Knowledge · Spondylodiscitis

Spondylodiscitis: Infection of the Disc and Vertebra

In spondylodiscitis, the intervertebral disc and adjacent vertebral bodies become inflamed — usually from bacteria that reach the spine via the bloodstream. The condition is becoming more common, affects primarily older and immunocompromised individuals, and is often recognised late because it initially resembles ordinary back pain. When diagnosed in time, most patients can be cured.

Research focus · Pathogen identification first · New AO Spine classification · Surgery when it improves survival and recovery

Warning signs — seek prompt assessment: worsening back pain together with fever or rigors, severe rest and night pain, recent infections or procedures, new-onset paralysis or sensory disturbances, bladder or bowel dysfunction. In neurological deficits or signs of sepsis: call 144 or go to the nearest emergency department.

The key message: Historically spondylodiscitis was almost always managed conservatively — with antibiotics and immobilisation. Evidence from recent years paints a different picture: in many patients, early surgery is associated with lower mortality, fewer recurrences, and faster recovery. What is decisive is a structured assessment of pathogen, stability, neurology, and abscesses — the basis for this is provided by the new AO Spine classification for primary spinal infections, which I contribute to within the AO Spine Knowledge Forum Trauma & Infection.

What is spondylodiscitis?

Spondylodiscitis is an infection of the intervertebral disc and the adjacent vertebral bodies; when the vertebral bone is primarily affected, it is also termed vertebral osteomyelitis. In most cases bacteria reach the spinal segment via the bloodstream — originating for example from skin, urinary tract, or dental infections, infected catheters, or endocarditis. Less commonly the infection arises after spinal procedures or by direct spread from adjacent structures.

The most common pathogen is Staphylococcus aureus. The lumbar spine is most frequently affected. The infection can extend into the spinal canal and form an epidural abscess that compresses nerves or the spinal cord — or weaken the load-bearing structures to the point where vertebrae collapse and the spine becomes unstable.

The condition is increasing: the population is ageing, more people live with diabetes, renal disease, cancer treatment, or immunosuppressive medication, and medical procedures involving catheters and implants have become more frequent. Special forms such as tuberculous spondylodiscitis follow their own rules; our Knowledge Forum has developed separate treatment recommendations for these.

Bigdon SF et al. Global Spine J. 2025. PMID 39852953 · Scherer J et al. Global Spine J. 2026. PMID 42533326

Typical symptoms — and why the diagnosis often comes late

What makes spondylodiscitis insidious is its unremarkable presentation at the outset:

Worsening, deep back pain — often also at rest and at night, unlike the usual load-related low back pain
Fever is frequently absent: only about half of patients have an elevated temperature — absent fever does not rule out the diagnosis
Systemic symptoms: fatigue, loss of appetite, night sweats, weight loss
Tenderness on percussion and pressure over the affected segment, pain-limited range of motion
Neurological deficits such as weakness, sensory disturbances, or bladder/bowel dysfunction — signs of nerve compression, usually from an abscess

Because the initial symptoms are non-specific, several weeks not infrequently pass between the first symptoms and the diagnosis. With back pain accompanied by fever, rest and night pain, recent infections, or elevated inflammatory markers, spondylodiscitis should therefore be actively considered.

Diagnosis: pathogen and extent determine treatment

The workup pursues two goals: to confirm and characterise the infection — and to identify the causative pathogen so that targeted treatment is possible.

Laboratory: inflammatory markers (especially CRP) are almost always elevated and also serve as a monitoring tool
Blood cultures: ideally before starting antibiotics — they identify the pathogen in a relevant proportion of patients
MRI: the standard investigation. It shows inflammation of the disc and vertebrae early, as well as abscesses in the spinal canal or musculature
CT: supplements assessment of bony destruction and serves surgical planning
Targeted biopsy: if the pathogen remains unknown and the patient is stable, CT-guided or open sampling is sought before starting empirical antibiotic therapy
Source investigation: depending on the situation, a search for the infectious source, including cardiac ultrasound to rule out endocarditis

A principle from international guidelines: in stable patients without sepsis and without neurological deficits it is worthwhile waiting for pathogen identification before starting antibiotics — a too-early, non-targeted therapy can obscure the diagnosis without controlling the infection.

Berbari EF et al. Clin Infect Dis. 2015 (IDSA guideline). PMID 26229122

The new AO Spine classification: a shared language for treatment

For spinal infections, what has long been standard for vertebral fractures was missing: an internationally uniform system describing the injury — here the infection — in such a way that transparent treatment decisions follow. Which spondylodiscitis can be treated conservatively, which requires surgery? Without a shared language, these decisions were made very differently from institution to institution.

The AO Spine Knowledge Forum Trauma & Infection has therefore developed the AO Spine classification for primary spinal infections — a new system intended to standardise decision-making worldwide and improve communication between treatment teams. Additionally, we have validated the mSISS (Modified Spinal Instability Spondylodiscitis Score) — a tool that answers the central question of mechanical stability in spondylodiscitis in a structured way. I am involved in both projects within the Knowledge Forum — the criteria on which I base my consultations here in Bern are the same ones being researched and implemented internationally.

