumented spinal fusion surgery is associated with increased risk of infection compared to nonfusion surgery. However, there are growing laboratory and clinical data to suggest that titanium-based implants may have improved resistance to bacterial colonization than traditional stainless steel, and may be appropriate for spinal stabilization even in the setting of acute infection. Selection of appropriate graft material to promote arthrodesis without serving as a host environment for further bacterial seeding is also critical. Last, various minimally invasive techniques for surgical debridement, instrumented stabilization, and fusion have become available that may serve to improve clinical outcomes and reduce surgical morbidity compared to conventional open surgical modalities.

Posterior Approach

Posterior decompression for spinal infection is primarily reserved for evacuation of isolated epidural abscesses without involvement of the bony anterior spinal column or intervertebral discs. Epidural abscesses, particularly in the thoracic and lumbar spine, preferentially occur dorsal to the thecal sac, and therefore are amenable to laminectomy for decompression and drainage. The extent of the laminectomy ideally does not involve the facet joints, so as to prevent iatrogenic destabilization. Reports of limited interlaminar decompression with epidural abscess fenestration have been described; however, the benefit of this minimal approach over standard laminectomy has not been demonstrated. The additional insertion of an epidural suction-irrigation catheter at the time of surgery, for postoperative continuous washout, has also been reported to have beneficial results.

Posterior decompression alone is not recommended in the setting of osteomyelitis, discitis, or osteodiscitis. Laminectomy with removal of the posterior tension band further destabilizes the spine in patients with already impaired anterior column support. Posterior decompression in the setting of vertebral osteomyelitis has resulted in unfavorable outcomes related to deformity progression, increased instability, and neurological deterioration.

With the advent of pedicle screw-rod fixation, a single-stage posterior approach for decompression, debridement, and instrumented stabilization may be an appropriate alternative surgical modality (Figure 47-5 A-C). Various posterior approaches for accessing anterior thoracic and lumbar pathology are available. Costotransversectomy, lateral extracavitary, and transpedicular techniques allow access to the anterior spinal column via a posteriorly based approach. With these techniques, debridement of varying degrees of the anterior column may be performed, although complete vertebrectomy via a solely posterior approach is technically challenging, given limited visualization of the anterior aspect of the thecal sac.

After debridement of infected, necrotic tissue, anterior column reconstruction may be achieved using either stackable or expandable interbody devices that are designed to be inserted from a posterior approach (Figure 47-6 A-B). Particularly in the thoracic spine, a unilateral single nerve root may be ligated to facilitate insertion of an interbody cage. Again, however, limited exposure via a posterior approach may restrict the size of interbody graft that can be inserted, thereby presenting potential risk for graft subsidence, kyphosis, or nonunion. Supplemental posterior fixation with a pedicle screw-rod construct provides instrumented stabilization, and thereby prevents progressive sagittal deformity as well as facilitates arthrodesis. A single-stage posterior approach for debridement, decompression, and stabilization may be particularly suited for medically compromised patients with osteomyelitis who may not tolerate a thoracotomy for anterior exposure.

Anterior Approach

Anterior procedures to surgically treat osteomyelitis have become increasingly popular since Hodgson first reported anterior debridement and fusion for spinal tuberculosis in 1960, and have since become the standard treatment for pyogenic osteomyelitis as well. Because the pathology is generally ventral, an anterior approach allows for thorough debridement of infected and necrotic tissue, and drainage of psoas or paravertebral abscesses. With an anterior approach, it is possible to completely remove all necrotic tissue until bleeding, well-vascularized bone is encountered, and to decompress the ventral thecal sac. Anterior spinal column reconstruction with an interbody graft for arthrodesis, anterior column support, and restoration of sagittal alignment is also best attained from an anterior approach. Spinal fixation for stabilization and to facilitate arthrodesis can also be performed from an anterior approach (Figure 47-7 A-D).

Various techniques for anterior debridement, vertebral column reconstruction, and instrumented stabilization have been described. Hodgson’s original description of an anterior procedure to treat spinal tuberculosis involved anterior debridement and autologous strut grafting without instrumentation. While initial reports demonstrated successful clinical outcomes, subsequent studies have observed loss of correction, progressive deformity, and pseudarthrosis without the use of instrumentation. As a result, various developments in device technology for spinal reconstruction and stabilization have been made to improve upon these findings. The uses of anterior instrumentation with or without posterior supplemental fixation in single-stage or two-stage procedure have evolved as modern modalities for the treatment of vertebral osteomyelitis.

