Brucellosis of the Spine (Spinal Brucellosis)

Spinal brucellosis is the most common form of osteoarticular brucellosis, a zoonotic infection caused by Brucella species. In endemic regions it is one of the most important causes of spondylodiscitis.

The organisms. Each species has its own animal host:

• B. melitensis (goats and sheep) - the most common and most virulent, responsible for more than 90% of spinal cases
• B. abortus (cattle) - milder disease, often from occupational exposure
• B. suis (pigs) - variable severity, and a less common cause of spinal disease
• B. canis (dogs) - a rare cause of human disease

Where it is found. The highest-risk regions are the Mediterranean basin (Spain, Italy, Greece), the Middle East, Central Asia and Latin America; the disease is also endemic in the Indian subcontinent and sub-Saharan Africa. There are about 500,000 new cases a year worldwide.

How often the skeleton is involved. Osteoarticular disease occurs in 6-12% of all brucellosis, and spondylitis is its most common manifestation, accounting for 50-60% of skeletal brucellosis. Spinal disease affects men twice as often as women, with a peak incidence at 40-60 years.

The history that makes the diagnosis. In non-endemic countries most cases are imported, so the travel, dietary and occupational history is essential. Ask specifically about unpasteurised dairy and animal contact.

The organism. Brucella is a small, gram-negative coccobacillus: non-motile, non-spore forming, with no exotoxins or plasmids. It is a facultative intracellular organism that survives within macrophages, and it grows slowly.

Transmission. Ingestion is the most common route:

• Ingestion of unpasteurised milk, cheese or ice cream
• Direct contact with infected animals or their products through skin abrasions - farmers, vets and abattoir workers, raw meat handling, and a high risk when assisting at animal births
• Inhalation of aerosols in laboratories, abattoirs and on farms
• Vertical or sexual transmission, rarely

Seeding the vertebra. The organisms spread haematogenously from the primary infection, usually intestinal, and lodge in the vertebral metaphyseal vessels. These segmental, metaphyseal end-arteries arborise in the richly vascular anterosuperior endplate region, so seeded organisms lodge there first. That is why erosion of the anterosuperior corner, the Pedro Pons sign, is the earliest radiographic change, before the disc space narrows. The infection starts as osteomyelitis of the vertebral endplate and spreads to the adjacent disc and vertebra, producing spondylodiscitis.

Why it destroys less bone. The inflammation is granulomatous, with microabscess formation, and less destructive than pyogenic infection. The corner lesion therefore tends to repair with reactive sclerosis and osteophyte ("parrot-beak") formation rather than the collapse and gibbus of tuberculosis.

View larger

Where in the spine. The lumbar spine is involved in 70-80% of cases, L4-L5 most often (40-50%) and L3-L4 next. Cervical disease is rare (less than 5%), and 10-20% of patients have multifocal involvement.

View larger

The systemic illness. The fever is undulant, classically rising and falling ("Malta fever"), with night sweats, malaise, fatigue, weight loss and anorexia. Hepatosplenomegaly is found in 30-50%. The presentation may be subacute or chronic.

The spine. Back pain is insidious in onset, progressive and worse at night. There is tenderness over the affected vertebrae, paravertebral muscle spasm, and reduced range of motion, particularly in extension.

Neurological involvement. Radiculopathy from root compression is the most common form. An epidural abscess can compress the cord, and cauda equina syndrome is rare but serious. The prognosis is better than that of tuberculous myelopathy.

View larger

Sacroiliitis. Concurrent sacroiliitis occurs in 20-30% of spinal brucellosis and can cause buttock pain, a positive FABER test and sacroiliac joint tenderness. In an endemic region, spine plus sacroiliac involvement should raise the suspicion of brucellosis.

Against pyogenic infection and tuberculosis. Brucellar spondylitis is more indolent than pyogenic spondylodiscitis, and fever is less prominent. Against tuberculosis, the systemic symptoms are more prominent while destruction, kyphotic deformity and neurological involvement are less severe; paraspinal abscesses are less common and smaller, and the response to medical treatment is better.

The principle. Most patients respond well to antibiotics alone, and surgery is required in only 10-15%, unlike tuberculosis, where it is more commonly needed.

Antibiotic TherapySurgical IndicationsConservative Management

Combination, never monotherapy. Monotherapy relapses in 10-40% and fails 2.6 times as often as combination therapy, so a single drug is never acceptable, whatever the duration. Brucella survives in macrophages, so the drugs need intracellular penetration as well as good bone penetration.

The choice of regimen. Level I evidence prefers doxycycline plus an aminoglycoside. The all-oral doxycycline-rifampicin regimen is widely used for compliance, a pragmatic choice rather than a better one. Rifampicin substitutes for the aminoglycoside only when injectables are impractical, and the higher relapse rate is then accepted and counselled.

