review articles a surgical revisitation of pott distemper...

The Spine Journal 3 (2003) 130–145

1529-9430/03/$ – see front matter © 2003 Elsevier Science Inc. All rights reserved.PII: S1529-9430(02)00410-2

Review Articles

A surgical revisitation of Pott distemper of the spine

Larry T. Khoo, MD, Kevin Mikawa, MD, Richard G. Fessler, MD, PhD*

Institute for Spine Care, Chicago Institute of Neurosurgery and Neuroresearch, Rush Presbyterian Medical Center, Chicago, IL 60614, USA

Received 21 January 2002; accepted 2 July 2002

Abstract

Background context:

Pott disease and tuberculosis have been with humans for countless millennia.Before the mid-twentieth century, the treatment of tuberculous spondylitis was primarily supportiveand typically resulted in dismal neurological, functional and cosmetic outcomes. The contemporarydevelopment of effective antituberculous medications, imaging modalities, anesthesia, operativetechniques and spinal instrumentation resulted in quantum improvements in the diagnosis, manage-ment and outcome of spinal tuberculosis. With the successful treatment of tuberculosis worldwide,interest in Pott disease has faded from the surgical forefront over the last 20 years. With the recentunchecked global pandemic of human immunodeficiency virus, the number of tuberculosis and sec-ondary spondylitis cases is again increasing at an alarming rate. A surgical revisitation of Pott dis-ease is thus essential to prepare spinal surgeons for this impending resurgence of tuberculosis.

Purpose:

To revisit the numerous treatment modalities for Pott disease and their outcomes. Fromthis information, a critical reappraisal of surgical nuances with regard to decision making, timing,operative approach, graft types and the use of instrumentation were conducted.

Study design:

A concise review of the diagnosis, management and surgical treatment of Pott disease.

Methods:

A broad review of the literature was conducted with a particular focus on the differentsurgical treatment modalities for Pott disease and their outcomes regarding neurological deficit, ky-phosis and spinal stability.

Results:

Whereas a variety of management schemes have been used for the debridement and recon-struction of tuberculous spondylitis, there has also been a spectrum of outcomes regarding neurolog-ical function and deformity. Medical treatment alone remains the cornerstone of therapy for the ma-jority of Pott disease cases. Surgical intervention should be limited primarily to cases of severe orprogressive deformity and/or neurological deficit. Based on the available evidence, radical ventraldebridement and grafting appears to provide reproducibly good long-term neurological outcomes.Furthermore, recurrence of infection is lowest with such techniques. Posterior operative techniquesare most effective in the reduction and prevention of spinal deformity.

Conclusions:

Unlike historical times, effective medical and surgical management of tuberculousspondyitis is now possible. Proper selection of drug therapy and operative modalities, however, isneeded to optimize functional outcomes for each individual case of Pott disease. © 2003 ElsevierScience Inc. All rights reserved.

Keywords:

Pott disease; Tuberculosis; Spondylitis

Introduction

Tuberculosis has and continues to be the world’s mostprevalent and lethal infectious disease with over 3.5 milliondeaths yearly [1–6]. More individuals died from tuberculo-

sis in 1996 than in any previous year in history [7]. Thereare presently over 40 million active cases of tuberculosisglobally with 8 to 10 million new cases expected each year[8]. Based on a 1% to 2% incidence of skeletal involvement,World Health Organization figures estimate that there areover 800,000 active cases of tuberculous spondylitis [8,9].Within the United States itself, the prevalence of tuberculo-sis increased from 1986 to 1992. Increasing populations ofelderly nursing home patients, the homeless, SoutheastAsian and Central American immigrants and patients in-fected with the human immunodeficiency virus (HIV) have

FDA device/drug status: not applicable.Nothing of value received from a commercial entity related to this

research.* Corresponding author. Chicago Institute of Neurosurgery and Neu-

roresearch, 2515 North Clark Street, Suite 800, Chicago, IL 60614, USA.Tel.: (773) 388-7759; fax: (773) 935-2132.

E-mail address

: [email protected] (R.G. Fessler)

L.T. Khoo et al. / The Spine Journal 3 (2003) 130–145

131

all contributed to this increasing prevalence [10–14]. De-spite improved screening and treatment programs, tubercu-losis persists as a significant health problem in the UnitedStates [6,8]. In African nations, the reported rate of caseshas increased by over 200% in the last 5 years becauseof the region’s unchecked HIV pandemic [3,15]. In sub-Saharan Africa alone, conservative estimates show that thereare approximately 25 million cases of active HIV with up to50 million cases worldwide [16]. Of HIV cases in develop-ing nations, approximately 10% to 15% will develop activetuberculosis. Up to 60% of these HIV-positive tuberculosispatients will have skeletal involvement, a figure that is fargreater than the usual 1% to 2% incidence seen in HIV-neg-ative patients [17,18]. From these epidemiological data, it isclear that a global resurgence of Pott disease is upon us.

Historical perspectives

“To attend to a distemper from its beginning through along and painful course, to its last fatal period, without evena hope of being able to do anything which shall be reallyserviceable, is of all tasks the most unpleasant” (Sir PercivalPott, 1779)[19]. Although Dr. Pott is credited with this earlygrim description of spondylitis, tuberculosis as a disease hasafflicted humans since the dawn of time [19]. From theArene Candide cave in Liguria, Italy, the remains of a 15-year old male buried in the first half of the fourth millen-nium BC were found to have tubercular lesions of the ver-tebral column [20]. Similarly, evidence of tuberculousspondylitis has been found in the mummified remains ofboth Egyptian kings and slaves alike from as early as 3400BC [21]. In the Sanskrit writings of ancient India, numerousreferences to tuberculosis under the eponym “Yakshma” arereplete in the Rig, Atharva Vedas (3500 to 1800 BC) andSusruta Samhitas (1000 to 600 BC) [21,22]. It is humblingto note that the Vedic physicians’ prescription of high alti-tude, fresh air, rest, immobilization and constitutional sup-plementation was the same as that used by “modern” West-ern doctors until the 1940s [23].

One millennium later, Hippocrates in his work,

On Artic-ulations

, vividly described both the deforming kyphus anddraining fistulae in patients afflicted by “Pott’s Distemper.”Galen subsequently recognized the relationship betweenthese tubercular changes and spinal deformity and pub-lished his observations in the second century AD [24]. Assuch, both Hippocratic and Galenic physicians used a novel“extension” bench to control spinal deformities throughmanual reduction techniques (Fig. 1). In their respectivesixteenth- and seventeenth-century publications, Diechampin France and Severino in Italy both postulated that the pro-gressive spinal deformity was responsible for the paraplegiaobserved in these patients [8]. With the advent of germ the-ory, Robert Koch correctly identified

M. tuberculosis

in1892 as the microbial culprit responsible for this wake ofhuman suffering [23]. In 1944, the introduction by Schatzand Waksman of streptomycin emancipated thousands from

the sanitoriums and also made surgical intervention feasibleby lowering the mortality rate [25–29]. In an effort to assessthe efficacy of these new medical and surgical treatments,the Medical Research Council (MRC) was founded in 1963to direct prospective research on the epidemiology, diagno-sis and treatment of tuberculosis [26,30–37].

Pathophysiology

Tuberculosis is a chronic, insidious and recurrent bacte-rial infection caused by

M. tuberculosis

,

Mycobacteriumbovis

or

Mycobacterium africanum

. M tuberculosis is themost common cause of tuberculosis and appears as an“acid-fast bacillus” with appropriate staining on micro-scopic examination. The lungs are most commonly infectedbecause

M tuberculosis

is a strict aerobe and strongly de-

Fig. 1. For centuries, many of those who survived their bout with tubercu-losis would suffer from chronic gibbus deformities. Treatment of thekyphus was limited to primarily external manual reduction maneuversincluding this example of a novel extension, traction-type bench utilized byHippocratic school physicians of the middle-ages [32].

132


pendent on oxygen tension. Miliary spread of the bacilliwith tuberculoma formation can, however, occur in virtu-ally any organ [1,2,38–40]. The gross pathologic finding intuberculous spondylitis is one of granuloma formation con-sisting of exudative granulation tissue and interspersed ab-scesses. As the abscesses coalesce, regions of caseating ne-crosis result and appear as a yellow, opaque, “cheesy”substance. Microscopic examination of the granuloma re-veals a nodular pattern with central necrosis. Langerhans gi-ant cells with peripheral nuclei are common, although notpathognomonic [41]. Special staining of adequate patho-logic samples may expediently reveal a causative acid-fastbacilli in the case of

M. tuberculosis

or other fungus [17].Tuberculosis infections of the spine are usually the result

of seeding from an extraspinal source (eg, pulmonary infec-tion) and commonly manifest as paradiscal, anterior or cen-tral lesions [17]. In adults, there are four primary patterns ofinvolvement: 1) paradiscal, 2) anterior granuloma, 3) cen-tral lesions and 4) appendiceal type lesions [8,9,42]. Ofthese, the paradiscal lesion is the most common and repre-sents approximately half of all cases [41]. In the paradiscallesion, the primary focus of infection is in the vertebralmetaphysis. The enlarging granuloma subsequently erodesthrough the cartilaginous end plate and then begins to nar-row the disc space (Fig. 2). Anterior granulomas developbeneath the anterior longitudinal ligament and can span sev-eral vertebral segments. Although anterior granulomas typi-cally demonstrate less bony destruction and deformity thanparadiscal or central lesions, elevation of the periosteum canresult in bone devascularization with subsequent develop-ment of an abscess, necrosis and deformity. In central le-sions, the entire vertebral body is involved, thereby causingsignificant deformity as the body loses structural integritycausing one or more pathologic fractures. Patients with cen-tral lesions will commonly have two or three affected verte-brae. Infections of the atlas, axis and isolated neural archesare less frequent but more likely to cause neurological defi-cits [43,44]. Skip lesions in different stages of developmentoccur in approximately 10% of patients [41]. Lastly, the ap-pendiceal type of lesion propagates in the laminae, pedicles,articular facets and spinous processes, causing initial expan-sion followed by subsequent rupture and failure [8].

Abscesses, commonly found in spinal tuberculosis, followtissue planes but may extend into the spinal canal at any level.The periphery of the abscess can form thin or dense adhe-sions with visceral and vascular structures. Further, these le-sions may cause symptoms resulting from neurovascularcompression (eg, pain), hemorrhage and direct mass effect[4,20,21,26,30–37,39,45–48]. In the neck, abscesses are morecommon in children than adults. Such lesions are often symp-tomatic because of compression of the trachea and esophagus[49]. In the thorax, the abscess may invade the lung and cre-ate pleural and/or diaphragmatic adhesions with the lung. In-vasion of the femoral trigone occurs when the abscess travelsdown the psoas sheath from the lumbar region. Abscesses inthe sacral region can also invade the perineum or the gluteal

region through the greater sciatic foramen. A rupture of anabscess through the skin may create a draining sinus, therebyincreasing the risk of superinfection.

Fig. 2. Example of a tuberculoma of the lumbar spine at L4 and L5. Thesagittal gadolinium-enhanced T1-MRI image demonstrates spondyliticdestruction with increased signal and enhancement of the vertebral bodiesand disc space (A). Lateral and AP radiographs show the bony erosion andkyphoscoliotic deformity with intervertebral collapse and tilt that is typicalof the early gibbus (B, C). Plain radiographs, however, often underestimatethe full extent of the infection. Often, contrast-enhanced CT or MR imagesare needed to completely define the degree of tuberculous injection in theparavertebral soft tissue (D, E).


