Insights into cystic fibrosis microbiology from the European tobramycin trial in cystic fibrosis

Journal of Cystic Fibrosis 1 (2002) S203–S208

Insights into cystic fibrosis microbiology from the European tobramycin trial in cystic fibrosis J.R.W. Govan* Department of Medical Microbiology, University of Edinburgh Medical School, Teviot Place, Edinburgh EH8 9AG, Scotland, UK

Abstract The infection of the airways of cystic fibrosis patients by Pseudomonas aeruginosa is a complex, multistaged process that is associated with a deterioration of lung function. The complexity of the formation of biofilms and their interaction with the immune system means that treatment with antibiotics has been an uncertain science. Tobramycin nebuliser solution (TNS) is a novel formulation of the antibiotic tobramycin developed specifically for inhalation. A recent large trial comparing TNS with inhaled colistin provided an opportunity to define further the effect of antibiotic treatment on microbial infection. In the TNS group, the percentage of patients with a tobramycin minimal inhibitory concentration (MIC) 04 mg ly1 increased from 38 to 49%, and the percentage of patients with a colistin MIC04 mg ly1 remained at 55%. In the colistin group, the percentage of patients with a colistin MIC04 mg ly1 remained at 34%, whereas the percentage of patients with a tobramycin MIC04 mg ly1 decreased from 27 to 16%. Furthermore, clinical and bacterial response to TNS and colistin was independent of the MIC at baseline. Neither antimicrobial therapy was associated with infection by Burkholderia cepacia or other inherently resistant pathogens. We conclude that conventional measures of antimicrobial resistance may underestimate the effectiveness of tobramycin and colistin when delivered at the high concentrations achieved with the TNS formulation. 䊚 2002 European Cystic Fibrosis Society. Published by Elsevier Science B.V. All rights reserved. Keywords: TOBI; Tobramycin; Cystic fibrosis; Pseudomonas aeruginosa

1. Introduction ‘The microbiology of lung disease in patients with cystic fibrosis is a subspeciality unto itself’ w1x. The above statement, written more than a decade ago, applies equally to the complex (and almost unique) relationship of Pseudomonas aeruginosa with cystic fibrosis (CF) lung disease and to the treatment of CF pulmonary infections. The recently completed European TOBI䉸 trial w49x, which compared the use of inhaled tobramycin nebuliser solution (TNS) with colistin, provided a unique opportunity to investigate aspects of antimicrobial response and resistance in this population. It is important to stress at the outset that the European TOBI trial focused on the treatment of established P. aeruginosa infection—a situation of which it has also been said, ‘the bacteria are seldom if ever eradicated’. To illustrate the challenge faced by clinicians and microbiologists in assessing antibiotic therapy in such chronic *Tel.: q44-131-650-3164; fax: q44-131-650-6653. E-mail address: [email protected] (J.R.W. Govan).

infections, it is necessary to summarize what is known of the relationship between P. aeruginosa and the CF lung. P. aeruginosa infection of CF airways is a life-long process involving several stages, and it provides a striking example of bacterial adaptation w2,3x. Firstly, an initial and often asymptomatic stage is signalled by culture of typical nonmucoid colonial morphotypes of P. aeruginosa. At this stage, infection may be intermittent and involve different strains of P. aeruginosa. This initial stage is characteristically followed by the gradual emergence of mucoid variants of the colonizing strain, a rise in antipseudomonas antibodies w4x and lifelimiting chronic infection w5,6x. The time taken for transition varies from patient to patient, and mucoid P. aeruginosa may be the first organism cultured from sputum if bacteriological investigations are infrequent. In one well-documented episode, involving primary acquisition of P. aeruginosa from a hydrotherapy pool, transition to mucoid occurred within 3 months w7x. In vivo, this transition is associated with the presence of bacterial biofilms, bronchial plugs and life-long in-

1569-1993/02/$ - see front matter 䊚 2002 European Cystic Fibrosis Society. Published by Elsevier Science B.V. All rights reserved. PII: S 1 5 6 9 - 1 9 9 3 Ž 0 2 . 0 0 0 0 5 - X

