This Article Abstract Full Text (PDF) Free Letters to the Editor: Submit a response Alert me when this article is cited Alert me when Letters to the Editor are posted Alert me if a correction is posted Services E-mail this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Add to My File Cabinet Download to citation manager Rights and Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by SEGAWA, H. Articles by GUSTILO, R. B. Search for Related Content PubMed PubMed Citation Articles by SEGAWA, H. Articles by GUSTILO, R. B. Social Bookmarking                   What's this?

Infection After Total Knee Arthroplasty. A Retrospective Study of the Treatment of Eighty-One Infections*

HIROYUKI SEGAWA, M.D., DEAN T. TSUKAYAMA, M.D., RICHARD F. KYLE, M.D., DOUGLAS A. BECKER, M.D. and RAMON B. GUSTILO, M.D., MINNEAPOLIS, MINNESOTA

Investigation performed at Hennepin Country Medical Center, Minneapolis

	Abstract

Top Abstract Introduction Materials and Methods Results Discussion References

 
Background: The clinical presentation of an infection at the site of a total knee arthroplasty can be used as a guide to treatment, including the decision as to whether the prosthesis should be retained or removed. We reviewed the results of treatment of infection after total knee arthroplasty to evaluate the effectiveness of four treatment protocols based on the clinical setting of the infection. Methods: We retrospectively evaluated the results of treatment of eighty-one infections in seventy-six consecutive patients who either had an infection after a total knee arthroplasty or had multiple positive intraoperative cultures of specimens of periprosthetic tissue obtained during a revision total knee arthroplasty performed because of presumed aseptic loosening. The patients were managed according to one of four protocols. Five infections in five patients who had positive intraoperative cultures were treated with antibiotic therapy alone. Twenty-three early postoperative infections in twenty-one patients were treated with débridement, antibiotic therapy, and retention of the prosthesis. Twenty-nine late chronic infections in twenty-eight patients were treated with a delayed-exchange arthroplasty after a course of antibiotics. Seven acute hematogenous infections in six patients were treated with débridement, antibiotic therapy, and retention of the prosthesis. Seventeen infections in seventeen patients were not treated according to one of the four protocols. Sixteen late chronic infections were treated either with an arthrodesis (five infections) or with débridement, antibiotic therapy, and retention of the prosthesis (eleven infections). One acute hematogenous infection was treated with resection arthroplasty because of life-threatening sepsis. Results: The mean duration of follow-up was 4.0 years (range, 0.3 to 14.0 years). Eleven patients who had an arthrodesis, a resection arthroplasty, or an above-the-knee amputation after less than two years of follow-up were included in the study as individuals who had a failure of treatment. In the group of patients who were managed according to protocol, the initial course of treatment was successful for all five infections that were diagnosed on the basis of positive intraoperative cultures, five of the ten deep early infections, all thirteen superficial early infections, twenty-four of the twenty-nine late chronic infections, and five of the seven acute hematogenous infections. Only one of eleven prostheses in patients who had a late chronic infection that was not treated according to protocol was successfully retained after débridement. Conclusions: Our treatment protocols, which were based on the clinical setting of the infection, were successful for most patients. A major factor associated with treatment failure was a compromised immune status. Bone loss and necrosis of the soft tissues around the joint also complicated the treatment of these infections.

	Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

 
Infection is a frequent reason for failure after total knee arthroplasty^1,30. The prevalence of infection after primary total knee arthroplasty has been reported to range from 0.5 to 5 percent in series ranging in size from 821 to 13,478 arthroplasties^{4,14,27,29,33}. The rate of infection is higher among patients managed with revision total knee arthroplasty and those who have rheumatoid arthritis^4,39.

Several systems have been proposed for the classification and staging of infection^6,13,19,26, but none of these systems have been shown to be particularly useful as guides to treatment, especially with regard to whether the prosthesis should be removed or retained. In a previous study³⁶, we proposed a classification system that was based on the clinical presentation of infection after a total hip arthroplasty; the four types of infections included early postoperative infections, late chronic infections, acute hematogenous infections, and clinically inapparent infections associated with positive cultures of specimens obtained at the time of a revision operation. This system is based on what we have observed to be the common clinical presentations of prosthesis-related infections and focuses on the acuteness of the symptoms as well as the temporal relationship between the infection and the index arthroplasty. We now report on the outcome of treatment in patients who had an infection after a total knee arthroplasty and were managed according to one of four protocols based on these same clinical settings. The goals of treatment were eradication of the infection and maintenance of a functional extremity.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

 
Eighty consecutive patients (eighty-five knees) who had an infection after a total knee arthroplasty or had positive intraoperative cultures of specimens that were obtained during revision of a presumed aseptically loose total knee prosthesis were diagnosed and managed from 1980 to 1995. Fifty infections occurred after procedures that had been done at either Hennepin County Medical Center or Metropolitan Mount Sinai Medical Center. Twenty of these infections occurred from 1980 to 1987, but we were unable to determine the number of total knee arthroplasties performed during that time-period. From 1987 to 1995, thirty infections occurred following 695 primary and 133 revision procedures, for a rate of infection of 3.6 percent. The remaining thirty-five infections occurred after procedures that had been performed at other institutions. Three patients were lost to follow-up, and one had incomplete records. This left seventy-six patients (95 percent of the original group) with eighty-one infections in eighty-one knees. The mean duration of follow-up was 4.0 years (range, 0.3 to 14.0 years). We reviewed all of the charts and radiographs for these patients and recorded their status with respect to infection and function of the knee at the time of the most recent clinical evaluation (fifty-four patients) or on the basis of correspondence (twenty-two patients). Patients who had rheumatoid arthritis, diabetes mellitus necessitating treatment with insulin or oral medication, chronic renal insufficiency, or a history of malignant disease were classified as potentially immunocompromised. The index operation was defined as the clean, elective knee operation that immediately preceded the infection.

Infection was diagnosed when multiple intraoperative cultures showed growth of the same organism or organisms or when there was clinically apparent pus in the knee joint, or both. When the same pathogen or pathogens persistently or intermittently grew on culture of specimens from the same joint, the infection was considered to be a single ongoing infection, regardless of the number of attempts at treatment. However, if the patient had another infection in the contralateral knee or a subsequent infection with a different pathogen in the ipsilateral knee, the infections were considered to be separate.