133

Neurological deficits are a consequence of 1) direct neuralcompression by granulomas, abscesses, sequestra of vertebralbodies and discs (extra- and intradural extramedullary tuber-culomas), 2) direct invasion of the neural parenchyma (in-tramedullary intradural tuberculoma), 3) tuberculous meningi-tis, 4) pathological dislocation and subluxation of bonyelements, 5) vascular compromise (thrombosis, direct pres-sure) [8,17,24,38,50–52]. The granulomas can create neuro-logical compromise by transdural compression of neural ele-ments or by invading the dura. The majority of suchcompressive lesions are secondary to an admixture of purulentmaterial and osseous elements [53]. However, pure granulo-matous or abscess-only compression of neural elements with-out an associated bony lesion has also been reported [54,55].

Intramedullary tuberculomas, first described in 1830 bySere, are much rarer than their intracranial counterparts buthave the same histological appearance of central caseationsurrounded by an often intense pachymeningitic arachnoiditis[8,51]. Because of the predominance of anterior spinal ele-ment involvement, ventral neural compression is most com-mon, but isolated or concomitant posterior compression mayoccur. Also, abscesses or the sequestra of vertebral body ordisc in the epidural space may directly press on neural ele-ments resulting in neurological deficits. Tuberculous spinalmeningitis may present in three main forms: 1) tuberculomaarising in the pachymeninges, 2) secondary extension of in-tracranial basal meningitis and 3) secondary extension fromtuberculous meningitis [52]. Thick exudates are typicallyseen surrounding the meninges, cord and nerve roots with in-termingled fibrous bands of granulation tissue. At times, cys-tic formation with pockets of cerebrospinal fluid can occur asa result of these changes [56]. Vascular ischemia, infarction,delayed myelomalacia, gliosis and syringomyelia may alsoresult [57]. Isolated pathologic dislocation or subluxation ofthe bony elements is an infrequent cause of neurologicalcompromise [58–61]. In the setting of apparent disease quies-cence, late paraplegia may occur because of such vertebralbody failure. Paraplegia is far more common in tuberculosisthan in other forms of pyogenic spondylitis because of the in-creased frequency of neural arch involvement and of patho-logical fracture with neural compression [62,63].

Spinal lesions comprise 50% of cases of articuloskeletaltuberculosis, which comprise 3% of tuberculosis cases inHIV-negative and 60% of HIV-positive cases [8,9,17,18]. Ofspinal tuberculomas, 5% are in the cervical spine, 75% in thethoracic spine and 20% in the lumbar and sacral spine. Ky-photic deformity results from the collapse of anterior verte-bral elements and is most commonly seen in thoracic lesions.The loss of one vertebral body may produce as much as 32degrees of kyphosis, but often multiple levels are affected[39,49,64–66]. Kyphotic deformity progresses until the col-lapsing vertebral bodies meet anteriorly or the granulatingand caseating tissue mature into bone. By the first 18 monthsafter medical treatment, most of the vertebral collapse has oc-curred, and the ultimate degree of kyphotic deformity is pro-portional to the initial loss of the vertebral bodies as opposed

to the initial degree of radiographic kyphosis [17]. With cir-cumferential erosion of both the posterior arch and anteriorbody, translational instability can occur with resultant ante-rior, lateral, rotatory subluxation and a significantly increasedrisk of neurological compromise [3,62].

Clinical presentation

The clinical presentation of a particular patient will varydepending on the age and health status of the patient, the lo-cation of the infection, the stage of the disease and the ab-sence or presence of abscesses, sinus tracts and neurologicalcompromise. Systemic symptoms of weight loss, fever andmalaise usually precede spinal involvement for severalmonths. When the spine is involved, the most common pre-senting complaint is one of pain with an insidious onset asopposed to complaints of acute pain with pyogenic infec-tions [60]. Cervical involvement may present with com-plaints of torticollis, neck pain and stiffness, stridor anddysphagia. Neurological compromise reported in 10% to61% of patients with spinal tuberculosis may be manifestedas somatosensory changes, weakness and changes in boweland bladder function [60]. In underdeveloped countries,paraplegia, kyphosis, deformity and draining sinuses arecommon presenting complaints because of poor health-careaccess and patient neglect [32,67]. Even in industrializednations, the diagnosis of tuberculosis can be delayed for upto 2 years after the onset of symptoms [58]. A completephysical and ophthalmological examination may yield sys-temic signs of mycosis, spinal alignment deformity, subcu-taneous flank masses and evidence of cutaneous fungal in-fections and/or sinuses. The neurological examinationshould assess the function of cranial nerves, somatosensoryand motor systems and the rectal sphincter, and determinethe presence of myelopathy (ie, hyperreflexia, sustainedclonus and pathological reflexes, such as Babinski andHoffman signs). Classical physical findings include localtenderness, muscle spasm, restricted spinal motion, defor-mity and neurological deficit [3,4,26,30–37,46–48,62,68–72]. Pain is an excellent localizing sign with the most com-mon locations being the thoracic spine followed by the lum-bar, cervical and sacral regions [67,73]. Paraplegia is mostoften from thoracic and cervical spondylytic destruction[49,73]. Abscesses and fistulae are found in the groin andbuttocks and can be mistaken for other types of anorectalfistulae [38].

Clinical subgroups

In younger children, far more extensive disease with mil-iary spread is often found at the time of diagnosis. Al-though larger abscesses are more frequently encountered,there is a relatively low incidence of paraparesis (15% to20%) and paralysis in pediatric cases. This is in direct con-trast to the adult subgroup that has far more localized dis-ease but a much higher incidence of paraplegia (over 80%)

134


[1,25,45,49,67,74–76]. In the United States, Europe andMiddle East, where disease prevalence is relatively low,older adults are more commonly infected than children.However, in areas of higher prevalence (eg, Asia and Cen-tral America) children are more commonly infected thanadults [3,4,26,30–37,46–48,62,68–72]. In western coun-tries, spinal involvement is typically less severe with fewerneurological symptoms and a milder degree of deformity[6,75,77]. Also spondylitis without disc involvement hasbeen increasingly reported in industrialized nations in con-trast to the classic spondylodiscitis typically seen elsewhere[6], presumably because of better access to care and earlydiagnosis.

In immunocompromised patients with advanced HIV dis-ease, the clinical presentation can be masked by numerousother HIV-associated changes in the central and peripheralnervous system, including vacuolar myelopathy of the spinalcord, demyelinating neuropathy, dementia and coexistent in-fection with other organisms (cytomegalovirus, herpes, vari-cella, Jacob-crutzfeld, treponema, Candida, Cryptococcus,Mucormycosis, Toxoplasma) [18]. Instead of classic paraple-gia, deformity and pain, these individuals may present withprogressive ataxia, dementia, spastic paraparesis, myelopathyand sphincter disturbances [8,78,79]. Lastly, a distinct syn-drome of tuberculous spondylitis has been reported in intra-venous drug addicts, many of whom are HIV positive as well.In these patients, acute fever, intense back pain, weight loss,night sweats and rapidly evolving deficits with diffuse mil-iary spread was observed [24,80].

Evaluation and diagnosis

Laboratory investigations

In the laboratory evaluation the white blood cell count istypically normal and the erythrocyte sedimentation rate(ESR) is elevated above 20 mm/hour in 80% to 100% of pa-tients with spinal osteomyelitis [81]. However, 5% to 10%of children have an ESR greater than 20 mm/hour [82], andhealthy elderly patients and pregnant women have mildlyelevated ESRs [83]. The Mantoux test (tuberculin skin test)is typically positive, but anergic individuals will producefalse-negative results. Serological testing using enzyme-linked immunosorbent assay and agglutination titers canprovide additional information about tuberculosis and fun-gal infections. The definitive diagnosis is made with open orpercutaneous trephine needle biopsy of the bone and discusing computed tomography (CT) or fluoroscopic guidance[84,85]. Cultures and special stains of the biopsied speci-mens will identify the causative pathogen. However, suchisolation of the bacterium from clinical specimens can oftentake several weeks with sensitivities as low as 50% [86].Recent advances in polymerase chain reaction (PCR) tech-niques have opened exciting new avenues of diagnosis andtreatment. By amplifying species-specific DNA sequencesin even formalin-fixed paraffin-embedded samples, PCR is

able to rapidly detect and diagnose several strains of Myco-bacterium without the need for prolonged culture and spe-cial histological staining techniques [87]. Furthermore, PCRtechniques have been used to identify discrete genetic muta-tions in DNA sequences associated with drug resistance[88]. Rifampin resistance has been mapped to a small regionof an RNA polymerase (rPOB) gene, and ciprofloxacin andstreptomycin resistance has been localized to gyrase A (GyrA) and ribosomal protein SR (rPSL) genes [3]. As such,PCR techniques can now be used to specifically select drugtherapy based on the organism’s genetic makeup for eachcase [89].

Radiographic evaluation

The classic case of tuberculous spondylitis is describedas beginning anteriorly and advancing superiorly and inferi-orly to the disc space. However, hematogeneous seeding tothe end-plate subchondral metaphyseal bone is more likelybecause of its rich vasculature [90]. As the disease progresses,paravertebral abscesses develop, the disc space is destroyedand adjacent bodies become rarefied, resulting in kyphoticdeformity as the vertebral body collapses. Imaging studiesare used to visualize this process and include spine andchest plain films, radionuclide scanning, CT and magneticresonance imaging (MRI). The basic imaging analysis in-cludes spine and chest plain films to visualize advanced rar-efaction, bony deformity, disc space narrowing, anteriorvertebral collapse, vertebrae plana, kyphosis and abscesses.Rarefaction, vertebral edge lucency and loss of cortical mar-gins can be seen 2 to 4 weeks after spinal infection and mayinvolve multiple segments. Paravertebral shadows indicat-ing paravertebral swelling usually over two to three seg-ments is often seen as the illness progresses. Abscesses mayappear as widened paravertebral shadows with soft tissuecalcifications and loss of the psoas muscle shadow. Involve-ment of the disc space resulting in disc space narrowing onradiographs is commonly seen before vertebral body col-lapse (Fig. 2, A to C). With plain radiographs, the extent ofthe disease is often underestimated, requiring radionuclidescanning, CT and MRI with contrast enhancement to fullydemonstrate the extent of the tuberculomas and abscesses(Fig. 2, D and E). The differential diagnosis of granuloma-tous disorders of the spine includes sarcoidosis, Candida,Torulopsis, Aspergillus,

Coccidioides immitis

, Blastomyces,Actinomyces [91], Nocardia [91–93], atypical Mycobacte-rium [21,93–96], Brucella [97,98], nonvenereal Treponema(yaws) [99] and Echinococcus (hydatid disease) [100].