S204

J.R.W. Govan / Journal of Cystic Fibrosis 1 (2002) S203–S208

Table 1 Microbiological issues in the antibiotic treatment of CF lung infections Does antibiotic treatment reduce susceptibility of P. aeruginosa and prejudice further treatment? Are high MIC values at the start of treatment associated with poor clinical outcome? How do we define resistance in the context of CF airway infection and aerosolized antibiotics? Do changes occur in the lung microbial flora of treated patients, e.g. superinfections with B. cepacia and other inherently resistant CF pathogens?

tractable airway infection with the same strain of P. aeruginosa w3,8–10x. From the mid-1980s, a major therapeutic development in the management of pseudomonas infection in CF patients was the growing acceptance of initiating aggressive therapy at the first sputum culture of P. aeruginosa colonization; the aim was to clear the airways and delay or prevent subsequent colonization w11–13x. Strategically, early and aggressive therapy makes sense. At this stage, nonmucoid P. aeruginosa can be eradicated w11,14–16x, and reduction in sputum bacterial density can improve lung function w14,17x. Reduction in the pseudomonas population will also reduce the bacterial reservoir from which spontaneous alginate-producing mutants of P. aeruginosa arise, selected by their enhanced resistance to antibiotics and phagocytes w3,8x. Recent in vitro studies reported mucoid variants of P. aeruginosa PAO1 in biofilm cultures grown in the presence of hydrogen peroxide; the authors suggest that conversion to mucoid is a stressrelated response during exposure to oxygen radicals such as would be released from pulmonary neutrophils w18x. At high bacterial densities, cell-to-cell signalling has also been implicated in the development of pseudomonas biofilms w19x. Known as ‘quorum sensing’, this system relies on autoinducer molecules (homoserine lactones) produced by P. aeruginosa reaching sufficient concentrations to influence biofilm formation and other bacterial systems. Pseudomonas alginate is composed of the uronic acid b-D-mannuronate and its C-5 epimer a-L-guluronate w20x. As with algal alginates, pseudomonas alginate readily forms gels in the presence of physiological concentrations of Ca2q w21,22x. The viscosity of alginates is also enhanced by the presence of bronchial mucins and neutrophil DNA w23x. These properties help to explain the large gel-like microcolonies and biofilms seen in CF airways w23,24x and the term ‘frustrated phagocytosis’ when pulmonary neutrophils are faced with a large persistent target w21,25x. Whether biofilms are formed mainly as a result of selection of spontaneous mucoid variants, by induction in the whole population in response to environmental stress, or by cell-to-cell signalling remains unclear. It is also not clear how bronchial mucins, DNA, alginate and other bacterial polysaccharides contribute structurally to these sessile bacterial communities. What is clear is

that, in the CF lung, biofilms would provide a complex community of P. aeruginosa with formidable protection against pulmonary defences and antibiotic therapy. Having summarized the complex microbiology of P. aeruginosa and CF lung disease, we can appreciate that antibiotic therapy plays a major role in the management of CF lung disease, and probably influences the striking differences in the prevalence rates for P. aeruginosa and other CF pathogens in patients attending different CF centres w26x. However, the design and interpretation of antibiotic trials, in particular those involving chronic P. aeruginosa infection, presents formidable challenges and differs from antibiotic use in non-CF patients w27x. Table 1 highlights other major microbiological issues associated with the use of inhaled antibiotics in the treatment of chronic P. aeruginosa infection in CF patients. Defining sensitivity and resistance in the context of antibiotic therapy in CF is difficult and merits particular attention. In non-CF infections, the term ‘resistance’ is usually defined as failure to eradicate the bacterial pathogen and reduce or eliminate the clinical symptoms of infection. If these criteria are applied to established CF lung infections caused by P. aeruginosa, then almost all could be defined as resistant. However, if the major end point is clinical response, then what parameters of response are acceptable? Are serial measurements of inflammatory markers, such as C-reactive protein, a better way of measuring how the patients rather than the pathogen is ‘responding’ to therapy? w28x. In a progressive disease, such as CF, is it acceptable to achieve short-term or long-term improvement in lung function without eradication or even reduction in bacterial load? Is resistance an inevitable but acceptable price to pay for improved lung function? Finally, in laboratory tests, what is the appropriate sample population on which the definition of bacterial susceptibility and resistance is made? The antimicrobial breakpoint at the site of infection is typically defined as the highest level of antibiotic which can be achieved at that site without toxicity—a bacterial pathogen with a minimal inhibitory concentration (MIC) greater than the breakpoint is defined as resistant. In non-CF infections, antibiotic susceptibility is usually determined by testing one or more bacterial colonies picked from a primary culture plate; this in vitro sensitivity is then used to predict the likelihood of eradicating the whole population of infecting organisms

