Short-term outcomes of articular cement spacer in the treatment of prosthetic knee joint infections: a retrospective case series

INTRODUCTION

Osteoarthritis of the knee is a prevalent disease with a pooled global prevalence of 22.9% in individuals aged 40 years and above.[1] It is one of the five leading causes of disability in Singapore,[2] incurring significant costs to society. Joint replacement surgery in the form of a total knee arthroplasty (TKA) is the definitive treatment for patients who suffer from knee osteoarthritis and remain symptomatic despite a trial of conservative treatment. An increasing number of TKA surgeries is performed each year in Singapore, with each tertiary hospital performing up to 1,800 cases annually. Prosthetic joint infection is a devastating complication following TKA, with an incidence of up to 1%–2% and high mortality rates.[3] A two-stage revision arthroplasty is currently considered the gold standard treatment for chronic prosthetic joint infections.[4,5,6] However, there is significant morbidity and mortality associated with the procedure.[7,8] Recent studies have shown early promising results with the use of a single-stage explantation and insertion of a prosthetic and/or articulating cement spacer as definitive treatment for prosthetic joint infections in patients who may not be fit to undergo a two-stage revision arthroplasty.[5,9,10]

This case series aimed to evaluate the short-term outcomes in patients with multiple comorbidities (American Society of Anesthesiologist [ASA] Physical Classification Score ≥3] who were treated with single-stage revision arthroplasty. These patients had low premorbid functional demands and were deemed high risk for a two-stage procedure. These patients underwent an explantation of the primary TKA components and insertion of an articulating cement spacer as a definitive treatment for prosthetic joint infection of the knee.

METHODS

This study is a retrospective case series describing five patients who were treated at Tan Tock Seng Hospital, Singapore, for chronic prosthetic joint infection following a primary TKA. These patients were assessed by anaesthetists and considered to have high perioperative risk. They were offered a single-stage explantation and insertion of an articulating cement spacer as a definitive option, as well as a traditional two-stage revision arthroplasty. All five patients opted for the former in view of their high perioperative risk and relatively low functional demands.

A review of the patients’ demographics, clinical charts, operative notes and laboratory test results was performed after approval was obtained from the National Healthcare Group Domain Specific Review Board (ref no. 2021/00557)..

All five patients underwent complete synovectomy and explantation of implants followed by insertion of antibiotic-laden articulating cement spacers. Explantation was performed with single-sided reciprocating saws and osteotomes while taking care to minimise bone loss. The joint was then soaked in povidone-iodine for 3 min, after which it was irrigated with 9 L of normal saline using a pulsed lavage system. The surgical field was reinforced with new sterile drapes, and a new set of surgical instruments was utilised for the next stage of the surgery. The preparation of the cement spacers was standardised with COPAL^® knee moulds using low-viscosity Palacos cement, which contains gentamicin and 2 g of vancomycin powder for each of the femur and tibia components. The thickness of the tibial components was sized intraoperatively to achieve adequate soft tissue balancing. These cement spacers were then cemented onto the cut surfaces of the femur and tibia using moderate viscosity Palacos cement prepared with 0.5 g of vancomycin. The patients’ postoperative protocols were standardised. They were allowed partial weight-bearing for the first 6 weeks with a knee brace, followed by full weight-bearing thereafter.

CASE REVIEWS

Case 1

Our first patient is an 81-year-old man who was premorbidly home ambulant with a walking stick and community ambulant with a wheelchair. He has multiple medical comorbidities including hypertension, hyperlipidaemia, poorly controlled type 2 diabetes mellitus, ischaemic heart disease, congestive heart failure with an ejection fraction of 25%, systemic vasculitis, chronic hepatitis B and a previous subarachnoid haemorrhage, which significantly contributed to his current ambulatory status. He underwent a primary right TKA in 2009.

Eleven years after his right TKA, he presented with a 4-week history of worsening knee pain, fever and swelling. Physical examination revealed a large knee effusion with limited and painful range of motion (ROM) of 0°–90°. Initial investigations pointed to a prosthetic joint infection with a Musculoskeletal Infection Society (MSIS) score of 9. Cultures from the knee arthrocentesis and intraoperative specimens grew Escherichia coli. His ASA score was 4 and he was deemed high risk for surgery. He subsequently underwent a complete synovectomy, explantation of the primary TKA components and insertion of an articulating cement spacer. Intraoperatively, seropurulent fluid was noted in the joint with unhealthy synovium. The femur and patellar components were loose. He was treated with 6 weeks of intravenous ertapenem and is currently on long-term prophylaxis with oral trimethoprim/sulfamethoxazole.

At 1-year postsurgery, he complained of mild intermittent pain with an average pain score of 3–4 on the Visual Analogue Scale (VAS). Currently, he largely ambulates in a motorised scooter in the community and is able to use a walking frame for ambulation and transfers at home. His postoperative Oxford Knee Score is 13. His ROM is 10°–100°.

