Revision of an infected knee replacement usually requires staged surgeries with at least two operations and a prolonged period of antibiotic therapy being necessary to give the best chance of achieving a pain free and infection free or infection controlled knee.

In general terms, the likelihood of developing a deep infection in the knee following a knee replacement is low, with less than 1% developing a deep infection. There are some patient specific risk factors such as; diabetes, smoking, immunosuppression, rheumatological medications (corticosteroids, methotrexate, plaquenil), poor circulation and previous infection, which may potentially increase the risk of infection. If infection does occur following a knee replacement, it can be considered as belonging to one of four groups, (1) Occult or missed infection, not realised at time of revision, but with routine wound swabs taken at the time of revision subsequently growing organisms; (2) Early post-operative (within first four weeks following surgery); (3) Acute haematogenous (sudden onset in previously well functioning TKR, symptoms of less than four weeks duration, usually occurring a number of months or years down the track, and finally (4) Late chronic (more than four weeks following TKR and present for more than four weeks duration) infection which is well established at the time of initial diagnosis.

The initial management of the infected knee replacement is to obtain reliable microbiological specimens, preferably without the patient being on any antibiotics, either by joint fluid aspiration prior to surgery, or at the time of surgery, by taking multiple biopsies of the knee joint tissue including the inner lining of the knee (synovium) and if present, membrane from the margins or undersurface of the implants. Once reliable specimens have been obtained, appropriate antibiotic therapy can be commenced, to prevent further progression of infection both within and outside the knee joint, and plans made for a salvage type “joint washout” procedure if the infection is an early post operative or acute haematogenous infection. The need for repeat surgery is essential if infection is to be eradicated, as bacteria can forming a biofilm and adhere to the implant surfaces, and as the knee joint does not have a blood supply inside the joint, but rather bathes in synovial fluid, the antibiotics do not readily circulate within the knee joint, and bacteria can flourish and infection become locally overwhelming. Hence, a thorough wound and joint debridement is imperative, including removal of the synovial lining of the knee (synovectomy), removal of the tibial plastic liner to allow access to the back of the knee joint and copious irrigation of the knee joint with pulsatile jets of saline (lavage) before implanting a new tibial polyethylene component (polyethylene exchange) and drain tubes to allow for escape of any bleeding along with any residual infection with a view to decreasing significantly or removing entirely the infecting organism. This approach, in conjunction with ongoing antibiotic therapy, gives the best chance of eradication of infection and preservation of the knee replacement.

Occasionally, patients will require more than one joint washout, but if the infecting organism is aggressive and cannot be controlled after two or more properly performed washouts or if the infection has been present for a long time as in a late chronic infected knee replacement, or the implants have become loose, a more complex operation (staged revision) will be required, where all implants (metal, plastic, cement) are removed and a temporary spacer implant made of cement with impregnated antibiotics is constructed and implanted. Implant loosening in the setting of infection occurs where the bacteria are attacked and consumed by the patient’s immune system white blood cells, releasing enzymes after the white cells die. The enzymatic release and associated synovitis, results in a leeching effect at the bony margins of the knee prostheses, resulting in resorption of bone adjacent to the prosthetic implants and eventual loosening. Where necessary to perform a staged revision, the removal of implants and implantation of a temporary spacer is known as the first stage of a two stage revision procedure, and the spacer is made by injecting pressurised bone cement into plastic spacer moulds sized to fit the patients anatomy and mixing powerful antibiotics effective against the identified (or most suspected) organism into the bone cement. This construct generally allows some preservation of range of motion and weight bearing, which eventually allows for a better functional outcome when the definitive knee replacement is restored, as the weakness and stiffness associated with the traditional more simple rectangular block or “hamburger patty” antibiotic cement spacer, can be avoided. Patients will generally receive six weeks of intravenous antibiotic therapy, and six weeks of oral antibiotic therapy, and serial blood tests performed to monitor their response to therapy. The antibiotics are usually administered in hospital for the first two weeks and then patients can go home with a PICC (Peripherally Inserted Central Catheter) line, and the remaining four weeks of intravenous antibiotics administered by an automated pump, and monitored by home nursing staff. The antibiotics loaded into the cement implants also leech into the bone and knee joint, further increasing the concentration of antibiotics into the knee joint, improving the chances of successfully eradicating the infecting organisms.

After three months of antibiotic therapy, all antibiotics are stopped for about two weeks before undergoing a second stage revision procedure, where the temporary spacer is removed, and samples are looked at under a microscope by a pathologist (frozen section) during the surgery, to determine whether the infection is eradicated or not, and if so, the definitive implants are re-implanted. Otherwise, if ongoing infection is suspected, a repeat debridement is performed and another staged revision performed, with temporary antibiotic cement spacer implanted, and the above cycle repeated until infection is eradicated and the definitive implant can be safely re-implanted. Using this two-stage technique, there is an approximate 80-90% likelihood that infection will be eradicated. Alternatively, if a patient is too unwell to undergo two operations in close succession, revision of the infected TKR can be performed in a single stage, but there is a lower likelihood (50-75%) of eradicating the infection, and a greater likelihood of requiring longer term to indefinite antibiotic suppression therapy and or eventual recurrent infection and implant loosening.

If a patient is discovered post-operatively to have an occult (subclinical) infection, from wound swabs or specimens taken at the time of surgery, then appropriate antibiotic therapy and follow up with clinical examination, X-rays and serological testing with inflammatory parameters (including ESR – erythrocyte sedimentation rate, CRP – C reactive protein) is necessary. This is not a common occurrence, and usually occurs in the setting of revision knee replacement, where infection is not suspected, but routine swabs taken at the time of a revision procedure come back 3 or 4 days after surgery with a positive result. It is possible that the swabs are a false positive or contaminant, and erring on the side of caution with antibiotic therapy and close clinical follow up as a precaution is advised. If the patient later develops a clinically obvious infection, then treatment as outlined earlier is appropriate.