For patients this means: the recommendation for or against surgery is not based on any individual surgeon's instinct, but on a systematic assessment of pathogen, neurology, abscess formation, bony destruction, and stability.

AO Spine Primary Infection Classification System — classification overview and toolkit · Schömig F et al. Global Spine J. 2026. PMID 42189054

Treatment: from the old to the new paradigm

Antibiotics are always the foundation

Every spondylodiscitis is treated with antibiotics — ideally targeted at the identified pathogen. Treatment typically lasts around six weeks: a large randomised trial showed that six weeks is non-inferior to twelve weeks of therapy. Depending on pathogen, immune status, and clinical course, longer treatment may be required in individual cases. Monitoring is done jointly with the infectious disease team; progress is assessed clinically and via inflammatory markers.

Bernard L et al. Lancet. 2015. PMID 25468170

When conservative treatment is sufficient

For some patients, treatment without surgery is appropriate: when the pathogen is known, there are no neurological deficits, no relevant abscess is present, the spine is stable, and the infection responds to antibiotics. This includes appropriate analgesia, early mobilisation — with an orthosis if needed — and close clinical, laboratory, and imaging monitoring. If inflammatory markers, pain, stability, or neurology deteriorate, the strategy is reassessed.

The paradigm shift: what early surgery can achieve

Historically, surgery was performed only when there was no other option — with paralysis, sepsis, or structural collapse. The accumulated evidence of recent years challenges this restraint: a meta-analysis of nearly 11,000 patients from 21 studies found mortality of 8 versus 13 percent with conservative treatment among early-operated patients, around 40 percent fewer recurrences and treatment failures, and a mean hospital stay almost eight days shorter. Surgery achieves what antibiotics alone reach poorly: removal of infected, destroyed tissue and abscesses — and restores the stability needed for early pain-free mobilisation.

An important caveat: these data come predominantly from observational studies, not randomised comparisons — sicker or indeed fitter patients may preferentially have been operated on. Older systematic reviews found comparable outcomes by both pathways in correctly selected patients. This does not justify an automatic decision to operate; but equally not an automatic decision for conservative treatment. The current European consensus of neurosurgical societies moves in the same direction of a more active surgical stance for defined situations.

Thavarajasingam SG et al. Sci Rep. 2023. PMID 37730826 · Rutges JP et al. Eur Spine J. 2016. PMID 26585975 · Kramer A et al. Brain Spine. 2024 (EANS consensus). PMID 39866360

When surgery is clearly indicated

Neurological deficits from compression of the spinal cord or nerve roots, particularly with an epidural abscess — here time is of the essence,
Sepsis or uncontrollable infection despite targeted antibiotics,
Instability or progressive malalignment from destruction of vertebrae and disc — assessed in a structured manner using the mSISS,
No pathogen identification when blood cultures and biopsy yield no result and therapy needs a target,
Failure of conservative treatment with persistent pain, rising inflammatory markers, or inability to mobilise.

How surgery is performed

The core principles are removal of infected tissue (débridement), decompression of compromised nerves, and stabilisation of the affected segment — depending on the situation posteriorly, via small approaches with percutaneous screw-rod instrumentation under fluoroscopy, or combined with anterior column reconstruction. Contrary to a widespread concern, placing implants in the infected area is, with correct technique, established and safe. In our own series of 73 patients with erosive lumbar spondylodiscitis, single-stage posterior management with débridement, stabilisation, and PMMA cement interposition was safe and effective — one procedure instead of several, with early mobilisation.

After surgery, targeted antibiotic therapy continues; mobilisation and progressive loading begin early and are monitored clinically and by laboratory values.

Deml MC et al. Bioengineering. 2022. PMID 35200426

My approach

Spinal infections are, alongside trauma and osteoporotic fractures, my third research focus — the treatment principles derive directly from this work:

Think of it early: Back pain with fever, night and rest pain, or elevated inflammatory markers is investigated in a targeted manner — MRI and blood cultures before valuable time is lost.
Pathogen before therapy: In stable patients the pathogen is identified before antibiotics are started — targeted treatment outperforms empirical treatment.
Structured classification: Neurology, abscess, destruction, and stability are assessed using the new AO Spine classification and the mSISS — the same criteria we develop and validate in the Knowledge Forum.
Interdisciplinary treatment: Therapy is developed jointly with infectious disease and — where needed — other disciplines; the duration of antibiotics is guided by evidence, not habit.
Surgery with a clear goal: When the criteria are met, early and decisive — débridement, decompression, stabilisation, as tissue-sparing as possible and single-stage where feasible. Hesitation worsens the starting point in this condition.

What the research shows — and what I am working on

The new AO Spine classification for primary spinal infections

Despite increasing incidence, a globally accepted classification system for primary spinal infections was lacking until recently. The AO Spine Knowledge Forum Trauma & Infection has closed this gap: the new classification standardises the description of the infection and the resulting decision-making — analogous to the established AO classifications for spinal injuries. The toolkit is freely available and is being implemented internationally.