J.R.W. Govan / Journal of Cystic Fibrosis 1 (2002) S203–S208 Table 2 Baseline demographics w1x

Number of patients Log10 CFU mly1a Tobramycin MIC 04 mg mly1, n (%)b Colistin MIC 04 mg mly1, n (%)c

S205

Table 3 Patients with highest tobramycin P. aeruginosa MIC04 mg ly1 TNS

Colistin

53 6.3"1.5 20 (39) 13 (25)

62 6.8"1.3 31 (52) 19 (32)

a

No significant differences were detected between the groups. One TNS patient missing CFU. c Two TNS and two colistin patients missing MIC. b

which may exceed 109 colony-forming units mly1. In many non-CF infections, the antibiotic sensitivity patterns of multiple colonial samples is relatively uniform. This is not the case in the chronically colonized CF lung, where a range of MIC values can be observed even if infection involves a single strain of P. aeruginosa w14,29,30x. Furthermore, in CF, drug delivery is handicapped by the presence of bronchial plugs and the location of P. aeruginosa within the protective matrix of bacterial biofilms w3,24x. Finally, in vitro susceptibility testing is exposed to many variables; these range from the growth state of the pathogen (log or stationary phases), inoculum density, culture medium and whether testing is based on antibiotic impregnated disks, agar dilution or microbroth dilution methodologies. 1.1. Use of inhaled tobramycin and colistin in the treatment of P. aeruginosa infection The design and baseline demographics relating to the European TOBI trial are presented by Hodson et al. elsewhere (this supplement and Ref. w31x). This paper will focus on the microbiological aspects of the trial. Bacterial culture, storage, and susceptibility testing (microbroth dilution) were designed to replicate those used in the US trials of intermittent inhaled tobramycin in patients with CF w32,33x. MIC values based on laboratory identification of different colonial morphotypes were developed from a previously described scheme w34x. In an attempt to deal with the heterogeneity of the P. aeruginosa population discussed earlier, two MIC values were calculated. Firstly, the highest MIC value observed when representatives of each colonial morphotype were examined (the maximum MIC), and secondly, the MIC value obtained from the colonial morphotype, which represented the majority of the P. aeruginosa population (the maximumdensity MIC). 1.2. Baseline demographics P. aeruginosa demographics at the start of the trial are outlined in Table 2. These data confirm that the target populations for the two treatment groups are comparable. Namely, 20 (39%) of the TNS-treated

Baseline Week 4

TNS (%) (ns45)

Colistin (%) (ns47)

38 49

55 55

patients harboured P. aeruginosa, with the highest MIC values exceeding the resistance breakpoint of 4 mg ly1; similarly, 19 (32%) of the colistin-treated patients were infected by organisms, with the highest MIC of colistin greater than 4 mg ly1. 1.3. Change in MIC values Comparative analysis of MICs at the start and end of therapy indicated that patients whose highest tobramycin MIC exceeded 4 mg ly1 rose in the TNS-treated patients from 38 to 49%, but remained at 55% in those treated with colistin (Table 3). In patients treated with colistin, colistin MICs above 4 mg ly1 remained stable at 34%; interestingly, tobramycin MICs reduced from 27 to 16% (Table 4). 1.4. MIC value and clinical response Fig. 1a,b present the data from individual patients, plotting the mean relative change in lung function (FEV1% predicted) against the baseline maximum MIC of tobramycin (Fig. 1a), and the maximum density tobramycin MIC value at baseline (Fig. 1b). In both MIC assessments, clinical change during treatment is not predicted by the MIC values at baseline. When the results for colistin MICs at baseline were plotted against clinical outcome, a similar outcome was revealed: namely, colistin MIC at baseline did not predict clinical outcome (data not shown). 1.5. 28-Day change in P. aeruginosa log10 CFU mly1 (intent-to-treat patients) The sputum density of P. aeruginosa (log10 CFU) during 28-day treatment with tobramycin or colistin was reduced by 0.9 and 0.6, respectively. The differences were not significant and may primarily reflect killing of planktonic forms of P. aeruginosa existing outside the protection of the biofilm. Table 4 Patients with highest colistin P. aeruginosa MIC04 mg ly1

Baseline Week 4

TNS (%) (ns45)

Colistin (%) (ns47)

27 16

34 34

J.R.W. Govan / Journal of Cystic Fibrosis 1 (2002) S203–S208

S206

Fig. 1. FEV1% response by baseline MIC: (a) Maximum tobramycin MIC; (b) Maximum density tobramycin MIC.