Technetium 99m pyrophosphate is the most commonscintigraphic study, because it is the most sensitive. How-ever, it is not specific [101]. Gallium scintigraphy can beperformed in early stages of disease and is useful for wholebody screening. CT is excellent for defining the anatomy ofbony destruction, extension into the spinal canal, posteriorelement involvement and formation of paravertebral ab-scess. As a single modality, however, MRI is the modalityof choice, because it provides superior anatomic detail of


135

the thecal sac, neural structures and paravertebral areas[40]. CT myelography can show similar details, but scintig-raphy cannot. Myelography also allows for collection ofcerebrospinal fluid for analysis to evaluate neurological in-volvement. With gadolinium, MRI can discriminate be-tween abscesses and granulation tissue and can define a softtissue mass and the extent of bony destruction. Noncon-trasted CT scans provide useful and complementing render-ing of bone detail, especially for surgical planning.

Medical management

In general, chemotherapy and adequate rest and nutritionare the cornerstones of therapy for spondylitis caused bytypical and atypical Mycobacterium, fungi, actinomyces,nocardia, brucella, echinococcus and nonvenereal tre-ponema organisms. The overall goals of medical treatmentinclude 1) recovery and/or maintenance of neurologicalfunction, 2) recovery and/or maintenance of mechanicalspine stability, 3) diagnosis and elimination of the causativepathogen and 4) functional return to activities of daily livingas soon as possible. Medical and conservative managementalone for cases without significant paraplegia, deformity orinstability has yielded good results in countless patients. InTuli’s classic series of over 200 cases without neurologicaldeficit, resolution of disease with good outcomes was ob-tained in 94% of patients with antibiotic treatment alone [9].This “middle-path” of reserving surgery only for medicalfailures is further substantiated by findings of the MRC,who found 93% favorable outcomes in patients treated withonly a 6- or 9-month regimen of antibiotics [46,47]. How-ever, the true meaning of “favorable” must be interpreted inlight of the health-care resources available and the func-tional expectations of the patient and physician.

Antimicrobial regimens

The modern era of antituberculous therapy began in 1943with Waxsman’s discovery of streptomycin [28]. Within 5years after its introduction, mortality rates in conservativelymanaged patients decreased by nearly 70% across the globe[102]. Isoniazid (INH) and p-aminosalicylic acid (PAS) weresubsequently introduced in 1952 with excellent clinical effi-cacy in eliminating the risk of disease dissemination andpreventing chronic abscess and sinus formation [103].These drugs were also successfully used without surgerywhen patients were kept immobilized in orthoses at bed restfor long periods [104].

A second major evolution in tuberculosis treatment cameabout in Nigeria when a shortage of beds forced physiciansto use INH and PAS on an ambulatory basis. Although cos-metic deformity was problematic, over 96% of their patientswere cured with this regimen [105]. The MRC subsequentlyconducted controlled prospective studies on the effect of thelength of bed rest, plaster of Paris jacket orthoses and theaddition of streptomycin to standardized INH and PAS regi-

mens. The 5- and 10-year results confirmed the effective-ness of the ambulatory regimen and could demonstrate nobenefit of the other adjunctive measures [34,36]. With theaddition of rifampin, the MRC altered their recommenda-tions to the following front-line agents: INH, rifampin,pyrazinamide, streptomycin and ethambutol. Second-lineagents include ethionamide, cycloserine, kanamycin, capre-omycin and PAS. The latest MRC studies have demon-strated that isoniazid and rifampin given for 6 to 9 monthswas as effective as isoniazid and paraaminosalicylic acid orethambutol given for 18 months [31,33]. However, signifi-cant controversy continues to surround the optimal length oftherapy. In particular, several authors have criticized thelarge degree of bone resorption and kyphosis seen with onlyshort-course 6- and 9-month regimens [106–108]. To pre-vent recurrence, prevent delayed kyphosis and to eradicate apaucibacillary disease with slow-growing bacilli, triple-drug therapy (INH, rifampin, ethambutol) for 18 months orfour-drug therapy (add pyrazinamide) has been advocatedby these groups. For the cases typically seen in Western in-dustrialized nations, a 6-month, three-drug regimen withINH, rifampin and pyrazinamide should be used for mostpatients with drug-sensitive infection [109].

Isoniazid is administered at a typical dose of 5 mg/kg/day up to 300 mg/day and should be supplemented with py-ridoxine to prevent peripheral neuropathy. Rifampin 10 mg/kg/day up to 600 mg/day is concurrently given with iso-niazid. Pyrazinamide should also be prescribed at a dose of15 to 30 mg/kg/day up to 2.0 g/day for up to 2 months onlyto prevent hepatotoxicity [4,26,30–37,46–48]. Becausepain, inflammation and bone resorption are of particularconcern during the initial phase of therapy, several groupshave advocated the use of nonsteroidal anti-inflammatoryagents. The antiprostaglandin effect of these agents isthought to decrease nonspecific synovial inflammation andbone resorption as well [62,69,110]. Baseline laboratory val-ues for hepatic enzymes, bilirubin, serum creatinine, BloodUrea Nitrogen (BUN), Complete Blood Count (CBC), andplatelets should be obtained initially, and these values maybe monitored at 1 or 3 months or more frequently dependingon particular clinical situations. Directly observed therapy isan important component of medical management, because pa-tient nonadherence to a prescribed treatment is a major causeof therapeutic failure and of the emergence of resistant or-ganisms. In this multi–drug resistant era, drug susceptibilityshould be tested before starting treatment and again 3 monthslater. When technically feasible, PCR should be performedearly on to allow for rapid identification of resistant strainsand timely adjustment of the drug regimen. If the organismsare resistant to isoniazid, then ethambutol 15 to 25 mg/kg/dayup to 2.5 g/day should be added to the chemotherapy.

Operative indications and timing

With the indisputable efficacy of medical treatmentalone, the proper role of surgery in the treatment of tubercu-

136


lous spondylitis is a subject of ongoing controversy [40]. Onthe one hand, Hodgson and Stock [111] advocated surgicaldebridement of tuberculous spinal lesions as soon as possi-ble after the diagnosis has been made. From their experience,they observed that earlier surgical extirpation was technicallyeasier, prevented delayed deformity, shortened the patient’slength of hospitalization and reduced the incidence of relapse.On the other hand, the prospective trials comparing the out-comes of surgical intervention with that of medical treatmentalone have not substantiated such a philosophy. Indeed, presentMRC recommendations are that surgery be reserved only fordiagnostic biopsy, spinal instability, severe deformity, myelop-athy, severe sepsis, significant abscesses and open draining si-nuses [4,26,30–37,46–48]. As a useful heuristic, the spinal sur-geon should consider three main questions for every case oftuberculous spondylitis. First, what is the rationale for the oper-ation (eg, to treat infection, neurological deficit, deformity painor functional restriction)? Second, when should the operationbe performed (ie, early, during medical treatment or after)? Fi-nally, what is the best operation or combination of proceduresto accomplish these goals?

Treatment of neurological deficits

For cases of mild neurological deficit (eg, paraparesis) oreven early paraplegia, several studies have demonstrated theeffectiveness of medical therapy without surgery [112,113].Lin et al. found that the majority of patients with only mildneurological symptoms would improve without surgical de-compression [114]. Of 89 consecutive patients with clearparaplegia who were treated with antibiotics and bedrestalone, Pattison [63] demonstrated that 85% returned to anormal life with full activity. Similarly, Moon et al. [69]conservatively treated 75 such patients with favorable out-comes seen in 95% of them. However, more recent prospec-tive data would suggest that the prognosis of patients withmoderate to severe neurological impairment is significantlyimproved by early surgical decompression and reconstruc-tion. When Lin et al. considered all patients with neurologi-cal deficit; they found that there was an improvement rate of94% with early surgery versus only 79% with nonsurgicalmanagement [114]. Patients with deficits that were man-aged according to the “middle-path” regimen who subse-quently had surgery obtained a similar overall success rateof 78.5% as well [9,61,115]. From a review of the literature,patients with less severe neurological deficit and those un-dergoing earlier surgery fare better than patients with severedeficits and later surgery [9,38,67,116–119]. Ultimately,numerous studies have emphasized that neurological out-come is directly correlated with rigid bony fusion regardlessof whether it is spontaneous or surgically induced [38,120,121].As such, some authors have advocated only a fusion procedurefor patients with mild deficits and reserved aggressive decom-pression for cases of severe paraplegia [45,49].

The overall prognosis for neurological recovery aftertimely surgical intervention is generally good but can often

be slow and prolonged [67]. Patients with active caseatingdisease and nonosseous abscesses typically recover fasterthan those with a chronic hard, bony kyphus and long-stand-ing neural compression [45]. Hodgson and Stock [111] un-derscored this finding in their study that demonstrated a di-rect correlation between the duration of neurologicalsymptoms before surgery and the time to recovery postoper-atively. Other poor prognosticators for neurological im-provement include direct infiltration of the pachymeningesand radiographic evidence of spinal cord atrophy [67,118].Surgical morbidity and complications also increase for de-layed procedures, because late fibrosis makes operative de-bridement more difficult as well [24].

Treatment of kyphosis

The residual kyphosis that is seen after ambulant chemo-therapy has always been a source of concern. In Konstam’soriginal study of ambulant chemotherapy, only 75% devel-oped a clear radiographic bony fusion with 49% having 0 to10 degrees of residual kyphosis and 18% having more than30 degrees of kyphosis [105]. Data from other MRC studieshave also echoed these findings with nonsurgical regimensyielding a solid fusion in 65% to 79% of cases [31,36].From a review of their experience, Moon et al. found thatthe kyphotic progression in these patients was typically ar-rested within the first 6 months of drug treatment and wasnot influenced by the overall length of therapy beyond that[3,62,68–70]. A controlled study by the Medical ResearchCouncil of England compared the results of inpatient versusoutpatient medical treatment and found a slightly greatermean total vertebral loss and angulation in patients treatedon an ambulant basis [36]. Further, use of orthoses and plas-ter casts did not seem to affect the ultimate kyphotic angula-tion in either group. When compared with adult cases as awhole, more advanced cases of tuberculous spondylitis andchildren with posterior element involvement are more likelyto be complicated by significant residual kyphosis after am-bulant chemotherapy [30,31,69,122,123]. Rajasekaran andShanmugasundaram [124] prospectively analyzed the ky-photic angulations after medical treatment in over 90 pa-tients for 6 years. From the controlled nonsurgical arm ofthe study, these authors were able to extrapolate a unique for-mula to predict the final gibbus angle after treatment within a90% degree of accuracy [125]. From the formula

Y

�

5.5+30.5

X

(

Y

�

final angle of the deformity,

X

�

initial loss of vertebralbody height), patients with an excessive predicted

Y

valuewere surgically reconstructed. Cases with multiple levels ofvertebral involvement, thoracic lesions, active growth andskeletal immaturity were also predictive of an increasedamount of posttreatment kyphosis [8,115].

Based on these considerations, an attempt should bemade to estimate the amount of kyphosis expected aftermedical treatment alone. When this figure is excessivelyhigh or there is evidence of significant progression duringmedical treatment, surgical intervention to correct and pre-


137

vent deformity is reasonable. Absolute criteria of the degreeof acceptable deformity will vary tremendously from caseto case. Whereas a 20-degree deformity may be consideredacceptable to patients in developing nations with limitedmedical resources, few patients in Western countries wouldbe willing to accept such a severe long-term cosmetic defor-mity. However, operative correction of severe kyphosis isextremely technically demanding and carries a high risk ofneurological injury [116,126]. Reduction of significantchronic fixed deformities for cosmetic reasons should beperformed sparingly and with great caution [3]. In suchcases, it is best to operate earlier on while the kyphosis isstill mobile and not excessive.

