**3.2 The role of biofilms**

416 Recent Advances in Arthroplasty

a small number of cases meet the definition for prosthetic joint infection, and yet remain

**Reference 1 2 3 4 5 6 7 8 9** 

Staphylococcus aureus 23 21 14 23 38 22 42 19 17 **23.8**  Streptococcus spp 5 7 10 9 13 9 6 12 10.7 **9.4**  Enterococcus spp 3 6 7 3 0 1 3 10 6.3 **4.6**  Diptheroids 2 4 4 0.5 2 0.6 1 0 5.4 **2.6**  Gram-negative bacilli 6 28 0 6 11 8 5 29 16.1 **12.7**  Propionibacterium 0 0.4 3 1.5 0 0 0.3 2 0.9 **0.9**  Polymicrobial 33 0 12 12 6 19 15 0 0 **9.7**  Anaerobes 3 2 0 2 0 6 2 2 6.3 **3.0**  Other 1 4 0 2 0 3 0.3 19 2.7 **3.5**  Culture negative 5 0 0 11 10 12 8 2 1.8 **5.2**  1. (Moran, et al. 2007), 2. (Sharma, et al. 2008), 3. (Pandey, et al. 2000), 4. (Steckelberg & Osmon 2000), 5. (Pulido, et al. 2008), 6. (Berbari, et al. 1998), 7. (Bengtson & Knutson 1991), 8. (Fitzgerald, et al. 1977), 9.

Acquisition of prosthetic joint infection occurs by two mechanisms: direct inoculation and haematogenous seeding. Direct inoculation of the prosthesis may occur at the time of implantation or with manipulation of the arthroplasty and is thought to be the predominant mechanism of infection. In a study by Southwood et al the 50% infective dose (ID50) of *Staphylococcus aureus* required to induce infection with direct inoculation of the prosthesis was just 50 organisms. This compared to an intravenous inoculum dose of 100 000 organisms at the time of operation for bacteraemic seeding and infection of the prosthesis to occur. Southwood also demonstrated that three weeks after implantation of the prosthesis, the likelihood of bacteraemic seeding of the prosthesis was significantly reduced. In fact, in the rabbit model, the inoculum of intravenous bacteria required was near to the lethal dose(Southwood, et al. 1985). Nevertheless, haematogenous seeding remains an important cause of arthroplasty infections and it has been reported that up to 34% of patients with prosthetic joints in-situ developed deep infection of that prosthesis following an intercurrent

Whilst theoretically distinct, clinically there is significant overlap between both mechanisms of infection. The simplified view is that infection resulting from inoculation occurs within the first year of implantation whilst haematogenous infections occur later. However the clinical presentation of prosthetic joint infections acquired during the original operation

*112 248 81 578 63 462 357 42 112* 

13 31 48 30 21 19 17 24 33 **27.3** 

culture negative on standard microbiologic techniques.

Table 1. Microbiological isolates in reported literature (percent)

episode of *Staphylococcus aureus* bacteraemia (Murdoch, et al. 2001).

*Total number of isolates* 

Coagulase negative Staphylococcus

(McDonald, et al. 1989)

**3. Pathogenesis** 

**3.1 Acquisition of infection** 

The pathogenesis of prosthetic joint infections is intimately connected to the property of biofilm formation by microorganisms. The presence of this biofilm can have a critical effect on the likely success of treatment for a number of reasons. Bacteria can exist in two unique forms; the free living or planktonic forms characterised by rapid cellular division, and the stationary or sessile forms characterised by slower cellular division (Costerton 1999, Costerton, et al. 1995).

The sessile bacteria secrete an extracellular matrix or slime. Together the microorganisms and this matrix comprise what is known as 'the biofilm'. The abiotic matrix performs a number of functions including provision of anchorage onto structures to support the sessile colonies(Donlan & Costerton 2002). It also facilitates communication between bacteria within the biofilm. This communication termed 'quorum sensing', is analogous to the paracrine signalling in multicellular organisms and enables the bacteria to regulate their gene synthesis(Gristina & Costerton 2009). Importantly, the matrix can provide bacteria with protection from antimicrobial chemicals and from host defense mechanisms. This impairment of host defense mechanisms has been demonstrated in a number of in vitro models. For example, the extracellular slime produced by *Staphylococcus epidermidis* can inhibit the phagocytic activity of neutrophils(Shiau & Wu 1998).

The concentration of antibiotic required to inhibit the growth of bacteria in biofilms is higher than that required to kill free-living bacteria. The mean inhibitory concentration (MIC) of many antibiotics is higher with the sessile forms than corresponding planktonic forms. Studies have demonstrated up to a 1000 fold increase in the MIC to particular antibiotics for bacteria moving from the planktonic to the sessile phenotype (Amorena, et al. 1999, Jones, et al. 2001, Rose & Poppens 2009, Schwank, et al. 1998, Souli & Giamarellou 1998, Stewart & Costerton 2001). This poses a major challenge for clinicians interpreting the reported antibiotic susceptibility results of bacteria, as our standard laboratory antibiotic susceptibility testing uses only the planktonic forms of bacteria. Newer technologies including the Calgary Biofilm Device can enable antibiotic susceptibility testing of the sessile phenotype of bacteria, but at present these are limited to a research setting and are not widely available(Ceri, et al. 1999).