Four clinical settings were defined in this study. A patient was considered to have positive intraoperative cultures when the same organism grew on culture of at least two specimens that had been obtained at the time of a revision operation. The diagnosis before all of these revisions was aseptic loosening. A minimum of five cultures were performed for every revision. The number of patients who had a single positive culture was not recorded. However, no patient with a single positive culture (who, according to our protocol, would not have received any treatment for infection) subsequently returned because of an infection during the follow-up period. An early postoperative infection was defined as a wound infection (superficial or deep) that developed less than four weeks after the index operation. A superficial early postoperative infection was diagnosed when, on direct visualization of an infected wound at the time of the operation, there was no extension of inflammation into the joint. If there was any doubt regarding extension into the joint, the infection was considered to be deep and the joint was debrided. A late chronic infection was one that developed four weeks or more after the index operation and that had an insidious clinical presentation. An acute hematogenous infection was associated with a documented or suspected antecedent bacteremia and was characterized by an acute onset of symptoms (severe pain and swelling within forty-eight hours) in the affected joint with the prosthesis.

Each infection was initially treated according to the protocol appropriate for its clinical presentation. Patients who had positive intraoperative cultures were managed with intravenous administration of antibiotics for six weeks without operative intervention. In these patients, a revision prosthesis had already been implanted because of presumed aseptic loosening before the results of the intraoperative cultures were available. All revision operations (even those performed because of true aseptic loosening) were performed with tobramycin-impregnated cement (1.2 grams of tobramycin powder mixed with forty grams of cement).

The treatment for superficial early postoperative infections was débridement and a course of antibiotic therapy. Deep early postoperative infections were treated with débridement, placement of tobramycin beads (made with a mixture of 1.2 grams of tobramycin in twenty grams of polymethylmethacrylate), replacement of the polyethylene insert of the tibial component, retention of the prosthesis, and intravenous administration of antibiotics. Removal of the original polyethylene insert permitted greater access to the posterior region of the joint for débridement. The tobramycin beads were removed two weeks after the first débridement through a minimum incision without a second débridement. The duration of antibiotic therapy, arbitrarily determined on the basis of our previous experience, was two weeks for superficial infections and four weeks for deep infections.

Late chronic infections were treated with débridement, removal of all prosthetic components and bone cement, and placement of a tobramycin-impregnated cement spacer and tobramycin beads^5,15,36. The spacer, a circular disk, was placed within the joint space. Approximately ninety antibiotic-impregnated cement beads, each about six millimeters in diameter, were placed into the joint space and into the intramedullary canals of the femur and tibia. Antibiotics were administered intravenously for six weeks. After completion of antibiotic therapy, the erythrocyte sedimentation rate and the level of C-reactive protein were determined, and the patient was observed for two more weeks. If these tests showed no evidence of inflammation and there was no clinical evidence of recurrent infection, a delayed-exchange arthroplasty was performed. Multiple cultures of specimens obtained during the revision operation were performed to confirm eradication of the infection. The tobramycin beads and the spacer were removed, and a posterior stabilized knee implant was inserted and fixed with antibiotic-impregnated bone cement (1.2 grams of tobramycin mixed with forty grams of cement).

Acute hematogenous infections were treated with débridement, replacement of the polyethylene insert, retention of the prosthesis if it was not loose, and intravenous administration of antibiotics for six weeks.

Overall, fifty-nine patients (sixty-four infections) were managed according to a protocol described in the present report. Seventeen patients (seventeen infections) were managed according to another method. Of these, five patients who had a late chronic infection were believed to be poor candidates for reimplantation and were managed with removal of the prosthesis followed by arthrodesis. Another eleven patients who had a late chronic infection were initially managed with retention of the prosthesis; five were believed to be poor candidates for a revision arthroplasty because of the poor soft-tissue envelope surrounding the infected knee, and the other six were believed to be marginal candidates for multiple operations because of poor general health. The remaining patient had life-threatening sepsis in association with an acute hematogenous infection and was managed with removal of the implant. The results of treatment of these seventeen infections are also reported.

Antibiotics were chosen empirically on the basis of our knowledge of the most likely pathogens associated with each clinical setting. Vancomycin, the most reliably active antibiotic against coagulase-negative staphylococci, which are the most frequent pathogens associated with an occult infection, was prescribed after a revision operation until the results of intraoperative cultures were known. Antibiotic treatment of early postoperative infections was directed against coagulase-positive staphylococci and gram-negative bacilli. Antibiotic treatment of late chronic infections was withheld until intraoperative specimens were obtained for culture. If the pathogens were unknown, vancomycin and often a second antibiotic directed against gram-negative bacilli were administered. The antibiotics were later adjusted on the basis of the results of the cultures.

The rate of success after the first course of treatment was determined for each protocol. If the initial treatment failed, we recorded the number and type of subsequent courses of treatment that were attempted. A course of treatment was defined as a protocol that included an operation or the administration of antibiotics, or both, that was initiated with the intention of eradicating the infection. In some patients, eradication was achieved only after multiple courses of treatment. A successful result was defined as no clinical evidence of infection at least two years after the end of antibiotic therapy and a functional knee with a total joint prosthesis in place at the time of the latest follow-up. For a knee to be considered functional, there had to be no or only slight pain with walking (with or without the use of a cane) and no radiographic findings, such as progressive osteolysis or movement of the prosthetic components, that indicated a need for immediate or impending operative intervention.

Postoperative and follow-up radiographs were evaluated according to the system of the Knee Society¹⁰. Radiographic loosening was defined as a complete radiolucent line of at least two millimeters in width at the bone-cement or prosthesis-bone interface or as evidence of migration or subsidence of a component.

