Cutibacterium acnes Shoulder Periprosthetic Infection

Cutibacterium acnes (formerly Propionibacterium acnes) is the most common causative organism in shoulder periprosthetic joint infection, responsible for approximately 30-60% of all shoulder PJIs and overtaking Staphylococcus aureus and the coagulase-negative staphylococci. Its indolent, low-virulence nature means it frequently presents with normal inflammatory markers and no systemic features, which makes it a diagnostic trap. Examiners test it because it illustrates the principle that normal bloods do not exclude infection.

The organism. A Gram-positive anaerobic bacillus and a skin commensal of the deep hair follicles and sebaceous glands of the upper trunk. It is slow-growing, requiring 14-21 days of anaerobic culture; it forms biofilm on implant surfaces; and its virulence is low, with minimal local inflammation and typically no systemic response.

Who gets it. Males are affected significantly more than females, likely because of the greater pilosebaceous density of the upper trunk, shoulder and proximal arm. The other risk factors are prior shoulder surgery, haematoma and prolonged operative time. Recognition has increased as improved culture techniques and longer incubation have revealed the true prevalence.

Inoculation. The shoulder region has the highest density of pilosebaceous units harbouring C. acnes on the human body, and the organism resides deep within them, where standard chlorhexidine or povidone-iodine skin preparation does not fully eradicate it. It can therefore be inoculated into the surgical field despite standard skin preparation: intra-operative contamination rates of 15-30% have been reported from tissue cultures taken during primary shoulder arthroplasty.

Biofilm. Once in contact with an implant surface, metal or polyethylene, C. acnes forms a biofilm, a structured extracellular polymeric matrix that protects the bacteria from both the host immune response and antibiotic penetration. Within it the organism exists in a metabolically quiescent state that is resistant to cell-wall active antibiotics, and the minimum biofilm eradication concentration (MBEC) can be 100-1000 times the planktonic minimum inhibitory concentration (MIC). This is why antibiotics alone rarely cure established implant-related infection, and why surgical intervention with component removal or exchange is usually required for eradication.

View larger

Prevention. The reservoir is the dermis rather than the surface, so contamination is the problem to attack before it becomes an infection, and the examinable point is what actually reduces it:

• Topical benzoyl peroxide, applied for several days pre-operatively, is the single most effective measure: it penetrates the follicle and significantly lowers the C. acnes skin and dermal burden
• Standard chlorhexidine/alcohol skin prep is necessary but insufficient: even after a textbook prep, dermal punch-biopsy cultures remain positive in a substantial proportion of patients
• Hydrogen peroxide applied to the skin and dermis after the incision has been shown to reduce C. acnes culture positivity from the wound
• Standard prophylactic cefazolin covers C. acnes, which is penicillin- and cephalosporin-sensitive; some units add clindamycin or vancomycin in high-risk cases, though the evidence for the addition is limited
• Male sex cannot be modified, so the modifiable levers are the skin and dermal antiseptic protocol, meticulous technique, and minimising operative time and haematoma

The trap is assuming that a normal-looking surgical prep has sterilised the field. The dermal reservoir means the field is never truly sterile for C. acnes, which is why these adjuncts and a low threshold for intra-operative culture both matter.

History. Pain is the most common presenting symptom: persistent or worsening pain after shoulder arthroplasty, with progressive loss of range of motion, presenting months to years after the index surgery. There is typically nothing to suggest infection, with no erythema, warmth or drainage from the wound and no fever, rigors or malaise. Delayed diagnosis is common, with an average delay of months to years from symptom onset.

Examination. The wound is well healed with no erythema or sinus. There may be mild tenderness around the implant, though palpation is often unremarkable; range of motion is decreased, often with pain at the end ranges; component loosening may cause subtle instability; and the neurovascular examination is usually intact, with no specific deficit.

The trap. C. acnes shoulder PJI is virtually indistinguishable from aseptic loosening on clinical and basic radiographic assessment. Both present with gradual-onset pain, stiffness and possible radiographic lucency, and the distinction depends entirely on microbiological culture with prolonged anaerobic incubation. Tissue therefore goes for prolonged culture at every revision shoulder arthroplasty, regardless of pre-operative suspicion.