Anterior Approach with Anterior Fixation

An anterior approach allows for a single-stage, single-approach surgical treatment for debridement, decompression, arthrodesis, and stabilization. Initially, anterior procedures incorporated autologous strut grafting without instrumented stabilization, because of concern about placing a foreign body in a contaminated wound. As a result, patients were immobilized and maintained on prolonged bed rest postoperatively. Recently, however, numerous reports have described successful use of titanium-based implants in the setting of spinal infection without evidence of persistent infection or relapse. The use of anterior spinal fixation in combination with anterior debridement and grafting allows for early patient mobilization, thereby reducing the risk of complications associated with prolonged recumbency such as pneumonia, pulmonary embolism, decubitus ulcer, and muscle atrophy.

Dai et al reported on 22 patients treated with an anterior-only approach for the treatment of thoracic and lumbar osteomyelitis.1 Patients underwent an anterior debridement, interbody fusion with autologous graft, and anterior instrumented stabilization. Follow-up was for a minimum of 3 years, and there were no cases of residual or recurrent infection. ESR and CRP returned to normal levels within 4 to 10 weeks postoperatively.

With anterior column reconstruction, the investigators observed an improvement in kyphosis, with an average correction rate of 93.1%. Solid arthrodesis was achieved in all patients within 6 months, with only two patients requiring immobilization with an external orthosis. Significantly, there were no cases of implant failure and only three instances of mild graft subsidence.

Patients also demonstrated significant functional and neurological recovery. Eighteen patients were standing and ambulating within 1 week postoperatively. The remaining 3 patients were walking within 4 weeks. There were no cases of postoperative neurological deterioration. All patients with preoperative neurological deficits had complete recovery within 6 months except for one Frankel grade C patient who improved to a Frankel grade D.

The anterior-only approach provides the benefit of thorough debridement, reconstruction, fusion and stabilization in a single-stage, single approach. With a single surgical procedure, there is less morbidity associated with prolonged anesthesia, lengthy operative time, blood loss, and potential tissue injury in patients who are generally medically compromised and may be predisposed to poor wound healing. The addition of supplemental posterior fixation may result in longer constructs with further loss of spinal motion segments. Others are concerned that placing instrumentation in a contaminated wound results in formation of a biofilm on the implant surface layer that harbors bacteria and is poorly penetrable by antibiotics. Therefore, additional posterior spinal instrumentation may present an increased risk for persistent infection or recurrence.

Single-Stage Anterior and Posterior Procedure

A combined anterior and posterior procedure to treat vertebral osteomyelitis provides several benefits over a single anterior approach. Circumferential access to the spinal canal allows for complete neural decompression in patients who may have both ventral compression from retropulsed bone fragments and dorsal compression from epidural abscess or posterior spinal arch involvement.

Korovessis et al studied 24 patients with osteomyelitis treated with a single-stage anterior debridement, partial vertebrectomy, mesh cage and autologous bone graft, and supplemental pedicle screw fixation.2 Follow-up was for an average of 56 months, with all patients demonstrating complete resolution of infection. While three patients who were ASIA A on presentation remained ASIA A postoperatively, patients with incomplete spinal cord injuries improved an average of 1.4 Frankel grades postoperatively. Six patients with incomplete injuries preoperatively had full recovery of neurological function within 1 year of surgery. Eleven patients who were neurologicalally intact preoperatively returned to full premorbid functional and activity levels within 4 to 6 months after surgery. Visual analog pain scores improved postoperatively as well.

Combined anterior and posterior instrumentation provides an idealized biomechanical construct to treat advanced bony destruction, spinal instability, and deformity secondary to vertebral osteomyelitis. Anterior removal of necrotic tissue with anterior column reconstruction provides optimal load sharing and restoration of sagittal alignment in cases of vertebral height loss. Posterior supplemental fixation recreates the posterior tension band to restrict potential for long-term loss of sagittal plane correction.

