Fungal Osteomyelitis

Fungal osteomyelitis is bone infection caused by pathogenic fungi, typically in an immunocompromised host or after exposure to endemic fungi in specific geographic regions. It is rare, less than 1% of all osteomyelitis, but increasing as the immunocompromised population grows: HIV, transplant recipients and patients on chemotherapy. Candida is the most common cause, 40-60% of fungal osteomyelitis.

Why it is missed. The course is indolent and the symptoms nonspecific, so diagnosis is delayed, on average 6 months from symptom onset. Examiners expect you to recognise the risk factors, order the right investigation, which is tissue biopsy, and know how long treatment takes.

Mortality. Aspergillus mortality is quoted at 50-80%, against 10-20% for the other fungi with treatment. In the 180-case review of Aspergillus osteomyelitis, overall mortality was 25% (see the Evidence Base).

Routes to bone. The fungi use the same three routes as bacteria:

• Haematogenous spread, the most common (Candida, Cryptococcus)
• Direct inoculation, through trauma or surgery
• Contiguous spread from adjacent soft-tissue infection

The tissue response. Unlike bacterial osteomyelitis, fungal infection typically provokes granulomatous inflammation with tissue destruction, and imaging shows minimal periosteal reaction.

Who is at risk. Each risk factor points towards particular organisms:

• HIV/AIDS with a CD4 count below 200: Cryptococcus and Histoplasma
• Chemotherapy and haematological malignancy: neutropenia allows fungal invasion, by Aspergillus and Candida; an absolute neutrophil count below 500 is the Aspergillus risk
• Organ transplant: immunosuppression allows opportunistic fungi, with Aspergillus in the first 6 months after transplant
• Chronic oral steroids: cell-mediated immunity is impaired
• Diabetes mellitus: impaired neutrophil function, and a risk for Candida and mucormycosis
• IV drug use: Candida vertebral osteomyelitis
• Central lines, dialysis and TPN: a portal of entry for Candida and a risk of Candida bloodstream infection
• Soil exposure in endemic areas: Coccidioides, Blastomyces and Histoplasma

CandidaAspergillusEndemic FungiCryptococcus

The organism. C. albicans causes 50-70% of cases; C. tropicalis, C. glabrata and C. parapsilosis make up the rest. Candida is a normal commensal and becomes pathogenic when host defences are impaired.

How it reaches bone. Haematogenous spread from candidaemia, and the vertebra is the most common site. The host shapes the pattern: IV drug users develop it in the lumbosacral spine and the sternoclavicular joint, while patients on TPN or with central lines can seed any bone. The course is indolent, with back pain and a low-grade fever.

IV Drug User with Back Pain

Chronic back pain and a low-grade fever in an IV drug user should prompt consideration of Candida vertebral osteomyelitis. Blood cultures are positive in less than 50%, so tissue biopsy is essential.

Aspergillus is Aggressive

Aspergillus osteomyelitis carries substantial mortality, particularly in severely immunocompromised patients. Start voriconazole promptly, obtain tissue when feasible, and involve the relevant surgical team early for source control, decompression or stabilisation. Do not delay antifungal therapy in unstable or rapidly progressive invasive disease.

The organism. A ubiquitous environmental mould, acquired by inhaling conidia. A. fumigatus is the most common species (90%); A. flavus, A. niger and A. terreus account for the rest. It is angioinvasive, causing vascular thrombosis and necrosis. Bone destruction on imaging is aggressive, and mortality is high even with treatment.

Who gets it. The risk factors are these:

• Prolonged neutropenia, more than 10 days
• Allogeneic stem cell transplant
• Solid organ transplant, especially lung
• High-dose corticosteroids
• Chronic granulomatous disease (CGD)
• Advanced AIDS, less commonly

How it reaches bone. The site follows the route:

• Skull base by spread from the paranasal sinuses
• Vertebrae by haematogenous spread from pulmonary aspergillosis
• Ribs and sternum by contiguous spread from pulmonary involvement

View larger

Travel and exposure history are crucial to the diagnosis of the endemic mycoses, so always ask about exposure in an endemic area.