Surgical techniques

From the late 1800s until the early 1900s, thousands ofposterior laminectomies were performed worldwide for tu-berculosis. The outcomes of these operations were, how-ever, uniformly disappointing [64,74,116,127–129]. Seddon[130], in his 1934 review on the topic, openly condemnedlaminectomy for its dismal results, its inability to provideadequate access for safe debridement and its failure to arrestprogression of the deformity. Founded on the orthopedicprinciple that ankylosis of peripheral joints often led to re-mission of destructive synovitides, Hibbs [64] introducedthe concept of in situ posterior arthrodesis with autogenousbone grafting in 1911. Because of the lack of spinal instru-mentation at the time, such unsupported fusions proved in-effective in preventing delayed kyphosis and subsequentparaplegia and were thus largely abandoned as a surgicalsolution [131].

To address these surgical shortcomings, Capener [132]and Menard [76,133] used a lateral rachiotomy (eg, costo-transversectomy) approach to obtain an improved postero-lateral corridor for the debridement of ventral tuberculomasin the late 1920s (Fig. 3). Like laminectomy, however, pos-terolateral approaches also contributed to the instability ofthe spinal column. Transperitoneal and transthoracic surgi-cal exposures of the anterior spine were subsequently devel-oped in the 1930s to treat both tuberculosis and spondylolis-thesis [134,135]. On the other side of the globe, Ito et al.[129] concurrently reported his series of tuberculosis pa-tients who were successfully treated by anterior debride-ment and grafting. Many surgeons, however, openly op-posed such aggressive surgical debridement in light of theextremely high mortality (30% to 70%) that resulted fromsecondary infections [66,127,128]. The morbidity of opensurgery before antituberculous therapy was so high thatCalot [136] remarked in 1930, “The surgeon who, so far astuberculosis is concerned, swears to remove the evil fromthe very root, will find only one result awaiting him—thedeath of his patient.” Despite these grim outcomes, theknowledge gained from these early pioneering effortswould serve as the bedrock on which modern spinal surgerywould be built [66,76,133,137,138]. After the introduction

of streptomycin and INH in the 1950s, the surgical mortalityrate plummeted from 70% to 0% to 2.1%, thereby causing arenaissance of surgical treatment for Pott disease [139].

Abscess drainage and excision

In his original manuscript, Sir Percival Pott describedsuccessful drainage of a tuberculoma: “The remedy for thismost dreadful disease consists merely in procuring a largedischarge of matter.... and in maintaining such discharge untilthe patient shall have perfectly recovered the use of his legs”[19]. For the majority of cases of tuberculous spondylitis,such open drainage was generally ineffective and often le-thal. With the efficacy of modern antituberculous therapy,many patients now experience early resolution of theirsepsis but still demonstrate persistent large tuberculomasknown as “cold abscesses” [3,56,75,80,112]. Percutaneousaspiration or open drainage of these lesions was originallythought to speed patient recovery, to prevent abscess exten-sion and to minimize subsequent bony destruction and de-formity. After decades of experience, this technique hasbeen shown to be ineffective with an increased risk of mor-bidity and no benefit over chemotherapy alone for treatmentof tuberculous infection [3,62,69,70]. As such, simple aspi-ration or drainage of large abscesses is not indicated exceptfor purposes of obtaining a tissue diagnosis. CT-directed bi-opsies using a fine needle have proven quite effective inconfirming the causative organism, thereby obviating theneed for open spinal biopsy [140]. For rare cases where ag-gressive debridement is not possible because of either ex-tensive disease or poor medical condition of the patient, ab-

Fig. 3. Numerous corridors of access have been used in the surgical treat-ment of tuberculous spondylitis. The specific choice of approach should betailored to the location of the bulk of the abscess and tuberculoma. Formost cases of Pott’s disease, the pathology lies ventrally thereby necessi-tating an anterior approach for adequate debridement and reconstruction.These approaches are shown in italics.

138


scess drainage can be accomplished by means of a limitedtranspedicular, costotransverse or retroperitoneal approach[132,141].

More aggressive drainage and excision of infected fociwithout fusion was commonly performed worldwide forsimilar reasons [32,139]. By removing all bony and soft tis-sue sequestra and opening new vascular channels, exci-sional therapy was shown to reduce general toxemia and to-tal time to disease remission in the preantibiotic era [8,115].Somerville and Wilkinson [142] retrospectively studied 130patients treated with excisional therapy without fusion andcompared these results with 105 patients treated only withantituberculous therapy; they could demonstrate no differ-ence in outcome. In response to these findings, Tuli et al.[9,61,115] suggested that only cases with progressive bonydestruction, increasing size of abscess, new neurologicaldeficit, disease recrudescence and failure to respond to drugtherapy should undergo excisional therapy. Subsequent pro-spective MRC data have, however, shown no long-termbenefit of focal debridement and abscess evacuation overthat of ambulant chemotherapy alone [30,31]. Other authorssuch as Neville [143] actually reported poorer outcomes forpatients undergoing excisional therapy. As excisional ther-apy without fusion generally leaves a significant anteriorcolumn bony deficit, several studies have found an increasedrate of nonunion, progressive kyphosis, persistent drainingsinuses and delayed neurological deficits in this group ofpatients [3,65,89,124,143] Based on these considerations, itwould seem that excisional surgery without fusion shouldbe performed only in rare circumstances.

Anterior surgical techniques

Based on present MRC guidelines, radical ventral debri-dement, fusion and reconstruction of the vertebral columnremains the gold standard of surgical treatment for tubercu-lous spondylitis [4,26,30–37,46–48]. Although anterolat-eral decompression of the lumbar spine had previouslybeen reported by other surgeons [8], it was Hodgson andcoworkers’ [67,111,144] experience that led to populariza-tion of radical anterior debridement or what is now knownas the “Hong Kong” procedure. Although numerous vari-ants of the Hong Kong operation have been described, thehallmarks of the procedure are aggressive extirpation of allinfected foci, thorough neural decompression and solid ar-throdesis by means of bone grafting. Infected bone, caseouspus and sequestra must be aggressively extirpated back tothe level of healthy, bleeding bone. Leaving behind infectedsequestra can lead to disease recurrence and delayed defor-mity with neurological deterioration [131]. Decompressionfor cases with neurological deficit should be carried tothe level of dura, thereby increasing the chance of recovery[3,68]. In cases where only debridement is required, theposterior longitudinal ligament can be left intact [24,38].

The success of radical anterior operation and concurrentchemotherapy has been well substantiated both retrospec-

tively and prospectively in the literature. Better outcomesfor deformity, disease recurrence, neurological deficit anddisease elimination have been demonstrated [67,106,108,111,144–148]. The results of the MRC’s controlled long-term trial comparing the Hong Kong operation, simple de-bridement and medical treatment alone, demonstrated that1) anterior bony fusion occurs earlier and in a higher per-centage of patients undergoing the more radical operation(70% vs. 20% and 26%, respectively, at 5-year follow-up),2) kyphotic angulation was less common at 5 years, and 3)at 10 years, kyphotic angulation increased in the simple de-bridement group but stabilized and decreased in the radicaldebridement plus fusion group [26]. In their series of 412consecutive patients treated with this procedure, Hodgson etal. [144] had a mortality rate of 2.9% with no cases of de-layed paraplegia.

Surgical approaches

Tuberculomas of the anterior craniovertebral junctionand atlantoaxial region can be debrided through either tran-soral or extreme lateral approaches. These cases typicallyrequire concurrent occipitocervical fusion to prevent col-lapse, instability and delayed deformity in this biomechani-cally tenuous area of the spine [38]. Mid-cervical lesionsare typically treated by means of standard anterior cervicalapproaches with excellent results (eg, Cloward and South-wick-Robinson) [49] or combined anterior and posterior re-construction. Some authors have advocated approachesthrough the posterior cervical triangle, because abscesscavities in this region seem to have a propensity for dorsalextension [145]. For lesions of the cervicothoracic junc-tion, transsternal, transmanubrial, or lateral extracavitaryapproaches have been employed with varying degrees ofsuccess [8,149–151]. In the thoracic spine, radical debride-ment and reconstruction can be accomplished through ei-ther a transthoracic, extrapleural anterolateral or extendedposterolateral approach (eg, costotransversectomy, lateralextracavitary approach; Fig. 3) [26,38,152]. Extrapleuraltechniques afford the theoretical benefit of avoiding tuber-culous empyema, although there have been no data to sub-stantiate its superiority over standard thoracotomy [153].Because of the limited access afforded by posterolateral ap-proaches, there may be increased risks of blind neurovascu-lar and visceral injury, residual foci of abscesses and in-fected sequestra and a decreased rate of fusion because of asmaller graft area available for arthrodesis. Compared withtransthoracic approaches, costotransversectomy (lateral ra-chiotomy) yielded a lower rate of solid fusion (78% vs.95%) and a higher mortality rate (8% vs. 3%) [154]. How-ever, more recent literature would suggest that wider poste-rolateral exposures by means of lateral extracavitary ap-proaches might yield outcomes equivalent to that oftransthoracic approaches [152]. In such regions as the highthoracic spine, which are poorly accessed by means of tho-racotomy, a lateral extracavitary procedure provides an ef-


139

fective alternate approach. Larger degrees of kyphosis tendto allow better access through such posterolateral ap-proaches, with better visualization of the dura. For the lum-bar spine, the debridement and reconstruction is typicallyperformed through a lateral retroperitoneal approach to af-ford maximal extensile exposure of the vertebral columnand psoas musculature [67,111,144]. Anterior transperito-neal or retroperitoneal approaches used for lumbar inter-body fusion operations typically are too restrictive in theirexposure of the mid and upper lumbar spine to allow forsafe debridement and reconstruction [155]. On the otherhand, these approaches are well suited for access to the lowlumbar and lumbosacral regions [124,125,156]. Supple-mental posterior sequential instrumentation and fusion isoften necessary.

Fig. 3 summarizes the operative corridors provided bythese various approaches.

Tuberculosis patients undergoing surgery frequentlyhave a compromised immune system, depleted nutritionalreserve and an overall poor potential for wound healing. Assuch, wound closure in such cases should be performed inmultiple layers with interrupted nonabsorbable monofila-ment sutures to prevent chronic draining sinuses [38]. As in-fected exudates typically persist at the operative site forsome time after surgery, it is desirable to have well-vascu-larized surfaces nearby to allow for effective antibiotic con-centrations and maximal reabsorption of fluids. Many sur-geons thus preferentially debride tuberculomas anteriorly toallow for drainage of the operative site to either the pleuralor peritoneal surfaces [17,67,106,129].