An anterior-only procedure presents concern regarding long-term stability. Some have observed that an anterior procedure without posterior supplemental fixation results in poor sagittal correction and long-term increase in kyphosis. An anterior fusion alone may be appropriate for a single-level corpectomy with an intact posterior tension band. However, patients with multilevel involvement, significant bony endplate destruction, or disease that crosses the thoracolumbar junction may be predisposed to failure with an anterior-only construct. Particularly, patients with loss of the posterior tension band, either through extensive posterior spinal arch involvement such as in spinal tuberculosis, or from iatrogenic destabilization via laminectomy, may also require supplemental posterior instrumentation. Alternatively, excellent restoration of sagittal alignment with long-term maintenance of correction has been demonstrated with a combined anterior-posterior procedure.

A combined anterior-posterior procedure also creates an optimal mechanical environment for arthrodesis. With an anterior interbody fusion, the graft is placed under compressive rather than tensile forces, which facilitates arthrodesis. Posteriorly placed transpedicular instrumentation provides rigid immobilization in all three planes of rotation to effectively stabilize the spine and improve fusion.

Two-Staged Anterior-Posterior Procedure

Circumferential treatment of vertebral osteomyelitis can be performed as a single-stage procedure or in a two-staged fashion, with initial anterior debridement and then delayed posterior fixation. Staged spinal surgery has gained popularity for the treatment of various other complex spinal disorders such as deformity, trauma, and oncologic and rheumatologic conditions. The benefit of staged surgery is a shorter operative time and less blood loss for each individual procedure, which may be particularly relevant for patients with worse overall general health. A two-staged surgery allows for a convalescent period to bridge between the two procedures, in which the patients may have an opportunity to recover clinically and neurologicalally. Also, performing supplemental posterior instrumentation in a delayed manner allows for a longer course of antimicrobial therapy to further reduce the infected environment prior to implantation of hardware.

Dimar et al reported on 42 patients with osteomyelitis treated with anterior debridement and strut grafting, followed by delayed instrumented posterior spinal fusion at an average of 14.4 days after the initial procedure.3 Many patients were acutely ill at presentation requiring urgent treatment, but were in poor overall clinical status to undergo an extensive circumferential operation. Most were significantly debilitated from inadequate nutrition as well. For these patients, Dimar et al performed anterior debridement and anterior strut grafting urgently to thoroughly remove the infection and restore anterior column support. Patients were then immobilized in an external orthosis, continued on intravenous antibiotics, aggressively resuscitated nutritionally, and initiated on physical therapy. Delayed posterior spinal fusion was then performed on a semi-elective basis when patients were clinically stable to undergo a second procedure.

All patients had complete resolution of their infection with no evidence of recurrence. Patients with preoperative neurological deficits improved postoperatively. No significant deterioration in patients’ overall medical condition was observed as a result of the interoperative period. The average length of hospitalization, however, was prolonged in this series, with a mean length of stay of 24 days and range of 14 to 53 days.

Use of Instrumentation

The use of instrumentation in patients with spinal infection remains a controversial issue. Hardware placement for fusion operations in uninfected patients has been shown to increase postoperative infection rates. As a result, historically, there has been concern regarding placement of instrumentation in a known contaminated field. This is particularly an issue in complex spinal reconstructive procedures, which often represent longer operations with extensive muscle exposure and tissue devitalization. This is further compounded by a patient population who are typically older with multiple medical comorbidities, and who may be predisposed to poor wound healing and infection.

Concern for use of instrumentation in the setting of infection arises from the risk of bacterial colonization of the implant. With traditional stainless steel implants, a biofilm harboring bacteria develops around the material. This colonized surface layer is poorly penetrated by antibiotics, resulting in persistent infection. Laboratory studies, however, suggest that titanium implants may be less susceptible to bacterial seeding than stainless steel. Titanium, especially when smooth polished, may be more resistant to bacterial adherence than other materials.