Specimens of periprosthetic tissue (including the joint capsule, synovial lining, intramedullary material from curettage, granulation tissue, and bone fragments) were obtained during the revision procedure or when the prosthesis was removed because of an infection and were evaluated histologically for evidence of inflammation. Histological examination was not routinely done when frank pus was present, as it was in many patients who had an early postoperative infection. The criterion for acute inflammation (more than five polymorphonuclear leukocytes per high-power field) described by Mirra et al.²⁴ was used. Cultures of material obtained from the knee joint and by swabbing of the prosthesis as well as cultures of periprosthetic tissue were performed. A minimum of five cultures were performed for each patient, and all cultures were evaluated for aerobic and anaerobic growth. Cultures that showed growth in broth only were not considered positive. The duration of hospitalization was recorded, and the mean number of days was calculated for each category of infection.

Ninety-five percent confidence intervals were calculated for the rates of successful clinical results, with reference to exact binomial probability tables. Chi-square analysis was used to evaluate the differences in the results between the patients who were immunocompromised and those who were not.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

 
There were eighty-one infections in the seventy-six patients (twenty-three men and fifty-three women). Seventeen infections, including sixteen late chronic infections and one acute hematogenous infection, were not treated according to protocol. The mean age of the patients at the time of the diagnosis of infection was sixty-seven years (range, thirty-two to eighty-seven years). Five patients had two infections each. One of these patients had bilateral, simultaneous acute hematogenous infection. The other four patients had infections affecting different joints at different times. Thirty-two infections (40 percent) were in knees that had had a previous revision procedure, and seventy-five (93 percent) were in knees in which the prosthesis had been inserted with cement.

The mean duration of follow-up was 4.0 years (range, 0.3 to 14.0 years). Treatment was not considered successful until the patient had been followed for a minimum of two years after the completion of all operative and antibiotic treatment. Eleven patients who had an arthrodesis, a resection arthroplasty, or an above-the-knee amputation less than two years after treatment were included in the study as individuals who had a failure of treatment. Patients who continued to take suppressive antibiotics also were considered to have had a failure of treatment. The mean interval between the index operation and the diagnosis of infection was 1.3 years (range, seven days to thirteen years). Five infections were diagnosed on the basis of positive intraoperative cultures, ten were deep early postoperative infections, thirteen were superficial early postoperative infections, forty-five were late chronic infections, and eight were acute hematogenous infections.

The primary total knee arthroplasty had been performed because of osteoarthritis in fifty-one knees, rheumatoid arthritis in twenty-five, and posttraumatic arthritis in five.

Microbiological Findings
Staphylococci accounted for sixty-four (73 percent) of the eighty-eight isolates (Table I). Overall, aerobic gram-positive cocci accounted for eighty-four (95 percent) of the eighty-eight isolates and gram-negative bacilli accounted for four (5 percent). All of the gram-negative isolates came from late chronic infections, whereas all of the isolates from acute hematogenous infections were gram-positive cocci. Nine (11 percent) of the eighty-one infections were polymicrobial, usually a combination of gram-negative bacilli and gram-positive cocci. Two patients had negative cultures, and the infection was diagnosed clinically on the basis of pus in the knee joint at the time of the operation.

Positive Intraoperative Cultures
Five infections in five patients were detected on the basis of positive intraoperative cultures at the time of revision. These five revisions represented 4 percent of the 133 revisions that were performed at either Hennepin County Medical Center or Metropolitan Mount Sinai Medical Center between 1987 and 1995. The results of histological examination, available for all five patients, showed evidence of chronic, but not acute, inflammation. The treatment protocol (six weeks of antibiotic therapy without additional operations) was successful for all five patients. One patient was immunocompromised as a result of rheumatoid arthritis.

All components in these five patients had been inserted with cement. One tibial component demonstrated evidence of loosening on the follow-up radiographs. The mean duration of hospitalization was eight days (range, four to twelve days). The patients were followed for a mean of 3.9 years (range, 2.8 to 4.8 years).

Deep Early Postoperative Infections
There were ten deep early postoperative infections in ten patients (Table II). Eight infections were associated with a primary replacement performed with cement, and two were associated with a revision replacement performed with cement. The treatment protocol (débridement and intravenous administration of antibiotics for four weeks) was successful for only five of the ten patients. The protocol was successful for two of the three patients who were not immunocompromised and for three of the seven who were; this difference could not be shown to be significant, with the numbers available.

The mean time between the initial total knee arthroplasty and the onset of symptoms of the infection was seventeen days for the five patients who had a successful result and seventeen days for the five patients in whom the infection failed to respond to the treatment regimen.

The mean duration of hospitalization was forty days (range, two to 125 days). The patients were followed for a mean of 3.7 years (range, 0.3 to 8.8 years). Four patients were followed for less than two years; of these, two had an above-the-knee amputation (at 0.3 and 1.1 years) and two had an arthrodesis (at 0.5 and 1.0 year). None of the five patients who had a successful result had radiographic findings (such as evidence of migration of the implant or of progressive osteolysis) that would have indicated a need for immediate or impending operative intervention.

Superficial Early Postoperative Infections
There were thirteen superficial early postoperative infections in eleven patients. Nine infections were associated with a primary replacement performed with cement, and four were associated with a revision replacement performed with cement. The treatment protocol (débridement without opening the joint and oral administration of antibiotics for two weeks) was successful for all patients. Four patients (four infections) were immunocompromised.

The mean duration of hospitalization was fifteen days (range, one to thirty-three days). The patients were followed for a mean of 7.6 years (range, 2.1 to 14.0 years). At the time of the most recent evaluation, all eleven patients were able to walk with no or only slight pain (some with the assistance of a cane). None of the patients had radiographic findings (such as evidence of migration of the implant or progressive osteolysis) that would have indicated a need for immediate or impending operative intervention.

Late Chronic Infections
There were forty-five late chronic infections in forty-four patients (Fig. 1). One infection that had a failure of treatment with the protocol for a deep early postoperative infection subsequently was treated with the protocol for a late chronic infection. Histological examination of periprosthetic tissue, obtained at the time that the prosthesis was removed, revealed acute inflammation in association with nine (43 percent) of twenty-one infections.

View larger version (20K): [in this window] [in a new window]   Fig. 1 Diagram demonstrating the outcome of treatment of late chronic infections.

