As the vast majority of Bacillus species are non-pathogenic and ubiquitous in the environment, many clinical microbiologists and clinicians dismiss Bacillus species cultured from clinical specimens as contaminants. Occasionally, this results in a missed diagnosis and inappropriate clinical decision-making. This case is important, because it illustrates the importance of communication between clinicians and the clinical laboratory staff in determining the significance of culture results. This report aims to educate healthcare workers on Bacillus cereus joint infections and further endeavours to assist healthcare practitioners in distinguishing when this organism should be dismissed as a contaminant and when it should be considered as a pathogen.

Written, informed consent was obtained from the patient. This research was approved by the University of Pretoria, Faculty of Health Sciences, Research Ethics Committee (ethics reference number 133/2019).

In November 2018, a 32-year-old male was referred to a tertiary academic hospital in Pretoria, South Africa, with a non-resolving septic arthritis of his right knee. The patient presented to a secondary hospital 10 days prior with a tender, swollen right knee, with no history of trauma. He underwent an arthroscopy at that centre and received intravenous amoxicillin-clavulanic acid, with a suboptimal response. In addition, he developed symptoms suggestive of systemic lupus erythematosus (SLE), which included polyarthritis, xerostomia, Raynaud’s phenomenon, proteinuria and confusion. On examination he was haemodynamically stable with a pulse rate of 100 beats per minute, a blood pressure of 117/81 mmHg and a temperature of 36.5 °C. He had asymmetric polyarthritis, involving the right elbow, left wrist, right knee and both ankles. The right knee was the worst affected, with swelling, erythema and tenderness on examination. An emergency arthroscopy of the right knee was performed, revealing a purulent effusion. Intravenous ceftriaxone (1 g twice daily) was started empirically. X-rays of all affected joints revealed no accompanying osteomyelitis. No other imaging of the joints was done. The diagnosis of SLE was confirmed based on a Systemic Lupus International Collaborating Clinics score¹ of five (anti-nuclear antibody positive, lupus nephritis class 3, arthritis, low C3 and neurologic SLE).

Based on the report, the patient was started on intravenous vancomycin 1 g twice daily. The dose was adjusted to achieve a target vancomycin blood trough level of 15 mg/mL – 20 mg/mL. Trough levels were monitored roughly every 3 days over the treatment period, and the dosage adjusted accordingly. No vancomycin-related adverse events occurred during this time. Vancomycin was continued for 4 weeks; over this period, the pain and swelling improved dramatically, inflammatory markers normalised and the patient regained mobility. Management of the SLE included prednisone, mycophenolate mofetil and chloroquine, with good response.

Bacillus species are Gram-positive, aerobic or facultative anaerobic sporulating bacilli, which are ubiquitous in the environment.³ Over 100 species are known to belong to the genus.³ Bacillus cereus is a common cause of food poisoning and occasionally causes opportunistic infections, usually in vulnerable hosts. These infections include ophthalmic infections, wound infections, septicaemia, endocarditis, meningitis, necrotising pneumonia and orthopaedic infections.^3,4 Important virulence factors of B. cereus include production of toxins and the formation of biofilms and spores.⁴ Except for B. cereus and B. anthracis, the genus Bacillus is rarely associated with disease.^3,5,6,7 Therefore, Bacillus species are often reported as contaminants, even when cultured from sterile specimens. This may result in a delay in correct diagnosis and inappropriate treatment.

Bacillus cereus grows well on most routine culture media such as blood agar (Figure 1), chocolate agar and MacConkey agar. Routine laboratory tests will reveal large boxcar-shaped, Gram-positive bacilli (Figure 2) that are catalase positive. However, identification to species level requires specialised techniques such as matrix-assisted laser desorption/ionisation mass spectrometry or molecular techniques.⁷

Musculoskeletal infections caused by B. cereus are rare but have been previously reported in literature.^4,5,6,7 Åkesson, Hedströum and Ripa.⁴ reported 12 cases of B. cereus orthopaedic infections, all occurring in post-operative or post-traumatic wounds, whilst Dubouix et al.⁵ reported 41 cases with B. cereus wound infections associated with open fractures. Gallo et al.⁶ reported two cases of B. cereus prosthesis-related septic arthritis, which were culture negative but identified using PCR-mass-spectrometric-technology and fluorescence in situ hybridisation of tissue. Ha et al.⁷ reported a single case of late prosthetic joint infection with B. cereus that occurred 13 years after total hip replacement surgery, confirmed by 16S ribosomal ribonucleic acid (rRNA) sequencing. In total, we found 56 cases reported over 25 years, highlighting the rarity of B. cereus musculoskeletal infections and emphasising the difficulty in making a definitive diagnosis. The scarcity of these infections may actually be the result of under-reporting, as Bacillus species are often reported as contaminants.

Bacillus cereus produces β-lactamases and is resistant to most β-lactam antibiotics, except carbapenems.³ Classically, Bacillus species are susceptible to vancomycin, aminoglycosides and fluoroquinolones, whilst resistance to erythromycin, tetracycline and even carbapenems have been reported.³ The empiric antibiotic of choice for invasive B. cereus infections is vancomycin.^3,7 In addition to administering antibiotics, it is imperative to obtain source control at the infected site, as B. cereus is known to form biofilm. The removal of implanted medical devices may be necessary.⁷

The concomitant SLE may be considered a notable risk factor for invasive B. cereus infection in this patient. It is uncertain if the Epstein–Barr virus viraemia played a significant role in predisposing the patient further to this infection. However, both SLE and Epstein–Barr virus infection are known to be immune–modulatory, down-regulating the innate and humoral systems and placing the patient at increased risk of opportunistic infections.^8,9 An additional risk factor was the preceding arthroscopy, which may have introduced spores into the joint space. To our knowledge, this is the first case of septic arthritis in an SLE patient where the organism was cultured from the joint specimen. The communication between the clinical and microbiology teams ensured that the organism was identified to species level and that antibiotic susceptibility testing was performed, resulting in a favourable outcome for the patient.