1.6. Superinfection with B. cepacia and other inherently resistant pathogens There was no emergence of B. cepacia or increased isolation of other inherently resistant Gram-negative pathogens, including Stenotrophomonas maltophila and Alcaligenes xylosoxidans, during the 4 weeks of inhaled tobramycin or colistin. There was a small increase in the prevalence of Aspergillus species, but this was not reflected in an increased reporting of acute bronchopulmonary aspergillosis. Of significance, isolation of B. cepacia did not increase over an extended treatment with inhaled tobramycin in ongoing North American trials w35x. 2. Discussion The observation that clinical and bacteriological response to tobramycin and colistin is independent of MIC value at the start of therapy is not novel; similar results have been observed in relation to the use of tobramycin in combination with azlocillin w36x and tobramycin with azlocillin and timentin w37x. It is also

well known that frequent use of antibiotics to treat chronic P. aeruginosa increases the risk of bacterial resistance, reinforcing the need for regular and expert microbiological monitoring w27,38x. However, resistance is not synonymous with a worse prognosis, even in those CF patients who receive a transplant w39,40x, and it has to be balanced with the benefits of aggressive therapy in a situation where, if untreated, lung disease is relentlessly progressive. The microbiological reasons behind this apparent anomaly are complex and are discussed under the headings given below. 2.1. Heterogeneity of the pseudomonas population As discussed earlier, the P. aeruginosa population present in CF airways is phenotypically heterogeneous, and this includes heterogeneity in MIC values at the beginning and end of therapy w14,29,30x. Underlying this heterogeneity are bacterial factors, such as the influence of chromosomal loci regulating antibiotic hypersusceptibility w41x, different muc alleles controlling alginate biosynthesis w3x, differences in alginate acetyl content and uronic acid block structure w20x and lipo-

J.R.W. Govan / Journal of Cystic Fibrosis 1 (2002) S203–S208

polysaccharide chemotype (ranging from smooth to rough 0-antigen deficient) w3x. Associated with, but not dependent on, these phenotypes are at least five colonial morphotypes of P. aeruginosa. Heterogeneity in P. aeruginosa also occurs in non-CF infections, but is best exemplified in the life-long infections encountered in the CF lung. In the European TOBI trial, the use of the highest MIC and highest-density MIC values was a novel attempt to address the problem of heterogeneity; however, at best these measurements are a compromise, with the diversity of the large bacterial population (around 108 mly1) remaining virtually unmeasurable. 2.2. Virulence attenuation of resistant organisms Mechanisms involved in decreasing bacterial susceptibility to antibiotics, either permanently or transiently, include changes in the antibiotic target, production of antibiotic inactivating agents, and reduced uptake of the antimicrobial agent or impermeability. Reduced permeability, which has been shown to be the major mechanism responsible for reduced susceptibility to inhaled tobramycin w42x, is typically adaptive or transient w43x. Such resistant organisms are deficient in the production of extracellular virulence determinants w22x and are attenuated in vivo w44x. 2.3. Virulence attenuation at sub MIC concentrations During most of the antibiotic era, the therapeutic benefits of antibiotics have focused on their ability to eliminate infectious disease by killing or inhibiting the growth of bacterial pathogens. It has long been recognized, however, that subinhibitory concentrations of antibiotics can suppress bacterial virulence w45x. Antibiotic-induced suppression of virulence is particularly relevant to CF lung disease where, with the exception of aerosol delivery, even the most potent antipseudomonal agents seldom reach MIC values and, at the site of infection, the bacteria are protected by biofilms, bronchial abscesses, and inflammatory exudates w2,23x. Aminoglycosides at subinhibitory concentrations inhibit production of iron-chelating siderophores w46x, in a manner similar to the suppression of bacterial proteases by ceftazidime and ciprofloxacin w14,22,47x. These effects are strain dependent and are independent of MIC values w22x. Most antibiotics have been found to inhibit adherence of P. aeruginosa to CF tracheal cells at subinhibitory concentrations w48x. Tobramycin and ciprofloxacin at sub MIC concentrations reduce alginate biosynthesis, arguably the most important bacterial virulence factor in CF airways. Indeed, when measured at the level of transcription using reporter genes, antibiotics reduce alginate biosynthesis in mucoid P. aeruginosa at concentrations as low as 5% of the MIC w7x.