Twenty-nine late chronic infections, in twenty-eight patients who had good bone stock and soft tissue in the region of the knee, were treated with a delayed-exchange arthroplasty (Table III). Twenty-four (83 percent) of these infections were eradicated after one course of treatment, and the other five (17 percent) necessitated additional treatment. Delayed-exchange arthroplasty was successful for twelve of thirteen patients who were not immuncompromised and twelve of fifteen patients who were immunocompromised; this difference could not be shown to be significant, with the numbers available. Of the five infections for which the initial treatment failed, two were successfully treated with a second attempt at delayed-exchange arthroplasty, one was treated with oral administration of suppressive antibiotics, and two were treated with an arthrodesis. One of the two arthrodeses was converted to an above-the-knee amputation because of persistent infection. Of the twenty-nine knees that initially were treated with a delayed-exchange arthroplasty, twenty-eight (97 percent) had eradication of bacteria and twenty-six (90 percent) had a functioning total knee prosthesis at the time of the latest follow-up (Table IV).

Eleven chronic infections were treated with delayed-exchange arthroplasty as a second procedure. Six of these infections originally had been treated with retention of the prosthesis, and five had been treated with a previous attempt at delayed-exchange arthroplasty. Six of the eleven infections were eventually treated successfully. Of the five infections for which the treatment was considered to have failed, one was treated with oral administration of suppressive antibiotics and four were treated with arthrodesis. One of the four arthrodeses was later converted to an above-the-knee amputation.

Delayed-exchange arthroplasty was performed at a mean of fifty-five days (range, forty-two to ninety-eight days) after removal of the implant. Our current practice is to perform the revision eight weeks after removal of the prosthesis, which is two weeks after the completion of antibiotic therapy. Patients did not have aspiration of the knee for culture before the revision operation. In two patients, the reimplantation was performed immediately after the completion of antibiotic therapy.

Overall, thirty-one (69 percent) of the forty-five knees had a successful result, defined as eradication of the infection, cessation of antibiotic therapy, and maintenance of a functional prosthesis (Table III). Of the fourteen knees that had a failure of treatment, ten ultimately had an arthrodesis, two were treated with suppressive antibiotics (administered for three and six years by the time of the most recent follow-up), one had a resection arthroplasty, and one had an above-the-knee amputation.

One femoral and six tibial components demonstrated evidence of loosening on the follow-up radiographs. The mean duration of hospitalization was thirty-four days (range, seven to 229 days); this included the duration of the first admission (for débridement and removal of the prosthesis) and that of the second admission (for revision after eradication of the infection). The mean duration of follow-up was 3.3 years (range, 0.4 to 8.5 years). Six patients were followed for less than two years; these individuals had an arthrodesis at 0.4, 0.5, 0.7, 0.9, 1.0, and 1.6 years. All patients who were classified as having a successful outcome were followed for a minimum of two years after the completion of all therapy.

Acute Hematogenous Infections
There were eight acute hematogenous infections in seven patients (Table V). Bacteremia was documented in three patients (four infections); a skin infection, in two; a lower respiratory infection, in one; and a urinary tract infection, in one. In patients who had an identified primary site of infection, the onset of the symptoms at that site preceded the symptoms in the knee joint by one to four days. Histological examination of the periprosthetic tissue revealed acute inflammation in association with three of five infections.

Seven of the eight acute hematogenous infections were associated with a prosthesis that had been inserted with cement and one, with a prosthesis that had been inserted without cement. Six infections were associated with a primary replacement and two, with a revision.

No component demonstrated evidence of loosening on the radiographs that were made at the time of the latest examination. The mean duration of hospitalization was thirty-one days (range, ten to seventy-two days). The patients were followed for a mean of 3.3 years (range, 0.3 to 5.5 years). The one patient who was followed for less than two years had an arthrodesis at 0.3 year.

Potential Factors Responsible for Failure
Forty-three patients (57 percent) were immunocompromised. Of these, twenty-three had rheumatoid arthritis, twelve had diabetes mellitus, eleven had renal failure, ten were taking corticosteroids (more than ten milligrams of prednisone daily), and three had malignant disease involving the breast, prostate, or tongue. These forty-three patients had forty-six infections, including one infection that was diagnosed on the basis of a positive intraoperative culture, seven deep early postoperative infections, four superficial early postoperative infections, twenty-eight late chronic infections, and six acute hematogenous infections. These infections represented 82 percent (twenty-three) of the twenty-eight infections for which the initial treatment had failed but only 43 percent (twenty-three) of the fifty-three infections for which the initial treatment was successful; this difference was significant (p < 0.01). Four of the five deep early postoperative infections for which the initial treatment had failed, sixteen of the twenty late chronic infections for which the initial treatment had failed, and all three of the acute hematogenous infections for which the initial treatment had failed were in immunocompromised patients. When only the patients who were managed according to protocol were considered, ten of the twelve infections for which the initial treatment had failed but only twenty-three (43 percent) of fifty-three infections for which it was successful were in immunocompromised patients; this difference was significant (p < 0.02). Other factors, such as the age of the patient, the type of organism, the number of previous operations, and the presence of underlying joint disease, were not found to have a significant association with the outcome, with the numbers available.

Final Result
Seventy-seven (95 percent) of the eighty-one infections were eradicated. Four infections, all of which were in immunocompromised patients, persisted. Of the sixty-four knees that were treated according to protocol, fifty-four (84 percent) ultimately were free of infection and had a functional prosthesis (Table IV). Of the seventeen knees that were not treated according to protocol, only five were free of infection and had a functional prosthesis. Thirteen infections ultimately were treated with an arthrodesis; two, with a resection arthroplasty; and three, with an above-the-knee amputation. Thirteen of the seventeen patients who were not managed according to protocol were immunocompromised.

Mechanical Loosening
Seven prostheses had radiographic evidence of mechanical loosening after two to nine years of follow-up. Two knees subsequently were revised because of aseptic loosening after the infection had been eradicated.