Grafting considerations

After debridement, grafting should be performed to recon-struct the anterior support column of the spine, promote solidbony arthrodesis and ensure a successful fusion. In an MRCprospective study of 119 patients, 58 underwent radical ante-rior debridement with autologous grafting, whereas the other61 had debridement alone. Fusion rates, nonunion rates andmean kyphotic angles were lower in the grafted group both at5- and 10-year follow-up [30]. Traditionally, iliac crest, rib orfibular autologous grafts have been employed with reliable re-sults [30,58,67,146,149,154,157–159]. More recently, metal-lic cage or mesh type devices containing nonstructural au-tograft have also been successfully used [152]. The ideal graftmust provide strong support while bony arthrodesis occurs.Fibular grafts provide excellent structural integrity but alsocontain a large area of cortical bone that is slowly remodeled,often requiring several years to fully incorporate [160]. Duringthis time, the poorly vascularized necrotic matrix of the corticalbone may shelter microbes from antibiotics in the bloodstream,thereby increasing the risk of persistent disease. For these rea-sons, Bradford and Daher [161] and Louw [162] have usedvascularized rib grafts for stabilization in children and demon-strated significantly earlier incorporation (mean, 8.5 weeks)and higher fusion rates. The ease of use and ready availability

of rib graft during thoracotomy made their use popular inadults as well (Fig. 4). Clinical results have, however, beenmixed in this population. In adults, Kemp et al. [149] demon-strated a high incidence of delayed rib fracture (32%), partialcollapse and pistoning of the struts into the adjacent end plates.When compared with full-thickness iliac crest grafts, rib seg-ments yielded a markedly lower long-term fusion rate (62% vs.94.5%) and a higher mean increase in kyphosis as well (20 de-grees). Similarly, Hodgson et al. [144], who exclusively usediliac crest grafts in Hong Kong, reported a far lower amount ofdelayed kyphosis and pseudoarthrosis. For these reasons, mostexperienced authors have advocated the use of full-thickness il-iac crest for reconstruction of large defects with significant ky-phosis [111,125,149]. The use of anterior compressive spinalinstrumentation as an adjunct to debridement and grafting hasbeen successfully described [163,164]. Others have cited poorbone stock, potential for progressive bone erosion from localdisease progression and fear of persistent hardware infection asreasons to avoid the use of anterior instrumentation [24,38].

Fig. 4. This 45 year-old male presented with severe posterior thoracic pain,radicular numbness, and early myelopathy. Parasagittal non-enhanced T1-MR images demonstrated a kyphus of the mid-thoracic spine with collapseof the intervertebral disc space and gibbus-type compression of the spinalcord (A, B). The patient was treated via a two-stage procedure with initialthoracotomy, radical ventral debridement, and autogenous rib grafting.Posterior hook-type instrucmentation was then applied. A lateral radio-graph shows the rib-grafts in place with the final construct (C). The patientmade a good neurological recovery with evidence of radiographic fusionand minimal kyphotic progession.

140


Posterior surgical techniques

Although many patients often improve initially after de-compressive laminectomy, neurological compromise typi-cally recurs without fusion as instability progresses. Bos-worth et al. [120] studied this phenomena in 14 patientsdecompressed with laminectomy with all unfused patientsdying. Laminectomy for tuberculous spondylitis should thusbe reserved only for rare cases of isolated posterior neuralarch disease with posterior epidural compression [43,112,130].For these cases, 60% to 90% can expect a good to fair resultwith those with lesser deficits, subacute disease and youngage faring the best [43,165]. Even for such cases, severalexperienced authors recommend a subsequent instrumentedfusion to prevent delayed kyphotic progression and neuro-logical deterioration [8,24,166]. For Pott disease, posteriorspinal operations thus continue to play a pivotal role in thetreatment of kyphosis for both prevention and correction ofthe deformity.

There is evidence that young children are at particularrisk of progressive kyphosis after successful anterior recon-struction because of persistence of growth in the posteriorspine despite anterior arrest. Posterior in situ arthrodesis ei-ther prophylactically or as a salvage procedure has beenadvocated by several authors to treat this skeletal growthimbalance. Schulitz et al. [167] followed 117 children oper-ated for Pott disease with either anterior or posterior nonin-strumented fusions. At 10 years, the anterior fusion grouphad a mean kyphotic increase of 12 degrees, whereas thecombined anterior-posterior fusion group actually experi-enced an improved correction of 7 degrees. On the otherhand, such authors as Bailey et al. [58], Fountain et al.[168], and Upadhyay et al. [148,169] have reported a muchlower incidence of late kyphotic progression in children af-ter successful anterior debridement and fusion. Of note,however, is that these authors typically included childrenwho were older than the cohort of Schulitz et al. Neverthe-less, it is prudent to follow children closely after anterior re-construction so that posterior fusion can be performed whenneeded to arrest progressive kyphosis [24,170].

In adults, posterior instrumented techniques have alsobeen used to prevent delayed deformity as well as to pro-mote fusion and prevent kyphotic progression after anteriorreconstruction. Several authors have expressed dissatisfac-tion with the ability of anterior grafting alone to correct pre-operative kyphosis and to prevent delayed deformity[3,62,69,70,124]. A significant loss of gibbus correction hasoften been observed within the first 6 to 24 months afterradical anterior debridement and fusion [171]. This has beenparticularly true for cases involving larger defects morethan two disc spaces in size where over 25% of patients mayhave late kyphotic angles over 20 degrees [124]. Thoraciclesions, marked preoperative kyphosis and the use of ribgrafting have also been associated with more severe late de-formity after anterior surgery [38]. For such cases, somesurgeons have advocated use of a two-staged approach with

an instrumented posterior fusion either preceded or fol-lowed by an anterior debridement and grafting [62,68,172].Moon et al. [71] reported on the use of posterior Harringtonrod instrumentation as a first-stage procedure in 39 adultswith multisegmental spondylitis and obtained excellentpostoperative reduction of the deformity with minor long-term loss of correction (less than 3 degrees).

Güven et al [173] have used single-stage pedicle-screwtype posterior instrumentation for cases of limited mono-segmental spondylitis without paraplegia or extensive de-struction and obtained good control of delayed kyphoticprogression as well (less than 3.4 degrees). Compared withprior hook-type or sublaminar wire-type constructs, trans-pedicular fixation has been shown to be biomechanically farsuperior with regard to rigidity, stiffness, torsional strengthand its ability to maintain the correction of deformity[65,71,72,106,152,173]. However, large postdebridementdefects should still be grafted after transpedicular fixation toensure adequate ventral column support. In such instances,posterior instrumented correction without anterior graftingcan lead to corporal collapse, failed union, pseudarthroses,hardware breakage and progressive kyphosis [3,71,72]. Useof distraction maneuvers without anterior support further in-creases the possibility of these complications. Hooks andsublaminar wire fixation should be used where transpedicu-lar fixation is not possible. Examples of this include tho-racic vertebrae with very small pedicles, severely os-teoporotic bone and incompetent pedicles from eithererosion or prior fracture. Long-segment constructs can oftenexert excessive lever-arm moments at the ends of the im-plants, thereby also increasing the risk of screw pullout. Useof hooks or sublaminar wires at the termini of the constructcan help to prevent this (Fig. 4).

Surgical correction of a severe kyphotic deformity(greater than 30 degrees) will often require posterior tech-niques that are more complex and technically demanding.Smith-Peterson et al. [174] first described a posterior closedwedge osteotomy in 1945 with V-shaped resection of a por-tion of the posterior arch and a subsequent closure in exten-sion. With this technique, Simmons [175] was able toachieve an average of 56 degrees of correction in kyphoticlumbar deformities. However, use of this type of closing os-teotomies necessitates anterior grafting, because a large de-fect is created ventrally after extension. In the classicGalveston one-stage posterior closing wedge osteotomy, amodified bilateral costotransversectomy approach to thespine is performed to allow for wedge-shaped resections ofthe vertebral arch, intervertebral disc and a portion of thebody centrum. Once this wedge of bone has been removed,the posterior and middle columns of the spine are preferen-tially shortened, thereby allowing for angular corrections of30 to 50 degrees [116]. Because there is bone-on-bone con-tact ventrally, the need for anterior reconstruction is obvi-ated unless there is a need for either debridement or decom-pression. The decancellation or corporal eggshell technique


141

of Thiranont and Netrawichien [89] can also be used toachieve excellent correction of kyphosis as well. As the pro-cedure involves transpedicular curettage and evacuation ofthe vertebral body’s cancellous bone in addition to excisionof the posterior elements, this operation carries additionalsurgical risks [173,176]. Surgical morbidity and mortalitycan be significant for these technically demanding proce-dures, with an 8% to 10% incidence of postcorrection neu-rological complication [3,173–175]. For these reasons, Pus-chel and Zielke [177] and Roy-Camille et al. [178] haveadvocated multiple-wedge osteotomies at several levels as asafer method of reducing severe rigid deformities by dissi-pating reduction forces over several levels.

For all such posterior closing-wedge osteotomies, mostauthors universally advocate concurrent internal fixation.For most cases, compressive posterior instrumentation witheither screws or hooks should be applied to maintain thecorrection and compress the anterior arthrodesis site [179].In the past, there have been concerns raised regarding theuse of metallic instrumentation in Pott disease for fear of apersistent focus of infection [180–182]. Whereas other bac-terium adhere readily to biologically inert surfaces, such astitanium, Mycobacterium typically is less adhesive and pro-duces less extra polysaccharide biofilm than other bacteria[182]. Many surgeons have reviewed their instrumentedPott cases and have not reported problems related to infec-tion persistence or reactivation [162,183–186].

Conclusion

As cases of HIV-related tuberculosis demonstrate excep-tionally high rates of spinal involvement, the prevalence ofPott disease will continue to increase rapidly across theglobe over the next decade. It is with a sense of irony thatspinal surgeons must now revisit our own history to applyrefined operative techniques to treat this new but also veryold disease. With the advances made in antituberculousdrugs over the last century, medical therapy remains thecornerstone of treatment for tuberculosis and tuberculousspondylitis. The effectiveness of ambulant chemotherapyhas been proven time and time again in numerous retrospec-tive and prospective studies. With the gains in laboratoryand radiological diagnosis made, tuberculosis and Pott dis-ease can now be treated in a more timely, effective and se-lective fashion. In contrast to the early twentieth century,spinal surgery as a discipline has also made major strides invisualization, hemostasis, corrective ability, anesthesia andantisepsis. Anterior debridement and reconstruction can beperformed safely and with low morbidity for patients withkyphosis and neurological deficit. The ability to stabilizeand correct the spine with rigid instrumentation has openedwhole new avenues of deformity management as well. As aresult of these innovations, operating on Pott disease is farless daunting than it was to our predecessors. Despite theserevolutionary improvements, surgery should still be used

sparingly as a second-line of defense. For cases of spondyli-tis uncomplicated by significant kyphosis, unresponsivesepsis, questionable diagnosis or neural deficit, early detec-tion and medical treatment should always be the mainstayof management. Only when progressive deformity, neuro-logical deficit or incapacitating pain/dysfunction ensueshould the spine surgeon be prepared to intervene.

References

[1] Garg RK. Tuberculosis of the central nervous system. Postgrad MedJ 1999;75:133–40.

[2] Garcia-Monco JC. Central nervous system tuberculosis. Neurol Clin1999;17:737–59.

[3] Moon MS. Controversies and a new challenge. Spine 1997;22:1791–7.[4] Medical Research Council. Five-year assessment of controlled trials

of short-course chemotherapy regimens of 6, 9 or 18 months’ dura-tion for spinal tuberculosis in patients ambulatory from the start orundergoing radical surgery. Fourteenth report of the Medical Re-search Council Working Party on Tuberculosis of the Spine. Int Or-thop 1999;23:73–81.