S207

2.4. Aerosol delivery provides high concentrations of antibiotic in the lung A final consideration in the interpretation of the European TOBI trial is the difficulty in defining ‘bacterial resistance’, when definitions have previously been based on non-inhaled administration. Based on parenteral administration, breakpoints of tobramycin and colistin have been set at 4 mg ly1. However, aerosol delivery can provide concentrations 100-fold higher than systemic concentrations w32x; thus, for inhaled antibiotics, new breakpoints for resistance seem appropriate. At present, too few patients harbouring ‘resistant’ P. aeruginosa have been examined to define such a breakpoint. Such a redefinition must also reflect the antibiotic-binding capacity of sputum. Since in vitro studies have shown that tobramycin concentrations of approximately 10 times the MIC of the bacteria are required to inhibit the growth of P. aeruginosa in sputum, a level of resistance for aerosol delivery of tobramycin may be )128 mg ly1 w33x. References w1x Stuttman HR, Marks MI. Pulmonary infection in children with cystic fibrosis. Sem. Resp. Med. 1987;2:166 –76. w2x Gilligan PH. Microbiology of airway disease in patients with cystic fibrosis. Clin. Microbiol. Rev. 1991;4:35 –51. w3x Govan JRW, Deretic V. Microbial pathogenesis in cystic fibrosis: mucoid Pseudomonas aeruginosa and Burkholderia cepacia. Microbiol. Rev. 1996;60:539 –74. w4x Pedersen SS, Espersen F, Hoiby N. Diagnosis of chronic Pseudomonas aeruginosa infection in cystic fibrosis by enzyme linked immunosorbent assays. J. Clin. Microbiol. 1987;25:1830 –6. w5x Kerem K, Corey M, Gold R, Levison H. Pulmonary function and clinical course in patients with cystic fibrosis after pulmonary colonisation with Pseudomonas aeruginosa. J. Pediatr. 1990;116:714 –9. w6x Pamukcu A, Bush A, Buchdahl R. Effects of Pseudomonas aeruginosa colonisation on lung function and anthropomorphic variables in children with cystic fibrosis. Pediatr. Pulmonol. 1995;19:10 –5. w7x Govan JRW, Nelson JW. Microbiology of lung infection in cystic fibrosis. Brit. Med. Bull. 1992;48:912 –30. w8x Pedersen SS, Hoiby N, Espersen F, Koch C. Role of alginate in infection with mucoid Pseudomonas aeruginosa in cystic fibrosis. Thorax 1992;47:6 –13. w9x Tummler B, Kiewitz C. Cystic fibrosis: an inherited susceptibility to bacterial respiratory infections. Mol. Med. Today 1999;5:351 –8. w10x Hutchison ML, Govan JRW. Pathogenicity of microbes associated with cystic fibrosis. Microb. Infect. 1999;1:1005 –14. w11x Littlewood JM, Miller MG, Ghoneim AG, Ramsden CH. Nebulised colomycin for early pseudomonas colonization in cystic fibrosis. Lancet 1985;I:865. w12x Valerius NH, Koch C, Hoiby N. Prevention of chronic Pseudomonas aeruginosa colonisation in cystic fibrosis by early treatment. Lancet 1991;338:725 –6. w13x Frederiksen B, Koch C, Hoiby N. Antibiotic treatment of initial colonisation with Pseudomonas aeruginosa postpones chronic

