The Infected Prosthetic Joint: When to Take it Out

Abstracts & Commentary

Synopsis: Prolonged administration of antibiotics may successfully suppress prosthetic joint infection in some patients.

Sources: Segreti J, et al. Prolonged suppressive antibiotic therapy for infected orthopedic prosthesis. Clin Infect Dis 1998;27:711-713; Karchmer AW. Salvage of infected orthopedic devices. Clin Infect Dis 1998;27:714-716. Editorial.

Segreti and colleagues at rush medical college accumulated 18 cases of patients with prosthetic hip or knee infections who were treated with prolonged suppressive antibiotic therapy without joint removal. The patients either refused removal or were thought unable to tolerate surgery. All of them had initial surgical debridement and 6-8 weeks of intravenous antibiotic therapy before being put on an oral agent known to be active against the identified pathogen. Staphylococcus aureus was recovered in eight cases—two were resistant to methicillin. Coagulase-negative staphylococci were recovered in seven cases. The other three included group B streptococci, Moraxella species, and Streptococcus pneumoniae. The oral antimicrobial regimens varied considerably and were given for an average of 49 months.

Of the 18 cases, three patients clearly failed antibiotic suppression and had their joints removed. In four cases, the patient quit antibiotic therapy after taking it for at least a year. One of those promptly relapsed within a month of completing a 22-month course of dicloxacillin. Three of the four failures involved infections due to methicillin-sensitive S. aureus; each had been treated with dicloxacillin. In only one case was dicloxacillin successful. The two methicillin-resistant S. aureus strains were treated with minocycline and rifampin and did well.

There were complications related to prolonged antimicrobial therapy in 22% of the patients. Four patients had Clostridium difficile diarrhea and two patients developed a drug rash, but the antibiotics were continued without apparent further adverse effects.

In an accompanying editorial, Karchmer reviews an additional series of patients with prosthetic joint infections. From these reports, he tries to draw some conclusions about the possible value of antimicrobial suppression. Some factors that seem to relate to a good outcome include whether the implant is stable, whether the infection is recognized early, and whether debridement is promptly performed. Also important is the lack of loosening of the prosthesis.

Comment By Alan D. Tice, MD, FACP

The article by Segreti et al adds another small series to the sparse literature about suppressing or possibly even curing infections of prosthetic joints without removing the hardware. This series and review, however, does provide some insight. Stability of the joint and pain are important factors to consider. In reviewing the series and advice from textbooks, it appears that the longer the infection is present, the less likely a successful outcome. It also appears that S. aureus is more difficult to eradicate or suppress than coagulase-negative staphylococci and possibly some of the other organisms.

The criteria for diagnosis of prosthetic joint infection also raises some questions. The recovery of coagulase-negative staphylococci may not necessarily represent an infection and may, instead, be the result of culture contamination. They are a frequent contaminant with cultures. Even when truly there, these organisms may attach to foreign material and set up a glycocalyx that allows them to live on the joint surface in relative harmony with the host.

It is interesting to note that the cases reported by Segreti et al were all treated with surgical debridement without removal of the prosthesis. How extensive this surgery should be and how important this is remains a question—particularly if there is no obvious necrotic material to serve as a further nidus of infection. The 6-8 week course of intravenous antibiotics raises another question. Is a prolonged course of IV therapy necessary if oral antibiotics are to be used for suppression? Would two weeks be enough?

It is also interesting that there was such a high failure rate with dicloxacillin against S. aureus. The minocycline/rifampin regimen seemed to be effective and even seemed to work against methicillin-resistant S. aureus. The recent reports about the value of a quinolone plus rifampin should be considered.1 Rifampin is a zwitterion with both a lipid-soluble and a water-soluble end. Its lipid solubility gives it the ability to penetrate white blood cells and then go on to penetrate the phagocytic vacuoles where microorganisms may be hiding. Dicloxacillin does not have the capacity for intracellular killing that rifampin does but the quinolones do.

Another question is how long oral antibiotics need to be given for an infection that is at least suppressed. If patients can tolerate long-term antibiotic therapy, it is advised that oral antibiotics be given for life, but, in this series, three of the four that stopped antibiotics after a year had not relapsed during the more than three years they were followed afterward. Unfortunately, there are no good laboratory tests to give us insight.

When to treat people with infected joint prostheses with limited surgical debridement and antibiotics remains uncertain. The apparent cure rates of 80-85% with removal and reimplantation (either early or late) of the prosthetic joint are difficult to compare with the series in which the joint is left in place—which has a success rate of 30-60% in other series. It should also be considered that the hosts in this population might not be as healthy; hence, infection might be more difficult to eradicate even if the joint is removed.

Other factors to consider in regard to joint replacement in the face of infection should include the host, the organism, and the duration of infection as well as the tolerance of oral antibiotics. If a prolonged course of oral antibiotics is opted for, one that contains rifampin would seem most reasonable.


1. Zimmerli W, et al. Foreign-body infection (FBI) study group. Role of rifampin for treatment of orthopedic implant-related staphylococcal infections. JAMA 1998; 279:1537-1541.