Case 2

Our second patient is an 81-year-old man who was premorbidly community ambulant without walking aids. He has multiple medical comorbidities, including poorly controlled type 2 diabetes mellitus, hypertensive heart disease with paroxysmal atrial fibrillation, hyperlipidaemia, stage 3B chronic kidney disease and peripheral arterial disease. He had undergone a primary left TKA in 2001.

He was admitted with a presenting complaint of 2 days of knee pain and swelling in 2018. Physical examination revealed a moderate knee effusion with severely limited ROM of 0°–30°. The MSIS score was 9, and the patient was treated initially with debridement, antibiotics and implant retention, in view of the 2-day history of knee pain. The joint aspirate and intraoperative cultures yielded Streptococcus dysgalactiae. Worsening of symptoms and rising inflammatory markers led to the decision to explant the prosthesis with insertion of articulating cement spacers 2 weeks later. Intraoperatively, there was copious haemopurulent collection with chalky tissue. The primary TKA components were not loose. He was treated with 6 weeks of intravenous cefazolin and is currently on long-term suppression with oral amoxicillin.

At 2 years 11 months postsurgery, he has no complaints of knee pain. He remains mostly wheelchair bound and is currently residing in a nursing home. He is able to perform simple transfers and accomplish simple activities of daily living like toileting independently. His postoperative Oxford Knee Score is 21. His ROM is 60°–90°.

Case 3

Our third patient is a 79-year-old man who was premorbidly community ambulant with a walking stick. He has multiple medical comorbidities including diabetes mellitus, hyperlipidaemia, previous cerebrovascular accident complicated by vascular dementia, paroxysmal atrial fibrillation and peripheral arterial disease. He had undergone a primary left TKA in 2011.

Ten years after his primary surgery, he was admitted with a presenting complaint of knee pain of uncertain duration as well as symptoms and signs suggestive of cellulitis. Physical examination revealed a moderate knee effusion with severely limited ROM of 20°–70°. The MSIS score was 7, and his joint aspirate grew Morganella morganii. He subsequently underwent an explantation of the primary TKA components and insertion of an articulating cement spacer. Intraoperatively, there was florid synovitis with turbid fluid and pus under the tibia component. The femoral and tibia components were not loose. He was treated with intravenous ertapenem after the surgery, but this was withheld in view of possible drug-induced liver injury.

At 9 months postsurgery, he has no complaints of knee pain. He is able to ambulate with a walking stick at home and a walking frame in the community. His postoperative Oxford Knee Score is 24. His ROM is 5°–70°.

Case 4

Our fourth patient is an 80-year-old man who was premorbidly community ambulant with a walking stick. He has multiple medical comorbidities including ischaemic heart disease with previous percutaneous coronary angioplasty, atrial fibrillation, diabetes mellitus, hypertension, hyperlipidaemia and a previous transient ischaemic attack. He had undergone a primary left TKA in 2014.

Six years after his primary surgery, he was admitted to General Medicine for type 1 respiratory failure and S. dysgalactiae bacteraemia. He also had a 4-week history of knee pain. Physical examination revealed a large knee effusion and severely limited ROM. The MSIS score was 9, and cultures were negative, likely as a result of the antibiotics started for his bacteraemia. He subsequently underwent explantation of the prosthesis and insertion of an articulating cement spacer. Intraoperatively, there was haemopurulent joint fluid and friable hypertrophic synovial tissue. The components were not loose. He was treated with 6 weeks of intravenous benzylpenicillin and is currently on a treatment dose of oral trimethoprim/sulfamethoxazole.

At 11 months postsurgery, he has no complaints of knee pain. He is ambulant with a walking stick at home and a walking frame in the community. His postoperative Oxford knee score is 27. His ROM is 10°–90°.

Case 5

Our fifth patient is a 75-year-old man who was premorbidly community ambulant without aids. He has multiple medical comorbidities including idiopathic dilated cardiomyopathy, hypertension, hyperlipidaemia and diabetes mellitus. He had undergone a primary right TKA in 2019.

One year after his primary surgery, he was admitted for a presenting complaint of 6 weeks of knee pain and swelling. Physical examination revealed a moderate knee effusion with a limited ROM of 0°–90°. There were no discharging sinuses. A joint aspiration was performed, which revealed a nucleated cell count of 42,700 with positively birefringent crystals seen. Gram stain revealed gram-positive cocci, and methicillin-susceptible Staphylococcus aureus was cultured from the aspiration specimen. Synovasure^® test was positive. In view of his history of cardiomyopathy and complaints of worsening effort tolerance since his TKA, a repeat echocardiogram was performed during the admission, which revealed an ejection fraction of 34%. Consequently, he underwent a single-stage explantation of the primary TKA components and insertion of an articulating cement spacer. Intraoperatively, there was a superior capsular defect and frank pus noted in the joint. The components were not loose. He was treated with 6 weeks of intravenous cloxacillin and is currently not on long-term antibiotics.

At 1 year 11 months postsurgery, he has no complaints of knee pain. He is ambulant with a walking stick. His Oxford Knee Score is 42. His ROM is 20°–100°.

