W. Costerton (Costerton et al., 1981). Similarly, Niels Høiby had observed that the aggregation of P. aeruginosa in the sputum of chronically infected CF patients was relevant to CF-associated lung infection compared with single-celled
bacteria (Høiby, 1977). In 1984, Costerton formally outlined the hypothesis that organisms like P. aeruginosa could behave similarly in human infections to the way they behaved in the environment. He further suggested that ‘glycocalyx-enclosed biofilms of P. aeruginosa BMS-777607 clinical trial or other bacteria have been identified in experimental or clinical infections arising from contaminated prostheses and in chronic refractory infections, such as endocarditis, osteomyelitis, and P. aeruginosa pneumonia associated with cystic fibrosis.’ (Costerton, 1984; Høiby et al., 1986). Clinicians may be more familiar with foreign body (implant) infections because of microbial attachment to a nonliving surface distinguished from biofilms associated with host tissues, or ‘native tissue infections’ (Lynch
& Robertson, 2008). These latter infections include Paclitaxel in vitro chronic lung infections of CF patients, chronic otitis media (OM), native valve (infectious) endocarditis (IE), and chronic wounds (Table 1). More broadly, we propose that BAI are ‘infections due to aggregated, pathogenic or opportunistic microorganisms encased in an exopolysaccharide matrix and recalcitrant to host defense mechanisms and antimicrobial treatment.’ The pathogenesis of many biofilm infections these also includes normal microbial flora of mucosal membranes or the skin, which gain access to an organ via foreign bodies and clinicians should suspect biofilm infections in such situations (Table 2).
BAI present significant challenges to current clinical practice guidelines because of the inherent difficulty in determining whether the infection is biofilm-related or is due to an acute infection with planktonic microorganisms. Therefore, functional, clinically relevant criteria would help to: (1) better distinguish BAI from acute planktonic infections, (2) obtain appropriate clinical samples, and (3) provide focus for the development of routine clinical tests. Criteria for biofilm infections have been previously proposed and modified, based on the initial Parsek–Singh criteria (Parsek & Singh, 2003; Hall-Stoodley & Stoodley, 2009) (Table 3). These criteria exemplify several characteristic features of BAI. The first two criteria include fundamental definitions of biofilms discussed earlier, such as association with a surface and aggregation. Whenever possible, sampling surfaces suspected of harboring biofilm microorganisms is preferred, even if fluid samples are also available. This is problematic, however, as it may involve invasive procedures such as biopsy, needle aspiration, or removal of an implant.