Reinfection
Of the fifty-nine knees that had been treated successfully, one subsequently became reinfected. This knee initially had been treated because of a late chronic infection with coagulase-positive staphylococci. Two years after a delayed-exchange arthroplasty, the knee became reinfected with coagulase-positive staphylococci as a result of hematogenous seeding after a thoracotomy.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

 
All infections around joint prostheses are not the same. The clinical presentation of the infection can serve as a useful guide to treatment, especially with regard to whether the prosthesis should be removed. Our treatment protocols, which are based on four clinical settings of infection, were successful for most patients. The prosthesis was successfully retained in all five knees with positive intraoperative cultures, in five of seven with an acute hematogenous infection, and in only five of ten with a deep early postoperative infection. Twenty-four (83 percent) of twenty-nine late chronic infections were successfully treated with a delayed-exchange arthroplasty. There were no instances of reinfection among the fifty-nine knees in which the prosthesis was successfully retained or replaced. One patient had a second infection as a result of hematogenous seeding after a thoracotomy.

The patients who had positive intraoperative cultures had been managed with direct-exchange arthroplasty because of presumed aseptic loosening. Once the results of the intraoperative cultures were known to be positive, the clinical issue that arose was whether the joint with the newly implanted prosthesis needed any treatment and whether that treatment should include an additional operation. It can be difficult to ascertain whether bacteria that grow on culture of specimens obtained during a revision operation represent infection or contamination^8,11. In order to minimize the chance that a laboratory contaminant would be inappropriately considered as a pathogen, we required the same pathogen to grow on culture of at least two intraoperative specimens and we disregarded cultures that were positive in only the broth medium³⁶. Other investigators have recommended recovery of bacteria from all specimens²² or have required at least five colonies per plate as a definition of a meaningful culture⁷.

All five infections that were diagnosed on the basis of positive intraoperative cultures responded to six weeks of antibiotic therapy, without removal of the newly implanted prosthesis. This treatment protocol warrants consideration as an option with a great potential benefit and relatively low risk. The routine use of antibiotic-impregnated cement also may be useful for the treatment of infection in such cases. A larger series of patients is needed in order to determine whether this protocol is appropriate for all patients who have positive intraoperative cultures of specimens obtained during revision procedures performed because of presumed aseptic loosening. An alternative would be to manage such patients with oral administration of antibiotics. The recent availability of quinolone antibiotic agents, which reach equivalent serum levels after oral and intravenous administration and which need to be taken only once a day, make the possibility of oral therapy especially attractive. The identification of patients who have positive intraoperative cultures but do not need antibiotic therapy is also an important goal for future investigation. Unfortunately, the evaluation of histological specimens has not been useful for the diagnosis of infection. In our previous study³⁶, only one (4 percent) of twenty-five infections that were diagnosed on the basis of positive intraoperative cultures and eight (27 percent) of thirty late chronic infections were associated with histological evidence of acute inflammation. It is not known whether this relative lack of an acute inflammatory response is due to the chronicity of the infection, the limited virulence of the pathogens involved, or unique aspects of the immediate environment of the implant.

Our definition of an early postoperative infection as one that occurs within four weeks after the operation is based on our experience, as well as that of others^18,25,31, that infections that become evident more than four weeks after the operation are not likely to respond to treatment that does not include removal of the prosthesis, unless the joint has been hematogenously seeded from another site of infection. Some authors have defined an early infection as one that occurs within two weeks after the arthroplasty and have reported good results in association with retention of the prosthesis when the onset of infection was within this period^6,35,37. In the present study, all of the early deep infections were diagnosed within two to four weeks after the operation. In general, the longer that an infection has been present, the more difficult it is to eradicate without removal of the prosthesis³⁶. In series ranging in size from five to thirty-five infections, the rate of success associated with the use of irrigation and débridement for the treatment of early postoperative infection has ranged from 0 to 83 percent^{6,18,25,28,34,35}. Especially impressive were the results of Mont et al.²⁵, who reported that a protocol of multiple débridements was successful for the treatment of ten of ten early postoperative infections (defined as those that occurred less than four weeks after the index arthroplasty) and ten of fourteen acute hematogenous infections. In the present study, the fact that the initial treatment failed for five of the ten deep early postoperative infections demonstrated that our treatment protocol for that type of infection was unsatisfactory. It is possible that early postoperative infections that occur less than two weeks after the index arthroplasty can be treated with retention of the prosthesis and a single débridement, whereas those that occur between two and four weeks can be treated with multiple débridements. The duration of antibiotic therapy also may be questioned. Our choice of four weeks of antibiotic treatment was based, in part, on the findings of our previous study³⁶, which indicated that four weeks was an adequate duration of treatment for infections at the site of a total hip arthroplasty. We currently prescribe six weeks of antibiotic therapy for all deep infections at the site of a joint prosthesis. The optimum duration of antibiotic therapy remains to be determined but probably depends on a number of variables, such as the immune status of the patient, the adequacy of operative débridement, the quality of the bone and surrounding soft tissue, and the virulence of the pathogen.

The results of the present study, especially those related to deep early postoperative infections, were not as good as the results of our previous study³⁵, in which the same four protocols were used for the treatment of infection after total hip arthroplasty. In these two series, the ages of the patients were comparable and the proportions of immunocompromised patients were similar. One reason for this difference may be related to problems with soft-tissue healing. The knee lies superficially beneath the skin and the fascial envelope and is covered by a limited amount of well vascularized muscle. The so-called watershed area of vascular supply to the skin that lies anteriorly in the path of the typical skin incision is a potential area of impaired wound-healing¹. A second factor may be that more of the joint is exposed to cement during a total knee arthroplasty than is exposed during a total hip arthroplasty. The difficulty of eradicating bacteria that are adherent to foreign bodies, such as bone cement, has been well described²⁴. A third possible reason is that the hole that is made in the distal part of the femur for the intramedullary guide during total knee arthroplasty provides a path for the bacteria to escape débridement. In addition, the results of treatment of deep early postoperative infections were not as good as the results of treatment of acute hematogenous infections, perhaps because the soft tissues had already recovered in patients who had an acute hematogenous infection.

The thirteen superficial early postoperative infections in the present study were easily treated with débridement of the wound and administration of antibiotics. Because of the risk of subsequent deep infection²¹, superficial infections should be treated intensively^23,31,38.

Delayed-exchange arthroplasty has been associated with good results when used for the treatment of chronic infections of the knee. The rate of success in the current study (83 percent; twenty-four of twenty-nine infections) is comparable with the rates reported by others^{12,16,19,32,40}.