Coccidioides immitis/posadasii

Endemic Region

Southwestern USA (Arizona, California, Nevada, New Mexico), Mexico

Exposure Source

Soil - dust inhalation

Blastomyces dermatitidis

Endemic Region

Mississippi/Ohio River Valley, Great Lakes

Exposure Source

Soil, decaying wood and organic matter - inhalation

Histoplasma capsulatum

Endemic Region

Ohio/Mississippi Valley, Central America

Exposure Source

Bird/bat droppings - cave exploration (spelunkers), chicken farmers

Geographic Distribution of Endemic Fungi

Organism	Endemic Region	Exposure Source
Coccidioides immitis/posadasii	Southwestern USA (Arizona, California, Nevada, New Mexico), Mexico	Soil - dust inhalation
Blastomyces dermatitidis	Mississippi/Ohio River Valley, Great Lakes	Soil, decaying wood and organic matter - inhalation
Histoplasma capsulatum	Ohio/Mississippi Valley, Central America	Bird/bat droppings - cave exploration (spelunkers), chicken farmers

Coccidioidomycosis (Valley fever). Filipino or African ancestry, pregnancy and immunocompromise all increase the risk. Bone is involved in 10-50% of disseminated cases, and the course is chronic and relapsing.

Blastomycosis. Bone is involved in 25-50% of extrapulmonary cases.

Histoplasmosis. Bone involvement is uncommon, less than 5% of disseminated cases.

Two species, two hosts. C. neoformans affects the immunocompromised, in HIV/AIDS. C. gattii can cause osteomyelitis in immunocompetent hosts in endemic regions, including Australia, the Pacific Northwest and tropical Asia; consider it in any patient with a lytic bone lesion mimicking tumour. Outcome is variable and depends on the immune status of the host.

The organisms side by side. Treatment by organism is set out under Management.

Symptoms. Pain is the most common symptom (90%): usually dull, aching and progressive, and it may be present for weeks to months before diagnosis. Swelling is variable and often minimal early, and motion is limited if the focus is near a joint. Systemic symptoms are a low-grade fever below 38.5°C that is often absent, night sweats (especially with the endemic mycoses), weight loss, malaise and fatigue.

How it differs from bacterial osteomyelitis. The course is indolent. Symptoms last months rather than weeks before diagnosis, the presentation is less acute, high fever is less common, and CRP and ESR are often only mildly elevated. Failure to respond to antibiotics is the key clue to a fungal cause. How the infection presents varies with the organism and the immune status of the host.

Examination. Locally there is tenderness over the affected bone, soft-tissue swelling that is often subtle, warmth less pronounced than in bacterial infection, reduced movement of adjacent joints and, in chronic cases, a draining sinus. Look beyond the bone for:

• A low-grade fever
• Signs of the underlying immunocompromise
• Lymphadenopathy, with the endemic mycoses
• Pulmonary findings, if the primary infection is in the lung
• Skin lesions of disseminated infection (Blastomyces, Cryptococcus)

When to think fungal. Index of suspicion is key, and any of these should prompt a fungal work-up:

• Osteomyelitis not responding to appropriate antibiotics
• Culture-negative bone infection
• An immunocompromised patient (HIV, transplant, chemotherapy) with a bone lesion
• Travel to an endemic area and bone symptoms
• An IV drug user with vertebral osteomyelitis

GMS and PAS stain all fungal walls; add mucicarmine for the cryptococcal capsule. The organisms divide into yeasts, moulds and one dimorphic exception.

Yeasts are round or oval cells:

• Candida: oval budding yeast plus pseudohyphae, which can form in tissue
• Cryptococcus: narrow-based budding yeast with a thick capsule (mucicarmine-positive, clear halo)
• Blastomyces: broad-based budding yeast
• Histoplasma: small intracellular yeast within macrophages
• Sporothrix: cigar-shaped yeast

Moulds form hyphae. Aspergillus and the Mucorales are told apart by septation and branching, a distinction that directs very different therapy: voriconazole versus amphotericin.

• Aspergillus: narrow, septate hyphae with acute-angle (45-degree) branching
• Mucorales: broad, aseptate (or sparsely septate), ribbon-like hyphae with wide-angle branching

The dimorphic exception. Coccidioides in tissue forms large spherules packed with endospores, neither a yeast nor a hypha, and they are pathognomonic.

View larger

View larger

View larger

View larger

The organism and the host. Mucormycosis (Rhizopus, Mucor and related Mucorales) is angioinvasive like Aspergillus, causing vascular thrombosis and tissue infarction and necrosis (black eschar). The classic host has diabetes with ketoacidosis, and neutropenia, transplant, and iron overload or deferoxamine also predispose. It reaches bone from rhino-orbital-cerebral disease (skull base, maxilla) or through a contaminated wound.

View larger

View larger

Diagnosis needs tissue. Beta-D-glucan and galactomannan are both negative with Mucorales, so serology does not help. The biopsy separates it from Aspergillus (see Telling the Fungi Apart on the Biopsy).