As a result, recently there has been increasing use of titanium-based spinal instrumentation in the surgical treatment of spinal osteomyelitis. The growing popularity of spinal instrumentation stems from an improved ability to restore sagittal alignment, maintain spinal stability, protect neurological function, and relieve pain. Spinal instrumentation also serves to reduce the risk of graft extrusion and to facilitate arthrodesis through rigid immobilization. Additionally, with internal spinal fixation, early patient mobilization is possible, thereby reducing the risk of complications associated with prolonged bed rest and use of external orthoses. Because of the general concern for increased infection with instrumentation, however, the indications for instrumented spinal stabilization in the setting of spinal infection must be clearly established prior to use.

Spinal instrumentation to treat vertebral osteomyelitis has evolved dramatically. Pedicle screw-rod fixation has become a standard method for supplemental posterior stabilization. More recently, device technology has developed titanium cages as a method for rigid anterior column support. Titanium cages offer several significant advantages over traditional anterior strut grafting with either tricortical iliac crest or rib autograft. Titanium cages provide immediate stability, and because of their rigidity can tolerate compressive forces. Titanium cages can be tailored for size and shape, and have a broad contact area for load distribution. They also have significant interface strength between the implant and the vertebral endplates to prevent extrusion or displacement. Newer expandable cages are designed to be inserted and then increased longitudinally in situ, thereby exerting corrective forces to restore sagittal alignment. Titanium cages also are engineered with a hollow mesh design to allow for packing of morselized bone graft within the cage and for bony ingrowth during the healing and fusion process.

Ruf et al examined 88 patients with vertebral osteomyelitis treated with anterior column reconstruction with a titanium mesh cage.4 Thirty-four cases involved placement of the cage in a single disc space. Twenty-eight cases replaced a single vertebral level. Twenty-three cases were two-level vertebral body replacements, and three cases involved three-level reconstructions. Kyphosis angles at the affected levels improved a mean 11.2° after surgery, with a minimal loss of correction of only 1.4°. Four patients with osteoporosis had evidence of cage settling, with three cases requiring revision surgery and additional posterior instrumentation. All patients demonstrated solid arthrodesis at last follow-up, with no recurrent infection.

Pee et al retrospectively reviewed 60 patients who underwent anterior debridement, posterior stabilization, and anterior column reconstruction with either tricortical iliac strut, titanium cage, or polyether ether ketone (PEEK) cage5. The titanium and PEEK cages were packed either with allograft bone chips, with autograft, or with mixed autograft/allograft for arthrodesis. Pee et al observed that the tricortical iliac strut group had an average 200 ml increase in operative blood loss compared to the titanium or PEEK cage groups. While there was no significant difference in postoperative fusion rate between iliac strut and cage groups, there was a significantly higher subsidence rate in the iliac strut group. Also, the mean time interval until subsidence was shorter in the autograft group compared to the cage group. They secondarily observed that patients with subsidence (regardless of graft type) had more pain and disability than those without subsidence. Therefore, the authors inferred that use of a cage may decrease the risk of this adverse outcome, although this was not proven to statistical significance in their series. Most relevant, however, was that all patients, regardless of iliac strut or cage, had normalization of ESR and CRP postoperatively, with complete resolution of infection at final follow-up.

While some may still have concern about the use of instrumentation in the setting of spinal infection, there is clearly a growing body of evidence that, titanium-based implants can be reasonably used with safety. Particularly, given the benefits of instrumentation with regards to spinal reconstruction, sagittal plane correction, stabilization, and early patient mobilization, a serious recommendation for use of instrumentation to treat vertebral osteomyelitis in select cases can be made.

Graft Type

The selection of graft type for arthrodesis in the setting of spinal infection is also controversial. The gold standard remains autologous bone graft, due to its ideal osteobiologic properties. Fresh autologous tissue also may allow rapid vascular ingrowth for effective antimicrobial delivery to the affected site and prevent the risk of persistent bacterial colonization. For this reason, vascularized grafts such as rotational rib grafts or free vascularized fibular grafts may be necessary for complicated revisions due to persistent infection or pseudarthrosis, although obtaining these grafts is technically demanding and time-consuming. Because of the morbidity associated with harvesting autologous bone graft, an alternative option is the use of allograft. Although allograft represents a devascularized foreign body, more recent studies have found that use of allograft struts in patients treated for vertebral osteomyelitis is otherwise safe and can be effective for arthrodesis.

