A healthy, 65-year-old woman presented to the emergency room with complaints of increased right knee redness, warmth, swelling, and pain 3 weeks after undergoing a primary right total knee arthroplasty (TKA). Physical examination revealed a swollen, red knee that was diffusely tender. Her passive range of motion was between 20 and 70 degrees, and she had significant pain throughout the arc of motion. The serum white blood cell (WBC) count was 8900/µL, with a differential finding of 85% neutrophils. The erythrocyte sedimentation rate (ESR) was 95 mm/hr, and the C-reactive protein (CRP) level was 166 mg/L. The knee was aspirated in the emergency department, and the fluid was sent for a WBC count and differential analysis and for aerobic and anaerobic bacterial and fungal cultures. The synovial cell count revealed 35,000 nucleated cells/µL, with a differential finding of 95% neutrophils.
The patient was diagnosed with an acute postoperative periprosthetic infection. Staged débridement with component retention was recommended. All antibiotics were withheld in anticipation of findings from the intraoperative tissue cultures. Six hours later, the patient was taken to the operating room for the first stage of a planned two-stage débridement. Her polyethylene insert was removed and sterilized, three tissue samples were taken for culture, a thorough débridement was performed, and the tibial insert was replaced. Antibiotic-loaded bone cement beads were placed, and the wound was closed.
Postoperatively, the patient was admitted to the surgical nursing unit, where she was continued on empirical parenteral antibiotics. An infectious disease consultation was sought. On postoperative day 2, her cultures grew methicillin-sensitive Staphylococcus aureus (MSSA). The antibiotic regimen was transitioned to rifampin and parenteral nafcillin.
On postoperative day 5, she was returned to the operating room for the second-stage débridement. The polyethylene insert and antibiotic beads were removed. Repeat irrigation and débridement were performed. A new tibial insert was placed and the wound closed. Postoperatively, she continued on rifampin and intravenous nafcillin for 6 weeks before being transitioned to oral trimethoprim/sulfamethoxazole. Her incision healed uneventfully. At clinical follow-up 5 months later, her knee was functioning well, there were no signs of infection, and levels of inflammatory markers had normalized. Oral antibiotics were discontinued, and routine postoperative follow-up continued. Examination 2 months later found that the levels of inflammatory markers remained normal.
Biofilm-associated microorganisms possess numerous resistance mechanisms and require local antibiotic concentrations on the order of 100 to 1000 times the minimum inhibitory concentration (MIC) for effective treatment.
Débridement with component retention is indicated for patients with periprosthetic infection less than 4 weeks in duration and those with prolonged postoperative wound drainage in whom a deep infection is suspected.
The diagnosis of periprosthetic joint infection is confirmed by analysis of serum levels of inflammatory markers and a synovial fluid cell count. Interpretation of these results depends on the timing of presentation (i.e., acute postoperative versus acute delayed or hematogenous infection).
Meticulous débridement is the keystone to success.
We think a staged débridement with interval use of antibiotic-loaded cement beads is superior to a single débridement to attain the local MICs necessary to kill residual biofilm-associated bacteria.
Combination therapy that includes rifampin should be employed for staphylococcal infections.
Antibiotic-loaded bone cement is considered safe. There have been no reports of systemic toxicity in patients receiving antibiotic-loaded cement beads alone.
Management of implant-related infections is difficult because of the presence of biofilm-associated pathogens, which possess several mechanisms for antibiotic resistance. Successful management of implant-related infections depends on understanding biofilm physiology and translating this knowledge into clinical practice. A comprehensive discussion of biofilm physiology is beyond the scope of this chapter, but some basic concepts about its formation and mechanisms of resistance are reviewed.
A biofilm develops in stages. During the initial attachment phase, free-flowing (i.e., planktonic ) bacteria adhere to a protein-coated surface. In the intermediate phase, small aggregates of bacteria known as microcolonies or cell clusters form and increase in size. During this process, the bacteria continuously secrete molecules known as autoinducers or pheromones , for which the bacteria also possess receptors. As the bacterial population grows, the autoinducer concentration increases until a threshold is reached. This threshold triggers activation of the autoinducer receptors, which alters gene transcription. Numerous phenotypic changes ensue, including the production of glycocalyx, which consists of proteins, extracellular DNA, and polysaccharides. This process of cell-to-cell communication is called quorum sensing . The bacteria residing in the biofilm are referred to as sessile (rather than planktonic), and they are phenotypically heterogeneous.