[5] Pertuiset E. Medical therapy of bone and joint tuberculosis in 1998.Rev Rheum Engl Ed 1999;66:152–7.

[6] Pertuiset E, Beaudreuil J, Liote F, et al. Spinal tuberculosis in adults.A study of 103 cases in a developed country, 1980–1994. Medicine(Baltimore) 1999;78:309–20.

[7] WHO. WHO/22. In: Press Release. (In press).[8] Sridhar K, Ramamurthi B. Granulomatous fungal and parasitic infec-

tions in the spine. In: Menezes AH, Sonntag V, Benzel E, editors.Principles of spinal surgery. New York: McGraw-Hill, 1996:1467–95.

[9] Tuli SM. Results of treatment of spinal tuberculosis by “middle-path” regime. J Bone Joint Surg 1975;57B:13–23.

[10] Powell KE, Brown ED, Farer LS. Tuberculosis among Indochineserefugees in the United States. JAMA 1983;249:1455–60.

[11] Reider H, Cauthen G, Comstock G, et.al. Epidemiology of tubercu-losis in the United States. Epidemiol Rev 1989;11:79–98.

[12] Schieffeibein CW, Snider DE. Tuberculosis control among home-less populations. Arch Intern Med 1988;148:1843–6.

[13] Selwyn P, Hartel D, Lewis V, et.al. A prospective study of the riskof tuberculosis among intravenous drug users with human immuno-deficiency virus infection. N Engl J Med 1989;320:545–50.

[14] Stead WW, Lofgren JP, Warren F. Tuberculosis as an epidermic andnosocomial infection among the elderly in nursing home. N Engl JMed 1985;312:1483–7.

[15] WHO. World tuberculosis toll on the rise. Asian Medical News 1991;3:9.[16] AIDS epidemic in Africa. Chicago Tribune, Dec. 7 2000:7.[17] Boachie-Adjei O, Squillante RG. Tuberculosis of the spine. Orthop

Clin North Am 1996;27:95–103.[18] Leibert E, Schluger NW, Bonk S, Rom WN. Spinal tuberculosis in

patients with human immunodeficiency virus infection: clinical pre-sentation, therapy and outcome. Tuber Lung Dis 1996;77:329–34.

[19] Pott P. Remarks on that kind of the lower limbs which is frequentlyfound to accompany a curvature of the spine, and is supposed to becaused by it. Together with its method of cure, to which are added,observations on the necessity and property of amputations in certaincases, and under certain circumstances. Medical classics. Baltimore,MD: Williams and Wilkins, 1936:271–323.

[20] Formicola V, Milanesi Q, Scarsini C. Evidence of spinal tuberculo-sis at the beginning of the fourth millennium BC, from Arene Can-dide Cave (Liguria, Italy). Am J Phys Anthropol 1987;72:1.

[21] Keers RY. Pulmonary tuberculosis: a journey down the centuries.London: Balliere-Tindall, 1978.

[22] Webb GB. Tuberculosis. New York: Hueber, 1936:20–4.[23] Leff A, Lester TW, Addington WW. Tuberculosis: a chemothera-

142


peutic triumph but a persistent socioeconomic problem. Arch InternMed 1929;139:1375–7.

[24] Currier BL. Spinal infections. In: Principles and techniques of spinesurgery. Baltimore, MD: Williams and Wilkins, 1998:567–603.

[25] Centers for Disease Control. Initial therapy for tuberculosis in theera of multi-drug resistance: recommendations of the AdvisoryCouncil for the Elimination of Tuberculosis. 1993;42:1–8.

[26] Medical Research Council. Medical Research Council WorkingParty on Tuberculosis of the Spine: a controlled trial of anterior spi-nal fusion and debridement in the surgical management of tubercu-losis of the spine in patients on standard chemotherapy: a study inHong Kong. Br J Surg 1974;61:853–66.

[27] Robitzek EHSIJ. Hydrazine derivative of isonicotinic acid in thetreatment of acute progressive tuberculosis. Am Rev Tuberculosis1952;64:402–10.

[28] Schatz A, Waksman SA. Effect of streptomycin and other antibioticsubstances upon

Mycobacterium tuberculosis

and related organism.Proc Soc Exp Biol Med 1944;57:244.

[29] Schroeder SA, Krupp MA, Tierney LMJ, et.al. Current medical di-agnosis and treatment. Norwalk, CT: Appleton and Lane, 1989:153–6,536–8,1020–2.

[30] Medical Research Council. A 10-year assessment of a controlled trialcomparing debridement and anterior spinal fusion in the management oftuberculosis of the spine in patients on standard chemotherapy in HongKong. Eighth report of the Medical Research Council Working Party onTuberculosis of the Spine. J Bone Joint Surg Br 1982;64:393–8.

[31] Medical Research Council. A controlled trial of ambulant out-pa-tient treatment and in-patient rest in bed in the management of tu-berculosis of the spine in young Korean patients on standard chemo-therapy a study in Masan, Korea. First report of the MedicalResearch Study in Pusan, Korea. First report of the Medical Re-search Council Working Party on Tuberculosis of the Spine. J BoneJoint Surg Br 1973;55:678–97.

[32] Medical Research Council. A controlled trial of anterior spinal fusionand debridement in the surgical management of tuberculosis of thespine in patients on standard chemotherapy: a study in two centres inSouth Africa. Seventh report of the Medical Research Council Work-ing Party on Tuberculosis of the Spine. Tubercle 1978;59:79–105.

[33] Medical Research Council. A controlled trial of plaster-of-Paris jack-ets in the management of ambulant outpatient treatment of tuberculo-sis of the spine in children on standard chemotherapy. A study in Pu-san, Korea. Second report of the Medical Research Council WorkingParty on Tuberculosis of the Spine. Tubercle 1973;54:261–82.

[34] Medical Research Council. A five year assessment of controlled tri-als of in-patient and out-patient treatment and of plaster-of-Parisjackets for tuberculosis of the spine in children on standard chemo-therapy. Studies in Masan and Pusan, Korea. Fifth report of theMedical Research Council Working Party on Tuberculosis of theSpine. J Bone Joint Surg Br 1976;58:399–411.

[35] Medical Research Council. Five-year assessments of controlled tri-als of ambulatory treatment, debridement and anterior spinal fusionin the management of tuberculosis of the spine. Studies in Bulawayo(Rhodesia) and in Hong Kong. Sixth report of the Medical ResearchCouncil Working Party on Tuberculosis of the Spine. J Bone JointSurg Br 1978;60:163–77.

[36] Medical Research Council. A 10-year assessment of controlled tri-als of inpatient and outpatient treatment of plaster-of-Paris jacketsfor tuberculosis of the spine in children on standard chemotherapy.Studies in Masan and Pusan, Korea. Ninth report of the Medical Re-search Council Working Party on Tuberculosis of the Spine. J BoneJoint Surg Br 1985;67:103–10.

[37] Medical Research Council. Medical Research Council investiga-tion: treatment of pulmonary tuberculosis with para-aminosalicylicacid and streptomycin: preliminary report. Br Med J 1949;2:1521–7.

[38] Currier BL, Eismont FJ. Infections of the spine. In: Herkowitz HN,Garfin, SR, Balderston, RA, editors. Rothman-Simeone, 4th ed.Philadelphia: W.B. Saunders, 1999:1207–58.

[39] Garst RJ. Tuberculosis of the spine: a review of 236 operated casesin an underdeveloped region from 1954 to 1964 [comments]. J Spi-nal Disord 1992;5:286–300.

[40] Garvey ATA, Eismont FJ. Tuberculous and fungal osteomyelitis ofthe spine. Seminar. Spine Surg 1996;8:126–41.

[41] Bullough PG, Boachie-Adjei O. Atlas of spinal disease. Philadelphia:Lippincott, 1988.

[42] Doub HP, Badgley CE. The Roentgen signs of tuberculosis of thevertebral body. Am J Roentgenol 1932;27:827–35.

[43] Babhulkar SS, Tayade WB, Babhulkar SK. Atypical spinal tubercu-losis. J Bone Joint Surg 1984;66B:239–42.

[44] Travlos J, DuTott G. Beware the posterior elements. J Bone JointSurg 1990;72B:722–3.

[45] Hsu LC, Cheng CL, Leong JC. Pott’s paraplegia of late onset. Thecause of compression and results after anterior decompression. JBone Joint Surg Br 1988;70:534–8.

[46] Medical Research Council. A comparison of 6 or 9 month course re-gime of chemotherapy in patients receiving ambulatory treatment orundergoing radical surgery for tuberculosis of the spine. Indian JTuberculosis 1989;36(suppl):1–21.

[47] Medical Research Council. A controlled trail of six month and ninemonth regimens of chemotherapy in patients undergoing radical sur-gery for tuberculosis of the spine in Hong Kong. Tubercle 1986;67:243–59.

[48] Medical Research Council. A 15-year assessment of controlled tri-als of the management of tuberculosis of the spine in Korea andHong Kong. Thirteenth report of the Medical Research CouncilWorking Party on Tuberculosis of the Spine. J Joint Surg Br 1998;80:456–62.

[49] Hsu LC, Cheng CL, Leong JC. Tuberculosis of the lower cervical spine(C2 to C7). A report on 40 cases. J Bone Joint Surg Br 1984;66:1–5.

[50] Lin TH. Intramedullary tuberculoma of the spinal cord. J Neurosurg1960;17:497–503.

[51] Tandon PN. Spinal tuberculosis. In: Ramaurthi B, Tandon PN, edi-tors. Textbook of neurosurgery. New Delhi: National Book Trust,1980:718–76.

[52] Wadia NH. Radiculomyelopathy associated with spinal meningiti-des (arachnoiditis) with special reference to the spinal tuberculosisvariety. London: Oxford University Press, 1956.

[53] Tandon PN, Pathak SN. Tuberculosis of the central nervous system.In: Spillane JD, editor. Tropical neurology. London: Oxford Univer-sity Press, 1973:37–62.

[54] Arseni C, Samitca DC. Intraspinal tuberculous granuloma. Brain1960;83:295.

[55] Johnston JDH, Ashbell TS, Rosomoff HL. Isolated intraspinal ex-tradural tuberculosis. N Engl J Med 1962;266:703–6.

[56] Dastur HM. A tuberculoma: review of some personal experiences.Part I—brain. Part II—spinal cord and its coverings. Neurol India1972;20:111.

[57] Netsky MA. Diffuse meningiomatosis arachnoidalfibrosis and sy-ringomyelia. Arch Neurol Physiatry 1957;78:553–60.

[58] Bailey HL, Gabriel M, Hodgson AR, Shin JS, Banta JV. Tuberculo-sis of the spine in children. Operative findings and results in onehundred consecutive patients treated by removal of the lesion andanterior grafting. J Bone Joint Surg Am 1972;54:1633–57.

[59] Martin NS. Pott’s paraplegia. A report on 120 cases. J Bone JointSurg 1971;53B:613–28.

[60] Quahes M, Martini M. Tuberculosis of the spine. Tuberculosis ofbone and joints. Berlin: Springer-Verlag, 1988:158.