S208

w14x

w15x

w16x

w17x

w18x

w19x

w20x w21x

w22x

w23x

w24x

w25x

w26x

w27x

w28x

w29x

w30x

w31x

J.R.W. Govan / Journal of Cystic Fibrosis 1 (2002) S203–S208 infection and prevents deterioration of pulmonary function in cystic fibrosis. Pediatr. Pulmonol. 1997;23:330 –5. Govan JRW, Doherty C, Glass S. Rational parameters for antibiotic therapy in patients with cystic fibrosis. Infection 1987;15:300 –7. Denton M, Wilcox MH. Antimicrobial treatment of pulmonary colonization and infection by Pseudomonas aeruginosa in cystic fibrosis patients. J. Antimicrob. Chemother. 1997;40:468 –74. Wiesemann HG, Steinkamp G, Ratjen F, Bauernfeid A, Pryzyklenk B, Doring G. Placebo-controlled double blind randomized study of aerosolized tobramycin for early treatment of Pseudomonas aeruginosa colonisation in cystic fibrosis. Pediatr. Pulmonol. 1998;25:88 –92. Regelman WE, Elliot GR, Warwick WJ, Clawson CC. Reduction in sputum Pseudomonas aeruginosa density by antibiotic improves lung function in cystic fibrosis more than do bronchodilators and chest physiotherapy. Am. Rev. Resp. Dis. 1990;141:914 –21. Mathee K, Ciofo O, Sternberg C, et al. Mucoid conversion of Pseudomonas aeruginosa by hydrogen peroxide: a mechanism for virulence activation in the cystic fibrosis lung. Microbiology 1999;145:1349 –57. Davies DG, Mparsek MR, Person JP, Iglewski BH, Costeron JW, Greenberg EP. The involvement of cell-to-cell signals in the development of a bacterial biofilm. Science 1998;280:295 – 8. Gacesa P. Bacterial alginate biosynthesis—recent progress and future prospects. Microbiology 1998;144:1133 –43. Govan JRW, Harris GS. Pseudomonas aeruginosa and cystic fibrosis: unusual adaptation and pathogenesis. Microbiol. Sci. 1986;3:302 –8. Govan JRW. Alginate biosynthesis and other unusual characteristics associated with the pathogenesis of Pseudomonas aeruginosa in cystic fibrosis. In: Donachie W, Griffith E, Stephens J, editors. Bacterial infections of respiratory and gastrointestinal mucosae. Oxford: IRL Press, 1988. p. 67 –96. Govan JRW. Alginate and antibiotics. In: Høiby N, Pedersen ¨ SS, Shand GH, Doring G, Holder IA, editors. Basel: Kargee, 1989. p. 88 –96. Lam J, Chan R, Lam K, Costerton JW. Production of mucoid microcolonies by Pseudomonas aeruginosa within infected lungs in cystic fibrosis. Infect. Immun. 1980;28:546 –56. Govan JRW, Glass S. The microbiology and therapy of cystic fibrosis lung infections. Antimicrob. Chemother. 1996;37:1155 –64. Bauernfeind A, Marks MI, Strandvik B, editors. Cystic fibrosis pulmonary infection: lessons from around the World. Basel: Birkhauser Verlag, 1996. The Cystic Fibrosis Trust’s Antibiotic Group. Antibiotic treatment for cystic fibrosis; Report of the UK CF Trust’s Antibiotic Group. Cystic Fibrosis, UK: Bromley, 2000. Glass S, Hayward C, Govan JRW. Serum C-reactive protein in assessment of pulmonary exacerbation and antimicrobial therapy in cystic fibrosis. J. Pediatr. 1988;113:76 –9. Seale TW, Thirkhill H, Tarpay M, Flux M, Rennert OM. Serotypes and antibiotic susceptibilities from single sputa of cystic fibrosis patients. J. Clin. Microbiol. 1979;9:72 –8. Thomassen MJ, Demko CA, Boxerbaum B, Stern RC, Kuchenbrod PJ. Multiple isolates of Pseudomonas aeruginosa with different antimicrobial susceptibility patterns from patients with cystic fibrosis. J. Infect. Dis. 1979;140:873 –80. Hodson ME, Gallagher CG, Govan JR. An open label randomised clinical trial of the efficacy and safety of tobramycin nebuliser solution or nebulised colistin sulphomethate sodium in patients with cystic fibrosis. Eur. Resp. J. in press.