Sixteen (36 percent) of the forty-five late chronic infections were in patients who were poor candidates for a delayed-exchange protocol. Eleven of these infections were treated with débridement and retention of the prosthesis, and five were treated with an arthrodesis. Factors commonly associated with infection after a total knee arthroplasty (but not after a total hip arthroplasty), such as massive bone loss, deficiency of the extensor mechanism, and instability of the joint, can compromise the revision operation, thereby necessitating an arthrodesis, resection arthroplasty, or above-the-knee amputation. Delayed-exchange arthroplasty as a second course of treatment is not as successful for infections at the site of a total knee arthroplasty as it is for those at the site of a total hip arthroplasty^2,9. Hanssen et al.¹⁷ reported only one successful result in a study of twenty-four patients (twenty-four knees) who were managed for reinfection after the reimplantation of a total knee prosthesis. Those authors recommended other, nonprosthetic salvage procedures rather than multiple attempts at reimplantation, particularly if reimplantation involved the use of either a stemmed prosthesis designed to be inserted with cement or a hinged prosthesis, both of which appeared to predispose the patient toward an unfavorable outcome such as amputation.

A prosthesis should not be retained in a patient who has a late chronic infection when the intent is to eradicate the infection⁶. We included eleven patients with late chronic infections who were managed with débridement alone in order to illustrate the poor outcome that results when the prosthesis is retained in patients who have such an infection. Only one of these eleven chronic infections was successfully treated without removal of the prosthesis. However, in selected instances in which a revision arthroplasty is not technically feasible or the patient's medical condition makes multiple operations hazardous, retention of the prosthesis in combination with suppressive antibiotic therapy may be a reasonable choice of management.

An acute hematogenous infection, which can occur at any time after the prosthesis has been implanted, has an abrupt onset of symptoms, including fever and acute pain and swelling of the affected joint. If treatment is initiated promptly, it is possible to eradicate the infection without removal of a stable prosthesis. Five of seven acute hematogenous infections in the current series were treated successfully with retention of the prosthesis. Although this result was similar to those reported by others^3,25, there were too few infections in this group to support any substantive conclusions. In some instances, it may be difficult to distinguish an acute infection from a chronic infection.

The classification of infection on the basis of the clinical setting allows for the prediction of the most likely pathogens. All five infections that were diagnosed on the basis of positive intraoperative cultures were caused by coagulase-negative staphylococci. Early postoperative and late chronic infections usually were caused by coagulase-positive staphylococci, coagulase-negative staphylococci, or gram-negative bacilli. Acute hematogenous infections were caused by gram-positive cocci. This knowledge is valuable when choosing antibiotic therapy empirically, before the results of cultures are available, or in the infrequent situation in which bacteria do not grow on culture despite a clinically apparent infection⁴.

One factor that we identified as contributing to the failure of treatment was a compromised immune status. Twenty-three (82 percent) of the twenty-eight infections for which the initial treatment failed were in immunocompromised patients compared with only twenty-three (43 percent) of the fifty-three infections for which the initial treatment succeeded; this difference was significant (p < 0.01). It was noted that immunocompromised patients were more likely to have infections that were difficult to treat (that is, deep postoperative infections, late chronic infections, and acute hematogenous infections), which may have accounted for the lower overall rate of success in that group.

The seventeen patients who were not managed according to protocol did poorly. Only one of these patients had a successful result after one course of treatment, and only five ultimately had a successful outcome. These patients were included in our study to illustrate several important points. First, most of these patients (thirteen of seventeen) were immunocompromised. This factor may have contributed to the assessment that they were not suitable candidates for delayed-exchange arthroplasty. Second, the results associated with retention of the prosthesis in patients who had chronic infections were poor, which is in agreement with the findings of previous studies^4,31. In retrospect, our decision to retain well fixed prostheses was incorrect. Third, in some instances, the need to control severe systemic effects of infection must override protocol considerations, as was the case for one patient who was in septic shock in association with an acute hematogenous infection, from whom the prosthesis was removed in order to ensure elimination of the site of infection. Although the patients who were managed according to protocol were more likely to have a successful result than those who were not, it must be recognized that the latter group consisted of a disproportionate number of immunocompromised patients. This selection bias could explain all or part of the observed difference between the patients who were managed in accordance with our protocols and those who were not.

We conclude that all infections around prosthetic joints cannot be treated in the same manner and that the classification of infection according to clinical setting serves as a relatively simple and reasonable guide for choosing an appropriate treatment regimen. It also provides a basis for the comparison of different treatment protocols. On the basis of a comparison of our data with those described in other reports in the literature, we have decided to modify our treatment of early postoperative infections to include repeat débridements, as advocated by Mont et al.²⁵. We also have increased the duration of antibiotic therapy for deep early postoperative infections from four to six weeks. We believe that comparison of alternative regimens for the clinical settings that we have described eventually will lead to improvement in the treatment of these very difficult infections.

	Footnotes

 
*No benefits in any form have been received or will be received from a commercial party related directly or indirectly to the subject of this article. No funds were received in support of this study.

Department of Orthopaedic Surgery, Niigata University School of Medicine, Niigata City 951-8510, Niigata, Japan.

Division of Infectious Diseases (D. T. T.) and Department of Orthopaedic Surgery (R. F. K.), Hennepin County Medical Center, 701 Park Avenue South, Minneapolis, Minnesota 55415.

Minneapolis Orthopedic and Arthritis Institute, 825 South 8th Street, Minneapolis, Minnesota 55404.