The mechanisms by which the bacteria in a biofilm achieve decreased susceptibility to the host immune system and antibiotics are not fully understood. It was originally thought that limited drug diffusion into the protective environment of the biofilm matrix caused antibiotic resistance. This does hold true for aminoglycosides because the diffusion coefficient in a biofilm is significantly limited, probably due to binding of the negatively charged biofilm to the positively charged aminoglycoside. However, for most other antibiotics, diffusion is not dramatically limited. The estimated effective diffusion coefficients of most antibiotics in biofilm are approximately 40% to 80% of those in pure water. It is improbable that this modest decrease in diffusion independently accounts for the reduced antibiotic susceptibility.
Factors such as local metabolic environment, bacterial replication rate, production of protective enzymes, and gene transfer likely play a larger role in bacterial resistance. Although antibiotics can physically penetrate the biofilm, they do not necessarily retain their antimicrobial properties. The local environment in a biofilm has a relatively low pH, high partial pressure of carbon dioxide (P co 2 ), high carbon dioxide level, low divalent cation and pyrimidine concentration, and low hydration level, which may have negative effects on antimicrobial activity. High local levels of enzymes such as lactamases, chelators, and aminoglycoside-modifying enzymes compromise the activity of antibiotics when they are at lower concentrations. Propagation of antibiotic resistance is further enhanced in a biofilm by horizontal gene transfer between microorganisms and by transcription regulation through quorum sensing. Many biofilm-associated bacteria enter a slow-growing state and are known as persisters . Phenotypic changes may result from quorum sensing and from local metabolic factors such as the lack of nutrients and oxygen and the accumulation of nitrates. Because antimicrobials are most effective at treating rapidly growing organisms, these metabolic changes decrease bacterial susceptibility. Numerous fungal species also produce biofilms and share many of the antimicrobial resistance mechanisms of their bacterial counterparts.
As a result of the numerous survival strategies employed by biofilm-associated microorganisms, antibiotic concentrations achieved by systemic therapy are less effective. Antibiotic concentrations 100 to 1000 times the minimum inhibitory concentration (MIC) typically required for planktonic microorganisms are needed to kill biofilm-associated pathogens.
Treatment options for periprosthetic infections include open débridement with component retention, one-stage exchange, two-stage exchange, arthrodesis, resection arthroplasty, amputation, and long-term antimicrobial suppression. There is no role for arthroscopic débridement of periprosthetic joint infections. Historically, débridement with component retention for the treatment of acute infections has had highly variable and usually poor results. Even the most thorough mechanical débridement and irrigation cannot reliably remove all biofilm material from a joint, and postoperative parenteral antibiotics do not reach the 100 to 1000 times MIC that is locally required to be effective against residual biofilm-associated bacteria. For this reason, it is advantageous to include depot antibiotic-loaded bone cement in the treatment regimen to provide the local concentrations necessary for the eradication of biofilm-associated organisms.
There are no published data on in vivo joint fluid antibiotic levels after implantation of high-dose antibiotic-loaded cement beads. Using two batches of low-dose antibiotic-loaded beads (i.e., 0.5 g of gentamicin and 2 g of vancomycin per 40-g batch of polymethylmethacrylate [PMMA]) for the treatment of total hip arthroplasty infections, Agnostakos and colleagues reported mean joint fluid gentamicin and vancomycin levels in drain fluid on postoperative day 1 of 80 µg/mL (range, 21 to 19 µg/mL) and 116 µg/mL (range, 12 to 371 µg/mL), respectively. Hsieh and co-workers reported in vivo joint fluid levels after implantation of high-dose antibiotic-loaded cement hip spacers (i.e., 4 g of vancomycin and 4 g of aztreonam per 40-g batch of PMMA). Mean drain fluid levels for vancomycin and aztreonam on postoperative day 1 were 1538 µg/mL and 1003 µg/mL, respectively. Elution characteristics are typically better for antibiotic-loaded cement beads than for spacers, likely because of the increased surface area of beads. We perform a two-stage débridement with interval use of high-dose antibiotic-loaded cement beads for the treatment of acute periprosthetic infections.
