The antibody response to P. aeruginosa in cystic fibrosis has been considered a marker of chronic infection, inflammation and tissue damage [22,23]. In the last three decades, several studies have shown the potential of antibody detection for early diagnosis of P. aeruginosa pulmonary infection [15-18] in order to initiate the early eradication therapy .
The standardization of a quantitative ELISA for detection of specific antibodies requires that the concentrations of all reagents and reaction times be optimized. It is important the measurement of enzymatic activity during the linear part of the reaction, when substrate concentration is much higher than enzyme concentration. The standardization of the assay involves the use of a positive standard from which a standard curve is prepared and included in each ELISA plate, with the reactivity of the samples being tested determined from the standard curve. The utilization of a standard curve in every reaction plate reduce the variability of the assay [24,25].
In our study, the rate of anti-P.aeruginosa IgG seropositivity and the concentration of this antibody were significantly higher in the patients chronically infected than those in the other CF groups, corroborating previous studies that showed high frequency of detection and elevated levels of this antibody in this phase of infection [18,26,27]. One serum sample from a CF patient with chronic infection by P. aeruginosa who was under antibiotic therapy gave a negative serological result for the pathogen. Some studies have shown that antibiotic treatment regimens may result in decrease of the antibody levels to P. aeruginosa [28-30].
A healthy control with no history of P. aeruginosa infection showed a positive serology to the pathogen. Previous reports showed the occurrence of false negative results [22,31-33]. Since P. aeruginosa is an environmental pathogen, it is possible that this healthy control have got in contact with this bacterium in a period next to the study, leading to a transient antibody response [23,26]. Some authors have also shown that the St-Ag:1–17 preparation contains antigens which are cross-reactive with other gram-negative bacteria, including Haemophilus influenzae [17,28], which is a common microorganism in the respiratory tract of healthy children and adults. Høiby et al.  and Pedersen et al.  observed, in healthy controls, an increase in the antibody levels over the age. They attribute these facts to the normal occurrence of cross reactions with other pathogens and to a possible acquired immunity over time, after patient contact with P. aeruginosa.
It has been known for many years that chronic P. aeruginosa infection is preceded by an intermittent colonization with the pathogen and previous reports showed an overlap in the antibody levels between chronically infected and intermittent colonized CF patients [16,26,28]. In the present study, we did not find statistical difference in the antibody titers between the intermittently colonized and the free of infection group. Indeed, the majority (66.7%) of intermittently colonized patients had negative serology to P. aeruginosa; however, it is well known that P. aeruginosa isolation in respiratory samples not necessarily indicates infection, unless there is a specific humoral response [8,34]. In some patients, the colonization can persist for 1 to 2 years without eliciting an humoral immune response , which can be due to the presence of transient or non pathogenic strains, to the inadequate immune response to the pathogen , or even to the adequate immune response of the patient through secretory IgA [23,28]. Besides, there is a great interchange of patients between these two groups when longitudinal microbiological data are observed . On the other hand, the positive antibody levels of 33.3% of the intermittently colonized patients in the present report may suggest a change in the colonization/infection status for chronically infected [16,17]. These results draw attention to limitations in the Leeds criteria for definition of colonization and infection status , which are based solely on microbiological cultures. Serology results, in conjunction with more frequent respiratory cultures, can play an important role in elaborating a better definition for colonization and infection, mainly for patients classified as free of infection and intermittently colonized [16,17,26,35].
Among the CF patients with negative microbiological respiratory culture, the rate of seropositivity was 2.9% and 11.1% in the never colonized group and the free of infection group, respectively. Such findings may suggest limitations in respiratory culture, which may not reflect the lung microbiota [11,13,14], or even recurrent infection [27,34]. Many of these patients are not sputum producers, making more difficult to obtain representative respiratory samples, which may lead to false negative results. So, these groups constitute an important focus of interest for detection of serum anti-Pseudomonas antibodies .
A large variation in the performances of immunological assays used for the diagnosis of P. aeruginosa infection in CF patients has been observed , which is expected and is related to several factors including the antigen chosen, the criteria used for selection of the patients and healthy controls, the criteria adopted for classification of the colonization/infection status, the immune status of the patients at time of blood collection, the antibiotic treatment regimen, the intrinsic properties of the serological methods used for antibody detection, the antigen preparation, the lack of an appropriate reference standard and the method for calculating the cut-off value [18,30,36]. Here, we considered that culture results of one year prior to the serological analysis were not enough to provide a consistent classification of P. aeruginosa colonization/infection status. So, we used the 12 last culture results, seeking to optimize this classification. For classification according to serological results, we fixed negative and positive cut-off values through a ROC analysis, based on the results of a healthy population. This may help to avoid bias when determining the accuracy of the method [30,36]. The St-Ag:1–17 antigen was chosen by the fact of being a commercialized antigen; besides, the studies with the St-Ag:1–17 has been continuous over the last 30 years. Other antigens, like exotoxin A, elastase and alkaline protease are quorum-sensing regulated antigens  and require a higher bacterial load to elicit the immune response. Since St-Ag:1–17 constitutes a pool of 64 antigens from 17 different P. aeruginosa serotypes, the immune response is earlier in relation to these other antigens [17,23].