AO Spine Primary Infection Classification System — overview and downloads

Making instability measurable: the mSISS

Whether spondylodiscitis mechanically destabilises the spine is one of the key questions for the surgical indication. In a multicentre study we adapted and validated the Spinal Instability Spondylodiscitis Score — as a reproducible instrument that structures and standardises this assessment.

Schömig F et al. Global Spine J. 2026. PMID 42189054

From research to practice: Knowledge Forum recommendations

In a review article by the Knowledge Forum we appraised the most impactful studies on pyogenic spondylodiscitis and translated them into clinical recommendations — from diagnostics through pathogen identification to the choice between conservative and operative treatment. Separate recommendations exist for tuberculous spinal infection.

Bigdon SF et al. Global Spine J. 2025. PMID 39852953 · Scherer J et al. Global Spine J. 2026. PMID 42533326

Early surgery: mortality, recurrence, and hospital stay

The meta-analysis by Thavarajasingam and colleagues (21 studies, 10,954 patients) found mortality of 8 versus 13 percent with early surgical treatment, a reduction in recurrences and treatment failure of around 40 percent, and a hospital stay shorter by an average of nearly eight days. Because randomised trials are lacking, bias from patient selection remains possible — but the consistency of the findings across endpoints supports the trend towards a more active surgical stance, which the European EANS consensus also reflects. The older systematic review by Rutges, which rated both pathways as equivalent, marks the historical starting point of this development.

Thavarajasingam SG et al. Sci Rep. 2023. PMID 37730826 · Kramer A et al. Brain Spine. 2024. PMID 39866360 · Rutges JP et al. Eur Spine J. 2016. PMID 26585975

Six weeks of antibiotics is usually sufficient

The French randomised trial by Bernard and colleagues demonstrated in pyogenic vertebral osteomyelitis that six weeks of antibiotic therapy is non-inferior to twelve weeks. Together with the IDSA guideline it forms the basis of today's shorter and more targeted antibiotic regimens.

Bernard L et al. Lancet. 2015. PMID 25468170 · Berbari EF et al. Clin Infect Dis. 2015. PMID 26229122

Single-stage management of erosive spondylodiscitis

In our Bern series of 73 patients with pyogenic erosive lumbar spondylodiscitis, single-stage posterior management — débridement, stabilisation, and PMMA cement interposition — proved safe and effective. It avoids multi-stage procedures and enables the early mobilisation that is decisive for these often debilitated patients.

Deml MC et al. Bioengineering. 2022. PMID 35200426

Questions about your situation?

Whether an unclear diagnosis, ongoing antibiotic therapy, or the question of whether a recommended operation is really necessary: describe your situation and send — where available — your findings, in particular MRI images, laboratory trends, and microbiological results. After review I will summarise my assessment in understandable terms — including as a second opinion.

Request assessment Contact

Written and medically reviewed by Sebastian Bigdon, MD · Last updated: 6 October 2026 · The sources used are cited with PubMed references in the section «What does the research say?».

Frequently Asked Questions

Can spondylodiscitis be cured?

Yes, in most cases. The prerequisites are timely diagnosis, targeted antibiotic therapy of sufficient duration, and — where necessary — surgery. The later the infection is recognised and the more severe the co-morbidities, the higher the risks. This is why unclear, worsening back pain with fever, night pain, or elevated inflammatory markers should be investigated promptly.

Does spondylodiscitis always require surgery?

Not always. When the spine is stable, the pathogen is known, there are no neurological deficits, and no relevant abscess is present, treatment with antibiotics and monitoring may be sufficient. More recent data show, however, that early surgery is associated with lower mortality, fewer recurrences, and shorter hospital stay in many patients. The decision is made in a structured and individual manner.

How long do I need antibiotics?

Usually around six weeks. A large randomised trial showed that six weeks is non-inferior to twelve weeks of treatment. Depending on the pathogen, immune status, implants, and clinical course, a longer course may be necessary. Duration is agreed jointly with the infectious disease team.

How dangerous is spondylodiscitis?

It is a serious condition: untreated or recognised too late it can lead to sepsis, permanent nerve damage, or spinal instability; mortality rates of up to around 20 percent are reported in study cohorts. With timely diagnosis and consistent treatment, however, the infection heals in most patients.

Where does the infection in the spine come from?

In most cases bacteria reach the disc and vertebral body via the bloodstream — originating for example from skin, urinary tract, or dental infections, catheters, or endocarditis. The most common pathogen is Staphylococcus aureus. Older age, diabetes, immune deficiency, renal disease, and prior procedures are predisposing factors.

What is an epidural abscess?

A collection of pus in the spinal canal between the bone and the dural membrane. It can compress the spinal cord or nerve roots and rapidly cause paralysis. An epidural abscess with neurological deficits is an emergency and is generally an indication for urgent surgery.

How long does recovery take?

Antibiotic therapy usually lasts six weeks; full recovery often takes three to six months — depending on the extent of the infection, treatment, and general condition. Inflammatory markers, symptoms, and imaging are monitored during follow-up. Many patients regain a good activity level; patience during the rehabilitation period is, however, part of the treatment.

Questions about your condition?

I take the time to review your imaging and provide a clear, written assessment – including as a second opinion.

Request appointment