A potentially exciting option is the use of bone morphogenetic protein (BMP) to promote fusion. Recombinant human BMP (InFuse, Medtronic, Memphis, TN, USA; OP-1, Stryker Biotech, Hopkinton, MA, USA) is a synthetic osteoinductive agent that has been demonstrated in both animal and clinical models to result in increased fusion rates. The use of BMP in the setting of spinal infection, however, has not been widely clinically explored and currently represents an off-label and contraindicated use of the product. Laboratory studies in animal models, however, show that BMP retains its osteoinductive properties even in the setting of acute or chronic infection. Interestingly, in experimental models, BMP in combination with antibiotics results in more rapid healing than BMP alone. This may be secondary to increased angiogenesis caused by BMP-stimulated osteoblast-derived vascular endothelial growth factor. Therefore increased vascular ingrowth not only facilitates osteogenesis, but also leads to increased local antibiotic delivery to better eliminate infection.

Limited clinical studies have investigated the use of BMP to promote fusion in patients with vertebral osteomyelitis. However, a handful of studies evaluating the use of BMP placed either in structural allograft or in a titanium cage with supplemental spinal fixation demonstrate successful bony fusion, without recurrence of infection or evidence of complication related to BMP use. While the use of BMP in the setting of spinal infection is still not FDA approved, there is a small body of preclinical and clinical evidence to suggest that BMP may be beneficial in promoting early fusion in vertebral osteomyelitis. Certainly, further testing is warranted prior to any recommendation for the clinical use of BMP for spinal infection can be made.

Minimally Invasive Surgery

In many instances, spinal infections are successfully treated with conservative medical therapy. When surgical intervention is required, however, operative treatment often requires extensive procedures consisting of radical debridement with spinal reconstruction and stabilization. Perioperative morbidity in the elderly population or in patients with significant medical comorbidities is of particular concern. Significant complications after complex spinal instrumentation procedures in patients with spinal infections are reported to be as high as 47%. Recently, minimally invasive surgical techniques have been developed as an alternative surgical modality for the treatment of a variety of spinal disorders. These techniques, coupled with novel device technology, have provided measures for performing many of the same types of open decompressive and fusion procedures, albeit through tissue-sparing approaches. As a result, decreased blood loss, less postoperative pain, shorter hospitalization, and earlier return to function have been observed.

Thoracoscopic Spinal Surgery

An anterior approach for vertebral osteomyelitis allows for direct visualization and access to the primary pathology. Anterior exposure of the thoracic and thoracolumbar spine via a conventional open approach, however, requires a thoracotomy and potential splitting of the diaphragm. Significant morbidity is associated with a standard thoracotomy, including chronic postoperative pain and respiratory compromise. To circumvent this issue, video-assisted thoracoscopic surgical techniques have been applied for minimally invasive treatment of thoracic spinal disease. While thoracoscopic spinal surgery is already widely used for basic procedures such as thoracic discectomy and sympathectomy, more recently this technique is being incorporated in operations to treat more complex pathology like scoliosis, trauma, tumors, and infection. Through multiple ports, exposure of the thoracic and thoracolumbar junction as well as anterior debridement, partial corpectomy, interbody cage placement, and spinal stabilization are possible.

The literature reporting thoracoscopic treatment of vertebral osteomyelitis is limited. Muckely et al described three patients with thoracic osteomyelitis that underwent thoracoscopic partial corpectomy, anterior reconstruction, and anterior spinal fixation.6 The improvement in kyphotic angle among the three patients ranged from 6° to 15° with no evidence of loss of correction at a minimum of 22 months of follow-up. There were no cases of recurrence and no instances of graft or hardware failure. Of note, one patient in the series was ambulating as early as postoperative day one. Amini et al presented a case report of a 70-year-old patient who developed osteodiscitis after a T11-12 discectomy.7 The patient was treated with a thoracoscopic vertebrectomy, allograft strut, and anterior instrumentation. At 1-year follow-up, the patient demonstrated solid fusion without evidence of recurrent disease.

Percutaneous Technology

Open posterior exposure of the spine consists of a midline incision with dissection of the musculature away from the bony elements. This approach allows for access to the dorsal spine for decompression as well as the necessary anatomy for placement of instrumentation and fusion bed preparation. Extensive muscle dissection and prolonged retraction, however, can result in tissue ischemia, denervation, scarring, and postsurgical dead space with increased risk of blood loss, infection, chronic pain, and delay to functional recovery.