DISCUSSION

Prosthetic joint infection following a primary TKA is a devastating complication. Lum et al. reported a 1-year mortality rate of 4.3% and a 5-year mortality rate of 21.6%.[11] Mortality is also significantly higher in patients who undergo revision arthroplasty for prosthetic joint infections compared to those who undergo aseptic revision arthroplasty.[12]

A two-stage revision arthroplasty is the current gold standard with the highest success rate as per the Delphi criteria and leads to the best functional outcomes in patients who are fit to undergo such a procedure.[4,5,6,13] The first stage involves explantation of the primary arthroplasty components, extensive debridement of infected tissue and insertion of an antibiotic cement spacer. This is coupled with intravenous antibiotic therapy. The second stage is performed at a later date after eradication of infection and involves removal of the cement spacer and implantation of revision arthroplasty components.

Definitive single-stage revision arthroplasty has also been described as a treatment option for prosthetic joint infections, with proposed advantages of lower morbidity and mortality, reduced healthcare costs, improved function and increased health-related quality-adjusted life-years.[14] The reported success rate varies between studies, but has generally been found to be comparable to the two-stage revision when performed on appropriately selected patients.[15,16,17,18] However, numerous contraindications and lower success rate in poor hosts sometimes limit the utility of single-stage revision arthroplasty.[15,18,19]

Hence, many patients are not ideal candidates for both two-stage and single-stage revision arthroplasty as they have high perioperative risks due to old age and significant medical comorbidities. Various studies have sporadically described outcomes in the unplanned use of long-term prosthetic and/or articulating cement spacers. Choi et al. found that 83.3% (15/18) of patients with unplanned retention of prosthetic spacers were pain free and had good functional scores for up to 6 years without necessarily requiring further surgical intervention.[6] Marson et al. also found that 39% (9/23) of patients had declined the second stage of revision arthroplasty and opted to retain their prosthetic spacers with good satisfaction and functional outcomes.[7]

More recently, promising results have been shown in the single-stage explanation with insertion of a prosthetic and/or articulating cement spacer as definitive treatment for prosthetic joint infection of the hip and knee. A case series reported that 79.3% (23/29) of patients with planned prosthetic spacer retention remained infection free and had satisfactory functional outcomes at 16.8 months follow-up.[10] A large multicentre in the United States demonstrated non-inferiority in the clearance of infection, reoperation and overall success rate (78.9%) in single-stage functional prosthetic spacers compared to traditional two-stage revision arthroplasty (70.8%).[5] The patients in the above studies received functional prosthetic spacers with a cemented cobalt chrome primary total knee femoral implant with an all-polyethylene tibia component. A Chinese study also showed no significant difference in reinfection rates between patients who received a prosthetic and/or definitive cement spacer and those who underwent a two-stage revision arthroplasty.[9] The patients in this study received either articulating cement spacers or resterilised primary TKA components. Definitive cement spacers have also been described in prosthetic joint infection of the shoulder. A systematic review by Alessio-Mazzola et al. found a high rate of infection eradication and moderate-to-good objective and subjective result.[20]

This study similarly shows good early outcomes for patients who underwent a single-stage explantation of the primary TKA components and insertion of an articulating cement spacer as definitive treatment for prosthetic joint infection of the knee. These patients were selected to receive a definitive cement spacer due to age (mean age of 79.2 years) and their significant medical comorbidities resulting in high perioperative risks. Antibiotic-impregnated articulating cement spacers were created using moulds and were loosely cemented to the femur and tibia, respectively. The attending surgeons considered the single-stage revision arthroplasty using a primary total knee femoral implant with an all-polyethylene tibia component, but recognised the increased cost and poor host factors as limitations. There were also concerns of inadequate elution of antibiotics with a single-stage revision using actual femoral and tibial implants.

All five patients remained infection free up to the point of follow-up and were able to function with some level of independence. There have been no spacer-related complications such as loosening, migration or dislocation of fracture. Figure 1 shows the postoperative knee radiographs of the five patients. Four patients had no complaints of knee pain, while one patient complained of mild pain with a VAS score of 3–4. Three patients are able to ambulate with aids, while two patients remain wheelchair bound, though they are able to weight-bear for transfers and activities of daily living. This surgical option has its limitations with most patients not being able to return to their full premorbid function. However, it is still an option in patients with high surgical risk.

Figure 1.

Postoperative knee radiographs in anteroposterior and lateral views of (a) Case 1; (b) Case 2; (c) Case 3; (d) Case 4; and (e) Case 5.

While this study is limited by the small number of patients and a relatively short postoperative follow-up for some patients, it is the first to describe the early outcomes in patients who underwent a single-stage explantation of the primary TKA components and insertion of an articulating cement spacer as definitive treatment for prosthetic joint infection of the knee in Singapore. This may prompt local orthopaedic surgeons to consider such a procedure for elderly patients with high surgical and anaesthetic risks for the traditional two-stage revision arthroplasty. Future studies may also look into comparing outcomes including but not limited to morbidity and mortality rates between patients who underwent a single-stage explantation and insertion of an articulating cement spacer, and patients who underwent the traditional two-stage revision arthroplasty.

Conflicts of interest

There are no conflicts of interest.

Funding Statement

Nil.