Introduction

Periprosthetic joint infection (PJI) is a serious complication of total joint arthroplasty, occurring in 1-2% of cases [1,5]. PJI is associated with significant morbidity and healthcare costs [1,2]. Late-onset PJIs, presenting more than 24 months after the index surgery, are often hematogenous in origin, with the oral cavity being a potential source [3,4]. Periodontitis, a chronic inflammatory condition affecting the supporting structures of teeth, has been associated with systemic inflammation and an increased risk of PJI [6,7]. Streptococcal species are the second most common pathogens in PJIs, but Group B Streptococcus (GBS), though only accounting for 5-6% of these, being particularly virulent [4,8,9]. We present a case of late-onset GBS PJI associated with severe periodontal disease, successfully treated with revision knee surgery after dental clearance. The case highlights the importance of identifying and addressing potential sources of infection in the management of PJI.

Case Presentation

A 73 year old, obese (BMI 34.1kgm-2) female self-referred to a private orthopaedic clinic with a painful, swollen and unstable left knee. She had a one-year history of these progressively worsening symptoms that was affecting her ability to mobilise without walking aids.

She had undergone uneventful left total knee replacement (TKR) surgery nine years prior and had no ill effects until one year prior, after she had spent one month in a private hospital where she was being treated for coronary artery disease. She underwent successful cardiac stent placement during that admission. Following discharge, her knee became painful and swollen. She denied any traumatic or systemic illnesses which may have accounted for the symptoms.

Her past medical history included hypertension, atrial fibrillation and coronary artery disease. On direct questioning, she denied suffering with diabetes mellitus, periodontal disease, autoimmune, malignant or immune deficiency conditions, and potential risk factors for periprosthetic joint infections (PJI)).

On examination, she walked with an antalgic gait and had a left Varus thrust. For the last three months, she reported, that she used a Zimmer frame to mobilize. Prior to that she walked unaided. There was a well healed midline knee surgical scar and no obvious stigmata of infection. On palpation, there was a grade 2 effusion, medial joint line tenderness, mild warmth, and significant Varus-valgus instability on direct testing. The range of motion (ROM) was limited to 10° to 87° flexion as measured with a manual goniometer. Her vital signs were all within normal range.

Based on the original history and clinical examination, a provisional diagnosis of TKR failure due to aseptic loosening was considered. Radiographs ordered at the time (Figure 1) confirmed tibial component loosening. Plans were made for an elective, single-stage revision total knee replacement (TKR) three months later, pending the results of routine pre-operative serological, cardiology and anaesthetic testing and pending her insurance company’s approval.

Figure 1: Lateral and Antero-Posterior Views of the Loose Left Total Knee Replacement. Arrows are highlighting areas of loosening around the tibial component

However, six weeks later, when the patient returned to the clinic for routine pre-operative MRSA screening swabs, she was noted to be septic. Her temperature was 38.1°Celsius, heart rate was 126 beats-per-minute (bpm), respiratory rate was 28 breaths- per-minute, and oxygen saturation on room air was 90% and her blood pressure was164/102bpm. Her left knee was now tensely swollen, and very hot to the touch. She had limited knee ROM with pain throughout the range. The neurovascular assessment was normal. The skin was intact and there was no evidence of sinuses or ulcers. Arthrocentesis was performed under aseptic conditions and more than 100ml of frank pus with mild blood staining was aspirated from the left knee (Figure 2). It was sent to the nearest laboratory for urgent Gram Staining, microscopy, culture, and sensitivity investigations.

Figure 2: Pus sample aspirated from the left knee.

Immediately, the definitive diagnosis was revised to septic loosening due to late onset PJI. The management priorities therefore changed. The priority was to surgically remove the infected prosthesis and the biofilm before the patient developed septic shock. The second priority was to accurately identify the pathogenic organism, and its origin, to help guide the therapeutic regime. The patient was therefore admitted to hospital and started empirically on Vancomycin intravenously because the Gram Stain of the aspirated pus revealed 45-50 WBC’s/OIF and Gram-positive cocci, thereby confirming bacterial PJI. Blood investigations (Figure 4) were also suggestive of infection. The patient’s first surgery (day 0) revealed infected medial subcutaneous soft tissues and a sinus tract leading to the biofilm covered medial proximal tibial bone and the loose tibial implant (Figure 3). All knee replacement implants were removed, biofilm debrided, and samples were sent for microbiology tests. A static antibiotic-impregnated cement spacer was fashioned with 160mg VersaBond™ AB with additional 4 grams of Vancomycin powder and implanted before closure and application of a hinged knee brace locked in extension.