PREFERRED by Level I evidence

Drugs

Doxycycline 100mg BD + Streptomycin 1g IM OD (or gentamicin)

Duration

2-3 weeks aminoglycoside, doxycycline for a minimum of 3 months, typically 6 months

Notes

Skalsky meta-analysis: doxycycline-rifampicin fails 2.8x more often (RR 2.80, 95% CI 1.81-4.36), driven by relapse. Gentamicin non-inferior to streptomycin

Triple therapy - severe or focal disease

Drugs

Doxycycline + Rifampicin + aminoglycoside

Duration

2-3 weeks aminoglycoside, then oral to 6+ months

Notes

Outperformed doxycycline-streptomycin in the meta-analysis (RR 2.50), though on only two trials. Reasonable for spondylitis with epidural or neurological involvement

WHO all-oral regimen (pragmatic, not superior)

Drugs

Doxycycline 100mg BD + Rifampicin 600-900mg OD

Duration

Minimum 6 weeks, 3-6 months for spondylitis

Notes

Chosen for compliance and where injectables are impractical - NOT because it works better. Accept and counsel the higher relapse rate

Pregnancy

Drugs

Rifampicin + TMP-SMX

Duration

Throughout pregnancy

Notes

Avoid doxycycline and aminoglycosides

Antibiotic Regimens for Spinal Brucellosis

Regimen	Drugs	Duration	Notes
PREFERRED by Level I evidence	Doxycycline 100mg BD + Streptomycin 1g IM OD (or gentamicin)	2-3 weeks aminoglycoside, doxycycline for a minimum of 3 months, typically 6 months	Skalsky meta-analysis: doxycycline-rifampicin fails 2.8x more often (RR 2.80, 95% CI 1.81-4.36), driven by relapse. Gentamicin non-inferior to streptomycin
Triple therapy - severe or focal disease	Doxycycline + Rifampicin + aminoglycoside	2-3 weeks aminoglycoside, then oral to 6+ months	Outperformed doxycycline-streptomycin in the meta-analysis (RR 2.50), though on only two trials. Reasonable for spondylitis with epidural or neurological involvement
WHO all-oral regimen (pragmatic, not superior)	Doxycycline 100mg BD + Rifampicin 600-900mg OD	Minimum 6 weeks, 3-6 months for spondylitis	Chosen for compliance and where injectables are impractical - NOT because it works better. Accept and counsel the higher relapse rate
Pregnancy	Rifampicin + TMP-SMX	Throughout pregnancy	Avoid doxycycline and aminoglycosides

Duration. In the meta-analysis, treatment for six weeks or more beat shorter courses. Spondylitis needs 3-6 months, guided by clinical and radiological response.

Treatment Duration

Minimum 3 months, but most experts recommend 6 months for spondylodiscitis. Shorter courses have unacceptable relapse rates (up to 30%). Continue until clinical cure and normalisation of inflammatory markers.

Monitoring response. Successful treatment brings clinical improvement within 2-4 weeks.

• Clinical - the fever resolves in 1-2 weeks, and pain improves
• CRP and ESR - should fall by 2-4 weeks; check the CRP every 2-4 weeks
• Serology - the SAT may remain positive, but the 2-ME test should become negative
• MRI - at 3 months if the response is good, earlier if there is concern, and repeated at 3-6 months if the response is uncertain

When the initial disease was extensive, follow-up MRI provides an anatomical endpoint.

View larger

View larger

Absolute indications. Progressive neurological deficit despite antibiotics, spinal cord compression with myelopathy, spinal instability (rare), and a large epidural abscess not responding to antibiotics.

Relative indications. A significant paraspinal abscess, failed medical treatment with persistent infection, unacceptable pain despite adequate treatment, and progressive deformity, which is rare in brucellosis. The abscess size that should trigger drainage is one of the open questions below.

Approaches.

• Posterior - decompression, debridement and instrumented fusion. A one-stage posterior approach provides debridement, neural decompression and interbody support when disease is accessible from behind
• Anterior - radical debridement and strut grafting
• Combined - for severe cases requiring anterior column support, such as severe anterior-column destruction reconstructed with posterior pedicle fixation plus anterior retroperitoneal debridement and interbody fusion

How it differs from tuberculosis surgery. Less aggressive debridement is often sufficient, the healing response to antibiotics is better, less structural grafting is typically required, and instrumentation fails less often.

View larger

View larger

View larger

Immobilisation. Bed rest in the initial phase if pain is significant, then an LSO or TLSO brace for lumbar involvement until the pain improves, typically 4-6 weeks.

Supportive care. Analgesia (NSAIDs, paracetamol, and opioids if needed), physiotherapy for deconditioning, nutritional support, and DVT prophylaxis during immobilisation.

Outcome. With appropriate antibiotics the cure rate is 95%. Early diagnosis improves outcomes, and treatment delay is associated with a higher complication rate; neurological involvement worsens the prognosis, and compliance with the prolonged antibiotic course is crucial.