When to suspect it. Always suspect C. acnes in:

• Any male patient with a painful shoulder arthroplasty and normal inflammatory markers
• Any revision shoulder arthroplasty, even if the pre-operative work-up is negative for infection
• Any "aseptic loosening" that progresses more rapidly than expected
• Persistent pain after arthroscopic or open shoulder surgery with implants in situ

Serum markers. CRP and ESR are frequently normal or only mildly elevated, and the FBC is frequently normal too, so do not be reassured by normal values: they are useful if positive and meaningless if negative. Interleukin-6 may be more sensitive than CRP for low-grade PJI but is not universally available. Procalcitonin is not sensitive for low-virulence infection and is not recommended for C. acnes screening.

Radiographs. AP, axillary lateral and scapular-Y views of the shoulder. Look for radiolucent lines at the bone-cement or bone-implant interface, component migration or subsidence and osteolysis; periosteal reaction is rare in C. acnes. The findings are often subtle and nonspecific, with only mild radiolucent lines or component loosening and no overt sign of infection, and radiographic loosening in C. acnes PJI is indistinguishable from aseptic loosening on X-ray alone. This indolent picture is a hallmark of the organism and contributes to the delay in diagnosis.

Joint aspiration. Ultrasound- or fluoroscopy-guided aspiration is preferable, because a dry tap is common without imaging guidance. Send the fluid for cell count and differential, Gram stain, and aerobic and anaerobic culture with a specific request for 14-21 day incubation. The synovial white cell count thresholds for the shoulder are lower than those for the hip and knee, as set out in the MSIS adaptation above.

Synovial biomarkers. Alpha-defensin performance varies with the assay format. The laboratory ELISA, in 105 painful shoulder arthroplasties in which 63% of the infections were C. acnes, gave a sensitivity of 75% (95% CI 53-90) and specificity of 96% (90-99); the lateral-flow point-of-care version, in a separate prospective cohort, gave sensitivity 60% and specificity 83%, and its authors concluded it "does not appear to be useful in predicting shoulder PJI". So a positive laboratory alpha-defensin is worth a lot, and a negative one does not exclude C. acnes: a quarter of infections were missed even by the better-performing assay. Leukocyte esterase is weaker still in this joint, with sensitivity 50% (95% CI 21-79) and specificity 87%, and is often unreadable because shoulder aspirates are scanty or blood-stained; it is useful to rule in and useless to rule out. Both tests are less sensitive in the shoulder than the hip and knee figures they are famous for.

Advanced imaging. CT assesses component position, version, bone stock and the pattern of osteolysis for revision planning. MRI with gadolinium may show a periprosthetic fluid collection, synovitis or abscess in low-grade PJI, and metal artefact reduction sequences are essential. Nuclear medicine (PET-CT, labelled leukocyte scan) has limited evidence specific to C. acnes shoulder PJI and may be used in equivocal cases.

Tissue culture. The single most important investigation is tissue culture with prolonged anaerobic incubation, because bloods, imaging and even aspiration may all be negative. At revision surgery, send a minimum of 5 tissue samples from different periprosthetic sites, and request sonication of any explanted components: it disrupts the biofilm and significantly increases culture yield compared with periprosthetic tissue culture alone.

Susceptibility. C. acnes is an anaerobe but is resistant to metronidazole, a classic exam trap; the agents it is usually susceptible to are set out in the table. Rifampicin is used as a combination agent for its biofilm-penetrating properties and never as monotherapy, because resistance develops rapidly.

Recurrence. The recurrence rate after revision is generally lower than for more virulent organisms, but recurrence may present late and with the same indolent pattern, so long-term clinical and microbiological surveillance is essential. Any recurrence of pain after revision should prompt repeat investigation with aspiration and prolonged culture.

Prognosis. The favourable factors are the organism itself, with its low virulence, one-stage or two-stage revision with appropriate antibiotics, adequate bone stock, a single organism and an immunocompetent patient. The poor ones are multi-organism infection, extensive bone loss requiring allograft, multiple prior revisions, an immunocompromised patient, and inadequate culture that leaves the organism unidentified. Properly managed with surgical revision and targeted antibiotics, C. acnes PJI has better outcomes than Staphylococcus aureus or fungal PJI.