w32x Ramsey BW, Pepe MS, Quan JM, et al. Intermittent administration of inhaled tobramycin in patients with cystic fibrosis. N Engl. J. Med. 1999;340:23 –30 (17 authors plus members of the Cystic Fibrosis Inhaled Tobramycin Study Group). w33x Burns JL. Clinical definition of susceptibility and resistance to tobramycin. Pediatr. Pulmonol. 1997;14(Suppl):139 –40. w34x Wolter JM, Kotsiou G, McCormick JG. Mixed morphotype testing of Pseudomonas aeruginosa cultures from cystic fibrosis patients. J. Med. Microbiol. 1995;42:220 –4. w35x Burns JL, Van Dalfsen JM, Shawar RM, et al. Effect of chronic intermittent administration of inhaled tobramycin on respiratory microbial flora in patients with cystic fibrosis. J. Infect. Dis. 1999;179:1190 –6. w36x Smith AL, Ramsey BW, Hedges DL, et al. Safety of aerosol tobramycin administration for 3 months to patients with cystic fibrosis. Pediatr. Pulmonol. 1989;7:265 –71. w37x McLaughlin FJ, Matthews WJ, Strieder DJ, et al. Clinical and bacteriological responses to three antibiotic regimens for acute exacerbations of cystic fibrosis: ticarcillin–tobramycin, azlocillin–tobramycin and azlocillin–placebo. J. Infect. Dis. 1983;147:559 –67. w38x Saimon L, Prince A. Microbial resistance. In: Bauernfeind A, Marks MI, Strandvik B, editors. Cystic fibrosis. Pulmonary infections: lessons from around the World. Basel: Birhauser Verlag, 1996. p. 51 –64. w39x Aris RM, Gilligan PH, Neuringer IP, Gott KK, Rea J, Yankaskas JR. The effects of panresistant bacteria in cystic fibrosis patients on lung transplantation outcome. Am. J. Respir. Crit. Care Med. 1997;155:1699 –704. w40x Burns JL, Ramsey BW, Fiel SB. In vitro antibiotic susceptibility testing of CF Pseudomonas isolates does not predict FEV1 response to combination intravenous antibiotic therapy (abstr). XIIIth International Cystic Fibrosis Congress, Stockholm, Sweden, 2000, p. 174. w41x Fyfe JM, Govan JRW. Chromosomal loci associated with antibiotic hypersusceptibility in pulmonary isolates of Pseudomonas aeruginosa. J. Gen. Microbiol. 1984;130:825 –34. w42x Macleod DL, Nelson LE, Shawar RM, et al. Aminoglycosideresistance mechanisms for cystic fibrosis Pseudomonas aeruginosa isolates are unchanged by long-term, intermittent, inhaled tobramycin treatment. J. Infect. Dis. 2000;181:1180 – 4. w43x Barclay ML, Begg EJ, Chambers ST, Thornley PE, Pattemore PK, Grimwood K. Adaptive response to tobramycin in Pseudomonas aeruginosa lung infection in cystic fibrosis. J. Antimicrob. Chemother. 1996;38:853 –8. w44x Gerber AU, Craig WA. Aminoglycoside-selected subpopulations of Pseudomonas aeruginosa: characterization and virulence in normal and leukopenic mice. J. Lab. Clin. Med. 1982;100:671 –81. w45x Lorian V. Some effects of subinhibitory concentrations of antibiotics on bacteria. Bull. NY Acad. Med. 1975;51:212 –59. w46x Morris G, Brown MRW. Novel modes of action of aminoglycosides antibiotics against Pseudomonas aeruginosa. Lancet 1998;I:1359 –61. w47x Dalhoff A, Doring G. Interference of ciprofloxacin with the expression of pathogenicity factors of Pseudomonas aeruginosa. In: Adam D, Halm H, OpferKuch W, editors. The influence of antibiotics on host parasite relationship. Berlin: Springer, 1985. p. 246 –55. w48x Wolter JM, Kotsiou G, McCormick JG. Mixed morphotype testing of Pseudomonas aeruginosa cultures from cystic fibrosis patients. J. Med. Microbiol. 1995;42:220 –4. w49x Hodson ME, Gallagher CG, Govan JRW. A randomised trial of nebulized tobramycin or colistin in cystic fibrosis. Eur. Respir. J. 2002;20:658-664.