View larger

View larger

Global epidemiology:

Brucellosis is the commonest bacterial zoonosis worldwide, with the heaviest burden across the Mediterranean basin, Middle East, Central Asia, the Indian subcontinent, sub-Saharan Africa and Latin America. Incidence is highly heterogeneous: many high-income countries that have eradicated bovine brucellosis (much of Western Europe, North America, Japan, Australia and New Zealand) now see almost exclusively imported cases in travellers and migrants, whereas hyperendemic regions report population incidence orders of magnitude higher. There is no implant registry for spinal infection; the relevant "registries" are national notifiable-disease surveillance systems (e.g. national public-health agencies and WHO/FAO/OIE reporting), which capture confirmed human cases and inform regional risk.

Side-by-side guidance:

Diagnostic and biosafety considerations:

Brucella is a hazard-group/risk-group 3 organism; laboratories must be alerted to the clinical suspicion so that cultures are handled in appropriate containment and incubated for the extended period required. In non-endemic settings, local laboratories may have limited experience, and serology (Rose Bengal screening with confirmatory agglutination/ELISA) plus molecular testing through reference laboratories is the mainstay. A clear travel, dietary and occupational history is the single most important step that prompts the correct test.

High- vs limited-resource practice variation:

• High-resource settings: MRI is readily available for early diagnosis and monitoring; automated blood-culture systems and PCR improve and accelerate confirmation; infectious-diseases input guides prolonged therapy and surgery is reserved for neurological deficit, instability or abscess unresponsive to antibiotics.
• Limited-resource / hyperendemic settings: Diagnosis often rests on serology and plain radiographs (Pedro Pons sign), with MRI and PCR less accessible; the doxycycline + rifampicin oral regimen is favoured for outpatient feasibility and adherence despite a higher relapse risk than aminoglycoside-containing regimens; the major public-health lever is control of animal reservoirs and pasteurisation of dairy.

Related pages: Tuberculosis of the Spine (Pott's Disease) is the diagnosis this one is mistaken for and the comparison every examiner wants - the discriminators from the 219-case series carded above are a prolonged course, thoracic level, absent fever, spinal deformity, neurological deficit and large paravertebral or epidural masses favouring tuberculosis, against the single-level lumbar involvement, preserved disc height and preserved vertebral architecture of brucellosis. Pyogenic Discitis and Vertebral Osteomyelitis completes that triad and is suggested instead by diabetes, intravenous drug use, recent spinal surgery, preceding bacteraemia and a markedly raised white count. Spinal Epidural Abscess is the complication that turns a medically treated infection into a surgical one and the reason for the single operation in Solera's 35 patients, with Cauda Equina Syndrome the deficit to exclude on every review. Systematic Approach to Spine Imaging and Systematic Approach to Infection Imaging cover the MRI protocol on which all of this is judged. Fungal Spine Infections and Atypical Mycobacterial Infection are the rest of the granulomatous, culture-negative differential - a chronic spondylodiscitis with negative routine cultures is brucellar, tuberculous, mycobacterial or fungal until the laboratory is asked the right question. Orthopaedic Antibiotic Therapy and Antibiotic Pharmacology and Resistance carry the intracellular penetration that makes doxycycline and the aminoglycosides work against an organism living inside macrophages, and explain why monotherapy relapses. Osteomyelitis Pathophysiology with Biofilm Formation and Intervertebral Disc Biology explain why the disc is involved early here and late in tuberculosis, and Thoracolumbar Fractures covers the deformity and instability that follow untreated destruction.

Optimal regimen. Meta-analysis evidence (Skalsky, BMJ 2008) shows that doxycycline-rifampicin relapses more than aminoglycoside-containing regimens, yet the all-oral combination remains widely used for adherence in outpatient and limited-resource settings. The trade-off between efficacy and feasibility is unresolved for spinal disease specifically.

Treatment duration. No randomised evidence defines the ideal duration for brucellar spondylitis. Practice ranges from a minimum of 3 months to 6 months or longer, individualised to clinical response, abscess and inflammatory-marker normalisation rather than a fixed endpoint.

Role and timing of surgery. Thresholds for operating on an epidural or paravertebral abscess without progressive neurology are debated. Many collections resolve with antibiotics alone, but the size and response that should trigger drainage are not standardised.

Monitoring cure and defining relapse. Serological titres can persist after microbiological cure, so distinguishing true relapse from residual antibody or non-infective mechanical back pain is difficult. The relative value of MRI, inflammatory markers and antibody assays for follow-up is not firmly established.

Empirical therapy in dual-endemic regions. Where tuberculosis and brucellosis coexist, whether to await tissue or molecular confirmation or to start empirical therapy when both are plausible remains a clinical judgement, given the very different drug regimens and durations.