View larger

Treatment is amphotericin and urgent radical surgery, not voriconazole. First-line is liposomal amphotericin B with early, aggressive surgical debridement, because the infarcted tissue is avascular and drugs cannot reach it. Mortality is high.

• Reverse the underlying state: correct ketoacidosis, reduce immunosuppression, stop deferoxamine
• Isavuconazole or posaconazole for step-down or salvage
• Mucorales are intrinsically resistant to voriconazole, so breakthrough mucormycosis can occur in a patient on voriconazole prophylaxis

Global Epidemiology

Fungal osteomyelitis accounts for under 1% of all osteomyelitis but is rising worldwide with expanding immunocompromised populations and global travel. Geography drives the organism:

• Candida - the leading fungal cause globally; linked to candidaemia, central venous catheters, parenteral nutrition, and injecting drug use.
• Aspergillus - worldwide environmental mould; bone disease clusters in haematology/transplant centres but also occurs after trauma or surgery in immunocompetent hosts.
• Endemic mycoses are geographically anchored: Coccidioides (southwestern USA, Mexico, parts of Central/South America), Blastomyces and Histoplasma (Mississippi/Ohio River valleys, Great Lakes, and focally in Africa/Asia), Talaromyces marneffei (Southeast Asia in advanced HIV).
• Cryptococcus gattii - historically associated with tropical/subtropical regions and eucalyptus, with well-described endemicity in Australia and a notable Pacific Northwest (Vancouver Island) outbreak; unlike C. neoformans it can affect immunocompetent hosts and occasionally causes lytic bone lesions mimicking tumour.

Major Guidelines Side by Side

Registry & Surveillance Notes

• No arthroplasty-style registry captures fungal osteomyelitis; the evidence base is pooled international case reviews (e.g. the International Osteoarticular Mycoses Study Consortium series for Candida and Aspergillus).
• National antifungal-resistance surveillance (e.g. CDC, ECDC/EUCAST, and the WHO Fungal Priority Pathogens List, 2022) increasingly tracks azole-resistant A. fumigatus and emerging multidrug-resistant Candida (incl. C. auris) - relevant when empirical azole therapy fails.

High- vs Limited-Resource Practice Variation

Across all settings, involve infectious diseases early for organism-directed therapy, duration, and management of drug interactions and toxicity.

Related pages: Osteomyelitis Pathophysiology is the parent framework, and the contrast is the point - fungal bone infection follows the same routes (haematogenous, contiguous, direct inoculation) but declares itself over months, with an indolent course and normal or near-normal inflammatory markers that make it look like anything except infection. Fungal Spine Infections covers the vertebral disease that dominates the adult series on this page, and Atypical Mycobacterial Infection with Tuberculosis of the Spine (Pott's Disease) and Brucellosis of the Spine are the rest of the chronic granulomatous, culture-negative differential - the single most useful habit on all four pages is to tell the laboratory what you suspect, because none of these organisms grows on a standard five-day plate. Periprosthetic Joint Infection is where Candida arrives on an implant and where the decision to remove the prosthesis is effectively made by the diagnosis, with Biofilm Formation explaining why. Septic Arthritis Pathophysiology matters because Candida septic arthritis coexisted with osteomyelitis in 21% of the largest series. Orthopaedic Antibiotic Therapy and Antibiotic Pharmacology and Resistance carry the azole pharmacology, the therapeutic drug monitoring that voriconazole requires and the interactions that make it difficult in transplant patients. Systematic Approach to Infection Imaging covers the MRI appearances, and Diabetic Foot Infections is the setting in which an indolent, non-healing osteomyelitis most often turns out to be fungal.

Reading the response rates together

Two outcome figures on this page look as though they compare, and they do not. The Candida series reports a complete response in 32%; the non-Aspergillus mould review reports an overall response in 85.8%. Taken at face value this says Scedosporium does better than Candida, which is almost certainly false. Three things separate them:

1. Different endpoints. Complete response is a strict endpoint; overall response includes partial responses. They are not the same measurement and cannot be placed side by side.
2. Different publication pressures. Both are reviews of published cases, but the rare-mould literature is smaller and newer, so a successfully treated Scedosporium osteomyelitis is a publishable case and a failure often is not.
3. Different mechanisms, and therefore different patients. Mould osteoarticular infection is 54.5% direct inoculation - often a healthy child with a puncture wound and one surgically accessible focus. Candida osteomyelitis is two-thirds haematogenous, in patients with lines, prostheses and comorbidity, and frequently in more than one bone.

What both datasets agree on, and what should be carried forward, is the direction of the surgical signal: combined medical and surgical management outperformed antifungal therapy alone in every series here, and relapse is driven by inadequate duration - months, not weeks.