Recently, percutaneous technology has been developed to perform a variety of spinal procedures including discectomy, spinal decompression, interbody fusion, and instrumented stabilization. These minimally invasive techniques have been incorporated in the treatment of spinal infections as well. Nagata et al applied a technique for percutaneous excision of lumbar disc herniations as a method for aspiration and drainage of pyogenic spondylodiscitis.8 Under local anesthesia, a percutaneous trocar that is 5.4 mm in diameter is inserted under intraoperative fluoroscopy into the affected level. Through this trocar, specialized forceps and a motor-driven shaver are used to curette the infected disc and endplate, which are then removed piecemeal. After debridement, large volume irrigation is flushed through the trocar. Finally, a small suction drainage tube is left in the disc space and the trocar is removed to allow for postoperative continuous antibiotic suction-irrigation.

Nagata et al performed this procedure in 23 patients with spondylodiscitis.8 The causative organism was identified through cultured tissue removed during curettage in 53% of patients. Ninety-one percent of patients had immediate relief of back pain after surgery, with 43% ambulating without pain within 3 days of surgery. All patients were followed for a minimum of 2 years with only one patient requiring a repeat operation for recurrent infection. No vascular or neurological complications were encountered as a result of the procedure. Three of 6 patients with preoperative neurological deficits, however, continued to have mild sensory impairment at last follow-up, and therefore, this procedure is not recommended for patients with significant bony destruction, epidural abscess, or neurological compromise.

Instrumentation for spinal fixation of unstable pathologic fractures and deformity correction has also been advanced by percutaneous technology. Standard open placement of pedicle screws requires extensive dissection of the posterior musculature to expose the necessary anatomic landmarks for screw insertion and connecting rod placement. Cannulated pedicle screws, however, have been introduced that allow for placement of screws over a guidewire percutaneously inserted under fluoroscopic imaging. With this technique, multilevel fixation can be performed with multiple separate stab incisions for each screw placement (Figure 47-8 A-E). Novel technology has been developed to allow for introducing a connecting rod through the screw heads via an additional separate stab incision. Due to its low operative morbidity, percutaneous stabilization may have a particularly beneficial role as supplemental posterior fixation for patients undergoing primary anterior debridement and spinal reconstruction.

Prognosis

With earlier diagnosis and better medical and surgical intervention, clinical outcomes and prognosis from spinal infection are improving. Mortality rates from pyogenic osteomyelitis were once as high as 25% to 71%. Depending on patient age and comorbidities, mortality rates for treated pyogenic osteomyelitis are now as low as 5% to 16%. Ninety-one percent of patients are estimated to recover uneventfully with either medical therapy or combined medical and surgical intervention. Similar improvement in mortality has been observed for patients with epidural abscesses. Dandy in 1926 reported a mortality rate of 83% for patients with spinal epidural abscesses. Now, with prompt surgical intervention and improved antibiotic therapy, the mortality rate ranges from 5% to 32%.

The prognosis for neurological recovery generally depends on the duration and severity of neurological impairment prior to intervention. Patients with epidural abscesses that are treated within 24 hours of onset of deficits have a better prognosis for recovery of function. Rigamonti et al observed that only 10% of patients with severe neurological deficits who were treated within 24 hours of onset of symptoms had poor neurologic outcomes, compared to 47% of patients treated more than 24 hours after onset of symptoms having a poor neurological outcome.9 Additionally, patients with complete paralysis, especially if present for more than 36 hours, do not generally recover, despite any intervention.

Chronic pain is a potential long-term complication associated with osteomyelitis, and may be multifactorial. Some studies have found that patients that underwent surgical intervention are actually less likely to have chronic back pain than those treated with antibiotics alone. Better restoration of sagittal alignment and spinal stabilization with surgery may account for this difference in outcome. However, 36% of patients treated only with medical therapy do recover without any long-term disabling back pain. This observation may be due to less severe bony destruction in patients treated nonsurgically, or to spontaneous fusion that occurs from the inflammatory response. Spontaneous bony ankylosis forms in 35% of patients; this, however, may require 6 to 24 months to occur. Deformity is another potential complication that may contribute to pain and long-term dysfunction. Deformity is more common with spinal tuberculosis, especially when occurring at the thoracic or thoracolumbar spine, or when involving more than 50% of one or more vertebral bodies. Good clinical outcomes, however, are demonstrated with surgical intervention. With an anterior decompression and fusion, 94% of patients with spinal tuberculosis recover normal neurological function, with a fusion rate of 92% at 5 years.