[61] Tuli SM, Srivastava TP, Varma BP, Sinha GP. Tuberculosis ofspine. Acta Orthop Scand 1967;38:445–58.

[62] Moon MS, Ha KY, Sun DH, et al. Pott’s paraplegia—67 cases. ClinOrthop 1996;323:122–8.

[63] Pattison PRM. An account of the treatment of 89 consecutive pa-tients. Paraplegia 1986;24:77–91.

[64] Hibbs RA. An operation for progressive spinal deformities. NYMed J 1911;93:1013–6.


143

[65] Pomerantz M, Brown J. The surgical management of tuberculosis.Semin Thorac Cardiovasc Surg 1995;7:108–11.

[66] Swett PP, Bennett GE, Street DM. The initial lesion, the relative in-frequency of extension by contiguity, the nature and type of healing,the role of the abscess, and the merits of operative and non-opera-tive treatment. J Bone Joint Surg 1940;22:878–94.

[67] Hodgson AR, Yau A, Kwon JS, et al. A clinical study of one hundredconsecutive cases of Pott’s paraplegia. Clin Orthop 1964;36:128–50.

[68] Moon MS. [The tuberculosis situation in the world]. MRC. Con-trolled trial of short-course regimens of chemotherapy in the ambula-tory treatment of spinal tuberculosis. Results at three years of a studyin Korea. Twelfth report of the Medical Research Council WorkingParty on Tuberculosis of the Spine. J Joint Surg Br 1993;75:240–8.

[69] Moon MS, Kim L, Woo YK. Conservative treatment of tuberculosisof the thoracic and lumbar spine in adults and children. Internat Or-thop (SICOT) 1987;11:315–22.

[70] Moon MS, Rhee SK, Kang YK. Harrington rods in treatment of ac-tive spinal tuberculosis with kyphosis. J Western Pacif Orthop As-soc 1986;2:53–8.

[71] Moon MS. Combined posterior instrumentation and anterior inter-body fusion for active tuberculosis kyphosis of dorsal and lumbarspine. Spine 1995;20:1910–6.

[72] Moon MS, Yoo YK, Lee KS, et al. Posterior instrumentation andanterior interbody fusion for tuberculosis kyphosis of dorsal andlumbar spine. Spine 1995;20:1910–6.

[73] Dobson J. An analysis of the results of conservative treatment andof the factors influencing the prognosis. J Bone Joint Surg 1951;33B:517–31.

[74] Clark JJ. The treatment of paraplegia in tuberculosis of the spine.Some suggestions and a review of recent work. Practitioner 1903;71:407–19.

[75] Janssens JP, de Haller R. Spinal tuberculosis in a developed coun-try: a review of 26 cases with special emphasis on abscesses andneurological complications. Clin Orthop 1990;257:67–80.

[76] Menard V. Causes de la Parapiegic daws le mal de Pott. Son treat-ment chirvagical par l’ouver directe du foyer tuberculous des verte-bres. Rev d’orthop 1894;5:47–64.

[77] La Rocca H. Spinal sepsis. In: Rothman RH, Simeone FA, editors.The spine, 2nd ed. Philadelphia: W.B. Saunders, 1982:767–74.

[78] Virus infections of the nervous system. XIII. TS Srinivasan Endow-ment Lecture. Madras, India, 1993.

[79] Price RW, Navia BA. Infections in AIDS and in other immuno-compound patients. In: Kennedy PGE, Johnson RE, editors. Infec-tions of the nervous system. London: Butterworth, 1987:248–73.

[80] Forlenza SW, Axelrod JL, Grieco MH. Pott’s disease in heroin ad-dicts. J Am Med Assoc 1979;241:379–80.

[81] Schulak DJ, Rayhack JM, Lippert FG, et al. The erythrocyte sedimen-tation rate in orthopaedic patients. Clin Orthop 1982;167:197–202.

[82] Hullinger N, Robinson S. A study of the erythrocyte sedimentationrate for well children. MRU 1953;42:304–10.

[83] Bottinger L, Molin L. Normal erythrocyte sedimentation rate andage. Br Med J 1967;2:85–8.

[84] Fyfe IS, Henry AP, Mulhulland RC. Closed vertebral biopsy. JBone Joint Surg 1983;65B:14–143.

[85] Milk CA, Zinreich J. Percutaneous trephine bone biopsy of the tho-racic spine. Spine 1985;10:737–40.

[86] Daniel TM. The rapid diagnosis of tuberculosis: a selective review.J Lab Clin Med 1990;116:277–82.

[87] Brisson-Noel A, Gicqual B, Lecossier D, et al. Rapid diagnosis oftuberculosis by amplification of mycobacterial DNA in clinicalsamples. Lancet 1989;334:1069–71.

[88] Berk RH, Yazici M, Atabey N, et al. Detection of mycobacterium tu-berculosis in formaldehyde solution-fixed, paraffin-embedded tissueby polymerase chain reaction in Pott’s disease. Spine 1996;21:1991–5.

[89] Thiranont N, Netrawichien P. Transpedicular decancellation closedwedge vertebral osteotomy for treatment of fixed flexion deformityof spine in ankylosing spondylitis. Spine 1993;18:2519–22.

[90] Waldvogel FA, Vasey H. Osteomyelitis: the past decade. N Engl JMed 1980;303:360–75.

[91] Pialat J, Bancel B, Etienne J, Massini B, Lucht F. Lumbar osteomy-elitis of actinomycotic or atypical mycobacterial origin? Ann Pathol1986;6:217–20.

[92] Awad I, Bay JW, Petersen JM. Nocardial osteomyelitis of the spinewith epidural spinal cord compression—a case report. Neurosurgery2001;15:254–6.

[93] Laurin JM, Resnik CS, Wheeler D, Needleman BW. Vertebral os-teomyelitis caused by

Nocardia asteroides

: report and review of theliterature. J Rheumatol 1991;18:455–8.

[94] Collert S, Petrini B, Wilkman K. Osteomyelitis caused by

Mycobac-terium avium

. Acta Orthop Scand 1983;54:449–51.[95] Marchevsky AM, Damsker B, Green S, et al. The clinicopathologi-

cal spectrum of non-tuberculous mycobacterial osteoarticular infec-tions. J Bone Join Surg Am 1975;67:925–9.

[96] Vincurova M, Maar D, Scwartz E. [Mycobacteria infection at thetransverse process of the 2nd lumbar vertebra caused by

Mycobacte-rium fortuitum

]. Z Orthop 1984;122:278–80.[97] Cordero M, Sanchez I, Roy V, Weisdorf D. Brucellar and tubercu-

lous spondylitis. A comparative study of their clinical features. My-cobacterial infections following bone marrow transplantation: a 20year retrospective review. J Bone Joint Surg 1991;73:100–3.

[98] Tekkok IH, Berker M, Ozcan OE, Ozgen T, Akalin E. Brucellosis ofthe spine. Neurosurgery 1993;33:838–44.

[99] Sengupta A. Musculoskeletal lesions in yaws. Clin Orthop 1985;194:193–9.

[100] Baysefer A, Gonul E, Canakci Z, Erdogan E, Aydogan N, Kayali H.Hydatic disease of the spine. Spinal Cord 1996;34:297–300.

[101] Merkel D, Fitzgerald RH, Brown ML. Scintigraphic evaluation inmusculoskeletal sepsis. Orthop Clin North Am 1984;15:401–16.

[102] Arnott G, Delfosse JM. Acute spinal epidural abscess. Neurology1985;213:343–4.

[103] Bosworth DM, Wright HA, Fielding JW. The treatment of bone andjoint tuberculosis: effect of 1-isonicotinyl-2-isopropylhydrazine. JBone Joint Surg 1952;34A:761–71.

[104] Kaplan CJ. Conservative therapy in skeletal tuberculosis: an appraisalbased on experience in South Africa. Tubercle 1959;40:355–68.

[105] Konstam PG, Blesovsky A. The ambulant treatment of spinal tuber-culosis. Br J Surg 1962;50:26–38.

[106] Chen WJ, Chen CH, Shih CH. Surgical treatment of tuberculousspondylitis. Acta Orthop Scand 1995;66:137–42.

[107] Griffith DLI. Short-course chemotherapy in the treatment of spinaltuberculosis. J Bone Joint Surg [Br] 1986;68:158.

[108] Upadhyay SS, Saji MJ, Yau CMC. Duration of antituberculous che-motherapy in conjunction with radical surgery in the management ofspinal tuberculosis. Spine 1996;21:1898–903.

[109] Haas DW, Des Prez RM. Mycobacterium tuberculosis. In: Man-dell GL, Gennett JE, Dolin R, editors. Principles and practice ofinfectious diseases. New York: Churchill Livingstone, 1995:2213–43.

[110] Dekel S, Francis MJO. Hematogeneous osteomyelitis and staphylo-coccus aureus. Postgrad Med 1981;6:131–4.

[111] Hodgson AR, Stock FE. Anterior spinal fusions: a preliminary com-munication on the radical treatment of Pott’s disease and Pott’sparaplegia. Br J Surg 1964;456:266–75.

[112] Garceau GJ, Brady TA. Pott’s paraplegia. J Bone Joint Surg 1950;32A:87–95.

[113] Kaplan CJ. Pott’s disease in South African Bantu children. An anal-ysis of results and comparison with Lancashire figures. Br J Tuber-culosis 1952;46:209–13.

[114] Lin TH. Tuberculous spondylitis in adults. J Bone Joint Surg 1960;67A:1405–13.

[115] Tuli SM. Tuberculosis of the skeletal system. New Delhi: JaypeeBrothers Medical Publishers, 1991.

[116] Adams JC. Technique, dangers, and safeguards in osteotomy of thespine. J Bone Joint Surg [Br] 1952;34:225–32.

144


[117] Adendorff JJ, Boeke EJ, Lazarus C. Tuberculosis of the spine: resultsof management of 300 patients. J R Coll Surg Edinb 1987;32:152–5.

[118] Govender S, Charles RW, Naidoo KS, et al. Results of surgical de-compression in chronic tuberculous paraplegia. S Afr Med J 1988;74:58–9.

[119] Kohli SB. Radical surgical approach to spinal tuberculosis. J BoneJoint Surg 1967;49B:668–73.

[120] Bosworth DM, Pietra AD, Farrell RF. Streptomycin in tuberculousbone and joint lesions with mixed infection and sinuses. J BoneJoint Surg 1950;32A:103–8.

[121] Bosworth DM, Pietra AD, Rahilly G. Paraplegia resulting from tu-berculosis of the spine. J Bone Joint Surg 1953;35A:735–40.

[122] Dickson JAS. Spinal tuberculosis in Nigerian children: a review ofambulant treatment. J Bone Joint Surg 1967;49B:682–94.

[123] Friedman B. Chemotherapy of tuberculosis of the spine. J BoneJoint Surg 1966;48A:451–74.

[124] Rajasekaran S, Shanmugasundaram TK. Progression of kyphosis intuberculosis of the spine treated by anterior arthrodesis. J Bone JointSurg 1989;71A:1314–23.

[125] Rajasekaran S, Shanmugasundaram TK. Prediction of the angle ofgibbus deformity in tuberculosis of the spine. J Bone Joint Surg 1987;69A:503–9.