Indications and Contraindications
Numerous factors influence the outcomes of periprosthetic joint infection treatment, most notably infection duration, host status, débridement technique, and antibiotic regimen. These infections are classified as acute or chronic.
Acute infections can be subclassified as acute postoperative or acute delayed (hematogenous) infections. Acute postoperative infections manifest within 4 weeks after surgery. They are likely caused by wound colonization at the time of surgery or by superficial wound infections spreading to the periprosthetic space. Hematologic seeding of a recently replaced joint may also be a pathway for infection early after surgery, but this mode of infection is difficult to verify early postoperatively. Acute delayed (hematogenous) infections arise in a previously well-functioning joint, and the onset of symptoms is relatively sudden. Because a remote source of infection is often difficult to identify, better terms may be acute delayed infection for cases with no obvious remote source and acute hematogenous infection for cases with an identified source.
Chronic infections are usually low-grade, indolent infections that are thought to originate at the time of surgery, but because of a small inoculum or low virulence of the organism, the onset of symptoms is delayed. These infections typically manifest within months or the first few years after surgery with a progressive deterioration in function and increase in pain. Acute infections with a missed or delayed diagnosis longer than 4 weeks after surgery should be considered chronic.
A fourth category includes periprosthetic infections with two or more positive culture results for the same organism taken at the time of revision for presumed aseptic failure. The significance of this finding is determined after weighing other clinical factors.
Prolonged duration of infection is associated with biofilm maturation and the potential for osteomyelitis. Débridement with component retention is indicated for patients with acute postoperative or acute hematogenous infections only. Traditionally, these are cases of infections lasting less than 4 weeks after surgery. However, some clinicians have provided evidence of improved outcomes if the débridement is performed within 3 weeks or less of infection onset. When taking a history from the patient, it is important to determine the exact duration of symptoms and recent risk factors for acute hematogenous spread to accurately categorize the infection as acute or chronic. It is also important to get a sense of how the joint has functioned since the primary surgery and whether there were postoperative wound healing problems to asses the risk of chronic infection.
The diagnosis of an acute postoperative infection can be challenging because there is considerable variability in the postoperative progression after joint replacement. Historically, results of inflammatory markers (i.e., erythrocyte sedimentation rate [ESR] and C-reactive protein [CRP]) have been difficult to interpret because these studies are nonspecific and levels are elevated in the early postoperative period in response to surgery. Synovial white blood cell counts are also elevated because of the postoperative inflammatory response.
One study reviewed joint aspirations performed within 6 weeks after primary total knee arthroplasty (TKA) and evaluated the diagnostic performance of the synovial white blood cell count and differential count, the serum level of CRP, and the ESR. It found that the optimal synovial white blood cell cut-off value was 27,800 cells/µL, giving a sensitivity of 84% and specificity of 99%. Unlike its use in chronic or acute delayed infection, the synovial cell differential was less useful in this case, with an optimal value of 89% neutrophils that resulted in a sensitivity of 84% and specificity of only 69%. The ESR values were not helpful. The infected group had a mean ESR of 80 mm/hr (range, 38 to 140 mm/hr), and the noninfected group had a mean ESR of 75 mm/hr (range, 1 to 140 mm/hr). The CRP values were significantly different for the two groups. The infected group had a mean of 171 mg/L (range, 29 to 490 mg/L), and the noninfected group had a mean of 88 mg/L (range, 4 to 380 mg/L), but there was considerable variability in the range.
