Pseudomonas aeruginosa is the major pathogen in cystic fibrosis (CF) patients and is an indicator of poor prognosis in CF patients, especially from the onset of the chronic stage when colonies become mucoid and variant phenotypes emerge. Early detection is essential given the success of early aggressive eradication therapy [6, 7]. Therefore, the most prevalent detection and identification methods, i.e. culture and (real-time) PCR, should be optimized to achieve the highest sensitivity.
West et al.  reported that specific P. aeruginosa antibodies were detectable between 6 and 12 months prior to the first positive culture for P. aeruginosa from respiratory samples. These findings suggest that culture may miss P. aeruginosa in the early stages of colonization. Also at later stages, culture can miss the emerging P. aeruginosa phenotypic variants such as the pyoverdine negative mutants, the slowly growing variants, the small colony variants and the auxotrophs, which do not grow on standard media [9, 10]. Therefore, the development of improved culture methods and/or of molecular methods is warranted, not only for early detection but also for follow up of colonized patients. However, although several molecular assays for the detection of Pseudomonas species have been described (e.g., [11, 13–19, 22–26]), surprisingly few studies have compared selective and nonselective culture methods with the different molecular methods that have been described for the detection of P. aeruginosa directly from clinical samples.
The studies comparing sensitivity of culture and species-specific PCR for the detection of P. aeruginosa from sputa of CF patients indicate comparable efficiency of both methods [8, 16], with slightly higher sensitivity for PCR in some studies [12, 18] or clearly higher sensitivity for PCR [13, 26]. We used the PCR format published by De Vos et al.  in combination with optimized DNA-extraction methods and used in addition real-time PCR to increase PCR sensitivity further. However, using a sputum dilution series of P. aeruginosa, and in accordance to most studies, we found no difference in sensitivity between any of the three culture methods and the most sensitive molecular method, i.e. DNA-extraction with easyMAG protocol Generic 2.0.1 and proteinase K pretreatment combined with any of the three probe-based real-time PCRs. In our hands, culture was more sensitive than PCR and SybrGreen based real-time PCR and the difference was even more pronounced when not optimal DNA-extraction methods were used. It should be noticed that we found no difference between selective and nonselective culture methods, but this may be due to the fact that no bacteria, other than P. aeruginosa in the two P. aeruginosa positive patients, could be cultured from the sputa of the 8 CF patients. As shown in other studies and confirmed here, the pretreatment of the sample and the DNA-extraction protocol strongly influence the sensitivity of the PCR [27, 28]. The most sensitive molecular detection method was obtained using the easyMAG Generic 2.0.1 protocol with proteinase K pretreatment in combination with real-time PCR with the TaqMan probe or the HybProbes. Previous studies showed already that the easyMAG extractor is one of the most sensitive and reliable methods for DNA-extraction [29–31]. An additional advantage of automated DNA-extraction like easyMAG might be the lower sample processing variability .
Because both approaches, i.e. culture and (real-time) PCR, have important advantages as well as drawbacks [14, 20, 32, 33], in our opinion, both should be or can be combined. PCR technology has the potential to detect the fastidious P.aeruginosa variants, which are not detected by the routinely used classical culture procedures [9, 10], whereas culture yields a complete genome that can be used for e.g. phenotypic susceptibility testing and whole genome based genotyping techniques like RAPD, PFGE and AFLP . Indeed, several of the published studies indicate that there are instances of culture positive PCR negative samples [11, 12, 15] as well as culture negative PCR positive samples [11–13, 18, 19], whereby P. aeruginosa infection can only be reliably demonstrated when both approaches are combined.