Conclusion

The reportedly growing number of spinal infections may become an increasingly significant health care problem. Spinal infections are predisposed to occur in the elderly and those who are medically compromised. As a result, spinal infections are often a complex medical condition with multiple contributing factors, and are therefore challenging to manage. The importance of successfully treating spinal infections is all the more relevant given the potential for significant associated medical complications, neurological compromise, functional impairment, and chronic disability.

Spinal infections represent a wide-ranging spectrum of pathologic involvement and therefore are not amenable to simple treatment algorithms or protocols. General principles dictate that early diagnosis with identification of the pathogenic organism is critical for successful medical therapy with eradication of infection and minimized complications. Patients who are neurologicalally intact and clinically stable can generally be managed nonsurgically. Patients who present with neurological deficits, clinical deterioration despite medical therapy, or evidence of spinal instability, deformity, or chronic pain may necessitate surgical intervention. Currently, a variety of surgical treatment options are available ranging from minimally-invasive techniques to radical debridement with complex spinal reconstruction and instrumented stabilization.

Developments in diagnostic imaging, antimicrobial therapy, and surgical techniques are advancing our therapeutic capabilities and improving clinical outcomes. Better understanding of the disease response to surgical interventions, particularly the use of instrumentation and osteobiologic agents, are reshaping current treatment paradigms. As a result, we are witnessing greatly reduced morbidity and mortality. With a multidisciplinary approach committed to early and aggressive management of spinal infections, we can continue to expect even better patient outcomes.

References

1. Dai L.Y., Chen W.H., Jiang L.S. Anterior instrumentation for the treatment of pyogenic vertebral osteomyelitis of thoracic and lumbar spine. Eur. Spine J.. 2008;17(8):1027-1034.

2. Korovessis P., Repantis T., Iliopoulos P., Hadjipavlou A. Beneficial influence of titanium mesh cage on infection healing and spinal reconstruction in hematogenous septic spondylitis: a retrospective analysis of surgical outcome of twenty-five consecutive cases and review of literature. Spine. 2008;33(21):E759-E767.

3. Dimar J.R., Carreon L.Y., Glassman S.D., Campbell M.J., Hartman M.J., Johnson J.R. Treatment of pyogenic vertebral osteomyelitis with anterior debridement and fusion followed by delayed posterior spinal fusion. Spine. 2004;29(3):326-332. discussion 32

4. Ruf M., Stoltze D., Merk H.R., Ames M., Harms J. Treatment of vertebral osteomyelitis by radical debridement and stabilization using titanium mesh cages. Spine. 2007;32(9):E275-E280.

5. Pee Y.H., Park J.D., Choi Y.G., Lee S.H. Anterior debridement and fusion followed by posterior pedicle screw fixation in pyogenic spondylodiscitis: autologous iliac bone strut versus cage. J. Neurosurg. Spine. 2008;8(5):405-412.

6. Muckley T., Schutz T., Schmidt M.H., Potulski M., Buhren V., Beisse R. The role of thoracoscopic spinal surgery in the management of pyogenic vertebral osteomyelitis. Spine. 2004;29(11):E227-E233.

7. Amini A., Beisse R., Schmidt M.H. Thoracoscopic debridement and stabilization of pyogenic vertebral osteomyelitis. Surg. Laparosc. Endosc. Percutan. Tech.. 2007;17(4):354-357.

8. Nagata K., Ohashi T., Ariyoshi M., Sonoda K., Imoto H., Inoue A. Percutaneous suction aspiration and drainage for pyogenic spondylitis. Spine. 1998;23(14):1600-1606.

9. Rigamonti D., Liem L., Sampath P., et al. Spinal epidural abscess: contemporary trends in etiology, evaluation, and management. Surg. Neurol.. 1999;52(2):189-196. discussion 97