Diagnosis of acute delayed infections relies on the clinical presentation and laboratory studies. Patients with an elevated serum ESR (>30 mm/hr) or elevated CRP level (>10 mg/L) should undergo joint aspiration. The aspirate is sent for microbiologic cultures, synovial fluid analysis, crystal detection, and white blood cell count with a differential cell count. Although the optimal synovial cell count and cell differential values vary among studies, a synovial cell count of more than 2500 cells/µL with more than 60% to 65% neutrophils typically indicates infection (see “Future Considerations”).
In addition to patients with a confirmed acute periprosthetic joint infection, patients with prolonged postoperative wound drainage after the index procedure should be considered for treatment with early (≤7 days) débridement. All antibiotics should be withheld until intraoperative tissue cultures have been obtained. The host status of all patients undergoing débridement with component retention should be optimized aggressively, including management of edema, diabetes, nutrition, and tobacco use.
There are conflicting reports about the outcomes of treating staphylococcal species. Many have reported inferior outcomes for retention débridement used for treating methicillin-resistant Staphylococcus aureus (MRSA). Others have found no relationship between the causative organism and outcome.
In vitro studies have shown poor efficacy for vancomycin monotherapy and have improved efficacy with rifampin used in combination therapy for the treatment of staphylococcal isolates in biofilms. Rifampin inhibits bacterial DNA-dependent RNA polymerase and is bactericidal against sessile staphylococcal species. Because resistance develops quickly, rifampin must be given in combination with another antibiotic. Promising results for the treatment of biofilm-associated infections have been reported when rifampin is used in combination with beta-lactam antibiotics or fluoroquinolones for sensitive organisms and with daptomycin, linezolid, minocycline, tigecycline, or quinupristin/dalfopristin for methicillin-resistant organisms. Clinical and in vitro results have been poor when rifampin was combined with antimicrobial.
A randomized, controlled trial compared 18 patients receiving ciprofloxacin-rifampin combination therapy with 15 patients receiving ciprofloxacin-placebo therapy for periprosthetic staphylococcal infections. There was a 100% cure rate in the ciprofloxacin-rifampin combination therapy group and a 58% cure rate in the ciprofloxacin-placebo group. In our retrospective series, only one of eight patients with a staphylococcal infection failed treatment. One of the three patients with MRSA infections failed treatment. All patients with staphylococcal infections received rifampin in combination therapy. Another retrospective study employing rifampin in combination therapy for most staphylococcal infections undergoing débridement with component retention ( n = 34) did not find the infecting organism predictive of outcome.
Many of the studies reporting inferior outcomes for débridement with component retention for periprosthetic staphylococcal infection did not regularly use rifampin in a combination treatment regimen or did not discuss its use in their articles. In light of this, we think that staphylococcal infections, including MRSA, are not a contraindication to a two-stage débridement using antibiotic beads. Waiting for aspirate culture results before surgical decision making delays surgery, allows further biofilm maturation, and potentially reduces overall success rates.
The following equipment is needed for treatment of an acute periprosthetic infection:
Padded lamina spreader
Rongeurs and curettes
Implant-specific polyethylene extraction tool or osteotome
6 to 9 L of irrigation solution
Betadine (povidone-iodine) or chlorhexidine gluconate
Small, sterile scrub brush or toothbrush
Clean set-up for closure
If staged débridement with antibiotic-loaded cement beads is planned, the following materials are needed:
One batch of PMMA
One large-diameter Prolene suture
Antibiotic powder (e.g., 3.6 g of tobramycin, 3 g of vancomycin, 2 g of cefazolin)
Débridement can be performed as a one- or two-stage procedure. The two-stage procedure includes placement of antibiotic-loaded cement beads at the time of the first procedure. The initial débridement technique is the same for both procedures. The two-stage technique described here has been previously reported.
If the organism is unknown, all preoperative antibiotics are withheld until tissue cultures are obtained.
The previous surgical approach is used. If more than one previous skin incision exists, the most lateral incision should be used. If synovial fluid has not been sent for culture, immediate inoculation of synovial fluid into aerobic and anaerobic blood culture flasks is preferred to standard swab cultures.
The joint space is exposed with a medial parapatellar arthrotomy ( Fig. 34.1 ).