	References

Top Abstract Introduction Materials and Methods Results Discussion References

 

1. Ayers, D. C.; Dennis, D. A.; Johanson, N. A.; and Pellegrini, V. D., Jr.: Common complications of total knee arthroplasty. J. Bone and Joint Surg., 79-A: 278-311, Feb. 1997.[Free Full Text]
2. Backe, H. A., Jr.; Wolff, D. A.; and Windsor, R. E.: Total knee replacement infection after 2-stage reimplantation. Results of subsequent 2-stage reimplantation. Clin. Orthop., 331: 125-131, 1996.
3. Bengtson, S.; Blomgren, C.; Knutson, K.; Wigren, A.; and Lidgren, L.: Hematogenous infection after knee arthroplasty. Acta Orthop. Scandinavica, 58: 529-534, 1987.[Medline]
4. Bengtson, S., and Knutson, K.: The infected knee arthroplasty. A 6-year follow-up of 357 cases. Acta Orthop. Scandinavica, 62: 301-311, 1991.[Medline]
5. Booth, R. E., Jr., and Lotke, P. A.: The results of spacer block technique in revision of infected total knee arthroplasty. Clin. Orthop., 248: 57-60, 1989.
6. Borden, L. S., and Gearen, P. F.: Infected total knee arthroplasty. A protocol for management. J. Arthroplasty, 2: 27-36, 1987.[Medline]
7. Dietz, F. R.; Koontz, F. P.; Found, E. M.; and Marsh, J. L.: The importance of positive bacterial cultures of specimens obtained during clean orthopaedic operations. J. Bone and Joint Surg., 73-A: 1200-1207, Sept. 1991.[Abstract/Free Full Text]
8. Dupont, J. A.: Significance of operative cultures in total hip arthroplasty. Clin. Orthop., 211: 122-127, 1986.
9. Elia, E. A., and Lotke, P. A.: Results of revision total knee arthroplasty associated with significant bone loss. Clin. Orthop., 271: 114-121, 1991.
10. Ewald, F. C.: The Knee Society total knee arthroplasty roentgenographic evaluation and scoring system. Clin. Orthop., 248: 9-12, 1989.
11. Fitzgerald, F. H., Jr.; Peterson, L. F. A.; Washington, J. A., II; Van Scoy, R. E.; and Coventry, M. B.: Bacterial colonization of wounds and sepsis in total hip arthroplasty. J. Bone and Joint Surg., 55-A: 1242-1250, Sept. 1973.[Abstract/Free Full Text]
12. Goldman, R. T.; Scuderi, G. R.; and Insall, J. N.: 2-stage reimplantation for infected total knee replacement. Clin. Orthop., 331: 118-124, 1996.
13. Gristina, A. G., and Kolkin, J.: Current concepts review. Total joint replacement and sepsis. J. Bone and Joint Surg., 65-A: 128-134, Jan. 1983.[Free Full Text]
14. Grogan, T. J.; Dorey, F.; Rollins, J.; and Amstutz, H. C.: Deep sepsis following total knee arthroplasty. Ten-year experience at the University of California at Los Angeles Medical Center. J. Bone and Joint Surg., 68-A: 226-234, Feb. 1986.[Abstract/Free Full Text]
15. Gruninger, R. P.; Tsukayama, D. T.; and Wicklund, B.: Antibiotic impregnated PMMA beads in bone and prosthetic joint infections. In Orthopaedic Infections, pp. 66-74. Edited by R. B. Gustilo, R. P. Gruninger, and D. T. Tsukayama. Philadelphia, W. B. Saunders, 1989.
16. Hanssen, A. D.; Rand, J. A.; and Osmon, D. R.: Treatment of the infected total knee arthroplasty with insertion of another prosthesis. The effect of antibiotic-impregnated bone cement. Clin. Orthop., 309: 44-55, 1994.
17. Hanssen, A. D.; Trousdale, R. T.; and Osmon, D. R.: Patient outcome with reinfection following reimplantation for the infected total knee arthroplasty. Clin. Orthop., 321: 55-67, 1995.
18. Hartman, M. B.; Fehring, T. K.; Jordan, L.; and Norton, H. J.: Periprosthetic knee sepsis. The role of irrigation and débridement. Clin. Orthop., 273: 113-118, 1991.
19. Insall, J. N.; Thompson, F. M.; and Brause, B. D.: Two-stage reimplantation for the salvage of infected total knee arthroplasty. J. Bone and Joint Surg., 65-A: 1087-1098, Oct. 1983.[Abstract/Free Full Text]
20. Insall, J. N.; Dorr, L. D.; Scott, R. D.; and Scott, W. N.: Rationale of the Knee Society clinical rating system. Clin. Orthop., 248: 13-14, 1989.
21. Johnson, D. P., and Bannister, G. C.: The outcome of infected arthroplasty of the knee. J. Bone and Joint Surg., 68-B(2): 289-291, 1986.
22. Kamme, C., and Lindberg, L.: Aerobic and anaerobic bacteria in deep infections after total hip arthroplasty. Differential diagnosis between infectious and non-infectious loosening. Clin. Orthop., 154: 201-207, 1981.
23. McLaren, A. C., and Spooner, C. E.: Salvage of infected total knee components. Clin.Orthop., 331: 146-150, 1996.
24. Mirra, J. M.; Amstutz, H. C.; Matos, M.; and Gold, R.: The pathology of the joint tissues and its clinical relevance in prosthesis failure. Clin. Orthop., 117: 221-240, 1976.
25. Mont, M. A.; Waldman, B.; Banerjee, C.; Pacheco, I. H.; and Hungerford, D. S.: Multiple irrigation, débridement, and retention of components in infected total knee arthroplasty. J. Arthroplasty, 12: 426-433, 1997.[Medline]
26. Morrey, B. F.; Westholm, F.; Schoifet, S.; Rand, J. A.; and Bryan, R. S.: Long-term results of various treatment options for infected total knee arthroplasty. Clin. Orthop., 248: 120-128, 1989.
27. Poss, R.; Thornhill, T. S.; Ewald, F. C.; Thomas, W. H.; Batte, N. J.; and Sledge, C. B.: Factors influencing the incidence and outcome of infection following total joint arthroplasty. Clin. Orthop., 182: 117-126, 1984.
28. Rand, J. A., and Bryan, R. S.: Reimplantation for the salvage of an infected total knee arthroplasty. J. Bone and Joint Surg., 65-A: 1081-1086, Oct. 1983.[Abstract/Free Full Text]
29. Rand, J. A.; Bryan, R. S.; Morrey, B. F.; and Westholm, F.: Management of infected total knee arthroplasty. Clin. Orthop., 205: 75-85, 1986.
30. Rand, J. A.: Alternatives to reimplantation for salvage of the total knee arthroplasty complicated by infection. J. Bone and Joint Surg., 75-A: 282-289, Feb. 1993.[Free Full Text]
31. Rasul, A. T., Jr.; Tsukayama, D. T.; and Gustilo, R. B.: Effect of time onset and depth of infection on the outcome of total knee arthroplasty infections. Clin. Orthop., 273: 98-104, 1991.
32. Rosenberg, A. G.; Haas, B.; Barden, R.; Marquez, D.; Landon, G. C.; and Galante, J. O.: Salvage of infected total knee arthroplasty. Clin. Orthop., 226: 29-33, 1988.
33. Salvati, E. A.; Robinson, R. P.; Zeno, S. M.; Koslin, B. L.; Brause, B. D.; and Wilson, P. D., Jr.: Infection rates after 3175 total hip and total knee replacements performed with and without a horizontal unidirectional filtered air-flow system. J. Bone and Joint Surg., 64-A: 525-535, April 1982.[Abstract/Free Full Text]
34. Schoifet, S. D., and Morrey, B. F.: Treatment of infection after total knee arthroplasty by débridement with retention of the components. J. Bone and Joint Surg., 72-A: 1383-1390, Oct. 1990.[Abstract/Free Full Text]
35. Teeny, S. M.; Dorr, L.; Murata, G.; and Conaty, P.: Treatment of infected total knee arthroplasty. Irrigation and débridement versus two-stage reimplantation. J. Arthroplasty, 5: 35-39, 1990.[Medline]
36. Tsukayama, D. T.; Estrada, R.; and Gustilo, R. B.: Infection after total hip arthroplasty. A study of the treatment of one hundred and six infections. J. Bone and Joint Surg., 78-A: 512-523, April 1996.[Abstract/Free Full Text]
37. Wasielewski, R. C.; Barden, R. M.; and Rosenberg, A. G.: Results of different surgical procedures on total knee arthroplasty infections. J. Arthroplasty, 11: 931-938, 1996.[Medline]
38. Weiss, A. P., and Krackow, K. A.: Persistent wound drainage after primary total knee arthroplasty. J. Arthroplasty, 8: 285-289, 1993.[Medline]
39. Wilson, M. G.; Kelley, K.; and Thornhill, T. S.: Infection as a complication of total knee-replacement arthroplasty. Risk factors and treatment in sixty-seven cases. J. Bone and Joint Surg., 72-A: 878-883, July 1990.[Abstract/Free Full Text]
40. Windsor, R. E.; Insall, J. N.; Urs, W. K.; Miller, D. V.; and Brause, B. D.: Two-stage reimplantation for the salvage of total knee arthroplasty complicated by infection. Further follow-up and refinement of indications. J. Bone and Joint Surg., 72-A: 272-278, Feb. 1990.[Abstract/Free Full Text]