There are a number of postulated mechanisms for the apparent resistance of biofilm residing bacteria to the effects of antibiotics. Firstly, the antibiotic may be deactivated at the surface of the biofilm. Secondly, the altered nutritional and biochemical environment within the biofilm may alter the activity of the antibiotics. Thirdly, antibiotics, particular cell wall active antibiotics such as betalactam antibiotics, rely on rapid growth and reproduction of the microorganism for their effect. These antibiotics are effective against the planktonic phenotype but have limited efficacy against the sessile phenotype as cellular turnover is greatly reduced. Finally the sessile forms act as 'spore-like' structures, which may act as a

Infection in Primary Hip and Knee Arthroplasty 419

day 2 and may remain elevated for 3 weeks (Aalto, et al. 1984, Larsson, et al. 1992, Shih, et

The search for other biochemical markers of infection has included interleukin 6 (Il-6), tumour necrosis factor α (TNF-α) and procalcitonin C. Il-6 and TNF-α are cytokines released by monocytes and macrophages in the setting of infection (Bottner, et al. 2007). Procalcitonin is a precursor of calcitonin, and has been shown to be a specific marker of bacterial sepsis (Fernandez Lopez, et al. 2003). In a review by Bottner et al of 78 patients undergoing revision arthroplasties Il-6, TNF-α and procalcitonin were all significantly elevated in patients with confirmed septic loosening. The sensitivity and specificity respectively of Il-6 was 95% and 87%, TNF-α 43% and 94% and procalcitonin 33% and 98%. (Bottner, et al. 2007). Il-6 is elevated in the post-operative period for primary arthroplasty however, in a study by Shah et al, Il-6 was shown to return to normal levels within 2 days of the operation. Therefore there is potential diagnostic utility of Il-6 over CRP and ESR in the early post-operative period if infection is suspected, particularly in

Synovial fluid characteristics can be used to assist in diagnosis of prosthetic joint infections. In a study by Trampuz and colleagues, the leucocyte count was significantly higher in patients with prosthetic joint infection with a median of 18.9 x 103/µL (range, 0.3 to 178 x 103/µL) compared to a median leucocyte count of 0.3 x 103/µL (range, 0.1 to 16 x 103/µL) in patients with aseptic loosening. Using receiver operating characteristic (ROC) curves the authors found a synovial total white cell count 1.7 x 103/µL and a leucocyte differential of greater than 65% neutrophils had a sensitivity and specificity of 94%, 88% and 97%, 98%

Plain radiographs lack sensitivity and specificity in diagnosing septic arthroplasty. Findings such as lucency around the prosthesis can be noted in both septic and aseptic loosening situations (Figure 1 A-D). In early infection plain radiographs are frequently normal (Miller

Technetium-Methylene Diphosphonate (MDP) bone scintigraphy is a sensitive test for prosthetic joint infection (Figure 1 E-G), but lacks specificity, as it does not differentiate between aseptic and septic loosening(Ghanem, et al. 2009). The bone scan can also remain positive for a year following primary arthroplasty. Bone scan does have a high negative predictive value therefore bone scans potentially can be used to exclude infection in the setting of a painful prosthetic joint (Smith, et al. 2001). Similar findings have been documented with newer modalities such as 18F-Fluoro-deoxyglucose positron emission tomography (FDG-PET) (Delank, et al. 2006, Zoccali, et al. 2009). A recent meta-analysis of FDG-PET reported a sensitivity of 82.1% and specificity of 86.6% for the presence of prosthetic joint infection, and hence this may be a useful test if available(Kwee, et al.

Computer tomography (CT) and magnetic resonance imaging are not considered useful imaging modalities due to artefact from the metal prosthesis interfering with interpretation of imaging findings. However newer CT scanners can minimise this effect and may be useful in detecting abnormalities of the soft tissues in periprosthetic infections (Figure 2 A-

E) but do not diagnose periprosthetic bone abnormalities well (Cyteval, et al. 2002)

al. 1987).

the first 21 days (Shah, et al. 2009).

respectively (Trampuz, et al. 2004).

**4.3 Radiological studies** 

2005).

2008).

nidus for later relapse of infection (Costerton 1999, Stewart & Costerton 2001, Trampuz, et al. 2003, Zimmerli 2006).

The properties of the biofilm alter with time; with age many biofilms become increasingly resistant to antibiotics. Monzon et al demonstrated the efficacy of vancomycin against *Staphylococcus epidermidis* decreased as a biofilm aged. This phenomenon was not consistent with all antibiotics; the activity of rifampicin and tetracyclines was not altered (Monzon, et al. 2002). Using Ribosomal RNA Fluorescence In Situ Hybridization studies, Poulson et al assessed the growth rate of biofilms and demonstrated that the cellular turnover was significantly higher in younger biofilms compared to established biofilms (Poulsen, et al. 1993). This finding could account for the difference to antimicrobial susceptibility observed. Implant factors are also recognised to play a role in the pathogenesis of infection. Biochemical properties of prosthetic material influences bacterial adhesion and may impair host immune responses. For example, methyl methacrylate cement has been shown to inhibit complement and lymphocyte activity (Panush & Petty 1978, Petty 1978).