[126] Yau AMC, Hsu LCS, O’Brien JP, Hodgson AR. Tuberculous kypho-sis: correction with spinal osteotomy, halopelvic distraction, and ante-rior and posterior fusion. J Bone Joint Surg [Am] 1974;56:1419–34.

[127] Eisberg CA. Diagnosis and treatment of surgical disease of the spinalcord and its membranes. Philadelphia: W.B. Saunders, 1916:312–4.

[128] Ely LW. Joint tuberculosis. New York: Wood, 1911:125.[129] Ito H, Isuchiya J, Asami G. A new radical operation for Pott’s dis-

ease. Report of ten cases. J Bone Joint Surg 1934;16:499–515.[130] Seddon HJ. Pott’s paraplegia: prognosis and treatment. Br J Surg

1934;22:769–99.[131] Hallock H, Jones B. Tuberculosis of the spine. J Bone Joint Surg

1954;36A:219.[132] Capener N. The evolution of lateral rachiotomy. J Bone Joint Surg

1954;36B:173–9.[133] Menard V. Etude pratique sur le mal de Pott. Paris Reviews in Or-

thopedics 5:47–64:1900.[134] Burns BH. An operation for spondylolisthesis. Lancet 1933;i:1933.[135] Capener N. Spondylolisthesis. Br J Surg 1932;19:374.[136] Calot T, as quoted by Tuli, Refer. 3. Acta Chir Scand 1930.[137] Vincent E. Contribution a la chirurgie rachidienne. Due drainage

vertebral dans le mal de Pott. Rev de Chir 1892;12:273–94.[138] Wassiliew MA. Operative Behandlung de paraplegia bie tubercu-

loser spondylitis. Arch F Chir 1909;88:845–60.[139] Kondo E, Yamada K. End results of focal debridement in bone and joint

tuberculosis and its indications. J Bone Joint Surg 1957;39A:27–31.[140] Mondal A. Cytological diagnosis of vertebral tuberculosis with fine-

needle aspiration biopsy. J Bone Joint Surg 1994;76A:181–4.[141] Weinberg JA. The surgical excision of psoas abscesses resulting

from spinal tuberculosis. J Bone Joint Surg 1957;39A:17–27.[142] Somerville EW, Wilkinson MC. Girdlestone’s tuberculosis of bone

and joints, 3rd ed. London: Oxford University Press, 1965.[143] Neville CH. Is surgical fusion still desirable in spinal tuberculosis?

Clin Orthop 1971;75:179–84.[144] Hodgson AR, Stock FE, Fang HSY, et al. Anterior spinal fusion: the

operative approach and pathological findings in 412 patients withPott’s disease of the spine. Br J Surg 1960;48:172–8.

[145] Hodgson AR. An approach to the cervical spine (C3–C7). Clin Or-thop 1965;39:129–34.

[146] Hodgson AR. Report on findings and results in 300 cases of Pott’sdisease treated by anterior fusion of the spine. J West Pacific OrthopAssoc 1965;1:3–10.

[147] Pun WK, Chow SP, Luk KDK, et al. Tuberculosis of the lumbosac-ral junction. J Bone Joint Surg 1990;72:675–9.

[148] Upadhyay SS, Sell P, Saji MJ, Sell B, Yau AC, Leong JC. 17-year pro-spective study of surgical management of spinal tuberculosis in chil-

dren. Hong Kong operation compared with debridement surgery forshort- and long-term outcome of deformity. Spine 1993;18:1704–11.

[149] Kemp HBS, Jackson JW, Jeremiah JD, Cook J. Anterior fusion of thespine for infective lesions in adults. J Bone Joint Surg 1973;55B:715.

[150] Khoo L, Samudrala S. Transclavicular/transmanubrial approach/closure. In: Fessler ED, Sekharil RG, editors. Atlas of neurosurgicaltechniques: cranial and peripheral nerve, spine and endoscopic sur-gery. New York: Thieme (in press).

[151] Whitecloud TS, LaRocca SH. Fibular strut graft in reconstructivesurgery of the cervical spine. Spine 1976;1:7–11.

[152] Dietze DD, Fessler RG, Jacob RP. Primary reconstructions for spi-nal infections. J Neurosurg 1997;86:981–9.

[153] Guirguis AR. Pott’s paraplegia. J Bone Joint Surg 1967;49B:658–67.[154] Kirkaldy-Willis WH, Thomas TG. Anteior approaches in the diag-

nosis and treatment of infections of the vertebral bodies. J BoneJoint Surg 1965;47A:87–110.

[155] Khoo L, Samudrala S, Shim R, Fessler R. Complications during an-terior surgery of the lumbar spine: an anatomically based study andreview. Neurosurgical Focus (www.neurosurgery.org/journals/online_i/dec99) 1999.

[156] Rajasekaran S, Shanmugasundaram TK. Tuberculous lesions of thelumbosacral region. A 15-year follow-up of patients treated by am-bulant chemotherapy. Spine 1998;15:1163–7.

[157] Allen AR, Stevenson AW. Follow-up notes on articles previously pub-lished in the journal: a 10-year follow-up of combined drug therapy andearly fusion in bone tuberculosis. J Bone Joint Surg 1967;21:178–87.

[158] Fellander M. Paraplegia in spondylitis: results of operative treat-ment. Paraplegia 1975;13:75–88.

[159] Wiltberg BR. Resection of vertebral bodies and bone grafting forchronic osteomyelitis of the spine. J Bone Joint Surg 1952;34A:215–8.

[160] Enneking WF, Burchardt H, Puhl JJ. Physical and biological aspectsof repair in dog cortical-bone transplants. J Bone Joint Surg 1975;57A:237–52.

[161] Bradford DS, Daher YH. Vascularized rib grafts and stabilization ofkyphosis. J Bone Joint Surg 1986;68B:357–61.

[162] Louw JA. Spinal tuberculosis with neurological deficit. Treatmentwith vascularized rib grafts, posterior osteotomies, and fusion. JBone Joint Surg 1994;72B:686–93.

[163] Kostuik JP. Anterior spinal cord decompression for lesions of thethoracic and lumbar spine. Techniques, new methods of internal fix-ation, results. Spine 1983;8:512–31.

[164] Yilmaz C, Selek HY, Gurkan I, Erdemli B, Korkusuz Z. Anterior in-strumentation for the treatment of spinal tuberculosis. J Bone JointSurg [Am] 1999;81:1261–7.

[165] Kumar K. A clinical study and classification of posterior spinal tu-berculosis. Int Orthop 1985;9:147–52.

[166] Kemp HBS, Jackson JW, Shaw NC. Laminectomy in paraplegia dueto ineffective spondylosis. Br J Surg 1974;61:66–72.

[167] Schulitz KP, Kothe R, Leong JC, Wehling P. Growth changes ofsolidly fused kyphotic bloc after surgery for tuberculosis. Compari-son of four procedures. Spine 1997;15(22):1150–5.

[168] Fountain SS, Hsu LCS, Yau ACM, Hodgson AR. Progressive ky-phosis following solid anterior spine fusion in children with tuber-culosis of the spine. J Bone Joint Surg 1966;57A:1104–7.

[169] Upadhyay SS, Sell P, Saji MJ, Sell B, Hsu LC. Surgical manage-ment of spinal tuberculosis in adults. Hong Kong operation com-pared with debridement surgery for short and long term outcome ofdeformity. Clin Orthop 1994;302:182.

[170] Jenkins DHR, Hodgson AR, Yau ACM, et al. Stabilisation of thespine in surgical treatment of severe spinal tuberculosis in children.Clin Orthop 1975;110:69–80.

[171] Kim NH, Lee HM, Suh JS. Magnetic resonance imaging for the di-agnosis of tuberculous spondylitis. Spine 1994;19:2451–5.

[172] Lee EY, Han MS. A study of influences of the anterior interverte-bral fusion upon the correctability of kyphosis in tuberculousspondylitis. J Korean Orthop Assoc 1968;3:31–40.


145

[173] Güven O, Kumano O, Yalcin S, et al. A single stage posterior ap-proach and rigid fixation for preventing kyphosis in the treatment ofspinal tuberculosis. Spine 1994;19:1039–43.

[174] Smith-Peterson MN, Larson CB, Aufranc OE. Osteotomy of thespine for correction of flexion deformity in rheumatoid arthritis. JBone Joint Surg 1945;27:1–11.

[175] Simmons EH. Kyphotic deformity of the spine in ankylosingspondylitis. Clin Orthop 1977;128:65–77.

[176] Wu SS, Hwa SY, Lin LC, et.al. Management of rigid posttraumatickyphosis. Spine 1996;21:2260–7.

[177] Puschel J, Zielke K. Korrekturooperation bei Bechterew-Kyphose:Indikation, Technik, Ergebnisse. Z Orthop 1982;120:338–42.

[178] Roy-Camille R, Henry P, Saillant G, Doursounian L. Chirurgie desgrandes cyphoses vertebrales de la spondylarthrie ankylosante. RevRhum 1987;54:261–7.

[179] Kostuik JP. Ankylosing spondylitis: surgical treatment. In: Fry-moyer JW, editor. The adult spine: principles and practice, 2nd ed.Philadelphia: Lippincott-Raven, 1997:845–70.

[180] Gristina AG, Oga M, Webb LX, Hogwood CD. Bacterial adherenceand pathogenesis of osteomyelitis. Science 1985;228:990–3.

[181] Gristina AG, Costerton JW. Bacterial adherence and glycocalyx andtheir role in musculoskeletal infection. Orthop Clin North Am 1985;15:517–35.

[182] Oga M, Sugioka Y, Hobgood CD, et al. Surgical biomaterials anddifferential colonization by

Staph. epidermidis

. Biomaterials 1988;9:285–9.

[183] Bakin MN, Kovalenko KN, Mal’Chenko OV. Harrington’s distrac-tor for reconstruction of the spine in children with tuberculosisspondylitis [Russian]. Probl-Tuberk 1992;40:7–8.

[184] Lavrov VN. Extra focal instrumental fixation of the spine in the sur-gical treatment of tuberculous spondylitis. Probl-Tuberk 1991;8:52.

[185] Oga M, Arizono T, Takasita M, Sugioka Y. Evaluation of the risk ofinstrumentation as a foreign body in spinal tuberculosis. Spine1993;18:1890–4.

[186] Ye Q. The application of Dick instrumentation in the field of spine sur-gery. Chung-kuo-I-Hsueh-Ko-Hsuch-Yuan-Hsuch-Pao 1999;15:39.

FourHundred SixtyYears Ago in Spine . . .

At the age of 29, Andreas Vesalius pub-lished what is generally called

The Fabrica

. This ency-clopedic work was the most thorough text ever publishedon human anatomy. It not only added to but also cor-rected the errors of established works, such as those of

Galen. His insistence that cadaver dissection be an essen-tial part of the education of physicians fueled a contro-versy that lasted for centuries. The extraordinary beautyof the illustrations not only served to establish the au-thority of the science but also became treasures of art.Who should be credited with the original illustrations isan unresolved topic of scholarly debate.

References

[1] Vesalius A. De humani corporis fabrica libri septem. Basileae, ex.Off. Ioannis Oporini, 1543.

[2] O’Malley CD. Andreas Vesalius of Brussels. Berkeley: Universityof California Press, 1964.

review articles a surgical revisitation of pott distemper...

Documents