CiteULike

Connotea

Del.icio.us

Facebook

Technorati

Twitter

This article has been cited by other articles:

				G. Bori, A. Soriano, S. Garcia, C. Mallofre, J. Riba, and J. Mensa Usefulness of Histological Analysis for Predicting the Presence of Microorganisms at the Time of Reimplantation After Hip Resection Arthroplasty for the Treatment of Infection J. Bone Joint Surg. Am., June 1, 2007; 89(6): 1232 - 1237. [Abstract] [Full Text] [PDF]

				Q. Cui, W. M. Mihalko, J. S. Shields, M. Ries, and K. J. Saleh Antibiotic-Impregnated Cement Spacers for the Treatment of Infection Associated with Total Hip or Knee Arthroplasty J. Bone Joint Surg. Am., April 1, 2007; 89(4): 871 - 882. [Full Text] [PDF]

				A. Trampuz, K. E. Piper, A. D. Hanssen, D. R. Osmon, F. R. Cockerill, J. M. Steckelberg, and R. Patel Sonication of Explanted Prosthetic Components in Bags for Diagnosis of Prosthetic Joint Infection Is Associated with Risk of Contamination J. Clin. Microbiol., February 1, 2006; 44(2): 628 - 631. [Abstract] [Full Text] [PDF]

				J. M. Leone and A. D. Hanssen Management of Infection at the Site of a Total Knee Arthroplasty J. Bone Joint Surg. Am., October 1, 2005; 87(10): 2335 - 2348. [Full Text] [PDF]

				R. Trebse, V. Pisot, and A. Trampuz Treatment of infected retained implants J Bone Joint Surg Br, February 1, 2005; 87-B(2): 249 - 256. [Abstract] [Full Text] [PDF]

				W. Zimmerli, A. Trampuz, and P. E. Ochsner Prosthetic-Joint Infections N. Engl. J. Med., October 14, 2004; 351(16): 1645 - 1654. [Full Text] [PDF]

				R. J. Sierra, R. T. Trousdale, and M. W. Pagnano Above-the-Knee Amputation After a Total Knee Replacement: Prevalence, Etiology, and Functional Outcome J. Bone Joint Surg. Am., May 28, 2003; 85(6): 1000 - 1004. [Abstract] [Full Text] [PDF]

				D. T. Tsukayama, V. M. Goldberg, and R. Kyle Diagnosis and Management of Infection After Total Knee Arthroplasty J. Bone Joint Surg. Am., January 21, 2003; 85(90001): S75 - 80. [Full Text] [PDF]

This Article Abstract Full Text (PDF) Free Letters to the Editor: Submit a response Alert me when this article is cited Alert me when Letters to the Editor are posted Alert me if a correction is posted Services E-mail this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Add to My File Cabinet Download to citation manager Rights and Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by SEGAWA, H. Articles by GUSTILO, R. B. Search for Related Content PubMed PubMed Citation Articles by SEGAWA, H. Articles by GUSTILO, R. B. Social Bookmarking                   What's this?

Stanford University Libraries' HighWire Press (TM) assists in the publication of JBJS Online.
Copyright © 1999 by The Journal of Bone and Joint Surgery, Inc. Online ISSN: 1535-1386.
The Journal of Bone and Joint Surgery®, JBJS®, JB&JS®, and eJBJS® are registered in the U.S. Patent and Trademark Office.